The Role of Plant-Derived Quaternary Ammonium Compounds, Including Carnitine and Choline-O-Sulfate, on the Biology of the Plant

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

2015 The oler of plant-derived quaternary ammonium compounds, including carnitine and choline-O- sulfate, on the biology of the plant pathogen Pseudomonas syringae and its interactions with the host species Phaseolus vulgaris Michael David Millican Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Agriculture Commons, Microbiology Commons, and the Plant Pathology Commons

Recommended Citation Millican, Michael David, "The or le of plant-derived quaternary ammonium compounds, including carnitine and choline-O-sulfate, on the biology of the plant pathogen Pseudomonas syringae and its interactions with the host species Phaseolus vulgaris" (2015). Graduate Theses and Dissertations. 14948. https://lib.dr.iastate.edu/etd/14948

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 Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected].

The role of plant-derived quaternary ammonium compounds, including carnitine and choline-O-sulfate, on the biology of the plant pathogen Pseudomonas syringae and its interactions with the host species Phaseolus vulgaris

Michael D. Millican

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 Beattie, Major Professor Thomas Bobik Qijing Zhang Byron Brehm-Stecher Larry Halverson

Iowa State University

Ames, Iowa

2015

DEDICATION

To my parents, Doug and Mary Millican. You have selflessly provided me the opportunity to pursue my dreams and no words could express my admiration and appreciation for you and all that you have done for me. And to my wife, Sarani Rangarajan, for your unwavering support I would like to thank you from the bottom of my heart, but for you my heart has no bottom.

iii

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...... v

ABSTRACT………………………………...... vi

CHAPTER 1 GENERAL INTRODUCTION ...... 1

Introduction ...... 1 Thesis Organization ...... 2 Literature Review ...... 3 References ...... 13

CHAPTER 2 CARNITINE RELEASED FROM GERMINATING PHASEOLUS VULGARIS SEEDS PROMOTES THE GROWTH OF PSEUDOMONAS SYRINGAE AND REQUIRES DHCAB FOR REMOVAL OF THE INTERMEDIATE ACETOACETATE FOR CATABOLISM ...... 26

Abstract to ...... 26 Introduction ...... 27 Methods and Materials ...... 30 Results of the ...... 37 Discussion ...... 50 References ...... 55

CHAPTER 3 CHOLINE-O-SULFATE CATABOLISM IN PSEUDOMONAS SYRINGAE REQUIRES A SULFATE EXPORTER THAT MAY FUNCTION TO RELIEVE SULFATE TOXICITY AND IS ENCODED IN A LOCUS WITH CHOLINE-O-SULFATE SULFATASE ...... 60

Abstract to ...... 60 Introduction ...... 61 Methods and Materials ...... 64 Results of the ...... 66 Discussion ...... 75 References ...... 84

CHAPTER 4 PLANT TISSUE-SPECIFIC BENEFITS OF QUATERNARY AMMONIUM COMPOUNDS TO PSEUDOMONAS SYRINGAE DURING LEAF AND SEED COLONIZATION ...... 89

Abstract to ...... 89 Introduction ...... 90 Methods and Materials ...... 95 Results of the ...... 98 Discussion ...... 105 References ...... 113

CHAPTER 5 CONCLUSIONS AND FUTURE DIRECTIONS ...... 118

References ...... 121

APPENDIX A SUPPLEMENTARY TABLES AND FIGURES FOR CHAPTER II ...... 123

APPENDIX B SUPPLEMENTARY TABLES AND FIGURES FOR CHAPTER III ...... 131

APPENDIX C SUPPLEMENTARY FIGURES FOR CHAPTER IV ...... 140

ACKNOWLEDGMENTS

I would like to thank my advisor, Gwyn A. Beattie, for her endless patience and thoughtful guidance over these years. I would also like to thank my committee members,

Thomas Bobik, Qijing Zhang, Byron Brehm-Stecher, and Larry Halverson for their guidance and support throughout the course of this research.

In addition, I would also like to thank my wife, without whom this dissertation would be an jarrbelled un reedabel dizazter. I want to thank my friends and family who have supported me all these years. Lastly, but certainly not least, I would like to thank Regina

McGrane for her thoughtful input and countless discussions.

ABSTRACT

Knowledge of the chemicals available during plant-microbe interactions is useful for promoting the growth of beneficial bacteria and predicting pathogen populations and thus the probability of disease. Pseudomonas syringae B728a is able to utilize plant-derived quaternary ammonium compounds (QACs), such as L-carnitine, choline, and choline derivatives, as osmoprotectants and nutrient sources. We explored in detail the biological roles of carnitine and choline-O-sulfate in P. syringae cells, including in culture and during host colonization. Although genes involved in carnitine catabolism were identified in other organisms, the roles of these genes have not been fully elucidated. We analyzed conflicting reports of the carnitine to glycine betaine degradation pathway and provided evidence for a modified pathway based on the growth and accumulation of metabolic intermediates in catabolic mutants. Although carnitine is converted to glycine betaine for accumulation as a compatible solute in B728a, we found that carnitine itself can function as a compatible solute in P. syringae. This was based on both the ability of a carnitine catabolic mutant to derive osmoprotection from carnitine, and the accumulation of carnitine in this mutant as detected using MALDI-MSI. Although we did not detect evidence for carnitine production by bean leaves, we discovered that carnitine is released during bean seed germination, specifically during radicle emergence and elongation, and this carnitine strongly contributed to the growth of B728a on germinating seeds. In exploring the genes involved in the catabolism of choline-O-sulfate (COS), we confirmed that a predicted COS sulfatase was required for catabolizing COS as a nutrient and osmoprotectant, but did not impact P. vulgaris leaf or

vii seed colonization, indicating little to no available COS in bean plants. However, we found unexpected evidence that a sulfate exporter is critical to the use of COS as a nutrient or osmoprotectant; this supports a model in which this exporter removes sulfate released from the cleavage of COS and prevents accumulation to toxic levels. Although COS was not a significant source of sulfur during colonization of the plant host, we used transcriptome data to elaborate a likely metabolic network highlighting multiple organosulfur sources that supply leaf surface bacteria with sulfur. Lastly, we systematically evaluated the roles of a broad set of QACs, including choline, phosphorylcholine, glycine betaine, carnitine, and

COS, in the colonization of leaves and germinating seeds by B728a using a collection of metabolic and transporter mutants. P. syringae derives osmoprotection from choline while colonizing the leaf surface and our studies demonstrated that choline was the most abundant and influential QAC present on P. vulgaris leaves. During colonization of the spermosphere,

P. syringae derives nutrients from choline released throughout seed development, as demonstrated by a compositional analysis of seed exudates. Collectively, these results demonstrate a key role for QACs as a driver of bacterial population dynamics on seeds and leaves, and thus support the possibility of exploiting QAC availability for the establishment of targeted beneficial microbes on plants.

CHAPTER 1. GENERAL INTRODUCTION

Introduction

Pseudomonas syringae pv. syringae B728a is a well characterized foliar pathogen and epiphyte. P. syringae is generally limited in nutrients on leaf surfaces and is constantly subjected to fluctuations in UV radiation, temperature, and water availability, with water limitation being particularly detrimental to establishing and/or maintaining populations. P. syringae utilizes compatible solutes for tolerance of water limitation encountered on leaves.

Compatible solutes can either be synthesized de novo, such as trehalose and N- acetylglutaminylglutamine amide (NAGGN), or taken up from the environment as osmoprotectants. Quaternary ammonium compounds (QACs) have been demonstrated to be particularly effective in functioning as osmoprotectants as well as nutrient sources in organisms from all domains of life. P. syringae contains an array of transporters to import

QACs, some of which function exclusively under osmotic stress, indicating P. syringae has evolved to exploit environments rich in QACs. Choline, glycine betaine, and carnitine have been demonstrated to function as nutrients and osmoprotectants in P. syringae. Carnitine catabolism has been investigated in Pseudomonas aeruginosa, although the exact nature of the catabolic pathway is unclear. In this work we will investigate the catabolic pathway that

P. syringae, and likely all Pseudomonads, utilize to degrade carnitine; plants are not known to produce abundant carnitine, so we will evaluate the role of carnitine catabolism in plant- microbe interactions to understand why a plant pathogen has evolved to retain this ability.

Furthermore, because P. syringae contains specific transport proteins for choline-O-sulfate

(COS), we will examine if P. syringae is able to utilize COS as a nutrient source and osmoprotectant. We will investigate the contributions of COS, a potential sulfur source, to plant colonization and which genes in the sulfur acquisition network are active during colonization as an indication as to which types of sulfur sources P. syringae encounters on leaf surfaces. Finally, we will investigate the relative impact of choline, glycine betaine, carnitine, COS, and phosphorylcholine during colonization of the plant host Phaseolus vulgaris to broaden our understanding of how microbes utilize QACs during colonization of distinct plant tissues, namely leaves and seeds.

Thesis Organization

This dissertation is organized into five chapters. Chapter 1 provides a review of relevant literature relating to adaptations of Pseudomonas syringae to water limitation stress, utilization of carnitine in bacteria and sulfate assimilation in Pseudomonads. Chapter 2 evaluates the carnitine degradation pathway in P. syringae and examines the impact of carnitine catabolism on P. syringae colonization of common bean (Phaseolus vulgaris). In this study, I collaborated with Adam Klein and Young-Jin Lee, Ph.D. candidate and

Professor, respectively, of the Department of Chemistry at the Iowa State University.

Specifically, they contributed to the work by collecting and analyzing the mass spectrometric data for two independent experiments: compatible solute accumulation in cell samples and compositional profiling of seed exudates. Chapter 3 investigates the utilization of choline-O- sulfate by P. syringae and the unusual requirement of sulfate export coupled to degradation of a sulfate ester. Chapter 4 explores the relative impact of plant-derived quaternary ammonium compounds on P. syringae colonization of P. vulgaris. Adam Klein and Young-

Jin Lee also contributed to the compositional analysis of seed exudates here. Chapter 5 describes my conclusions and possible avenues of future work.

Literature Review

Pseudomonas syringae, a plant pathogen and model epiphyte.

Pseudomonas syringae was first isolated from diseased lilac (Syringa vulgaris) by

MW Beijerinck in 1899 (112). The organism was later named and characterized in 1905 by

CJJ van Hall (47, 112). Over the years, P. syringae has been established as a model organism in plant-microbe interactions and extensively studied for its pathogenesis and ecology. P. syringae is a gram-negative species that is comprised of over 50 pathovars (55), which are defined based on host specificity. P. syringae causes a wide variety of diseases from spots, specks, and blights on leaves, fruits, flowers and shoots, to galls and cankers on woody tissues. Although the majority of P. syringae strains are foliar pathogens and must gain entry into the interior space of the leaf in order to cause disease, common entry points into the apoplast are wounds or cracks on the leaf surface, stomata and the hydathodes (79,

82, 110). Although P. syringae can directly immigrate via these entry points, P. syringae generally first establishes populations on the leaf surface before inciting disease symptoms.

Thus, it has been extensively studied as a model organism on leaf surfaces, that is, in epiphytic sites (47, 48).

P. syringae adaptations to water limitation in epiphytic sites.

Organisms colonizing the phyllosphere are subjected to constant fluctuations in UV radiation, temperature, and water availability (76). Bacteria such as P. syringae have

4 evolved several mechanisms to adapt to water limitation, such as the production of exopolymeric substances (EPS), the formation of aggregates, and the accumulation of compatible solutes (36, 71, 86, 96).

The formation of aggregates and EPS production generally occur together, resulting in a biofilm-like structure on the leaf surface that helps protect the cells from desiccation stress. Alginate, the primary component of P. syringae EPS (93, 126), can absorb a large amount of water. Water limitation stimulates the production of alginate, resulting in the encapsulation of surrounding cells and the formation of bacterial aggregates (72). Larger aggregates of P. syringae are associated with increased epiphytic fitness, while being subjected to water limited conditions (84).

Water limitation on leaves is likely due to the combined stresses of a low water content and osmotic stress, which results from high concentrations of solutes in the available water. While EPS production and aggregation are useful strategies for dealing with low water availability in the long term, cells need other mechanisms to deal with osmotic stresses following immigration but prior to establishing aggregated population structures. Bacterial cells accumulate compatible solutes to counteract the efflux of water and prevent crenation of the membranes in the presence of hyperosmolarity (60, 71, 121). Compatible solutes are osmolytes that can be accumulated in the cytoplasm without interfering with biological processes (65). These osmolytes have an overall neutral charge, being either uncharged or zwitterions, and are comprised mainly of organic acids and saccharides (125). Beyond osmotolerance, compatible solutes also provide tolerance to both heat and cold stress, which may both occur in the phyllosphere. Temperature stress adaptation through compatible

5 solute accumulation has not been demonstrated in P. syringae but has been in Pseudomonas putida, Bacillus subtilis, and Listeria monocytogenes, among others (8, 49, 64).

To combat osmotic stress, P. syringae synthesizes de novo two major compatible solutes, N-acetylglutaminylglutamine amide (NAGGN) and trehalose. For the production of

NAGGN, P. syringae utilizes glutamine as a precursor. GgnB, an acetyltransferase, adds an acetyl group to one glutamine, and GgnA, an amidotransferase, adds an amide group to another glutamine. Finally, GgnC, a transpeptidase, links the acetylglutamine and glutamine amide, producing NAGGN (100). Elimination of NAGGN production in P. syringae pv. syringae B728a and P. syringae pv. tomato DC3000 reduced growth under osmotic stress, although B728a experienced a much greater reduction (59), indicating these highly related strains of P. syringae exhibit quantitatively distinct responses to osmotic stress. Trehalose, the other major compatible solute synthesized de novo by P. syringae, is a disaccharide produced through two pathways, which are mediated by TreYZ and TreS. TreYZ degrades maltoogliosaccharides, producing maltotriose or maltose and trehalose; TreS converts maltose to trehalose (39). In P. syringe DC3000, deletion of either TreS or TreYZ pathways eliminated trehalose production (39), reducing epiphytic fitness (40) and establishing the importance of trehalose production in epiphytic survival. Accumulation of both of these compatible solutes contributes to osmotolerance, leading to growth and survival in the phyllosphere.

P. syringae preferentially utilizes osmoprotectants, if they are available in the environment, to adapt to osmotic stress. Osmoprotectants are compounds that are taken up from the environment and either degraded to, or utilized directly as, compatible solutes (41).

Quaternary ammonium compounds (QACs) are a particularly interesting class of chemicals

6 that function as osmoprotectants; these include glycine betaine, carnitine, choline, and choline derivatives like choline-O-sulfate (COS) and phosphorylcholine (15, 16, 59, 81).

While both carnitine and COS have a limited ability to accumulate as a compatible solute

(45, 59, 88), glycine betaine functions as a compatible solute in all domains of life (5, 99). In

Pseudomonads, these QACs are generally converted to glycine betaine to enhance osmotolerance; glycine betaine can alternatively be catabolized as a nutrient source (74, 119).

P. syringae uses an extensive transport system for the import of QACs for osmoprotection and nutrient acquisition. The ABC transporter OpuC imports choline, glycine betaine, choline-O-sulfate, and carnitine under high osmotic stress, whereas the

Betaine/Carnitine/Choline Transporter (BCCT) BetT transports choline and acetylcholine under both high and low osmotic stress (23). Interestingly, P. syringae contains another

ABC transporter, CbcWV, which functions primarily at low osmolarity and is involved predominantly in nutrient acquisition. Thus, P. syringae has evolved to use QACs not only as osmoprotectants but also as nutrient sources.

In this thesis, I investigate the utilization of the QACs carnitine (Chapter II) and choline-O-sulfate (Chapter III) during P. syringae growth in culture and on plants, and generate models of how various QACs influence this colonization in a tissue-specific manner, that is, differentially on leaves versus seeds of the host P. vulgaris (Chapter IV).

Carnitine utilization in bacteria.

Carnitine is a QAC serving a myriad of biological functions in all domains of life.

Carnitine was originally extracted from muscle tissue (129) and functions in eukaryotic cells as a carrier molecule in fatty acid degradation (β-oxidation), facilitating the transport of acyl-

7 and acetyl- groups into and out of the mitochondria (98). Microorganisms are able to utilize carnitine as a nutrient source, osmoprotectant, and terminal electron acceptor (59, 117, 118).

Carnitine can be metabolized under both aerobic and anaerobic conditions. The two aerobic pathways have been investigated in animal-associated bacteria: one converts carnitine to glycine betaine and the other cleaves the nitrogen-carbon bond of the quaternary ammonium functional group to yield malic acid and trimethylamine (TMA). Although the L- carnitine to glycine betaine degradation pathway has been investigated in Pseudomonas aeruginosa (6, 118), discrepancies among the proposed pathways have not yet been resolved

(addressed in detail in Chapter II). Once L-carnitine is degraded to glycine betaine (6, 118), it is catabolized to glycine through a step-wise demethylation process in which GbcAB converts glycine betaine to dimethyl glycine, dimethyl glycine is catabolized to sarcosine through DgcAB, and finally sarcosine is converted to glycine by SoxBDAG (74). In the second aerobic catabolic pathway, carnitine is cleaved into TMA and malic acid; though initially described in Serratia marcescens, the genes involved have not yet been identified in this system (111). However, Acinetobacter sp. utilizes D- and L-carnitine as sole carbon sources (66, 83), and catabolism involves the enzymes CntA, which is a Rieske-type iron- sulfur oxygenase, and CntB, which is a reductase (128). The malic acid that is released enters the TCA cycle, whereas TMA is exported out of the cell. Orthologs to cntAB have been identified in several gut dwelling organisms such as Escherichia coli and Citrobacter,

Achromobacter and Sporosarcina spp. TMA produced from carnitine catabolism by gut bacteria is correlated with human cardiovascular health (68, 69), since TMA is oxidized in the liver to trimethylamine N-oxide, a marker in the blood for cardiac disease (11).

Anaerobic utilization of carnitine and crotonobetaine, a catabolic intermediate, has been extensively studied in E. coli due to the potential for metabolic engineering for the commercial production of L-carnitine (2, 89). Proteins encoded by the fixABCX operon facilitate the transfer of electrons to carnitine (32, 117). Both the fixABCX and the caiTABCDE operons are regulated by CaiF and are required for the catabolism and utilization of carnitine as a terminal electron acceptor in E. coli (17, 18, 33). The caiTABCDE operon is involved in the transport and conversion of carnitine to crotonobetaine

(31). CaiT, an antiporter, couples carnitine import with γ-butyrobetaine export (56, 116).

The remaining genes, caiABCDE, encode enzymes for the inter-conversion of carnitine, crotonobetaine, and γ-butyrobetaine, a network that facilitates the transfer of electrons to carnitine as a terminal electron acceptor (13, 34, 35).

Beyond acting as a nutrient source, carnitine can alleviate abiotic stress by serving as an osmoprotectant, and in some organisms, a compatible solute. In P. aeruginosa, carnitine is imported as an osmoprotectant by either a BCCT or ABC transporter (22, 24, 77) and converted to glycine betaine, which is then accumulated as a compatible solute (6, 118).

Some organisms, such as Bacillus subtilis, can utilize carnitine directly as a compatible solute and are unable to catabolize carnitine further (59). In Listeria monocytogenes, carnitine is imported via the ABC transporter OpuC and accumulated directly in response to osmotic stress (14, 113, 114); carnitine-derived osmotolerance is a critical factor in L. monocytogenes pathogenesis in mouse systems (104, 106), where carnitine also provides resistance to bile and temperature stresses. Specifically, the OpuC transporter is required for protection during exposure to bile from the small intestine and is co-regulated with the BilE

9 bile efflux system (105); opuC gene expression was increased at low temperatures allowing

L. monocytogenes to maintain growth at 4°C (4, 8, 67).

Chapter II of this thesis is focused on carnitine utilization by P. syringae. I will characterize the carnitine to glycine betaine degradation pathway in P. syringae, resolve discrepancies in the literature, and present a model of the pathway. I will also investigate the role of carnitine catabolism in the P. syringae colonization of the plant host P. vulgaris.

Sulfate metabolism in Pseudomonads.

Although sulfur is a critical component of proteins due to its integration into methionine and cysteine, the sources of sulfur for plant-associated bacteria have received little attention. However, organosulfur has been characterized as comprising over 95% of the total sulfur in the soil (107) and is found primarily as sulfate-esters and sulfonates (97).

Sulfate-esters are estimated to contribute to more than 50% of the total available sulfur in soil

(120), and the activity of sulfatases in the soil increase the availability of sulfur to plants to nearly as much as inorganic sulfur (101). P. putida mutants lacking a sulfatase activity showed a significantly reduced survival in soil environments, suggesting that the ability of microorganisms to utilize sulfate-esters can be critical to their survival (58). Chapter III focuses on the QAC, choline-O-sulfate, a sulfate-ester accumulated by some plant species

(123).

While alkylsulfate-esters are known to supply carbon to P. putida, arylsulfate-esters are not (29, 57). In Pseudomonas C12B, two types of alkylsulfatases act on primary alkylsulfate-esters; their optimum pH for activity, 6.1 and 8.3, coincides with the general range of environmental soil pH (7, 26). Secondary sulfate-esters are degraded by

Pseudomonas sp. for use not only as a sulfur source, but also as a carbon source via β- oxidation of the released carbon chains (75). The majority of the research on alkylsulfate- ester degradation has focused on the degradation of sodium dodecyl sulfate (SDS), a long- chain aliphatic sulfate-ester. SDS is often thought to be an environmental contaminant, though it is also found in uncontaminated soils and water (122), and is degraded by several soil bacteria including Pseudomonas sp. (70). The SDS sulfatase in P. aeruginosa, SdsA

(44), shows homology to the alkaline sulfatase of P. syringae C12B (61). It functions in the periplasm of the bacterium (50), suggesting that it works in concert with a sulfate transporter.

P. aeruginosa deletion mutants of atsRBC, an ABC-type transporter for aromatic sulfonates, were able to grow with SDS, but not hexylsulfate, as a sulfur source (54), suggesting that shorter chain sulfate-esters are transported into the cell for degradation, whereas longer chain sulfate-esters are more likely broken down into sulfite and the corresponding alcohol and transported separately. The second type of sulfate-esters is arylsulfate-esters, where the sulfate group is attached to an aromatic ring. Pseudomonas sp. use arylsulfatases to cleave arylsulfate-esters into sulfate and the corresponding phenol. Though several arylsulfatases have been identified in P. aeruginosa (46), including AtsA, which desulfurizes a number of arylsulfate-esters (9, 44), P. aeruginosa mutants lacking the atsA gene were unable to desulfurize arylsulfate-esters (28). AtsA is predicted to be a cytoplasmic Cys-type arylsulfatases since it contains the motif (C/S-X-P-R-X(4)-T-G) (44) but not a signal peptides

(28).

Another large group of organosulfurs in the soil are sulfonates, which are primarily used in the sulfate starvation responses of Pseudomonads. Sulfonates are characterized by a sulfur-carbon bond (R-CH2-SO3); sulfonates in the soil are derived from sulfolipids of plants

(12). Pseudomonas species can degrade both alkane sulfonates and aromatic sulfonates found in soil environments (109). Alkane sulfonates are transported via the ABC-type transporter encoded by ssuABC (30, 61). In the cytoplasm, SsuD, a monooxygenase, cleaves alkane sulfonates through incorporation of molecular oxygen (61, 63, 109) coupled to the oxidation of FMNH2 into FMN by the oxidoreductase SsuE (61, 63). Interestingly, the expression levels of ssuE were elevated in P. syringae colonizing the leaf surface (80), indicating that sulfonates are utilized in a sulfate starvation response (see Chapter III).

Although aromatic sulfonates share the AstRBC transporter with the alkyl sulfate-esters rather than with alkane sulfonates, they are still cleaved by SsuD (61, 63, 109, 115), coupled to the oxidation of FMNH2 to FMN by the reductase AsfA, through a ferrodoxin bridge protein, AsfB (115).

The sulfate produced from sulfate-esters and sulfonate degradation must either be assimilated by the cell through the sulfur assimilation pathways or exported from the cell

(123, 124). Even though inorganic sulfate is a preferred sulfur source in both prokaryotes and eukaryotes (52), sulfate is not largely abundant in the soil and on the plant host, making it less biologically available to P. syringae.

If it is available, many prokaryotes import sulfate through an ABC-type transporter, as demonstrated in E. coli (50, 62, 102, 103) and P. aeruginosa PAO1 (108), with four components: the permeases CysW and CysT, the ATP- binding subunit CysA, and the sulfate binding protein Sbp (1-4). Thiosulfate can also be used as a sulfur source by Pseudomonas species; it is used to synthesize cysteine (53) and is transported via the CysWTA system through CysP, a thiosulfate binding protein (1-4). CysM putatively reduces thiosulfate to

12 cysteine in E. coli (21, 25, 42). Following transport, sulfate is integrated into organic molecules through the sulfate assimilation pathway.

The first step in sulfate assimilation is the activation of sulfate into adenosine-5’- phosphosulfate (APS) by ATP sulfurylase (19, 53) a heterodimer of CysD and CysN.

Coupled with the hydrolysis of GTP by the regulatory G protein CysN, CysD catalyzes the cleavage of ATP and transfers the sulfate onto AMP to form APS (53, 87). APS is phosphorylated by the APS kinase, CysC, to phosphoadenosine-5’-phosphosulfate (PAPS)

(73), which is reduced to sulfite by CysH, a thioredoxin-dependent PAPS reductase (85, 92).

In E. coli, sulfite is reduced to sulfide by the sulfite reductase complex, CysIJ, a heme- protein, and a flavoprotein, respectively (10, 27, 43, 53, 92). In P. aeruginosa, PuiB is predicted to function as the CysJ flavoprotein (90). The sulfide is converted into either cysteine or methionine.

The cysteine synthase complex (17, 24) combines serine and sulfide to produce cysteine in E. coli (53, 78), where CysE acetylates serine (13) and CysK, an O-acetylserine sulfhydrylase, catalyzes the substitution of the acetyl group with sulfide to form cysteine (20,

21, 51, 94, 95, 127). Methionine is synthesized via either the cysteine methylation or the homoserine sulfhydrylation pathway, the latter being the predominant pathway in

Pseudomonads (37). In the cysteine methylation pathway in E. coli, homocysteine is synthesized from cysteine through a cystathionine intermediate by MetB and MetC; however, in P. aeruginosa the activity of MetB and MetC is decreased (38). In the homoserine sulfhydrylation pathway, homoserine is first converted to homocysteine. In P. aeruginosa,

MetA catalyzes homoserine to O-succinylhomoserine (1, 38), while in P. syringae homoserine is acetylated to O-acetylhomoserine by MetXW (1, 3). This intermediate is

13 combined by MetZ with sulfide to form homocysteine (1, 91). The methylation of homocysteine to methionine is catalyzed, in P. syringae, by either MetE or MetH, a function also predicted in P. putida (1). The complex network involved in the import and utilization of the various sulfur sources that are biologically available to Pseudomonads in the soil and on plants bears further investigation (Chapter III), especially in the context of the use of the

QAC choline-O-sulfate as a putative source of sulfur for plant-associated bacteria.

The major focus in this dissertation is to investigate the utilization of QACs as nutrients and as osmoprotectants. Our goal is to characterize carnitine as a carbon source by investigating the carnitine degradation pathway; choline-O-sulfate as a carbon, nitrogen, and sulfur source through evaluation of the P. syringae choline sulfatase. Furthermore, our guiding motivation in undertaking this work is to evaluate whether QACs carnitine, COS, choline, glycine betaine and phosphorylcholine impact plant-microbe interactions, and if so, whether they contribute as nutrients, osmoprotectants, or both. This will allow us to further our understanding of phyllosphere ecology and explicate how microbes benefit host-derived compounds at the plant-microbe chemical interface.

References

1. Alaminos M, Ramos JL. 2001. The methionine biosynthetic pathway from homoserine in Pseudomonas putida involves the metW, metX, metZ, metH and metE gene products. Archives of Microbiology 176:151-154.

2. Alvarez‐Vasquez F, Canovas M, Iborra JL, Torres NV. 2002. Modeling, optimization and experimental assessment of continuous L-carnitine production by Escherichia coli cultures. Biotechnology and Bioengineering 80:794-805.

3. Andersen GL, Beattie GA, Lindow SE. 1998. Molecular characterization and sequence of a methionine biosynthetic locus from Pseudomonas syringae. Journal of Bacteriology 180:4497-4507.

4. Angelidis AS, Smith GM. 2003. Role of the glycine betaine and carnitine transporters in adaptation of Listeria monocytogenes to chill stress in defined medium. Applied and Environmental Microbiology 69:7492-7498.

5. Ashraf M, Foolad M. 2007. Roles of glycine betaine and proline in improving plant abiotic stress resistance. Environmental and Experimental Botany 59:206-216.

6. Bastard K, Smith AAT, Vergne-Vaxelaire C, Perret A, Zaparucha A, De Melo- Minardi R, Mariage A, Boutard M, Debard A, Lechaplais C. 2014. Revealing the hidden functional diversity of an enzyme family. Nat Chem Biol 10:42-49.

7. Bateman TJ, Dodgson KS, White GF. 1986. Primary alkylsulfatase activities of the detergent-degrading bacterium Pseudomonas C12B - Purification and properties of the P1 enzyme. Biochemical Journal 236:401-408.

8. Bayles D, Wilkinson B. 2000. Osmoprotectants and cryoprotectants for Listeria monocytogenes. Letters in Applied Microbiology 30:23-27.

9. Beil S, Kehrli H, James P, Staudenmann W, Cook AM, Leisinger T, Kertesz MA. 1995. Purification and characterization of the arylsulfatase synthesized by Pseudomonas aeruginosa PAO during growth in sulfate-free medium and cloning of the arylsulfatase gene (atsA). European Journal of Biochemistry 229:385-394.

10. Belinsky MI. 1996. Hyperfine evidence of strong double exchange in multimetallic { Fe4S4 -Fe} active center of Escherichia coli sulfite reductase. Journal of Biological Inorganic Chemistry 1:186-188.

11. Bennett BJ, de Aguiar Vallim TQ, Wang Z, Shih DM, Meng Y, Gregory J, Allayee H, Lee R, Graham M, Crooke R. 2013. Trimethylamine-N-oxide, a metabolite associated with atherosclerosis, exhibits complex genetic and dietary regulation. Cell Metabolism 17:49-60.

12. Benson AA, Daniel H, Wiser R. 1959. A sulfolipid in plants. Proceedings of the National Academy of Sciences 45:1582-1587.

13. Bernal V, Arense P, Blatz V, Mandrand‐Berthelot M, Cánovas M, Iborra J. 2008. Role of betaine: CoA ligase (CaiC) in the activation of betaines and the transfer of coenzyme A in Escherichia coli. Journal of Applied Microbiology 105:42-50.

14. Beumer R, Te Giffel M, Cox L, Rombouts F, Abee T. 1994. Effect of exogenous proline, betaine, and carnitine on growth of Listeria monocytogenes in a minimal medium. Applied and Environmental Microbiology 60:1359-1363.

15. Boch J, Kempf B, Bremer E. 1994. Osmoregulation in Bacillus subtilis: synthesis of the osmoprotectant glycine betaine from exogenously provided choline. Journal of Bacteriology 176:5364-5371.

16. Boch J, Kempf B, Schmid R, Bremer E. 1996. Synthesis of the osmoprotectant glycine betaine in Bacillus subtilis: characterization of the gbsAB genes. Journal of Bacteriology 178:5121-5129.

17. Buchet A, Eichler K, Mandrand-Berthelot M-A. 1998. Regulation of the carnitine pathway in Escherichia coli: Investigation of the cai-fix divergent promoter region. Journal of Bacteriology 180:2599-2608.

18. Buchet A, Nasser W, Eichler K, Mandrand‐Berthelot MA. 1999. Positive co‐regulation of the Escherichia coli carnitine pathway cai and fix operons by CRP and the CaiF activator. Molecular Microbiology 34:562-575.

19. Bykowski T, van der Ploeg JR, Iwanicka-Nowicka R, Hryniewicz MM. 2002. The switch from inorganic to organic sulphur assimilation in Escherichia coli: adenosine 5 '-phosphosulphate (APS) as a signalling molecule for sulphate excess. Molecular Microbiology 43:1347-1358.

20. Campanini B, Speroni F, Salsi E, Cook PF, Roderick SL, Huang B, Bettati S, Mozzarelli A. 2005. Interaction of serine acetyltransferase with O-acetylserine sulfhydrylase active site: evidence from fluorescence spectroscopy. Protein Science 14:2115-2124.

21. Chattopadhyay A, Meier M, Ivaninskii S, Burkhard P, Speroni F, Campanini B, Bettati S, Mozzarelli A, Rabeh WM, Li L, Cook PF. 2007. Structure, mechanism, and conformational dynamics of O-acetylserine sulfhydrylase from Salmonella typhimurium: Comparison of A and B isozymes. Biochemistry 46:8315-8330.

22. Chen C, Beattie GA. 2007. Characterization of the osmoprotectant transporter OpuC from Pseudomonas syringae and demonstration that cystathionine-β-synthase domains are required for its osmoregulatory function. Journal of Bacteriology 189:6901-6912.

23. Chen C, Beattie GA. 2008. Pseudomonas syringae BetT is a low-affinity choline transporter that is responsible for superior osmoprotection by choline over glycine betaine. Journal of Bacteriology 190:2717-2725.

24. Chen C, Malek AA, Wargo MJ, Hogan DA, Beattie GA. 2010. The ATP‐binding cassette transporter Cbc (choline/betaine/carnitine) recruits multiple substrate‐binding proteins with strong specificity for distinct quaternary ammonium compounds. Molecular Microbiology 75:29-45.

25. Claus MT, Zocher GE, Maier THP, Schulz GE. 2005. Structure of the O- acetylserine sulfhydrylase isoenzyme CysM from Escherichia coli. Biochemistry 44:8620-8626.

26. Cloves JM, Dodgson KS, White GF, Fitzgerald JW. 1980. Purification and properties of the P2 primary alkylsulfohydrolase of the detergent-degrading bacterium Pseudomonas C12B. Biochemical Journal 185:23-31.

27. Crane BR, Getzoff ED. 1996. The relationship between structure and function for the sulfite reductases. Current Opinion in Structural Biology 6:744-756.

28. Dierks T, Miech C, Hummerjohann J, Schmidt B, Kertesz MA, von Figura K. 1998. Posttranslational formation of formylglycine in prokaryotic sulfatases by modification of either cysteine or serine. Journal of Biological Chemistry 273:25560- 25564.

29. Dodgson KS GFW, and J W Fitzgerald. 1982. Sulfatases of microbial origin. CRC Press, Boca Raton.

30. Eichhorn E, Van der Ploeg JR, Leisinger T. 2000. Deletion analysis of the Escherichia coli taurine and alkanesulfonate transport systems. Journal of Bacteriology 182:2687-2695.

31. Eichler K, Bourgis F, Buchet A, Kleber HP, Mandrand‐Berthelot MA. 1994. Molecular characterization of the cai operon necessary for carnitine metabolism in Escherichia coli. Molecular Microbiology 13:775-786.

32. Eichler K, Buchet A, Bourgis F, Kleber HP, Mandrand‐Berthelot MA. 1995. The fix Escherichia coli region contains four genes related to carnitine metabolism. Journal of Basic Microbiology 35:217-227.

33. Eichler K, Buchet A, Lemke R, Kleber H-P, Mandrand-Berthelot M-A. 1996. Identification and characterization of the caiF gene encoding a potential transcriptional activator of carnitine metabolism in Escherichia coli. Journal of Bacteriology 178:1248-1257.

34. Elssner T, Engemann C, Baumgart K, Kleber H-P. 2001. Involvement of coenzyme A esters and two new enzymes, an enoyl-CoA hydratase and a CoA- transferase, in the hydration of crotonobetaine to L-carnitine by Escherichia coli. Biochemistry 40:11140-11148.

35. Engemann C, Elssner T, Kleber H-P. 2001. Biotransformation of crotonobetaine to L (–)-carnitine in Proteus sp. Archives of Microbiology 175:353-359.

36. Fett WF, Dunn MF. 1989. Exopolysaccharides produced by phytopathogenic Pseudomonas syringae pathovars in infected leaves of susceptible hosts. Plant Physiology 89:5-9.

37. Foglino M, Borne F, Bally M, Ball G, Patte J. 1995. A direct sulfhydrylation pathway is used for methionine biosynthesis in Pseudomonas aeruginosa. Microbiology 141:431-439.

38. Foglino M, Borne F, Bally M, Ball G, Patte JC. 1995. A direct sulfhydrylation pathway is used for methionine biosynthesis in Pseudomonas aeruginosa. Microbiology-UK 141:431-439.

39. Freeman BC, Chen C, Beattie GA. 2010. Identification of the trehalose biosynthetic loci of Pseudomonas syringae and their contribution to fitness in the phyllosphere. Environmental Microbiology 12:1486-1497.

40. Freeman BC, Chen C, Yu X, Nielsen L, Peterson K, Beattie GA. 2013. Physiological and transcriptional responses to osmotic stress of two Pseudomonas syringae strains that differ in epiphytic fitness and osmotolerance. Journal of Bacteriology 195:4742-4752.

41. Garcia AB, Engler J, Iyer S, Gerats T, Van Montagu M, Caplan AB. 1997. Effects of osmoprotectants upon NaCl stress in rice. Plant Physiology 115:159-169.

42. Garvis SG, Tipton SL, Konkel ME. 1997. Identification of a functional homolog of the Escherichia coli and Salmonella typhimurium cysM gene encoding O-acetylserine sulfhydrylase B in Campylobacter jejuni. Gene 185:63-67.

43. Guillouard I, Auger S, Hullo MF, Chetouani F, Danchin A, Martin-Verstraete I. 2002. Identification of Bacillus subtilis CysL, a regulator of the cysJI operon, which encodes sulfite reductase. Journal of Bacteriology 184:4681-4689.

44. Hagelueken G, Adams TM, Wiehlmann L, Widow L, Kolmar H, Tummler B, Heinz DW, Schubert WD. 2006. The crystal structure of SdsA1, an alkylsulfatase from Pseudomonas aeruginosa, defines a third class of sulfatases. Proceedings of the National Academy of Sciences 103:7631-7636.

45. Hanson AD, Rathinasabapathi B, Chamberlin B, Gage DA. 1991. Comparative physiological evidence that β-alanine betaine and choline-O-sulfate act as compatible osmolytes in halophytic Limonium species. Plant physiology 97:1199-1205.

46. Harada T. 1964. Formation of sulphatases by Pseudomonas aeruginosa. Biochimica Et Biophysica Acta 81:193-&.

47. Hirano SS, Upper CD. 1990. Population biology and epidemiology of Pseudomonas syringae. Annual Review of Phytopathology 28:155-177.

48. Hirano SS, Upper CD. 2000. Bacteria in the leaf ecosystem with emphasis on Pseudomonas syringae—a pathogen, ice nucleus, and epiphyte. Microbiology and Molecular Biology Reviews 64:624-653.

49. Hoffmann T, Bremer E. 2011. Protection of Bacillus subtilis against cold stress via compatible-solute acquisition. Journal of Bacteriology 193:1552-1562.

50. Hryniewicz M, Sirko A, Palucha A, Bock A, Hulanicka D. 1990. Sultate and thiosulfate transport in Escherichia coli K-12 - Identification of a gene encoding a novel protein involved in thiosulfate binding. Journal of Bacteriology 172:3358-3366.

51. Huang B, Vetting MW, Roderick SL. 2005. The active site of O-acetylserine sulfhydrylase is the anchor point for bienzyme complex formation with serine acetyltransferase. Journal of Bacteriology 187:3201-3205.

52. Hummerjohann J, Küttel E, Quadroni M, Ragaller J, Leisinger T, Kertes MA. 1998. Regulation of the sulfate starvation response in Pseudomonas aeruginosa: role of cysteine biosynthetic intermediates. Microbiology 144:1375-1386.

53. Hummerjohann J, Kuttel E, Quadroni M, Ragaller J, Leisinger T, Kertesz MA. 1998. Regulation of the sulfate starvation response in Pseudomonas aeruginosa: role of cysteine biosynthetic intermediates. Microbiology-UK 144:1375-1386.

54. Hummerjohann J, Laudenbach S, Retey J, Leisinger T, Kertesz MA. 2000. The sulfur-regulated arylsulfatase gene cluster of Pseudomonas aeruginosa, a new member of the cys regulon. Journal of Bacteriology 182:2055-2058.

55. Joardar V, Lindeberg M, Jackson RW, Selengut J, Dodson R, Brinkac LM, Daugherty SC, DeBoy R, Durkin AS, Giglio MG. 2005. Whole-genome sequence analysis of Pseudomonas syringae pv. phaseolicola 1448A reveals divergence among pathovars in genes involved in virulence and transposition. Journal of Bacteriology 187:6488-6498.

56. Jung H, Buchholz M, Clausen J, Nietschke M, Revermann A, Schmid R, Jung K. 2002. CaiT of Escherichia coli, a new transporter catalyzing L-carnitine/γ- butyrobetaine exchange. Journal of Biological Chemistry 277:39251-39258.

57. K S Dodgson GFW, and J W Fitzgerald. 1982. Sulfatases of Microbial Origin. CRC Press, Boca Raton.

58. Kahnert A, Mirleau P, Wait R, Kertesz MA. 2002. The LysR-type regulator SftR is involved in soil survival and sulphate ester metabolism in Pseudomonas putida. Environmental Microbiology 4:225-237.

59. Kappes RM, Bremer E. 1998. Response of Bacillus subtilis to high osmolarity: uptake of carnitine, crotonobetaine and γ-butyrobetaine via the ABC transport system OpuC. Microbiology 144:83-90.

60. Kempf B, Bremer E. 1998. Uptake and synthesis of compatible solutes as microbial stress responses to high-osmolality environments. Archives of Microbiology 170:319-330.

61. Kertesz MA. 2000. Riding the sulfur cycle - metabolism of sulfonates and sulfate esters in Gram-negative bacteria. FEMS Microbiology Reviews 24:135-175.

62. Kertesz MA. 2001. Bacterial transporters for sulfate and organosulfur compounds. Research in Microbiology 152:279-290.

63. Kertesz MA, Wietek C. 2001. Desulfurization and desulfonation: applications of sulfur-controlled gene expression in bacteria. Applied Microbiology and Biotechnology 57:460-466.

64. Kets E, Galinski EA, De Wit M, De Bont J, Heipieper HJ. 1996. Mannitol, a novel bacterial compatible solute in Pseudomonas putida S12. Journal of Bacteriology 178:6665-6670.

65. Klähn S, Hagemann M. 2011. Compatible solute biosynthesis in cyanobacteria. Environmental Microbiology 13:551-562.

66. Kleber H, Seim H, Aurich H, Strack E. 1977. Utilization of trimethylammonium- compounds by Acinetobacter calcoaceticus (author's transl). Archives of Microbiology 112:201.

67. Ko R, Smith LT, Smith GM. 1994. Glycine betaine confers enhanced osmotolerance and cryotolerance on Listeria monocytogenes. Journal of Bacteriology 176:426-431.

68. Koeth RA, Levison BS, Culley MK, Buffa JA, Wang Z, Gregory JC, Wu Y, Li L, Smith JD, Tang WW. 2014. γ-Butyrobetaine is a proatherogenic intermediate in gut microbial metabolism of L-carnitine to TMAO. Cell Metabolism 20:799-812.

69. Koeth RA, Wang Z, Levison BS, Buffa JA, Sheehy BT, Britt EB, Fu X, Wu Y, Li L, Smith JD. 2013. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nature Medicine 19:576-585.

70. Kostal J, Suchanek M, Klierova H, Demnerova K, Kralova B, McBeth DL. 1998. Pseudomonas C12B, an SDS degrading strain, harbours a plasmid coding for degradation of medium chain length n-alkanes. International Biodeterioration & Biodegradation 42:221-228.

71. Kurz M, Burch AY, Seip B, Lindow SE, Gross H. 2010. Genome-driven investigation of compatible solute biosynthesis pathways of Pseudomonas syringae pv. syringae and their contribution to water stress tolerance. Applied and Environmental Microbiology 76:5452-5462.

72. Laue H, Schenk A, Li H, Lambertsen L, Neu TR, Molin S, Ullrich MS. 2006. Contribution of alginate and levan production to biofilm formation by Pseudomonas syringae. Microbiology 152:2909-2918.

73. Leyh TS, Taylor JC, Markham GD. 1988. The sulfate activation locus of Escherichia coli K12 - cloning, genetic and enzymatic characterization. Journal of Biological Chemistry 263:2409-2416.

74. Li S, Yu X, Beattie GA. 2013. Glycine betaine catabolism contributes to Pseudomonas syringae tolerance to hyperosmotic stress by relieving betaine- mediated suppression of compatible solute synthesis. Journal of Bacteriology 195:2415-2423.

75. Lijmbach GW, Brinkhui.E. 1973. Microbial degradation of secondary alkyl sulfates and secondary alkanols. Antonie Van Leeuwenhoek Journal of Microbiology 39:415- 423.

76. Lindow SE, Brandl MT. 2003. Microbiology of the phyllosphere. Applied and Environmental Microbiology 69:1875-1883.

77. Malek AA, Chen C, Wargo MJ, Beattie GA, Hogan DA. 2011. Roles of three transporters, CbcXWV, BetT1, and BetT3, in Pseudomonas aeruginosa choline uptake for catabolism. Journal of Bacteriology 193:3033-3041.

78. Mansilla MC, deMendoza D. 1997. L-cysteine biosynthesis in Bacillus subtilis: Identification, sequencing, and functional characterization of the gene coding for phosphoadenylylsulfate sulfotransferase. Journal of Bacteriology 179:976-981.

79. Mansvelt EL, Hattingh M. 1989. Scanning electron microscopy of invasion of apple leaves and blossoms by Pseudomonas syringae pv. syringae. Applied and Environmental Microbiology 55:533-538.

80. Marco ML, Legac J, Lindow SE. 2005. Pseudomonas syringae genes induced during colonization of leaf surfaces. Environmental Microbiology 7:1379-1391.

81. McNeil SD, Nuccio ML, Hanson AD. 1999. Betaines and related osmoprotectants: targets for metabolic engineering of stress resistance. Plant Physiology 120:945-949.

82. Melotto M, Underwood W, Koczan J, Nomura K, He SY. 2006. Plant stomata function in innate immunity against bacterial invasion. Cell 126:969-980.

83. Miura-Fraboni J, Kleber H-P, Englard S. 1982. Assimilation of γ-butyrobetaine, and D-and L-carnitine by resting cell suspensions of Acinetobacter calcoaceticus and Pseudomonas putida. Archives of Microbiology 133:217-221.

84. Monier J-M, Lindow S. 2003. Differential survival of solitary and aggregated bacterial cells promotes aggregate formation on leaf surfaces. Proceedings of the National Academy of Sciences 100:15977-15982.

85. Montoya G, Svensson C, Savage H, Schwenn JD, Sinning I. 1998. Crystallization and preliminary X-ray diffraction studies of phospho-adenylylsulfate (PAPS) reductase from E. coli. Acta Crystallographica Section D-Biological Crystallography 54:281-283.

86. Morris CE, Monier J-M. 2003. The ecological significance of biofilm formation by plant-associated bacteria. Annual Review of Phytopathology 41:429-453.

87. Mougous JD, Lee DH, Hubbard SC, Schelle MW, Vocadlo DJ, Berger JM, Bertozzi CR. 2006. Molecular basis for G protein control of the prokaryotic ATP sulfurylase. Molecular Cell 21:109-122.

88. Nau-Wagner G, Boch J, Le Good JA, Bremer E. 1999. High-affinity transport of choline-O-sulfate and its use as a compatible solute in Bacillus subtilis. Applied and Environmental Microbiology 65:560-568.

89. Obón J, Maiquez J, Cánovas M, Kleber H-P, Iborra J. 1999. High-density Escherichia coli cultures for continuous L (−)-carnitine production. Applied Microbiology Biotechnology 51:760-764.

90. Ochsner UA, Vasil ML. 1996. Gene repression by the ferric uptake regulator in Pseudomonas aeruginosa: Cycle selection of iron-regulated genes. Proceedings of the National Academy of Sciences 93:4409-4414.

91. Old IG, Phillips SEV, Stockley PG, Saintgirons I. 1991. Regulation of methionine biosynthesis in the Enterobacteriaceae. Progress in Biophysics & Molecular Biology 56:145-185.

92. Ostrowski J, Wu JY, Rueger DC, Miller BE, Siegel LM, Kredich NM. 1989. Characterization of the cysJIH regions of Salmonella typhimurium and Escherichia coli B - DNA sequences of cysI and cysH and a model for the siroheme F34S4 active center of sulfite reductase hemoprotein based on amino acid homology with spinach nitrite reductase. Journal of Biological Chemistry 264:15726-15737.

93. Penaloza-Vazquez A, Kidambi SP, Chakrabarty AM, Bender CL. 1997. Characterization of the alginate biosynthetic gene cluster in Pseudomonas syringae pv. syringae. Journal of Bacteriology 179:4464-4472.

94. Rabeh WM, Alguindigue SS, Cook PF. 2005. Mechanism of the addition half of the O-acetylserine sulfhydrylase-A reaction. Biochemistry 44:5541-5550.

95. Rabeh WM, Mather T, Cook PF. 2006. A three-dimensional homology model of the O-acetylserine sulfhydrylase-B from Salmonella typhimurium. Protein and Peptide Letters 13:7-13.

96. Ramey BE, Koutsoudis M, von Bodman SB, Fuqua C. 2004. Biofilm formation in plant–microbe associations. Current Opinion in Microbiology 7:602-609.

97. Ramos J-L (ed). 2004. Biosynthesis of Macromolecules and Molecular Metabolism. Kluwer Academic, New York, NY.

98. Rebouche CJ, Paulson DJ. 1986. Carnitine metabolism and function in humans. Annual Review of Nutrition 6:41-66.

99. Roeßler M, Müller V. 2001. Osmoadaptation in bacteria and archaea: common principles and differences. Environmental Microbiology 3:743-754.

100. Sagot B, Gaysinski M, Mehiri M, Guigonis J-M, Le Rudulier D, Alloing G. 2010. Osmotically induced synthesis of the dipeptide N-acetylglutaminylglutamine amide is mediated by a new pathway conserved among bacteria. Proceedings of the National Academy of Sciences 107:12652-12657.

101. Shan XQ, Chen B, Zhang TH, Li FL, Wen B, Qian J. 1997. Relationship between sulfur speciation in soils and plant availability. Science of the Total Environment 199:237-246.

102. Sirko A, Hryniewicz M, Hulanicka D, Bock A. 1990. Sulfate and thiosulfate transport in Escherichia coli K-12 - Nucleotide sequence and expression of the cysTWAM gene clutser. Journal of Bacteriology 172:3351-3357.

103. Sirko A, Zatyka M, Sadowy E, Hulanicka D. 1995. Sulfate and thiosulfate transport in Escherichia coli K-12 - Evidence for a functional overlapping of sulfate- binding and thiosulfate-binding proteins. Journal of Bacteriology 177:4134-4136.

104. Sleator R, Francis G, O'Beirne D, Gahan C, Hill C. 2003. Betaine and carnitine uptake systems in Listeria monocytogenes affect growth and survival in foods and during infection. Journal of Applied Microbiology 95:839-846.

105. Sleator RD, Wemekamp‐Kamphuis HH, Gahan CG, Abee T, Hill C. 2005. A PrfA‐regulated bile exclusion system (BilE) is a novel virulence factor in Listeria monocytogenes. Molecular Microbiology 55:1183-1195.

106. Sleator RD, Wouters J, Gahan CG, Abee T, Hill C. 2001. Analysis of the role of OpuC, an osmolyte transport system, in salt tolerance and virulence potential of Listeria monocytogenes. Applied and Environmental Microbiology 67:2692-2698.

107. Stevenson FJ. 1986. Cycles of Soil: Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrients. Wiley-Interscience, New York, NY.

108. Stover CK, Pham XQ, Erwin AL, Mizoguchi SD, Warrener P, Hickey MJ, Brinkman FSL, Hufnagle WO, Kowalik DJ, Lagrou M, Garber RL, Goltry L, Tolentino E, Westbrock-Wadman S, Yuan Y, Brody LL, Coulter SN, Folger KR, Kas A, Larbig K, Lim R, Smith K, Spencer D, Wong GKS, Wu Z, Paulsen IT, Reizer J, Saier MH, Hancock REW, Lory S, Olson MV. 2000. Complete genome sequence of Pseudomonas aeruginosa PAO1, an opportunistic pathogen. Nature 406:959-964.

109. Thysse GJE, Wanders TH. 1972. Degradation of alkane-1-sulfonates by Pseudomonas. Antonie Van Leeuwenhoek Journal of Microbiology and Serology 38:53-&.

110. Underwood W, Melotto M, He SY. 2007. Role of plant stomata in bacterial invasion. Cellular Microbiology 9:1621-1629.

111. Unemoto T, Hayashi M, Miyaki K, Hayashi M. 1966. Formation of trimethylamine from DL-carnitine by Serratia marcescens. Biochimica et Biophysica Acta (BBA)- General Subjects 121:220-222.

112. van Hall C. 1903. Bijdragen tot de Kennis der bakterieele Plantenziekten.(Beiträge zur Kenntnis der bakteriellen Pflanzenkrankheiten). JSTOR.

113. Verheul A, Glaasker E, Poolman B, Abee T. 1997. Betaine and L-carnitine transport by Listeria monocytogenes Scott A in response to osmotic signals. Journal of Bacteriology 179:6979-6985.

114. Verheul A, Rombouts FM, Beumer RR, Abee T. 1995. An ATP-dependent L- carnitine transporter in Listeria monocytogenes Scott A is involved in osmoprotection. Journal of Bacteriology 177:3205-3212.

115. Vermeij P, Wietek C, Kahnert A, Wuest T, Kertesz MA. 1999. Genetic organization of sulphur-controlled aryl desulphonation in Pseudomonas putida S-313. Molecular Microbiology 32:913-926.

116. Vinothkumar KR, Raunser S, Jung H, Kühlbrandt W. 2006. Oligomeric structure of the carnitine transporter CaiT from Escherichia coli. Journal of Biological Chemistry 281:4795-4801.

117. Walt A, Kahn ML. 2002. The fixA and fixB genes are necessary for anaerobic carnitine reduction in Escherichia coli. Journal of Bacteriology 184:4044-4047.

118. Wargo MJ, Hogan DA. 2009. Identification of genes required for Pseudomonas aeruginosa carnitine catabolism. Microbiology 155:2411-2419.

119. Wargo MJ, Szwergold BS, Hogan DA. 2008. Identification of two gene clusters and a transcriptional regulator required for Pseudomonas aeruginosa glycine betaine catabolism. Journal of Bacteriology 190:2690-2699.

120. Watwood ME, Fitzgerald JW, Gosz JR. 1986. Sulfur processing in forest soil and litter along an elevational and vegetative gradient. Canadian Journal of Forest Research 16:689-695.

121. Welsh DT. 2000. Ecological significance of compatible solute accumulation by micro-organisms: from single cells to global climate. FEMS Microbiology Reviews 24:263-290.

122. White GF, Russell NJ, Day MJ. 1985. A survey of sodium dodecyl-sulfate (SDS) resistance and alkylsulfatase production in bacteria from clean and polluted river sites. Environmental Pollution Series A-Ecological and Biological 37:1-11.

123. Wübbeler JH, Hiessl S, Meinert C, Poehlein A, Schuldes J, Daniel R, Steinbüchel A. 2015. The genome of Variovorax paradoxus strain TBEA6 provides new understandings for the catabolism of 3, 3′-thiodipropionic acid and hence the production of polythioesters. Journal of Biotechnology 209:85-95.

124. Wübbeler JH, Hiessl S, Schuldes J, Thürmer A, Daniel R, Steinbüchel A. 2014. Unravelling the complete genome sequence of Advenella mimigardefordensis strain DPN7T and novel insights in the catabolism of the xenobiotic polythioester precursor 3, 3′-dithiodipropionate. Microbiology 160:1401-1416.

125. Yancey PH. 2005. Organic osmolytes as compatible, metabolic and counteracting cytoprotectants in high osmolarity and other stresses. Journal of Experimental Biology 208:2819-2830.

126. Yu J, Peñaloza‐Vázquez A, Chakrabarty AM, Bender CL. 1999. Involvement of the exopolysaccharide alginate in the virulence and epiphytic fitness of Pseudomonas syringae pv. syringae. Molecular Microbiology 33:712-720.

127. Zhao C, Kumada Y, Imanaka H, Imamura K, Nakanishi K. 2006. Cloning, overexpression, purification, and characterization of O-acetylserine sulfhydrylase-B from Escherichia coli. Protein Expression and Purification 47:607-613.

128. Zhu Y, Jameson E, Crosatti M, Schäfer H, Rajakumar K, Bugg TD, Chen Y. 2014. Carnitine metabolism to trimethylamine by an unusual Rieske-type oxygenase

from human microbiota. Proceedings of the National Academy of Sciences 111:4268- 4273.

129. Zurbriggen E. 2000. L-carnitine: historical review. Annals of Nutrition and Metabolism 44:75-96.

CHAPTER 2. CARNITINE RELEASED FROM GERMINATING PHASEOLUS

VULGARIS SEEDS PROMOTES THE GROWTH OF PSEUDOMONAS SYRINGAE AND

REQUIRES DHCAB FOR REMOVAL OF THE INTERMEDIATE ACETOACETATE

FOR CATABOLISM

A paper to be submitted for publication

Michael D. Millican, Adam Klein, Young-Jin Lee, and Gwyn A. Beattie

Abstract

Though plants are not known for the production of abundant L-carnitine, the common epiphyte Pseudomonas syringae contains homologs to carnitine degradation loci cdhXCAB and dhcAB. We analyzed the conflicting reports of the carnitine degradation pathway and proposed an alternative pathway based on our results. Using deletion mutants, we characterized the cellular role of the genes Psyr_2917-2919, cdhCAB, and Psyr_3237-3238, dhcAB, in the L- carnitine degradation pathway. We proved that DhcAB metabolizes acetoacetate, a by-product of carnitine degradation, to acetoacetate-CoA and that DhcAB acts on acetoacetate as part of the carnitine, phenylalanine, and leucine degradation pathways. L-carnitine provides osmoprotection to both B728a and ∆cdhA, a mutant deficient in the first step of carnitine degradation. Analyzing B728a and ∆cdhA colonies grown under osmotic stress on solid MinAS with L-carnitine provided as an osmoprotectant, MALDI-MSI analysis revealed L-carnitine accumulates in ∆cdhA but is preferentially degraded to glycine betaine in B728a. We evaluated the impact of carnitine catabolism while colonizing the plant host, no reduction in epiphytic and apoplastic fitness was observed in ∆cdhA populations relative to B728a. However, we 27 ascertained that carnitine supports population growth while colonizing germinating bean seeds, specifically during radicle emergence and elongation. Compositional analysis of seed exudates determined that carnitine is transiently released during germination, reaching the highest abundance during radicle emergence and elongation, thus identifying carnitine catabolism as a potential factor contributing to the early establishment of P. syringae populations.

Introduction

Carnitine is a quaternary ammonium compound (QAC) found in and utilized by prokaryotes and eukaryotes. In bacteria, carnitine can function as a carbon source, nitrogen source, osmoprotectant, compatible solute, and terminal electron acceptor (25, 37, 38). Carnitine utilization was first described in Escherichia coli, where it involves the conversion of carnitine to

γ-butyrobetaine (CAI pathway) and requires the caiTABCDE and fixABCX operons (13, 14).

The CAI pathway has been of particular interest due to the possibility of exploiting its function in E. coli for the commercial production of L-carnitine through metabolic engineering (34). The

CAI pathway has also been shown to contribute to the fitness of Sinorhizobium meliloti, a plant- associated bacterium (18). Neither, Bacillus subtilis nor Listeria monocytogenes require catabolism of carnitine but rather accumulate carnitine for osmotolerance (23, 45). Both B. subtilis and L. monocytogenes transport L-carnitine through an ABC transporter OpuC and accumulate L-carnitine as a compatible solute (23, 39). Neither organism is able to degrade carnitine as a carbon source and thus they export it once it is no longer required as a compatible solute (23, 46). In Pseudomonas aeruginosa, carnitine functions as an osmoprotectant, but unlike B. subtilis and L. monocytogenes, P. aeruginosa converts carnitine to glycine betaine for

28 accumulation and then degrades glycine betaine as a nutrient source following alleviation of osmotic stress (11, 30, 48).

Pseudomonas syringae possesses two transporters for carnitine, OpuC and CbcWV, and these enable transport predominantly for osmoprotection and nutrient acquisition, respectively

(7, 9). The CbcWV transporter in P. syringae utilizes multiple substrate binding proteins (SBPs) to transport QACs such as choline, glycine betaine, choline-O-sulfate, and carnitine into the cell, with each SBP specifically binding a single QAC (9). The SBP specific for carnitine, CdhX

(Psyr_2916), is encoded in P. syringae B728a as part of a locus containing genes predicted to catabolize carnitine to glycine betaine (8, 9). These genes, cdhX, cdhC, cdhA, and cdhB, are similar to a locus in P. aeruginosa required for L-carnitine catabolism (49). The DhcAB complex, encoded elsewhere, may also contribute to carnitine catabolism, although reports addressing pathways for carnitine degradation in prokaryotes differ in the role that they attribute to this complex. In P. aeruginosa, though a dhcAB mutant was unable to catabolize carnitine

(49), an enzymatic assay demonstrated that DhcAB was not required to convert L-carnitine to glycine betaine (3, 49). Finally, in a review article of the role of carnitine in bacterial physiology, Meadows et al. (32) proposed that DhcAB converts glycine betaine-CoA to glycine betaine in the final step of the carnitine degradation pathway. In this work, we investigated the carnitine degradation pathway used by P. syringae B728a, and moreover examined why this strain, a plant pathogen, retains genes associated with carnitine catabolism even though plants are not known to be a major source of carnitine.

The presence and function of carnitine in plants has not been well characterized; however, carnitine has a major role in fatty acid metabolism in animal cells, where it helps transport acyl and acetyl groups across the membrane of cellular organelles (6, 17). The

29 classical model of carnitine function is that it binds the acyl groups of activated fatty acids and marks them for transport into peroxisomes or mitochondria for β-oxidation; this is designated the carnitine translocation pathway (6, 17). In animal and fungal cells, carnitine also binds free acyl and acetyl groups in the cytosol to prevent negative feedback of metabolic networks due to buildup of these metabolites (10).

The role of carnitine in plants may go beyond that observed in β-oxidation. Masterson et al. (31) demonstrated a divide between β-oxidation occurring in the peroxisome versus the mitochondria in pea seedlings, showing that inhibiting mitochondrial β-oxidation decreased chlorophyll synthesis; moreover, supplementing seedlings with carnitine stimulated production of chlorophyll. In Arabidopsis, a chloroplast-associated protein named A BOUT DE SOUFFLE

(BOU) was predicted to be a carnitine acyl carrier-like protein that facilitates the transport of acetyl groups into the chloroplast for chlorophyll synthesis (27). However, a more recent study demonstrates that BOU is not carnitine dependent but functions in transporting an undetermined organic acid (15). This recent report suggests that carnitine is not involved in transferring acetyl groups into the chloroplast for chlorophyll synthesis, which is consistent with the possibility that the primary role of carnitine may be similar in plants and animals, namely as a carrier molecule for fatty acids in β-oxidation. The Findings of Masterson et al. (30) can be explained through an increase in acetyl-CoA production due to increased carnitine leading to an increase in the rate of

β-oxidation (16, 26). Fatty acid oxidation is likely to be important in plants primarily during the germination of oil seeds, as these seeds rely on β-oxidation for energy production until photosynthesis becomes active.

The goals of this study were to (i) clarify the carnitine degradation pathway in P. syringae, (ii) investigate if P. syringae preferentially uses carnitine or the carnitine degradation

30 product glycine betaine as a compatible solute during osmoadaptation, and (iii) evaluate if the catabolism of plant-derived carnitine influences P. syringae colonization of epiphytic, apoplastic, and spermospheric sites of beans (Phaseolus vulgaris).

Materials and Methods

Bacterial strains, media, and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. P. syringae strains were routinely grown on solid King’s B medium (24) at 28°C preceding inoculation into

MinAS or into MinA broth (29), with amendments as indicated. Cells in liquid media were grown at 25°C with shaking. Antibiotics were added as needed at the following concentrations

(µg ml-1): kanamycin (Km), 50; rifampin (Rif), 100; tetracycline (Tet), 20; spectinomycin (Spc),

60; and cyclohexamide (Cyclo), 100. Cell growth was monitored in test tubes based on the optical density (OD) at 600 nm and in 96-well microtiter plates based on OD600 nm or OD405 nm/OD630 nm ratio to compensate for optical interference due to water condensation on the lid of the microtiter plate.

Construction of mutant strains.

Deletion mutants were constructed as described by Yu et al. (51). Briefly, two 1-kb fragments flanking the target locus were PCR amplified from B728a genomic DNA using primers listed in Appendix A Table A.1. A kan cassette containing a resistance gene flanked by two FLP recombination target (FRT) sites was PCR amplified from pKD13 using primers listed in Table A.1. The three fragments were joined together using splice-overlap-extension PCR, creating a fusion product consisting of a kan cassette flanked by two PCR-amplified regions

Table 1. Bacterial strains and plasmids used in this study.

Strain or plasmid Description or relevant genotype Reference or source P. syringae strain

B728a Wild Type; RifR Loper et al. (1987)

B728a deletion mutant lacking cdhA (Psyr_2918) encoding carnitine ΔcdhA Chen et al. (2013) dehydrogenase RifR B728a deletion mutant lacking cdhC (Psyr_2917) encoding a BKACE ΔcdhC Li et al. (2013) family enzyme; RifR S. Li and G. B728a deletion mutant lacking cdhB (Psyr_2919) encoding a ΔcdhB Beattie, thioesterase; RifR unpublished data S. Li and G. B728a deletion mutant lacking dhcAB (Psyr_3237-3238) encoding 3- ΔdhcAB Beattie, oxoacid CoA transferase subunits A and B; RifR unpublished data B728a deletion mutant lacking gbcAB (Psyr_4775-4776) encoding ΔgbcAB Li et al. (2013) glycine betaine catabolism complex; RifR

Complementation strain for deletion mutant lacking cdhA overexpressing ΔcdhA pNcdhA This work cdhA; KanR RifR

B728a triple deletion mutant lacking opuCA (Psyr_4248-4252), betT BT Chen et al. (2010) (Psyr_4827), and cbcXWV (Psyr_4709-4711) QAC transporters; RifR

Plasmids pTOK2T Suicide vector in P. syringae; TetR Chen et al. (2010) pN pME6041 with nptII promoter; KanR Chen et al. (2010) pNcdhA Overexpression vector expressing cdhA driven by PnptII; KanR This work Template for kan cassette flanked by FLP recombination target sites; Datsenko et al. pKD13 KanR (2000) pFLP2Ω Encodes Flp recombinase; SpcR Li et al. (2013) pRK2013 RP4 transfer functions for mobilization; KanR Ditta et al. (1980)

from B728a. The fused fragment was then ligated to SmaI-digested pTOK2T, and introduced into B728a via a tri-parental mating utilizing the helper plasmid pRK2013. Marker-exchange derivatives were identified as RifR KmR TetS and deletion of the target locus was confirmed by

PCR. The kanamycin resistance gene was removed by introducing pFlp2Ω, and excision was identified based on RifR KmS TetS in the colonies and was confirmed by PCR and sequencing.

The pFlp2Ω plasmid was cured by counter-selection on 20% sucrose and the cured deletion mutants were identified as RifR KmS TetS SpcS colonies.

Nutrient utilization assay.

Overnight cultures grown in MinAS medium were adjusted to an OD600 nm of 0.1. Cells in 1 ml of this suspension were washed three times with nanopure water and re-suspended in 1 ml of water, and 20 µl aliquots were added to 180 µl of MinA with either succinate, L-carnitine,

L-phenylalanine, L-leucine, or acetoacetate (each at 20 mM) in 96-well plates. The plates were then incubated at 25ºC in a microtiter plate shaker at 300 rpm and the OD600 nm was recorded at

45 min intervals using an automated microplate reader. The succinate, L-carnitine, L- phenylalanine, and L-leucine were purchased from Sigma-Aldrich (St. Louis, MO). The acetoacetate was prepared by complete hydrolysis of ethyl acetoacetate, purchased from Sigma-

Aldrich, as described previously (47).

Acetoacetate accumulation analysis.

B728a and ΔdhcAB were grown overnight in 200 ml MinAS, washed three times with ice-cold nanopure water, re-suspended in 50 ml MinA with 40 mM L-carnitine as a carbon source and incubated at 25ºC with shaking for 8 h. After incubation, the cultures were cooled on ice and the cells were washed three times with ice-cold nanopure water. After washing, the cells were re-suspended in 2 ml of ice-cold 80% ethanol, vortexed, allowed to reach room temperature, and centrifuged to remove cell debris. The supernatant was collected and analyzed using an Agilent Quadrapol Time of Flight mass spectrometer (QTOF-MS) (Wilmington, DE,

USA). 1 µL of sample was directly injected with a flow rate of 0.5 mL/min and mobile phase of

50:50:0.1 methanol, water, and formic acid. Samples were introduced into the mass

33 spectrometer using electrospray ionization and positive mode detection. Skimmer cone and fragmentor voltages were set to 40 V and 175 V, respectively, and mass range of m/z 50-1000 was collected. Extracted ion chromatograms were generated for the protonated, sodiated, and potassiated ions of acetoacetic acid and carnitine, using a mass window of ± 0.005 atomic mass units. The integrated areas for the three adducts of each metabolite were summed and reported as the total metabolite abundance.

Osmoprotection Assay.

Overnight cultures grown in MinAS medium were adjusted to an OD600 nm of 0.1. Cells in 1 ml of this suspension were washed three times with nanopure water and re-suspended in 1 ml of water. 20 µl aliquots were added to 180 µl of MinAS medium with either 0, 0.3, or 0.6 M

NaCl and with and without 1 mM L-carnitine amended in 96-well plates. The plates were then incubated at 25ºC in a microtiter plate with shaking at 300 rpm. OD600 nm was recorded at 45 min intervals using an automated microplate reader.

Compatible solute accumulation analysis based on MALDI –MSI of bacterial colonies.

Accumulation of glycine betaine, carnitine and N-acetylglutaminylglutamine amide

(NAGGN) in whole cells was analyzed using Matrix Assisted Laser desorption/ionization mass spectrometry imaging (MALDI-MSI). Cells were grown on thin agar films that had been created by pouring 15 ml of medium containing 1.5% agar into a 100-mm x 15-mm polystyrene petri plate. The medium was MinAS containing 1 mM L-carnitine and NaCl added at final concentrations of 0, 0.3, or 0.6 M. Colonies were excised from plates and placed onto microscope slides. The excised colonies were dried at 200 mtorr and 37°C for 1 h. A dihydroxybenzoic acid (DHB) matrix was applied to the surface of the colonies via sublimation

34 by placing 500 mg of DHB into the bottom of the condensation apparatus and placing the apparatus on a heating mantle. The microscope slide was taped to the bottom of the condenser and the apparatus was placed under vacuum (~150 mtorr). Ice water was added to the condenser and the apparatus was heated to 230°C. The DHB matrix was applied until a vapor layer was seen on the sides of the apparatus (5-7 min). The apparatus was removed from the heat and water was added to the ice bath to bring the sample to room temperature before it was returned to atmospheric pressure. The matrix-coated sample was removed and subjected to MALDI-MSI data acquisition, as described below. A raster step size of 100 µm was used and each spectrum was collected with 10 laser shots per scan and one scan per raster step. Orbitrap scans were acquired over the m/z range of 100 to 2000.

The MALDI-MSI data were acquired using a linear ion trap-orbitrap mass spectrometer with MALDI ion source (MALDI LTQ-Orbitrap Discovery; Thermo Scientific, San Jose, CA,

USA). The instrument was modified to use an external frequency tripled, diode-pumped

Nd:YAG laser operating at 355 nm (UVFQ; Elforlight Ltd., Daventry, UK). Laser energy of about 4-5 µJ/pulse was used at a 60-Hz repetition rate. The laser spot size was estimated as ~30

µm when determined from laser burn marks on a thin film of α-cyanohydroxycinnamic acid.

ImageQuest software (Thermo Scientific) was used to produce chemical images from MS imaging data sets. A mass tolerance of ±0.002 Da was used for generating Orbitrap MS images.

Epiphytic fitness assay.

An epiphytic fitness assay was performed to evaluate the ability of various strains to grow on leaf surfaces. Bacteria grown on solid King’s B medium for 48 h were suspended in water to a density of 2.5x105 cells ml-1. All plant studies used Phaseolus vulgaris cultivar Bush

Blue Lake 274 bean seeds. Bean plants were grown in 5-inch pots at a density of 10 seeds/pot until the primary leaves were fully expanded. The plants in each pot were inoculated by immersion of the leaves in the bacterial suspension for 1 min. After inoculation, the plants were immediately enclosed in a tent made of a plastic bag, with the leaves not touching the bag, and were left on the bench top for 24 h. To enumerate the bacterial populations, six leaves were removed from the plants inoculated with each bacterial strain at various times following inoculation (two leaves per pot at each time point). The leaves were placed individually into washing buffer (10 mM phosphate buffer, 0.1% peptone) and sonicated for 7 min. After mixing, the suspensions of recovered bacteria were serial diluted and plated onto solid King’s B medium amended with Rif and Cyclo and the enumerated cells expressed as colony forming units per gram of fresh leaf tissue.

Apoplastic fitness assay.

Bacteria were grown on solid King’s B medium as described for the epiphytic fitness assay and were suspended in sterile water containing 0.01% silwet L-77 (Lehle Seeds Round

Rock, Texas, USA) to a density of 2.5x105 cells ml-1. Bean plants with fully expanded primary leaves were inverted and submerged in a bacterial suspension inside a vacuum chamber. The chamber was then sealed and a vacuum drawn for 90 s; the vacuum was slowly released allowing the bacterial suspension to enter the apoplast of leaves as described previously (51). At each sampling time, two leaves were removed from each pot and placed in washing buffer containing

0.01% silwet. Whole leaves were then homogenized using sterile wood applicators and serial diluted, dilutions were plated, and bacterial populations were enumerated based on viable plate counts and expressed as colony forming units per gram of leaf tissue.

Chemical analysis of seed exudate.

Bean seeds were surface sterilized by submersion in 70% ethanol for 1 min and 10% bleach for 1 min, and repeatedly washed with sterile water until the bleach smell was no longer detected. Surface-sterilized seeds were then placed in 12-well tissue culture plates, with one seed and 1 ml of sterile water per well. The plates were subjected to a 16-h light/ 8-h dark photoperiod under plant growth lights, and the seeds were allowed to germinate for 0, 1, 2, and 3 days. At various stages of seed germination and growth, the liquid around the germinated seed was removed and diluted with water to a final volume of 1 ml; this seed solution was washed over the germinated seed three times to maximize recovery of the seed exudate. The collected exudates were analyzed using Liquid Chromatography Mass Spectrometry (LC-MS). Aqueous seed exudate samples were analyzed on an LC-MS 2020 (Shimadzu, Columbia, MD, USA) quadrupole mass spectrometer with a Zorbax HILIC Plus, 2.1x50 mm, 3.5 µm column (Agilent).

LabSolutions (Shimadzu Scientific Instruments, Columbia, MD) was used to define the acquisition parameters. Flow rate was set at 0.5 ml/min with a gradient elution as follows: 90%

MPB at 0 minutes, ramped to 75% at 1 minute and 20% at 3 minutes, held at 20% until 5 minutes, and re-equilibrated at 90% from 5 to 8 minutes. A drying line temperature of 180°C, heat block temperature of 300°C, and a nebulizing gas flow rate of 1.5 L/min was used. The auto sampler temperature was set at 4°C to minimize sample degradation. Data was collected using a Selected Ion Monitoring (SIM) scan event, with positive mode electrospray ionization that scanned the peaks of interest: choline (m/z 104.1), carnitine (m/z 162.1), and phosphorylcholine (m/z 184.1). Integration was performed on the SIM scan chromatograms and the peak area was used for quantification. The signal-to-noise ratio (S/N) was also measured for each sample to establish a limit of quantitation (LOQ). Peaks with an S/N ratio <10 were

37 considered below the LOQ. The final concentration was calculated for all samples, and the highest concentration that had S/N <10 was used for the LOQ.

Spermosphere fitness assay.

Bacterial inocula were prepared as described in the epiphytic fitness assay. Sand

(Quikrete Premium Play Sand, Lowes, Ames, IA, USA) was washed by allowing water to percolate, ~1.5 gallon/h, for 12 h. Excess water was decanted and the sand was dried at 120°C for 12 h, or until completely dry, and then autoclaved. The field capacity was determined by assessing the amount of water, based on weight, that remained associated with sand following saturation and drainage, with 100% field capacity expressed as g water per g dried sand. Bean seeds were submerged in the bacterial suspension for 1 min and planted in washed sterile sand watered to 60% field capacity and incubated at 25°C under constant light. Bacterial populations were enumerated over time by removing the seed from the sand, placing it in 3 ml of washing buffer and sonicating for 7 min followed by dilution plating as described in the epiphytic fitness assay, except that the results were expressed as the CFU per sample, where the sample was the entire germinating seed or developing seedling.

Results

The operons cdhCAB and dhcAB are required for P. syringae catabolism of carnitine as a sole carbon source.

Psyr_2916 encodes the substrate binding protein CdhX, which interacts with the transporter CbcWV to specifically import carnitine. We found that cdhX is co-transcribed with three other genes, Psyr_2917-2919 (Figure 1). These genes share synteny with the P. aeruginosa operon cdhCAB (Figure 2); the encoded proteins are conserved and show high similarity across

Lanes: 1 2 3 4 5 6 7 8 9 Primer pairs: AB CD EF GH IJ AD CF EH GJ

A B C D E F G H I J

cdhR cdhX cdhC cdhA cdhB

Figure 1. The carnitine locus in P. syringae B728a contains two operons. Primer locations within the carnitine locus are shown. The primer pairs that were used to amplify regions within each gene and across intergenic regions are shown, where the template DNA was either B728a genomic DNA (top) or cDNA created from B728a cells that were grown in the presence of 20 mM L-carnitine (bottom).

cdhR cdhX cdhC cdhA cdhB dhcR dhcA dhcB

P. syringae B728a 2915 2916 2917 2918 2919 3236 3237 3238

P. aeruginosa PA01 5389 5388 5387 5386 5385 1998 1999 2000

P. putida KT2440 0305 0304 0303 0302 0301 3122 3121 3120

P. fluorescens Pf0-1 5229 5230 5231 5232 5233 2077 2076 2075

Figure 2: Comparison of the carnitine loci to the genomes of related Pseudomonads. Conservation of the gene clusters based on analysis of protein sequence using BLASTp algorithm through the Pseudomonas Genome Database displays relatedness between pseudomonads at the protein level indicating related function.

various Pseudomonas sp. (Figure 2). Moreover, P. syringae has orthologs to dhcAB (Figure 2), which are also required for carnitine degradation in P. aeruginosa (49). To identify whether the

P. syringae genes cdhCAB and dhcAB are required for carnitine catabolism, we constructed deletions and designated the resulting mutants as ΔcdhA, ΔcdhB, ΔcdhC, and ΔdhcAB,

39 respectively. The genes cdhA and cdhB were previously identified as required for carnitine catabolism in P. aeruginosa (49), but cdhC was not examined even though cdhC is the second gene in an operon with cdhX, cdhA and cdhB (Figure 1).

When succinate was provided as a sole carbon source to B728a, ΔcdhC, ΔcdhA, ΔcdhB,

ΔdhcAB, and the QAC transporter-deficient mutant BT, the strains all demonstrated similar growth dynamics, indicating that deletion of these genes did not interfere with central metabolism (Figure 3A). In contrast, all of the mutants were deficient in growth on carnitine as a sole carbon source (Figure 3B), providing evidence that cdhA, cdhB, and also cdhC and dhcAB are required for utilization of carnitine.

Growth on L-carnitine A Growth on succinate B 1.0 1.0

0.8 0.8

0.6 0.6 B728a

600 nm cdhA 600 nm ΔcdhA 600 nm 600 nm cdhBΔcdhB O.D. O.D. 0.4 0.4

OD cdhCΔcdhC dhcABΔdhcAB 0.2 0.2 BT

0.0 0.0 0 10 20 30 40 50 0 10 20 30 40 50 TimeTime (h) (h) TimeTime (h) (h)

Figure 3. Deletion of cdhA, cdhB, cdhC, or dhcAB blocked growth on L-carnitine without interfering with central metabolism. Strains were grown in MinA medium with (A) 20 mM succinate and (B) 20 mM L-carnitine as the sole carbon source. Strain BT is deficient in carnitine uptake. Values are mean ± standard error of the mean (SE) (n=5). Results are representative of 3 replicated experiments (presented in Appendix A).

The growth deficiencies of ΔcdhC and ΔdhcAB were of particular interest because of the ambiguity in the literature on the roles of CdhC and DhcAB in the catabolism of carnitine. Three previously proposed pathways are shown in Figure 4. Our results do not fully support any of

40 these pathways. Pathways I and III omit roles for CdhC and CdhB, respectively, but our results indicate that each of these proteins are required for carnitine catabolism (Figure 3B). Similarly, the requirement for DhcAB contradicts the experimental evidence supporting pathway II, in which Bastard et al. (3) demonstrated enzymatically that CdhCAB are the minimum requirement for L-carnitine degradation to glycine betaine in P. aeruginosa. Our results demonstrate CdhC and DhcAB are required for carnitine degradation.

A B C Pathway I Pathway II Pathway III

Carnitine

CdhA CdhA CdhB CdhA

3-Dehydrocarnitine

DhcAB CdhC CdhC

3-Dehydrocarnitine CoA

AtoB CdhB DhcAB

Glycine betaine CoA

Unknown Thiolase

Glycine betaine

Figure 4. Proposed L-carnitine degradation pathways of P. aeruginosa PA01 pathways proposed by (A) Wargo et al. 2009 (49) (B) Bastard et al. 2014 (3) and (C) Meadows et al. 2015 (32).

DhcAB contributes to the catabolism of acetoacetate, an intermediate by-product of the carnitine degradation pathway.

DhcAB is annotated as subunits A and B of a 3-oxoacid CoA transferase (50), an enzyme that adds CoA onto an acid such as acetoacetate. Bastard et al. (3) showed that CdhC-mediated cleavage of 3-dehydrocarnitine releases acetoacetate, and thus that acetoacetate could be available as a substrate for DhcAB. Further support for a role for DhcAB in modifying

Carnitine Phenylalanine CdhA

3-dehydrocarnitine

CdhC

Acetoacetate Glycine betaine CoA DhcAB CdhB

Acetoacetyl-CoA Glycine betaine

Figure 5. Proposed alternative pathway for carnitine catabolism in Pseudomonas syringae B728a.

42 acetoacetate is that DhcAB in P. putida catalyzes the conversion of acetoacetate to acetacetyl-

CoA during phenylalanine and tyrosine degradation (35). Collectively, these reports and our findings are consistent with the possibility that 3-dehydrocarnitine cleavage releases acetoacetate that is subsequently modified by DhcAB (Figure 5), with the loss of this modification activity in

∆dhcAB preventing growth in the presence of carnitine. Moreover, DhcAB may metabolize acetoacetate in both the phenylalanine and carnitine catabolism pathways as a junction point; this function of DhcAB in carnitine catabolism would be overlooked in an enzymatic study (3).

A B C

0.8 0.45 0.6

0.7 0.40 0.5

0.6 0.35

0.4 0.5 0.30

600 nm 600 nm 0.4 0.25 0.3

0.3 0.20

OD 0.2 B728a 0.2 0.15 cdhAΔcdhA 0.1 0.1 0.10 dhcABΔdhcAB

0.0 0.05 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 0 20 40 60 Time (h) Time (h) Time (h)

Figure 6. DhcAB is involved in acetoacetate metabolism a central intermediate in leucine, phenylalanine, and carninitine catabolic pathways. Growth of B728a and the, ΔcdhA, and ΔdhcAB mutants in MinA medium with either (A) 20 mM L-leucine, (B) 20 mM L-phenylalanine, or (C) 20 mM acetoacetate, provided as carbon sources. Values are the mean ± SE (n=5). Results are representative of 3 replicated experiments (Appendix A).

To test whether DhcAB mediates acetoacetate catabolism, we examined the growth of

ΔdhcAB and B728a on L-phenylalanine, acetoacetate, and L-leucine, which is another amino acid predicted to progress through acetoacetate during catabolism (1). As indicated in Figure 6,

ΔdhcAB did not reach the same cell density as B728a during growth on L-phenylalanine, L- leucine, or acetoacetate; this supports a role for DhcAB in acetoacetate metabolism. As further evidence, cell lysates of ΔdhcAB and B728a that were exposed to L-carnitine for 8 h were analyzed using ESI QTOF mass spectrometry (Figure 7); ΔdhcAB accumulated more

43 acetoacetate and L-carnitine than B728a. Collectively, these results support our proposed pathway (Figure 5) and our prediction that DhcAB facilitates the metabolism of acetoacetate during carnitine utilization; the results suggest that blocking carnitine utilization at the step catalyzed by DhcAB causes feedback inhibition resulting in a growth-inhibiting buildup of metabolite of the carnitine degradation pathway in the cells.

A B

4e+6 25000 * * 20000 3e+6

15000

2e+6

10000 Total abundance abundance Total 1e+6 5000

0 0 B728a ΔdhcAB B728a ΔdhcAB Figure 7. DhcAB metabolizes acetoacetate, an intermediate by-product of the central carnitine catabolic pathway. Cells were grown in MinAS medium, washed, and incubated with 40 mM L-carnitine for 8 h before lysis and quantification of the (A) carnitine and (B) acetoacetate in the cell lysates by ESI QTOF mass spectrometry. Total abundance of acetoacetate and carnitine was calculated as the sum of the individual abundances of the protonated, sodiated, and potassiated adducts. Asterisks indicate differences between B728a and ΔdhcAB (p<0.05 by student’s t-test). Values are mean ± SEM (n=3).

Carnitine provides osmoprotection in hyperosmotic conditions and can function as a compatible solute.

To test whether P. syringae derives osmotic protection from carnitine under hyperosmotic conditions, B728a was grown in MinAS amended with 0, 0.3, and 0.6 M NaCl in both the presence and absence of 1 mM L-carnitine. Carnitine amendment reduced the duration of the lag phase in cells grown with 0.6 M NaCl, and to a lesser extent with 0.3 M NaCl (Figure

8), illustrating that carnitine functions as an osmoprotectant, as previously reported (7).

A 0 M NaCl B 0.3 M NaCl C 0.6 M NaCl 0.22 0.22 0.22

0.20 0.20 0.20

0.18 0.18 0.18

0.16 0.16 0.16

0.14 0.14 0.14

0.12 0.12 0.12 B728aB728a 405 nm/630 nm nm 405 nm/630 0.10 0.10 0.10 B728aB728a + + CAR 1 mM Carnitine !cdhA OD 0.08 0.08 0.08 ΔcdhA

0.06 0.06 0.06 Δ!cdhAcdhA + + CAR 1 mM Carnitine

0.04 0.04 0.04 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 TimeTime (h) TimeTime (h) TimeTime (h) Figure 8. Carnitine catabolism was not required for carnitine to function in osmoprotection. B728a and ΔcdhA were grown in MinAS with and without 1 mM L-carnitine (CAR) and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD ratios in microtiter plates. Values are means ± SE (n=3). Results are representative of 3 replicated experiments (Appendix A).

To evaluate whether B728a accumulated carnitine as a compatible solute, as in B. subtilis and L. monocytogenes (23, 45), or accumulated the degradation product glycine betaine as a compatible solute, as in P. aeruginosa, we grew B728a and ΔcdhA at increasing osmolarity with and without 1 mM L-carnitine. The mutant ΔcdhA exhibited similar growth to B728a (Figure 8), suggesting that L-carnitine can function directly as a compatible solute in B728a.

These data however, do not indicate whether the wild-type strain, B728a, accumulates carnitine or glycine betaine as its compatible solute. To answer this question, we grew B728a,

ΔcdhA, the glycine betaine catabolic mutant ΔgbcAB (29), and BT (8) on MinAS agar at 0, 0.3, and 0.6 M NaCl with 1 mM L-carnitine. Upon reaching their maximal size, colonies were analyzed using MALDI-MSI to determine the nature of the compatible solutes that accumulated in the cells in response to the hyperosmotic stress. MALDI-MSI revealed that ΔcdhA accumulated carnitine whereas B728a accumulated both carnitine and glycine betaine at 0.3 and

0.6M NaCl (Figures 9B and 9C). The observation of carnitine accumulation in B728a is likely due to either carnitine associated with the cell, as seen in BT (Figure 9), or due to feedback

45 inhibition of the carnitine degradation pathway due to the accumulation of glycine betaine. This observation indicates that L-carnitine can function as a compatible solute in B728a, but that it is preferentially degraded to glycine betaine for accumulation as a compatible solute in the wild- type strain.

A B728a ΔcdhA ΔgbcAB BT glycine betaine

carnitine

trehalose

NAGGN B glycine betaine

carnitine

trehalose

NAGGN

C glycine betaine

carnitine

trehalose

NAGGN

Figure 9. Carnitine can function as a compatible solute under osmotic stress, although it was not a preferred compatible solute. MALDI-MSI analysis of compatible solute accumulation in bacterial colonies grown in MinAS amended with 1 mM carnitine and various NaCl concentrations: (A) 0 M, (B) 0.3 M, and (C) 0.6 M NaCl. Images represent relative abundance across colonies for the indicated compounds based on the m/z values 162.112 (carnitine + H+), 118.087 (glycine betaine + H+), and 354.119 (NAGGN + K+). Colonies are not shown for strain BT at 0.6 M NaCl due to poor growth at this high level of osmotic stress. Results are representative of 3 replicated experiments.

Carnitine contributes to spermosphere colonization at distinct stages during germination.

Based on the function of carnitine in β-oxidation, we predict that carnitine is abundant in, and is released by, germinating seeds. To test this, we collected the exudates of bean seeds at various stages during incubation in sterile water, with these stages designated as 0 – 4 (Figure

10A). We analyzed them for their carnitine content using HPLC-MS. The carnitine abundance peaked at stages 1 and 2 and then decreased as the seedlings developed (Figure 10B).

A Stage 1 Stage 2 Stage 3 Stage 4

B 4.6 Carnitine

4.4

4.2

4.0

3.8 Log(abundance) Log(abundance) 3.6

3.4

3.2 0 1 2 3 4 5 Germination stage

Figure 10. Carnitine was released from germinating bean seeds and seedlings at distinct developmental stages. (A) Germinating seeds were placed into one of four developmental stages: stage 1, imbibition; stage 2, radicle emergence; stage 3, radicle elongation; and stage 4, lateral root formation/greening. (B) The abundance of carnitine in exudates was quantified by LCMS for seeds at each germination stage. Values are mean ± SE (stage 1, n=17; Stage 2, n=27; stage 3, n=27; and stage 4, n=23). Results are representative of 3 replicated experiments.

A A 9 AB 8 B AB C 7 C C B728aWT 6 ΔcdhAcdhA Log (CFU/sample) (CFU/sample) Log ΔpNcdhAcdhA ( pNcdhA) BTBT 5 Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure 11. Carnitine utilization contributed to B728a colonization of the spermosphere during imbibition and radical emergence. Bean seeds were inoculated with B728a, ΔcdhA, ΔcdhA (pNcdhA), and BT and population sizes were estimated at 0, 1, 2, and 3 days post-inoculation based on viable plate counts. Stages correspond to seed developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. Values are mean ± SEM (n=6). Results are representative of 3 replicated experiments (Appendix A). Within each time point, values indicated by the same letter do not differ significantly (p<0.05 based on an ANOVA).

To determine whether carnitine catabolism contributed to bean seed colonization, we evaluated the growth of the carnitine catabolic mutant ΔcdhA and its complement

ΔcdhA(pNcdhA) on seeds. All of the strains grew similarly on seeds while they were imbibing water (stage 1 in Figure 11), indicating that carnitine, if present, did not have a detectable impact on bacterial growth at this stage. In stage 2, however, ΔcdhA grew much less than B728a and expression of cdhA in ΔcdhA(pNcdhA) at least partially restored growth (Figure 11); these results demonstrate that carnitine is a major contributor to population development on seeds during radicle emergence. By stage 4, the ΔcdhA mutant had achieved a population size that was significantly larger than BT (Figure 11), indicating that an additional QAC was likely released

48 from the seedlings that contributed to growth. Moreover, differences in population size between

ΔcdhA and ΔcdhA(pNcdhA) were no longer significant, indicating that carnitine catabolism did not contribute to growth on seedlings during greening and lateral root formations.

The growth dynamics of B728a and ΔcdhA on germinating seeds correlated with the chemical analysis of the seed exudates. In particular, the bacteria appeared to exploit carnitine primarily after the seeds had imbibed water and the radicles were emerging, and this is at, and just after, the time at which maximal levels of carnitine were released (Figure 10). Once the seedlings had established lateral roots and initiated greening, the carnitine levels decreased (stage

4 in Figure 10) and the impact of carnitine on population development lessened based on the lack of differences between the wild type and ΔcdhA populations (stage 4 in Figure 11).

Carnitine does not significantly contribute to colonization of leaves.

In order to evaluate the contribution of carnitine to colonization of leaf surfaces, epiphytic fitness was evaluated for B728a, ΔcdhA, and ΔdhcAB on plants incubated for 24 h with leaf surface wetness and then 48 h of drying stress. The ΔcdhA mutant grew similarly to B728a throughout the incubation (Figure 12A), indicating that carnitine was not used as a nutrient during epiphytic colonization; this is consistent with the low expression of carnitine catabolic genes during epiphytic growth (Table 2) observed in a previous study (51). This test did not address the use of carnitine as an osmoprotectant since ΔcdhA derives osmoprotection from carnitine (Figure 8). The lack of carnitine use on leaves is likely due to a lack of carnitine availability. The carnitine catabolic genes, however, were up-regulated during the apoplastic growth of B728a in bean leaves (Table 2), suggesting that carnitine may be available in the intracellular spaces of leaves. However, ΔcdhA and ΔdhcAB do not differ from B728a in apoplastic colonization (Figure 12B).

A B

9 10

9 8

8 7 /g leaf) leaf) /g 7 cfu 6 B728aB728a B728aB728a

Log( cdhAΔcdhA 6 cdhAΔcdhA dhcABΔdhcAB dhcABΔdhcAB 5 5 wet dry

4 4 0 20 40 60 80 0 10 20 30 40 50 60 Time (h) Time (h)

Figure 12. Carnitine does not contribute to colonization of leaves by B728a. (A) B728a, ΔcdhA, and ΔdhcAB populations were monitored during epiphytic colonization during 24 h water replete conditions followed by 48 h of drying stress. (B) B728a, ΔcdhA, and ΔdhcAB populations were monitored during apoplastic colonization following vacuum infiltration. Populations were enumerated based on viable plate counts. Values are mean ± SEM (n=6). Results are representative of 3 replicated experiments (Appendix A).

Table 2. Fold changes relative to fluorescent intensities of carnitine transport and catabolism genes of P. syringae B728a1 Fluorescence Fold-change2 Intensity3 Epiphytic Apoplastic Psyr # Gene Product Name Basal Medium sites sites Quaternary Ammonium Compound (QAC) Transport Psyr_2917 cdhX ABC transporter, carnitine binding protein 75 1.9 1.8 Quaternary Ammonium Compound (QAC) Metabolism Psyr_2917 cdhC Beta-keto cleavage enzyme 56 1.5 2

Psyr_2918 cdhA 3-hydroxyacyl dehydrogenase 1.2 1.9

Psyr_2919 cdhB Thioesterase 53 1.1 2

Psyr_3237 dhcA 3-oxoacid CoA-transferase subunit A 162 16.4 5.5

Psyr_3238 dhcB 3-oxoacid CoA-transferase subunit B 211 13.8 6.7

______1 Data are derived from Gene Expression Omnibus accession no. GSE42544. 2The fold change values represent the change in transcript abundance in epiphytic and apoplastic environment as compared to abundance in basal medium HMM. Yu et al. 2013 3Fluorescent intensities across the B728a transcriptome ranged from 37 to 54119.

Discussion

This work clarified the L-carnitine degradation pathway in P. syringae, including identifying the likely function of DhcAB in the pathway. Given the similarity of the catabolic loci across the Pseudomonads, this pathway is likely shared by P. aeruginosa and other

Pseudomonads that have the cdhCAB locus. This study was the first to investigate the role of plant-derived carnitine in plant-bacteria associations. While we found no evidence for a role of carnitine in the leaf-associated growth of a foliar pathogen, we found strong evidence for a clear role in seed colonization. This is consistent with a physiological role for carnitine in fatty acid oxidation and a role for fatty acid oxidation in the germination of bean seeds. Importantly, we demonstrated that a single QAC released during germination could be a major driver of bacterial growth during seed colonization. Lastly, we refined our knowledge of carnitine as an osmoprotectant in P. syringae B728a by demonstrating that, while it can serve directly as a compatible solute, it generally is degraded to glycine betaine for accumulation, similar to in P. aeruginosa but in contrast to in B. subtilis and L. monocytogenes.

Our results demonstrated that carnitine is released by germinating seeds, is biologically available to seed-colonizing bacteria, and significantly enhances the growth of these bacteria.

Following imbibition of water into bean seeds, a pulse of carnitine is released, as observed in an analysis of seed exudates (Figure 10B); this pulse correlates with a significant reduction in the populations of ΔcdhA as compared to B728a (Figure 11). These data collectively demonstrate that carnitine is present and biologically available to the bacteria in the spermosphere. The

ΔcdhA populations at 2 days post inoculation (dpi) were similar to the populations of the QAC uptake-deficient strain BT, demonstrating that carnitine was the major QAC contributing to spermosphere growth during this stage of seed germination (Figure 11). Interestingly, P.

51 syringae contains a methyl accepting chemotaxis protein (MCP) (Psyr_4706), which is encoded by a gene that is co-localized with the genes for the ABC transporter CbcWV (Psyr_4710-4711); this protein may function in chemosensing of L-carnitine, consistent with a model in which the carnitine functions as a signal to enhance movement to the germinating seed (28).

We demonstrated that CdhA and CdhB are required for the catabolism of carnitine based on the fact that cdhA and cdhB deletions eliminated growth on carnitine as a carbon source

(Figure 4B), but this did not resolve the distinct roles of these proteins in the catabolic pathway.

Uanschou et al. (44) showed that several organisms with carnitine dehydrogenase activity possess a fused cdhAB hybrid gene encoding a carnitine hydrogenase (CdhA) joined through a small linker to a thioesterase (CdhB). This arrangement suggests that these proteins function as a complex that catalyzes the first step of the carnitine degradation pathway (44). Wargo et al. (49) initially proposed that the CdhAB complex catalyzes the first step of the pathway, namely the conversion of carnitine to 3-dehydrocarnitine. This possibility is supported by the presence of a fused cdhAB gene in Xanthomonas translucens and Rhizobium sp. (2). However, using these organisms, Arima et al. (2) demonstrated that cleavage of the thioesterase (CdhB) domain did not affect the ability of the CdhAB complex to convert carnitine to 3-dehydrocarnitine, demonstrating that the CdhB domain is not required for the first step of carnitine degradation.

Moreover, the predicted function of CdhB as a thioesterase is not consistent with the enzymatic function required for converting carnitine to 3-dehydrocarnitine, as this step is an oxidation of a

3-hydroxyl group rather than the transfer of a CoA moiety (41, 52).

Wargo et al. (49) omitted CdhC from their proposed carnitine degradation pathway in P. aeruginosa; however, we found that CdhC is required for carnitine catabolism in P. syringae.

Through a bioinformatic analysis, Bastard et al. (3) identified CdhC as a beta-ketoacid cleavage

52 enzyme, or BKACE. Using an enzymatic assay with whole cell extracts, they demonstrated that

CdhC acts specifically on 3-dehydrocarnitine, cleaving it into glycine betaine-CoA and acetoacetate. Furthermore, they confirmed that CdhA from P. aeruginosa PAO1 converts carnitine to 3-dehydrocarnitine by overexpressing the enzyme in E. coli and monitoring its

NADH consumption; this provided evidence for the first step of the pathways shown in Figures

3B, 3C and 7. They also showed that without CdhB, glycine betaine was not produced, confirming the third step of the pathways shown in Figures 4B and 5, and critically, that CdhC mediated the cleavage of 3-dehydrocarnitine to glycine betaine-CoA and acetoacetate, as shown in Figure 5. Upon the addition of CdhB into the reaction mixture, they detected an increase in released CoA using a colorimetric assay and the production of glycine betaine through analysis with mass spectrometry. By defining the functions of CdhC and CdhB, Bastard et al. (3) were able to assign function to all the genes in the cdhCAB operon. Our results are consistent with the pathway shown in Figures 4B and 5 and contradict those shown in Figures 4A and 4C.

Although DhcAB was not required for converting L-carnitine to glycine betaine in whole cell extracts of P. aeruginosa PAO1 (3), our results with B728a are consistent with genetic studies of PAO1 (49) that found that a dhcAB deletion blocks growth on carnitine. In a recent review article, Meadows et al. (32) proposed a carnitine catabolic pathway (Figure 4C) based on the pathway of Bastard et al. (3) (Figure 4B) in which they replaced CdhB with DhcAB for the last step and omitted CdhB entirely without explanation. A role for CdhB had been demonstrated both biochemically (3) and genetically based on the phenotype of a cdhB deletion mutant of PAO1 (49), and we further verified this role in P. syringae B728a (Figure 3B). These findings, however, did not resolve the actual role of DhcAB in carnitine catabolism. The protein sequences of DhcA and DhcB in a Blastp analysis showed their strongest similarity to be to β-

53 ketoacid CoA transferase subunits A and B, respectively. Functional studies on DhcA and DhcB in P. aeruginosa demonstrated a role in tyrosine and phenylalanine catabolism, and specifically in catalyzing the conversion of acetoacetate to acetoacetate-CoA (35). Based on the recent finding that acetoacetate is a by-product of CdhC-mediated cleavage of 3-dehydrocarnitine (3), we predicted that DhcAB uses acetoacetate from multiple catabolic pathways as a substrate, including pathways for phenylalanine and leucine. This prediction was strongly supported by our finding that deleting dhcAB blocked growth of P. syringae on phenylalanine and leucine as well as carnitine when provided as a sole carbon source (Figure 6). Furthermore, loss of dhcAB reduced growth on acetoacetate as a carbon source and increased accumulation of acetoacetate during growth on carnitine (Figure 7), thus providing strong evidence that the role of DhcAB in carnitine degradation in P. syringae, and likely other Pseudomonads, is to facilitate the removal of acetoacetate.

Interestingly, although carnitine had not previously been examined for a role in plant- microbe interactions due to limited investigations into its presence in plants, we found that it has a clear role during P. syringae colonization of seeds. Our data illustrate that carnitine was released from bean seeds during germination (Figure 10B); this presumably was due to the production of carnitine as a carrier for importing fatty acids into mitochondria for the oxidation of seed oils during germination. Our data also illustrate that this released carnitine was available to colonizing microorganisms and that it supported the proliferation of P. syringae B728a populations in the spermosphere (Figure 11). In contrast, we found that carnitine catabolism did not detectably contribute to colonization of leaves, either on leaf surfaces or in the interior

(Figure 12A and 12B), despite previous gene expression data (51) that indicated a low level of induction of the cdhCAB locus in epiphytic and apoplastic sites (Table 2). Given that carnitine

54 catabolism to glycine betaine is not required for osmoprotection (Figure 8C), the gene expression data are consistent with a role, albeit small, for carnitine as an osmoprotectant for P. syringae cells on or in leaves.

Previous studies have demonstrated an influence of carnitine on host-microbe interactions in animal systems, including L. monocytogenes, P. aeruginosa, and Staphylococcus aureus host- interactions (19, 37, 40, 45), but our results are the first to highlight carnitine as a plant-derived compound impacting the resident microbes. Carnitine has been examined for its cellular roles in other bacteria, including a few that are plant-associated. For example, carnitine can serve as an osmoprotectant and/or carbon source for S. meliloti (18, 43), Agrobacterium tumefaciens, X. translucens, Burkholderia cepacia, Brevibacterium linens, and even the cyanobacterial species

Aphanothece halophytica (2, 12, 20-22, 33). Moreover, in the phytopathogenic fungus

Magnaporthe grisea, a peroxisomal carnitine acetyl-transferase is critical to penetration hyphae during host invasion (4). Our findings indicate that, although plants do not produce as much carnitine as animals, they produce enough to impact the populations of colonizing bacteria, particularly on seeds.

These results raise the question of the breadth of plant species that produce carnitine at sufficient levels to influence colonizing microbes. Research into carnitine in plants has focused mainly on its role in fatty acid metabolism in Arabidopsis and peas (5), but carnitine has been found at detectable levels in the tissues of other plants as well (36). Another interesting question is whether P. syringae, which employs carnitine for multiple cellular roles, encounters carnitine during distinct stages of its lifecycle. For example, P. syringae B728a has been found to colonize pea aphids, an animal system, provoking the question of whether carnitine, as a common animal-associated compound, influences P. syringae fitness in aphids (42). Some final

55 questions are the breadth of seed-associated microbes that exploit carnitine and the extent to which carnitine is a driver for competition among these microbes. This question stems from one of our most surprising findings, namely that, despite the wealth of nutrients in the spermosphere, loss of the ability to utilize only one compound, carnitine, resulted in a detectable decrease in bacterial growth on seeds. This finding suggests that the ability to catabolize carnitine may provide a distinct fitness advantage during seed colonization, and thus that increasing the ability of a strain to exploit carnitine may be a useful strategy for increasing the establishment and thus efficacy of seedborne bacterial inoculants.

References

1. Aguilar J, Zavala A, Diaz-Perez C, Cervantes C, Diaz-Perez A, Campos-Garcia J. 2006. The atu and liu clusters are involved in the catabolic pathways for acyclic monoterpenes and leucine in Pseudomonas aeruginosa. Appl Environ Microbiol 72:2070-2079.

2. Arima J, Uesumi A, Mitsuzumi H, Mori N. 2010. Biochemical characterization of L- carnitine dehydrogenases from Rhizobium sp. and Xanthomonas translucens. Biosci, Biotechnol, Biochem 74:1237-1242.

3. Bastard K, Smith AAT, Vergne-Vaxelaire C, Perret A, Zaparucha A, De Melo- Minardi R, Mariage A, Boutard M, Debard A, Lechaplais C. 2014. Revealing the hidden functional diversity of an enzyme family. Nat Chem Biol 10:42-49.

4. Bhambra GK, Wang ZY, Soanes DM, Wakley GE, Talbot NJ. 2006. Peroxisomal carnitine acetyl transferase is required for elaboration of penetration hyphae during plant infection by Magnaporthe grisea. Mol Microbiol 61:46-60.

5. Bourdin B, Adenier H, Perrin Y. 2007. Carnitine is associated with fatty acid metabolism in plants. Plant Physiol Biochem 45:926-931.

6. Bremer J. 1977. Carnitine and its role in fatty acid metabolism. Trends Biochem Sci. 2:207-209.

7. Chen C, Beattie GA. 2007. Characterization of the osmoprotectant transporter OpuC from Pseudomonas syringae and demonstration that cystathionine-β-synthase domains are required for its osmoregulatory function. J Bacteriol 189:6901-6912.

8. Chen C, Li S, McKeever DR, Beattie GA. 2013. The widespread plant‐colonizing bacterial species Pseudomonas syringae detects and exploits an extracellular pool of choline in hosts. Plant J 75:891-902.

9. Chen C, Malek AA, Wargo MJ, Hogan DA, Beattie GA. 2010. The ATP‐binding cassette transporter Cbc (choline/betaine/carnitine) recruits multiple substrate‐binding proteins with strong specificity for distinct quaternary ammonium compounds. Mol Microbiol 75:29-45.

10. Constantin-Teodosiu D, Howell S, Greenhaff P. 1996. Carnitine metabolism in human muscle fiber types during submaximal dynamic exercise. J Appl Physiol 80:1061-1064.

11. D'Souza-Ault M, Smith L, Smith G. 1993. Roles of N-acetylglutaminylglutamine amide and glycine betaine in adaptation of Pseudomonas aeruginosa to osmotic stress. Appl Environ Microbiol 59:473-478.

12. Dalmastri C, Fiore A, Alisi C, Bevivino A, Tabacchioni S, Giuliano G, Sprocati AR, Segre L, Mahenthiralingam E, Chiarini L. 2003. A rhizospheric Burkholderia cepacia complex population: genotypic and phenotypic diversity of Burkholderia cenocepacia and Burkholderia ambifaria. FEMS Microbiol Ecol 46:179-187.

13. Eichler K, Bourgis F, Buchet A, Kleber HP, Mandrand‐Berthelot MA. 1994. Molecular characterization of the cai operon necessary for carnitine metabolism in Escherichia coli. Mol Microbiol 13:775-786.

14. Eichler K, Buchet A, Bourgis F, Kleber HP, Mandrand‐Berthelot MA. 1995. The fix Escherichia coli region contains four genes related to carnitine metabolism. J Basic Microbiol 35:217-227.

15. Eisenhut M, Planchais S, Cabassa C, Guivarc'h A, Justin AM, Taconnat L, Renou JP, Linka M, Gagneul D, Timm S. 2013. Arabidopsis A BOUT DE SOUFFLE is a putative mitochondrial transporter involved in photorespiratory metabolism and is required for meristem growth at ambient CO2 levels. Plant J 73:836-849.

16. Engel AG, Angelini C. 1973. Carnitine deficiency of human skeletal muscle with associated lipid storage myopathy: a new syndrome. Science 179:899-902.

17. Fritz IB. 1962. Carnitine and its role in fatty acid metabolism. Adv Lipid Res 1:285-334.

18. Goldmann A, Boivin C, Fleury V, Lecoeur L, Maille M, Tepfer D. 1991. Betaine use by rhizosphere bacteria: genes essential for trigonelline, stachydrine, and carnitine

catabolism in Rhizobium meliloti are located on pSym in the symbiotic region. Mol Plant Microbe Interact 4:571-578.

19. Gutierrez C, Abee T, Booth IR. 1995. Physiology of the osmotic stress response in microorganisms. Int J Food Microbiol 28:233-244.

20. Hanschmann H, Ehricht R, Kleber H-P. 1996. Purification and properties of l (−)- carnitine dehydrogenase from Agrobacterium sp. Biochim Biophys Acta-Gen Sub 1290:177-183.

21. Incharoensakdi A, Waditee R. 2000. Degradation of glycinebetaine by betaine- homocysteine methyltransferase in Aphanothece halophytica: effect of salt downshock and starvation. Curr Microbiol 41:227-231.

22. Jebbar M, Champion C, Blanco C, Bonnassie S. 1998. Carnitine acts as a compatible solute in Brevibacterium linens. Res Microbiol 149:211-219.

23. Kappes RM, Bremer E. 1998. Response of Bacillus subtilis to high osmolarity: uptake of carnitine, crotonobetaine and γ-butyrobetaine via the ABC transport system OpuC. Microbiol 144:83-90.

24. King EO, Ward MK, Raney DE. 1954. Two simple media for the demonstration of pyocyanin and fluorescin. J Lab Clinic Med 44:301-307.

25. Kleber H-P. 1997. Bacterial carnitine metabolism. FEMS Microbiol Lett 147:1-9.

26. Koves TR, Ussher JR, Noland RC, Slentz D, Mosedale M, Ilkayeva O, Bain J, Stevens R, Dyck JR, Newgard CB. 2008. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metabol 7:45-56.

27. Lawand S, Dorne A-J, Long D, Coupland G, Mache R, Carol P. 2002. Arabidopsis A BOUT DE SOUFFLE, which is homologous with mammalian carnitine acyl carrier, is required for postembryonic growth in the light. Plant Cell 14:2161-2173.

28. Lelandais‐Brière C, Jovanovic M, Torres GA, Perrin Y, Lemoine R, Corre‐Menguy F, Hartmann C. 2007. Disruption of AtOCT1, an organic cation transporter gene, affects root development and carnitine‐related responses in Arabidopsis. Plant J 51:154-164.

29. Li S, Yu X, Beattie GA. 2013. Glycine betaine catabolism contributes to Pseudomonas syringae tolerance to hyperosmotic stress by relieving betaine-mediated suppression of compatible solute synthesis. J Bacteriol 195:2415-2423.

30. Lucchesi GI, Lisa TA, Casale CH, Domenech CE. 1995. Carnitine resembles choline in the induction of cholinesterase, acid phosphatase, and phospholipase C and in its action as an osmoprotectant in Pseudomonas aeruginosa. Curr Microbiol 30:55-60.

31. Masterson C, Wood C. 2009. Influence of mitochondrial β‐oxidation on early pea seedling development. New Phytol 181:832-842.

32. Meadows JA, Wargo MJ. 2015. Carnitine in bacterial physiology and metabolism. Microbiol 161:1161-1174.

33. Mori N, Kasugai T, Kitamoto Y, Ichikawa Y. 1988. Purification and some properties of carnitine dehydrogenase from Xanthomonas translucens. Agric Biol Chem 52:249- 250.

34. Obón J, Maiquez J, Cánovas M, Kleber H-P, Iborra J. 1999. High-density Escherichia coli cultures for continuous L (−)-carnitine production. Appl Microbiol Biotechnol 51:760-764.

35. Palmer GC, Palmer KL, Jorth PA, Whiteley M. 2010. Characterization of the Pseudomonas aeruginosa transcriptional response to phenylalanine and tyrosine. J Bacteriol 192:2722-2728.

36. Panter R, Mudd J. 1969. Carnitine levels in some higher plants. FEBS Lett 5:169-170.

37. Qi Y, Miller KJ. 2000. Effect of low water activity on staphylococcal enterotoxin A and B biosynthesis. J Food Protect 63:473-478.

38. Rebouche CJ, Seim H. 1998. Carnitine metabolism and its regulation in microorganisms and mammals. Annu Rev Nutr 18:39-61.

39. Sleator RD, Wouters J, Gahan CG, Abee T, Hill C. 2001. Analysis of the role of OpuC, an osmolyte transport system, in salt tolerance and virulence potential of Listeria monocytogenes. Appl Environ Microbiol 67:2692-2698.

40. Smith LT. 1996. Role of osmolytes in adaptation of osmotically stressed and chill- stressed Listeria monocytogenes grown in liquid media and on processed meat surfaces. Appl Environ Microbiol 62:3088-3093.

41. Song F, Zhuang Z, Finci L, Dunaway-Mariano D, Kniewel R, Buglino JA, Solorzano V, Wu J, Lima CD. 2006. Structure, function, and mechanism of the phenylacetate pathway hot dog-fold thioesterase PaaI. J Biol Chem 281:11028-11038.

42. Stavrinides J, McCloskey JK, Ochman H. 2009. Pea aphid as both host and vector for the phytopathogenic bacterium Pseudomonas syringae. Appl Environ Microbiol 75:2230-2235.

43. Talibart R, Jebbar M, Gouffi K, Pichereau V, Gouesbet G, Blanco C, Bernard T, Pocard J. 1997. Transient accumulation of glycine betaine and dynamics of endogenous osmolytes in salt-stressed cultures of Sinorhizobium meliloti. Appl Environ Microbiol 63:4657-4663.

44. Uanschou C, Frieht R, Pittner F. 2005. What to learn from a comparative genomic sequence analysis of L-carnitine dehydrogenase. Monatsh Chem 136:1365-1381.

45. Verheul A, Glaasker E, Poolman B, Abee T. 1997. Betaine and L-carnitine transport by Listeria monocytogenes Scott A in response to osmotic signals. J Bacteriol 179:6979- 6985.

46. Verheul A, Rombouts FM, Beumer RR, Abee T. 1995. An ATP-dependent L-carnitine transporter in Listeria monocytogenes Scott A is involved in osmoprotection. J Bacteriol 177:3205-3212.

47. Walker P. 1954. A colorimetric method for the estimation of acetoacetate. Biochem J 58:699.

48. Wargo MJ. 2013. Homeostasis and catabolism of choline and glycine betaine: lessons from Pseudomonas aeruginosa. Appl Environ Microbiol 79:2112-2120.

49. Wargo MJ, Hogan DA. 2009. Identification of genes required for Pseudomonas aeruginosa carnitine catabolism. Microbiol 155:2411-2419.

50. Winsor GL, Lam DK, Fleming L, Lo R, Whiteside MD, Yu NY, Hancock RE, Brinkman FS. 2011. Pseudomonas Genome Database: improved comparative analysis and population genomics capability for Pseudomonas genomes. Nucleic Acids Res 39:D596-600.

51. Yu X, Lund SP, Scott RA, Greenwald JW, Records AH, Nettleton D, Lindow SE, Gross DC, Beattie GA. 2013. Transcriptional responses of Pseudomonas syringae to growth in epiphytic versus apoplastic leaf sites. Proc Natl Acad Sci U S A 110:E425- E434.

52. Zhuang Z, Song F, Zhao H, Li L, Cao J, Eisenstein E, Herzberg O, Dunaway- Mariano D. 2008. Divergence of function in the hot dog fold enzyme superfamily: the bacterial thioesterase YciA. Biochemistry 47:2789-2796.

CHAPTER 3. CHOLINE-O-SULFATE CATABOLISM IN PSEUDOMONAS

SYRINGAE REQUIRES A SULFATE EXPORTER THAT MAY FUNCTION TO

RELIEVE SULFATE TOXICITY AND IS ENCODED IN A LOCUS WITH

CHOLINE-O-SULFATE SULFATASE.

A paper to be submitted for publication

Michael D. Millican and Gwyn A. Beattie

Abstract

Pseudomonas syringae is able to degrade choline-O-sulfate (COS) as an osmoprotectant and a carbon, nitrogen, and sulfur source. The choline sulfatase is encoded by cosC is the first in an operon with a COS specific substrate binding protein cosX and the cos locus Psyr_0029-Psyr_0026. Deletion of CosC blocked catabolism of COS preventing its function as an osmoprotectant and nutrient source. Catabolism of COS did not impact P. syringae colonization of the host plant Phaseolus vulgaris, as demonstrated by comparing population sizes in epiphytic, apoplastic, and spermospheric fitness assays. The last gene in the cos locus and encoded in its own operon, CosE (Psyr_0026), is annotated as a Sul1 superfamily putative sulfate exporter. Upon deletion, ΔcosE was unable to utilize COS as carbon nitrogen or sulfur source, but was unaltered in its ability to utilize free succinate, ammonium, or free sulfate as nutrient sources. When ΔcosE was grown under osmotic stress

COS was toxic when provided as an osmoprotectant. Comparison of the ΔcosE catabolic phenotypes to sulfonate degradation systems, CosE likely functions in exporting sulfate release from cleavage of COS by CosC to avoid sulfate toxicity. This is the first report of

61 sulfate ester degradation coupled with sulfate exportation to avoid toxicity, a system proposed in sulfonate degradation.

Introduction

Choline-O-sulfate (COS) is a sulfated derivative of choline and belongs to a class of chemicals known as quaternary ammonium compounds (QACs). COS accounts for 30-45% of the sulfur in soil (31). Plants and some fungi produce COS through the sulfurization of choline using 3’-phosphoadenosine-5’-phosphosulfate as a sulfate donor, resulting in the production of COS and 3’-phosphoadenosine-5’-phosphate (33, 36). A wide variety of organisms, including plants, fungi, algae, lichens, and bacteria, utilize COS as a sulfate storage molecule, nutrient source, and/or compatible solute (1, 5, 9, 17, 18, 22). COS produced by plants and soil fungi could provide an opportunity for plant-associated microorganisms to gain access to sulfur.

Some organisms are additionally able to use COS as an osmoprotectant, and these include the plant species Limonium sp., fungal species Penicillium fellutanum, and bacterial species Bacillus subtilis, Sinorhizobium meliloti, and Pseudomonas putida (5, 19, 32, 34). B. subtilis is able to transport and accumulate COS during hyperosmotic stress through the

OpuC transporter (16). S. meliloti has the ability to not only transport COS but also catabolize it as a nutrient source. S. meliloti contains the operon betICBA, which encodes a

COS-specific sulfatase (BetC), as well as betaine aldehyde dehydrogenase (BetB) and choline dehydrogenase (BetA) (32). The sulfatase BetC hydrolyzes COS to choline and sulfate, and BetAB converts the resultant choline to glycine betaine. Glycine betaine can be accumulated as a compatible solute or degraded as a nutrient source. Like S. meliloti, P. putida is able to utilize COS as an osmoprotectant and nutrient source via conversion to

62 glycine betaine (34). The genetic arrangement of the P. putida genes required for this conversion, however, is quite different from that of the bet operon contained in the S. meliloti genome. In P. putida, the gene encoding COS sulfatase (BetC) is not co-transcribed with the choline degradation genes (betBA), but rather with a gene encoding a substrate binding protein (BetD) and a sulfate transporter (BetE) (16).

The role of COS in cell physiology has been examined in several soil-dwelling bacterial species, while the role of COS in plant-microbe interactions has been addressed only in S. meliloti. During nodulation of Medicago sativa, the betICBA operon was highly induced in bacteroids (28). In culture, the betICBA operon was induced by COS and even more strongly by choline, but was not induced by hyperosmotic conditions. This finding suggests activation of the betICBA operon in bacteroids for nutrition rather than osmoprotection; however, it is not clear whether the inducing compound was choline, COS, or both (28). This ambiguity leaves the question of whether COS has a role in this plant- microbe interaction unanswered. In our study, we evaluate whether COS has a role in a second plant-microbe interaction, namely that of the pathogen Pseudomonas syringae during colonization of the host Phaseolus vulgaris.

P. syringae imports QACs through a set of three transporters BetT, OpuC, and

CbcWV and uses them as osmoprotectants and nutrients (10, 11, 13). BetT is a BCCT type transporter that has a high capacity and specificity for choline (11). The ABC-type transporters OpuC and CbcWV have a broader range of substrates, including choline, glycine betaine, carnitine and COS (10, 13). BetT and OpuC activity is increased by hyperosmolarity, and the two transporters are responsible for importing QACs under hyperosmotic stress conditions for osmoprotection (10, 11). In contrast, the CbcWV ABC-

63 type transporter imports QACs under low osmotic conditions, suggesting a role for this transporter in nutrient acquisition (13). The P. syringae substrate binding protein (SBP)

CosX specifically binds COS and interacts with the CbcWV transporter for COS import (12).

Given the role of CbcWV primarily in nutrient acquisition, the existence of this COS-specific

SBP suggests that P. syringae uses COS as a nutrient source. The gene encoding the COS- specific SBP, cosX (Psyr_0028) is located near orthologs of the P. putida betCDE loci, which includes genes encoding a COS-specific sulfatase, BetC, and a sulfate transporter, BetE (12), hereafter referred to as the cos locus, CosC and CosE respectively in P. syringae. The betC gene in P. putida showed decreased expression under hyperosmolarity, suggesting that expression of a COS sulfatase from this genetic arrangement is more likely associated with

COS catabolism than osmoprotection (15, 16).

There has been limited work characterizing sulfur acquisition by P. syringae or any plant-associated bacterial species during colonization of a host plant. Andersen et al. (2) established that P. syringae requires methionine during adaptation to drying stress while colonizing the leaf surface, potentially due to decreased intracellular pools of methionine required for protein production rather than to a limitation in sulfur acquisition. While investigating P. syringae genes up-regulated during leaf colonization, Marco et al. (30) found that the gene ssuE was induced ~300-fold, although growth of an ssuE deletion mutant was not impaired in epiphytic fitness compared to the wild type, implying that sulfonates are not limiting to pathogen growth. The gene ssuE is involved in sulfonate metabolism and is regulated by a sulfate starvation regulon (30), implying a dearth of sulfate on the leaf. Based on these observations (30), P. syringae must obtain the requisite sulfur from other sources, possibly including COS. Plants and seedlings in the Plumbaginaceae family produce COS as

64 a compatible solute in response to osmotic shock and also to prevent sulfate toxicity while storing sulfur (19, 20), but the extent to which plants produce COS in the absence of water limitation and sulfate abundance has not been characterized.

In this study we evaluated the biological roles of COS in P. syringae. We determined how COS influences P. syringae in culture and during colonization of leaves and seeds of the host P. vulgaris by exploiting mutants deficient in COS transport and catabolism. We also investigated the sulfur metabolic networks that are active during leaf colonization. Our studies demonstrated that, like the COS sulfatase CosC, the putative sulfate transporter CosE was critical for COS catabolism and osmoprotection, with the growth dynamics of mutants suggesting that CosE functions in sulfate export to attenuate the toxicity of accumulated sulfate. Lastly, we generated a model predicting how P. syringae obtains sulfur during growth in planta.

Materials and Methods

Bacterial strains, media, and growth conditions.

The bacterial strains used in this study include B728a, the wild type strain, ΔcosC, a

B728a deletion mutant lacking cosC (Psyr_0029) encoding a sulfatse, and ΔcosE, a B728a deletion mutant lacking cosE (Psyr_0026) encoding a putative sulfate transporter. The deletion mutants were constructed as described in Chapter II and primers used for mutant contruction are listed in Table 1 in Appendix B. P. syringae strains were routinely grown on solid King’s B medium (26) at 28°C preceding inoculation into MinAS (27). Cells in liquid media were grown at 25°C with shaking. Antibiotics were added to the growth media as needed at the following concentrations (µg ml-1): kanamycin (Km), 50; rifampin (Rif), 100; tetracycline (Tet), 20; spectinomycin (Spc), 60; and cyclohexamide (Cyclo), 100. Cell

65 growth was monitored either in test tubes based on the optical density (OD) at 600 nm, or in

96-well microtiter plates, with shaking for aeration. Cell growth in microtiter plates was measured based on the OD600nm or OD405 nm/OD630 nm ratio to compensate for optical interference due to water condensation on the lid of the microtiter plate.

Synthesis of choline-O-sulfate.

Choline-O-sulfate was synthesized through a method adapted from Nau-Wagner et al

1999 (11). Choline was mixed with concentrated sulfuric acid in a 1:3 molar ratio. The mixture was heated to 110°C for 8 h. After heating, the reaction was quenched in 8 volumes of ice cold 95% ethanol on ice. Precipitate from quenching was collected via filtration and dissolved in hot 60% ethanol to create a supersaturated solution. The solution was allowed to slowly cool to room temperature to allow the COS to recrystallize (5). The recrystallization process was repeated at least two times to ensure all residual choline was removed. The final precipitate was collected via filtration, dried, and stored at -20°C.

Nutrient utilization assay.

Overnight cultures grown in MinAS medium were adjusted to an OD600nm of 0.1.

Cells in this suspension were washed three times with nanopure water and resuspended in an equal volume of nanopure water. In 96-well plates, 20 µl of inoculum was added to 180 µl of

MinAS or the modified media: (i) MinA lacking carbon source, MinA-C, (ii) MinA without nitrogen source, MinA-N, and (iii) MinA without a sulfur source, MinA-S, each with and without 20 mM choline-O-sulfate (final conc). The plates were incubated at 25°C in a microtiter plate shaker at 300 rpm and growth was monitored over time as OD600 nm or

OD405nm/630nm.

Osmoprotection Assay.

Inoculum was prepared as in the nutrient utilization assay, and 20 µl of inoculum was added to 180 µl of MinAS with either 0 M NaCl, 0.3 M NaCl, or 0.6 M NaCl, with and without 1 mM COS (final conc.) in 96-well plates. The plates were incubated and monitored as described in the nutrient utilization assay.

Plant colonization Assays.

Plants were inoculated and bacterial populations were monitored for growth on leaves

(epiphytic), in the leaf (apoplastic), and on the seed as described in Chapter II.

Statistical Analysis

All growth and osmoprotection assays in culture, epiphytic, apoplastic, and spermospheric fitness assays were performed in three replicate experiments. Differences observed and reported in this study were confirmed using a Student’s T-test with p ≤ 0.05.

Results

P. syringae B728a can utilize choline-O-sulfate as a sole carbon, nitrogen, and sulfur source.

The P. putida locus dedicated to COS utilization, betCDE (16), shares synteny with

B728a genes Psyr_0029-0026, here named the cos locus, except that the B728a locus has an additional gene (Figure 1A). Specifically, the putative sulfatase CosC (Psyr_0029, 76.9% identity), the COS specific SBP CosX (Psyr_0028, 70.5% identity) and a putative sulfate transporter, CosE (Psyr_0026, 76.7% identity) have orthologs in P. putida, while the hypothetical protein encoded in Psyr_0027, CosY, does not (Figure 1). We analyzed the

A cosC cosX cosY cosE P. syringae B728a 0029 0028 0027 0026

P. syringae DC3000 0165 0166 0167 0168

P. syringae 1448a 0030 0029 0028

P. aeruginosa PAO1 0031 0030 0029 betC betD betE P. putida KT2440 0077 0076 0075

B Sinorhizobium meliloti betI betC betB betA

Figure 1. Comparison of loci involved in COS metabolism in various prokaryotes. (A) Conservation of the genes within Pseudomonads except for homologs of Psyr_0027, which is only present in P. syringae B728a and P. syringae DC3000. (B) Sinorhizobium contains an alternative arrangement of COS utilization genes associated with choline degradation genes.

transcriptional structure of the cos locus using RT-PCR of RNA isolated from cells (as described in Chapter II) exposed to 20 mM COS for 4 h, and found that cosC, cosX, and cosY were co-transcribed, while cosE was transcribed independently (Figure 2).

To determine whether B728a can utilize COS as a nutrient source, growth of B728a was monitored when COS was provided as the sole carbon, nitrogen, or sulfur source; in all three cases, B728a growth with COS provided evidence of catabolism (Figure 3).

Comparing the growth of B728a on MinAS, which contains succinate, to growth on MinA-C with 10 mM COS as a sole carbon source demonstrated that the bacteria grew slower with

Lanes: 1 2 3 4 5 6 7 Primer pairs: AB CD EF GH AD CF EH

Primers: A B C D E F G H

cosC cosX cosY cosE

Transcripts:

Figure 2. The cos locus in P. syringae B728a contains two operons. Primer locations within the cos locus are shown. The primer pairs that were used to amplify regions within each gene and across intergenic regions are shown, where the template DNA was either B728a genomic DNA (top) or cDNA created from B728a cells that were grown in the presence of 20 mM L-choline-O-sulfate (bottom).

A B C

0.35 0.35 0.35 Succinate Ammonia Sulfate COS COS COS 0.30 No carbonx" 0.30 No nitrogenx" 0.30 No sulfurx"

0.25 0.25 0.25

0.20 0.20 0.20 405 nm/630 nm nm 405 nm/630 0.15 0.15 0.15 OD 0.10 0.10 0.10

0.05 0.05 0.05

0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 Time (h) MinASB728a MinASB728a MinASB728a Choline-cdhA O-sulfate Choline-cdhA O-sulfate Choline-cdhA O-sulfate NodhcAB carbon NodhcAB nitrogen NodhcAB sulfur

Figure 3. COS can function as a sole carbon, nitrogen, and sulfur source. Growth of B728a in (A) MinAS, MinA-C with COS, and MinA-C. (B) MinAS, MinA-N with COS, and MinA-N. (C) MinAS, MinA-S with COS, and MinA-S. Values are mean ± standard error of the mean (SEM) (n=3). Results are representative of 3 replicated experiments (Appendix B).

COS than with succinate (Figure 3A). When comparing growth in MinAS, which has ammonium as a sole nitrogen source, to MinA-N with 10 mM COS, B728a grew slower with

COS, but to a higher cell density, since COS provides more nitrogen than available in

MinAS, proving that COS can also serve as a nitrogen source in B728a. Comparison of

B728a growth on MinA-S with 10 mM COS to MinAS, which has MgSO4 as a sole sulfur source, demonstrated that COS served as a sole sulfur source (Figure 3C). B728a displayed a

2- diauxic growth curve during growth on COS, which can be attributed to residual SO4 remaining in the synthesized COS preparation (see Materials and Methods). The diauxic growth also indicated that COS is not the preferred sulfur source for B728a, as it is only utilized once the sulfate in the medium was exhausted, which implies COS catabolism is part of the sulfate starvation response.

CosC and CosE are required for B728a growth on COS.

To test whether CosC functions as a choline sulfatase, a deletion mutant, ΔcosC, was first constructed and compared to B728a for growth on COS. ΔcosC did not grow on COS as a sole carbon, nitrogen or sulfur source (Figure 4), indicating that it was no longer able to metabolize COS. When COS was provided as a sulfur source (Figure 4C), ΔcosC displayed similar growth to B728a in the first phase of the diauxic growth curve, but when B728a switched to COS metabolism, ΔcosC no longer grew, again demonstrating that CosC was essential to COS metabolism. The deletion mutant ΔcosE was also tested for its ability to utilize COS as a sole nutrient source. CosE is a putative sulfate transporter and thus does not have an explicit function predicted in COS catabolism, and cosE is transcribed separately from the cosCXY operon (Figure 2). However, we found that ΔcosE, like ΔcosC, exhibited a

A B C

0.35 0.35 0.35

1 mM Carnitine mM 1 + cdhA 0.30 0.30 0.30 !

0.25 0.25 0.25 B728a cdhA !

0.20 0.20 0.20 B728a + 10 mM COS 0.15 0.15 0.15 Carnitine mM 1 + B728a

405 nm/630 nm nm 405 nm/630 ΔcosC 0.10 0.10 0.10 B728a

OD ΔcosC + 10 mM COS 0.05 0.05 0.05 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 Time (h)

Figure 4. Deletion of cosC blocked growth on COS as a sole carbon, nitrogen, and sulfur source. Comparison between B728a and ΔcosC in (A) MinA-C with COS and MinA-C; (B) MinA-N with COS and MinA-N; (C) MinA-S with COS and MinA-S. Values are mean ± SEM (n=3). Results are representative of 3 replicated experiments (Appendix B).

A B C

0.35 0.35 0.35

1 mM Carnitine mM 1 + cdhA 0.30 0.30 0.30 !

B728a 0.25 0.25 0.25 cdhA !

0.20 0.20 0.20 B728a + 10 mM COS B728a + 1 mM Carnitine mM 1 + B728a

405 nm/630 nm nm 405 nm/630 0.15 0.15 0.15 ΔcosE 0.10 0.10 0.10 B728a

OD ΔcosE + 10 mM COS 0.05 0.05 0.05 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 Time (h)

Figure 5. Deletion of cosE blocked growth on COS as a sole carbon, nitrogen, and sulfur source. Comparison between B728a and ΔcosE in (A) MinA-C with COS and MinA-C; (B) MinA-N with COS and MinA-N; (C) MinA-S with COS and MinA-S. Values are mean ± SEM (n=3). Results are representative of 3 replicated experiments (Appendix B).

complete lack of growth on COS when provided as a sole carbon, nitrogen, or sulfur source

(Figure 5A, 5B and 5C), and like ΔcosC, grew during the first phase of diauxic growth when provided COS, likely containing a low level of sulfate, as the sulfur source (Figure 5C).

In P. syringae, COS functions as an osmoprotectant upon catabolism.

To determine whether COS functions as an osmoprotectant in P. syringae, we grew

B728a at low (0 M NaCl), moderate (0.3 M NaCl), and high (0.6 M NaCl) osmotic stress

71 with and without COS. Under low osmotic stress, COS addition did not impact B728 growth, but did contribute to a higher cell density during the stationary phase (Figure 6A), since COS contributed additional carbon available from MinAS. At moderate osmotic stress, adding COS resulted in a slight increase in the growth rate and, again, in a higher final cell density when compared to growth in MinAS without COS (Figure 6B). At high osmotic stress, the impact of COS on growth rate was even more pronounced than at moderate stress

(Figure 6C). These results illustrate that COS functions as an effective osmoprotectant for

B728a.

A B C 1.0 1.0 1.0

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4 405 nm/630 nm nm 405 nm/630

OD 0.2 0.2 0.2

0.0 0.0 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure 6. COS can function as an osmoprotectant in B728a. B728a grown in MinAS (black line) and MinAS with 1 mM COS (red line) added with hyperosmotic stress imposed (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD405 nm/630 nm in microtiter plates. Values are mean ± SE (n=3). Results are representative of 3 replicated experiments (Appendix B).

In order to determine if CosC influences COS-derived osmoprotection in P. syringae, we grew B728a and ΔcosC in MinAS with and without 1 mM COS at low, moderate, and high osmotic stress. When grown under low and moderate osmotic conditions, ΔcosC grew similar to B728a, both with and without COS (Figure 7A and 7B). When subjected to high osmotic stress, ΔcosC had a longer lag phase and did not reach as high a final optical density as B728a with or without COS and ΔcosC without COS added (Figure 7C), indicating that

1.0 A 1.2 B 1.0 C

1.0 0.8 0.8

0.8 0.6 0.6

600 nm 600 nm 0.6 B728aTime vs B728a MinAS 0M 0.4 0.4

OD B728a + 1 mM COS 0.4 Time vs B728a MinAS 0M +COS ΔcosC 0.2 0.2 Time vs betC MinAS 0M 0.2 ΔTimecosC vs betC + MinAS 1 mM 0M +COS COS 0.0 0.0 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60

Time (h)

Figure 7. COS catabolism was required for COS to function in osmoprotection. B728a and ΔcosC were grown in MinAS with and without 1 mM COS and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD600 nm in microtiter plates. Values are mean ± SE (n=3). Results are representative of 3 replicated experiments (Appendix B).

the accumulation of COS in ΔcosC is detrimental to cell growth under high, but not moderate, osmotic conditions.

To test if CosE was required for osmotolerance, the growth of ΔcosE was compared to B728a at increasing hyperosmotic stress with and without 1 mM COS. At low osmotic stress, ΔcosE and B728a displayed similar growth, but only B728a reached a higher final density when COS was present than when it was not (Figure 8A). In both moderate and high osmotic stress conditions, unlike B728a, the ΔcosE does not show enhanced growth with

COS (Figures 8B and 8C), confirming the loss of COS-mediated osmoprotection. Moreover, in both conditions, ΔcosE shows a detrimental impact of COS, with the extent to which COS was associated with a reduced final cell density amplified with increased osmotic stress. The data indicate that the deletion of cosE was detrimental to growth under both moderate and high osmotic stress in the presence of COS. Consolidating the data and compensating for differences in the inoculum (Figure 9) to enable comparisons between the ΔcosC and ΔcosE mutants highlighted the distinct impact of COS on the mutants and provided evidence that the loss of CosE was more toxic than the loss of CosC to cells grown in the presence of COS.

A B C 1.0 1.0 1.0

0.8 0.8 0.8

B728a 0.6 0.6 0.6 Time vs B728a MinAS 0M B728aTime vs B728a + MinAS1 mM 0M +COSCOS 600 nm 600 nm 0.4 0.4 0.4 ΔTimecosE vs betC MinAS 0M OD ΔTimecosE vs betC + MinAS 1 mM 0M +COS COS 0.2 0.2 0.2

0.0 0.0 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure 8. COS catabolism was required for COS to function in osmoprotection. B728a and ΔcosE were grown in MinAS with and without 1 mM COS and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD600 nm in microtiter plates. Values are mean ± SE (n=3). Results are representative of 3 replicated experiments (Appendix B).

A B C 0.35 0.35 0.35 B728a B728a B728a 0.30 !0029 0.30 !0029 0.30 !" " # $ !0026 !0026 !0026 0.25 0.25 0.25 0.20 0.20 0.20 0.15 405nm/630nm 405nm/630nm 0.15 0.15 0.10 OD 0.05 0.10 0.10

0.00 0.05 0.05 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 Time (h) D E F 0.24 0.24 0.24

0.22 0.22 0.22

0.20 0.20 0.20

0.18 0.18 0.18

0.16 0.16 0.16 0.14 0.14 0.14 B728aB728a 0.12 0.12 0.12 405nm/630nm 405nm/630nm 0.10 0.10 0.10 ΔcdhAcosC B728a B728a B728a

OD 0.08 0.08 0.08 !" " # $ !" " # $ !" " # $ ΔdhcABcosE 0.06 !" " # % 0.06 !" " # % 0.06 !" " # % 0.04 0.04 0.04 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 Time (h)

Figure 9. Compiled and correct data for comparison of B728a, ΔcosC, and ΔcosE suggests function of CosE as a sulfate exporter. Growth of strains compared in MinA with 10 mM COS provided as (A) sole carbon source, (B) sole nitrogen source, and (C) sole sulfur source. Growth of strains compared in MinAS with 1 mM COS with (D) 0 M NaCl, (E) 0.3 M NaCl, and (F) 0.6 M NaCl added. Data shown in panels A- C are compiled from experiments shown in Figures 3 – 5. Data shown in panels D-F are compiled from experiments shown in Appendix B Figure B.9 – B.11.

COS does not contribute to P. syringae colonization of bean leaves or seeds.

We evaluated the growth of the mutants in surface (epiphytic) and internal

(apoplastic) sites on leaves and on seeds of the host plant P. vulgaris (Figure 10). B728a and

ΔcosC populations were similar at all time points during growth in epiphytic sites, during both a 24 h period of wetness when osmotic stress should be low and during a subsequent 48 h period of leaf-surface drying (Figure 10A). Similarly, B728a and ΔcosC had comparable populations during 48 h of growth following introduction into leaf apoplastic sites (Figure

10B). Supporting our results, a previous study examining the expression of B728a genes in epiphytic and apoplastic sites of P. vulgaris leaves provided data showing that genes in the cos locus were significantly down-regulated during colonization of both the epiphytic and apoplastic sites (Table 2). B728a and ΔcosC populations were similar at all times after inoculation onto bean seeds in both sterilized potting soil in the laboratory (Figure 10C) and in soil with native microbes, and thus potentially sources of COS in addition to the plant, under field conditions (Figures 10C and 10D). Collectively, these data indicate COS does not influence the interaction between P. syringae and P. vulgaris.

Table 2. Fold changes relative to fluorescent intensities of choline-O-sulfate transport and catabolism genes of P. syringae B728a1 Fluorescence Fold-change2 Intensity3

Basal Epiphytic Apoplastic Psyr # Gene Product Name Medium sites sites

Psyr_0029 cosC sulfatase 1739 -2.5 -2.0 Psyr_0028 cosX BCC ABC transporter, substrate binding protein 1104 -1.8 -1.3 Psyr_0027 cosY hypothetical protein 1403 -1.2 -1.4 Psyr_0026 cosE sulfate transporter/anti-sigma factor antagonist 277 -1.6 -1.6

A B 8.5 10

8.0

7.5

8 7.0

6.5 7 /g leaf tissue) leaf /g /g leaf tissue) leaf /g cfu 6.0 B728a cfu 6 ΔcosC Log ( Log 5.5 ( Log 5 B728a 5.0 wet dry B728a B728a cosC ΔcosCcosC 4.5 4 0 20 40 60 80 0 10 20 30 40 50 60 Time (h) Time (h)

C D 10 7

9 6

8 5 /sample) /sample) cfu cfu 7 4 Log ( Log Log ( Log 6 B728a 3 B728a ΔcosCB728a ΔcosCB728a cosC cosC 5 2 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 2 4 6 8 10 Time (h) Time (h)

Figure 10. COS does not contribute to colonization of Phaseolus vulgaris by B728a. B728a and ΔcosC populations were monitored during (A) epiphytic colonization during 24 h water replete conditions followed by 48 h of drying stress, and (B) apoplastic colonization following vacuum infiltration. Bean seeds were inoculated with B728a and ΔcosC, and planted in (C) sand under laboratory conditions, or (D) in agricultural soil under field conditions. Population sizes were estimated based on viable plate counts. Values are mean ± SEM (n=6). Results are representative of 3 replicated experiments (Appendix B).

Discussion

Our data adds to a growing body of evidence that, among prokaryotes, distinct biological roles for COS are correlated with the genetic structure of their COS catabolic loci.

Moreover, our data supports the unusual phenomenon where the export of a toxic by-product

76 is coupled with catabolism of a nutrient; the coupling is manifest even in the co-localization of the import, catabolism, and toxic by-product export genes. Lastly, although COS may be associated with the growth of the plant-associated bacterial species S. meliloti during nodulation of its host, we did not find evidence for a role of COS in host colonization in the

P. syringae – P. vulgaris pathosystem under the conditions examined.

In this study, we investigated whether COS impacts the ecology of microorganisms and determined that COS catabolism is not essential for P. syringae colonization of P. vulgaris. First, we established the arrangement of the cos locus, then we demonstrated that catabolic genes within the cos locus are required for COS to function as an osmoprotectant and as a nutrient source for P. syringae. Although we did not detect an impact of the COS catabolic genes on epiphytic, apoplastic and spermosphere colonization, this only implies that P. syringae obtains sulfur from a source other than COS.

Bacteria can be grouped into three categories based on the function of COS and the arrangement of the COS utilization genes. Bacteria in the first category uses COS as a compatible solute, as characterized in B. subtilis (32). They contain no genes for COS catabolism, although they encode the OpuC transporter. Bacteria in the second category likely utilize COS as an osmoprotectant. These organisms contain an OpuC ortholog as well as the betCBA operon for COS catabolism to glycine betaine (34). Bacteria from the third category contain the cos locus, which encodes a sulfatase, CosC, a periplasmic binding protein, CosX, which interacts with CbcWV, and a sulfate permease, CosE. This operon is not induced by osmotic stress but rather by the presence of COS in P. putida (16), indicating that COS is primarily used as a nutrient source in organisms with this arrangement of COS related genes. In our study, RT-PCR and bioinformatics analysis revealed that P. syringae

77 has a high synteny to the third category of gene arrangements with the exception that

Psyr_0027, cosY, is encoded between cosX and cosE (Figure 1A). CosY, unique to P. syringae B728a and P. syringae DC3000 (Figure 1B), is annotated only as a hypothetical protein. Deletion of cosY causes a reduction in the efficiency of COS utilization as a nutrient and osmoprotectant (see Appendix B). We were unable to identify any mobile elements surrounding cosY and do not have a hypothesis as to its function. Apart from the presence of cosY, the cos locus of P. syringae B728a functions as demonstrated in P. putida and

Pseudomonas sp. ATCC19151 (16, 24), thereby suggesting that this is a common arrangement for COS utilization genes in the Pseudomonads.

The first gene in the cos locus in B728a, Psyr_0029, encodes CosC, a COS- specific sulfatase. Comparison of the CosC sequence to demonstrated choline sulfatases from P. putida and S. meliloti reveals several conserved motifs specific to choline sulfatases rather than arylsulfatases. In choline sulfatases, the amino acids from 101-106 contain a GKMHFV motif, whereas arylsulfatases contain a conserved GKWHL motif; and between 295 – 301, choline sulfatases contain SDHGDML, while arylsulfatases contain SDHGAWL (15). Upon deletion of cosC, P. syringae was unable to catabolize COS as a carbon, nitrogen, or sulfur source (Figure 4A, 4B, and 4C). However, the deletion strain ΔcosC retained a limited ability to derive osmoprotection from COS under moderate osmotic stress (Figure 7B). At high osmotic stress, ΔcosC was unable to utilize COS as a compatible solute and COS became slightly toxic (Figure 7C), in results similar to those observed in P. putida (16).

Since COS provided osmotic protection at high osmotic stress in B728a, our data establishes that cosC encodes a choline sulfatase critical for utilizing COS for osmoadaption.

Our results showed a surprising requirement for a predicted sulfate transporter, CosE, for growth on COS. Loss of cosE, which is transcribed independently of the cosCXY operon

(Figure 2), resulted in complete loss of growth on COS as a sole carbon, nitrogen or sulfur source, demonstrating that CosE function is as critical as CosC function for COS utilization

(Figure 5A, 5B and 5C). In P. putida, the expression of betC in ΔbetCDE mutant did not restore the ability of the mutant to grow on COS (16), further verifying that BetC alone is not sufficient for growth on COS. CosE is predicted to be a sulfate transporter (29), but ΔcosE was not deficient in sulfate import based on its similarity to B728a in the use of residual sulfate from COS synthesis (Figure 5C). However, CosE may function as a sulfate exporter.

When COS was provided as an osmoprotectant, ΔcosE grew to lower cell densities than

B728a and ΔcosC (Figure 9E and 9F), and this difference was amplified with increasing osmotic stress (Figure 9F). Since the difference in the growth of B728a and ΔcosC at each osmotic stress level reflected physiological changes associated with loss of COS degradation or conversion to glycine betaine, and thus with COS accumulation, the larger growth reduction by ΔcosE than ΔcosC in both growth rate and final density is consistent with the accumulation of a toxic compound other than COS. Cleavage of sulfate esters such as COS release sulfate (21). Under sulfate-limiting conditions, this sulfate may be assimilated, but under non-limiting conditions accumulated sulfate could be toxic. Our results are consistent with CosE functioning as a sulfate exporter that prevents the build-up of sulfate to toxic levels during growth on COS.

An alternative model for CosE function is as a sulfatase maturation enzyme that ensures the proper function of CosC, although this model has less supporting evidence than a role as a sulfate transporter. Sulfatase enzymes such as CosC require post-translational

79 modification of the active site cysteine to formylglycine (38). In P. syringae, the sulfatase maturation enzyme has not yet been identified. Comparison of CosE to a known Cys-type sulfatase maturation enzyme from Escherichia coli, YidF (6, 21), indicated little homology

(< 34% identity with < 7% query coverage). Some sulfatase enzymes show translational modification of an active site serine rather than cysteine. Comparison of CosE to a Ser-type sulfatase maturation enzyme from Klebsiella pneumoniae, AtsB (21, 40), also showed little homology (<30% identity with <7% query coverage). This lack of similarity of CosE to both types of known prokaryotic sulfatase maturation enzymes indicates that this is unlikely to be its function. Furthermore, CosE has little homology to radical SAM superfamily proteins.

However, Psyr_2646 shows homology to a radical SAM superfamily proteins, which have peptide modification activity as well as an anaerobic sulfate maturation enzyme (> 25% indentity and 53% query coverage) and additionally, analysis of protein interactions in P. syringae using the STRING database reveals that Psyr_2646 interacts with CosC (41); hypothetically, Psyr_2646 could be the Cys-type sulfatase maturation enzyme required for proper function of CosC.

CosE has been annotated as a sulfate transporter with an anti-sigma factor antagonist

(STAS) domain (44). The primary conserved domain in CosE is the SUL1 inorganic ion transporter superfamily, COG0659 (29), with an E-value of 2.48E-104 for 96% query coverage (29, 44) (Figure 11). CosE also contains an N-terminal domain (amino acids 10-

92) with a GLY sequence that is strongly conserved among sulfate transporter proteins, and a

C-terminal domain (amino acids 435-510) with homology to a STAS domain (29, 44). In sulfate transporters, the STAS domain has two major functions: facilitation of active transport through cleavage of ATP, and regulation of transport activity through

1 130 261 391 522 Protein sequence CosE (Psyr_0026)

Multi domains Sul1

Sulfate Sulfate transporter/ Domains transporter GLY Sulfate Transporter anti-sigma motif factor antagonist

Figure 11. CosE has homology to sulfate transporter in the Sul1 family. Schematic representation of domains identified in CosE (Psyr_0026) displaying the common domains present among sulfate transporters.

phosphorylation (3, 37, 39). The STAS domain is correlated with NTP binding and is found in the C-terminus of both sulfate transporters and anti-sigma factor antagonist proteins, which relieve transcriptional inhibition by anti-sigma factors. The presence of these domains in CosE suggests a function as a sulfate transporter; the growth patterns observed in Figures

5 and 8 cannot be explained by the inhibition of sulfate import through CosE. If CosE exports sulfate liberated from the hydrolysis of COS instead, ΔcosE grown in the presence of

COS would demonstrate the observed reduced growth due to sulfate accumulation.

Sulfate exporters have not yet been characterized in bacteria, although several studies suggest such export activity, particularly in the context of sulfonate catabolism (14, 43). An export function for a Sul1-type transporter was suggested to help deal with sulfate accumulation in the sulfide-oxidizing species Beggiatoa alba (8). Similarly, a sulfate export function was proposed to minimize sulfate accumulation during the breakdown of the sulfonate cysteate in Paracoccus pantotrophus, as this involves the release of sulfite and its subsequent oxidation to sulfate (35). Two putative sulfate exporters (PSEs) in Variovorax paradoxus, VPARA_05640 and/or VPARA_05760, were hypothesized to export sulfate

81 formed during sulfonate degradation; these were identified based on their localization near the sulfonate catabolism genes (45). Both VPARA_05640 and VPARA_05760 contain the same conserved domain, designated hypothetical conserved domain 698, COG2855 (29); its function is not known. The protein MIM_c23530 in Advenella mimigardefordensis also contains the hypothetical conserved domain 698; moreover, Tn5 inactivation of

MIM_c23530 prevented growth on sulfonate compounds (46, 47), suggesting a critical role to sulfonate utilization and consistent with a requirement for sulfate export in the utilization of a sulfated compound, as we observed. Although CosE and these PSEs do not share significant homology, CosE-mediated sulfate export is consistent with our finding that the cosE deletion is detrimental only when COS catabolism occurs. Our findings are the first to suggest a requirement for sulfate export for bacterial utilization of a sulfate ester, and indicate a need for further work investigating an export function during bacterial catabolism of sulfate esters and sulfonates.

Given our paucity of knowledge on the sources of sulfur supporting the growth of plant-associated bacteria, one objective this study was determining whether COS, a potential plant-derived sulfur source, impacts the ecology of P. syringae during colonization of its host plant, P. vulgaris. We did not detect an impact of the loss of COS catabolism, as caused by the loss of the CosC COS sulfatase, on the colonization of epiphytic or apoplastic sites on leaves or the colonization of seeds (Figure 10). A previous study demonstrated that during epiphytic colonization, cos locus gene expression was down-regulated (48), consistent with our results. In contrast, the expression of a gene involved in sulfonate metabolism, ssuE, was identified as one of only a few genes in B728a that was highly up-regulated on leaves, based on an in vivo expression approach (30, 48). Although ssuE was induced 300-fold on leaves,

exogenous sulfate

CysP 2.59 choline-O- thio- CosC CosX sulfate sulfate sulfate CysT -1.85 -2.5 -1.04

CysD -2.42

APS

CysH -1.90 * TauD aryl- taurine TauA * 1.92 2.58 AtsR AtsA sulfonates sulfite * 1.57 SsuD alkyl- 1.89 (e. g. sulfoquinovose) SsuA -1.83 CysJ 2.13 sulfonates 2.09 (e.g. methane sulfonate) sulfide

CysK -1.66

CysX cysteine Central regulator of sulfur -1.74 metabolism MetC -1.89 * CysB 1.08

homocysteine

MetE 1.12

Methionine MsrB methionine sulfoxide 1.48

Figure 12. Transcriptome data indicates multiple organosulfur sources contribute to epiphytic colonization. Gene expression data adapted from microarray analysis by Yu et al. 2013. Transcript changes are expressed as fold induction based on comparison to expression in basal medium utilizing sulfate as sole sulfur source. Gene Expression Omnibus accession no. GSE42544

an ssuE deletion mutant showed no reduction in epiphytic fitness (30, 48). The combination of these results raises the question of how P. syringae obtains sulfur during colonization. We

83 have presented previous data on the expression of B728a in epiphytic leaf sites (48) as a sulfur acquisition network in Figure 12. The genes tauD, ssuD, and atsA, which are responsible for taurine, alkylsulfonate, and arylsulfonate degradation respectively (4, 7, 23,

25, 42), were up-regulated. On the other hand, sulfate assimilation genes were down- regulated, with cysD, which encodes a protein critical to the first step of the sulfate assimilation pathway (4), showing a 2.4-fold decrease compared to in cells grown in a basal minimal medium (Figure 12). The expression values in this sulfur assimilation network indicate that the P. syringae cells on leaves under the conditions in the previous study, which were the same as those used in the current study, were limited for sulfur. This conclusion is based on three factors: (i) cysD expression was decreased, (ii) cysB, a master regulator for sulfur metabolism, was increased in expression, and (iii) sulfonate-based sulfur acquisition genes were up-regulated (23, 42). This in turn indicates that sulfate was not prevalent during epiphytic colonization under these conditions, and that sulfonates are likely a major source of sulfur. Sulfonates are most likely derived from plant sulfolipids that have leached to the leaf surface (7). This evidence supports a model of sulfur acquisition and incorporation by P. syringae utilizing multiple sources of organo-sulfur during epiphytic colonization, and lays a foundation for future work to understand the nutrient interplay at the plant-microbe interface.

In conclusion, the data presented here demonstrate that P. syringae can catabolize

COS as a carbon, nitrogen, and sulfur source, and during growth in hyperosmotic stress, as an osmoprotectant. These functions require the degradation of COS through the action of the choline sulfatase CosC. The data suggest that the sulfate transporter CosE functions as a sulfate exporter to relieve P. syringae of sulfate toxicity, and that COS does not influence P. syringae colonization of P. vulgaris as either a nutrient source or as an osmoprotectant.

Given the widespread distribution of P. syringae on plants of many species, P. syringae may have evolved a system for COS utilization to support its colonization of plants beyond its host, including ones known to produce COS, such as Limonium sp. (18). Alternatively, P. syringae may encounter COS produced from other organisms on leaves, such as epiphytic fungi (5, 9, 17). Although COS does not serve as a sulfur source to P. syringae during the colonization of P. vulgaris leaves or seeds, the gene expression data shown in Figure 12 strongly indicate that P. syringae cells on leaves acquire sulfur from multiple sulfated organic compounds. Further research into the sources of sulfur during plant-pathogen interactions could provide valuable insights in the epiphytic ecology helping to understand how microbes obtain micronutrients on the leaf surface.

References

1. Alexander M. 1977. Microbial Transformation of Sulfur. Introduction to Soil Microbiology. John Wiley & Sons, New York, NY:350-367.

2. Andersen GL, Beattie GA, Lindow SE. 1998. Molecular characterization and sequence of a methionine biosynthetic locus from Pseudomonas syringae. Journal of Bacteriology 180:4497-4507.

3. Aravind L, Koonin EV. 2000. The STAS domain—a link between anion transporters and antisigma-factor antagonists. Current Biology 10:R53-R55.

4. Beil S, Kertesz MA, Leisinger T, Cook AM. 1996. The assimilation of sulfur from multiple sources and its correlation with expression of the sulfate-starvation-induced stimulon in Pseudomonas putida S-313. Microbiology 142:1989-1995.

5. Bellenger N, Nissen P, Wood TC, Segel IH. 1968. Specificity and control of choline–O-sulfate transport in filamentous fungi. Journal of Bacteriology 96:1574- 1585.

6. Benjdia A, Dehò G, Rabot S, Berteau O. 2007. First evidences for a third sulfatase maturation system in prokaryotes from E. coli aslB and ydeM deletion mutants. FEBS Letters 581:1009-1014.

7. Benning C. 2007. Questions remaining in sulfolipid biosynthesis: a historical perspective. Photosynthesis Research 92:199-203.

8. Berg JS, Schwedt A, Kreutzmann A-C, Kuypers MM, Milucka J. 2014. Polysulfides as intermediates in the oxidation of sulfide to sulfate by Beggiatoa sp. Applied and Environmental Microbiology 80:629-636.

9. Catalfomo P, Block J, Constantine G, Kirk P. 1973. Choline sulfate (ester) in marine higher fungi. Marine Chemistry 1:157-162.

10. Chen C, Beattie GA. 2007. Characterization of the osmoprotectant transporter OpuC from Pseudomonas syringae and demonstration that cystathionine-β-synthase domains are required for its osmoregulatory function. Journal of Bacteriology 189:6901-6912.

11. Chen C, Beattie GA. 2008. Pseudomonas syringae BetT is a low-affinity choline transporter that is responsible for superior osmoprotection by choline over glycine betaine. Journal of Bacteriology 190:2717-2725.

12. Chen C, Li S, McKeever DR, Beattie GA. 2013. The widespread plant‐colonizing bacterial species Pseudomonas syringae detects and exploits an extracellular pool of choline in hosts. The Plant Journal 75:891-902.

13. Chen C, Malek AA, Wargo MJ, Hogan DA, Beattie GA. 2010. The ATP‐binding cassette transporter Cbc (choline/betaine/carnitine) recruits multiple substrate‐binding proteins with strong specificity for distinct quaternary ammonium compounds. Molecular Microbiology 75:29-45.

14. Cook AM, Denger K, Smits TH. 2006. Dissimilation of C3-sulfonates. Archives of Microbiology 185:83-90.

15. Cregut M, Durand M-J, Thouand G. 2014. The Diversity and Functions of Choline Sulphatases in Microorganisms. Microbial Ecology 67:350-357.

16. Galvão TC, De Lorenzo V, Cánovas D. 2006. Uncoupling of choline‐O‐sulphate utilization from osmoprotection in Pseudomonas putida. Molecular Microbiology 62:1643-1654.

17. Gravel RA. 1976. Choline-O-sulphate utilization in Aspergillus nidulans. Genetical Research 28:261-276.

18. Hanson A, Gage D. 1991. Identification and determination by fast atom bombardment mass spectrometry of the compatible solute choline-O-sulfate in Limonium species and other halophytes. Functional Plant Biology 18:317-327.

19. Hanson AD, Rathinasabapathi B, Chamberlin B, Gage DA. 1991. Comparative physiological evidence that β-alanine betaine and choline-O-sulfate act as compatible osmolytes in halophytic Limonium species. Plant Physiology 97:1199-1205.

20. Hanson AD, Rathinasabapathi B, Rivoal J, Burnet M, Dillon MO, Gage DA. 1994. Osmoprotective compounds in the Plumbaginaceae: a natural experiment in metabolic engineering of stress tolerance. Proceedings of the National Academy of Sciences 91:306-310.

21. Hanson SR, Best MD, Wong CH. 2004. Sulfatases: structure, mechanism, biological activity, inhibition, and synthetic utility. Angewandte Chemie International Edition 43:5736-5763.

22. Harada T, Spencer B. 1960. Choline sulphate in fungi. Journal of General Microbiology 22:520-527.

23. Imperi F, Tiburzi F, Fimia GM, Visca P. 2010. Transcriptional control of the pvdS iron starvation sigma factor gene by the master regulator of sulfur metabolism CysB in Pseudomonas aeruginosa. Environmental Microbiology 12:1630-1642.

24. Jovcic B, Venturi V, Topisirovic L, Kojic M. 2011. Inducible expression of choline sulfatase and its regulator BetR in Pseudomonas sp. ATCC19151. Archives of Microbiology 193:399-405.

25. Kertesz MA. 2000. Riding the sulfur cycle–metabolism of sulfonates and sulfate esters in Gram-negative bacteria. FEMS Microbiology Reviews 24:135-175.

26. King EO, Ward MK, Raney DE. 1954. Two simple media for the demonstration of pyocyanin and fluorescin. The Journal of Laboratory and Clinical Medicine 44:301- 307.

27. Li S, Yu X, Beattie GA. 2013. Glycine betaine catabolism contributes to Pseudomonas syringae tolerance to hyperosmotic stress by relieving betaine- mediated suppression of compatible solute synthesis. Journal of Bacteriology 195:2415-2423.

28. Mandon K, Østerås M, Boncompagni E, Trinchant JC, Spennato G, Poggi MC, Le Rudulier D. 2003. The Sinorhizobium meliloti glycine betaine biosynthetic genes (betICBA) are induced by choline and highly expressed in bacteroids. Molecular Plant-Microbe Interactions 16:709-719.

29. Marchler-Bauer A, Derbyshire MK, Gonzales NR, Lu S, Chitsaz F, Geer LY, Geer RC, He J, Gwadz M, Hurwitz DI, Lanczycki CJ, Lu F, Marchler GH, Song JS, Thanki N, Wang Z, Yamashita RA, Zhang D, Zheng C, Bryant SH. 2015. CDD: NCBI's conserved domain database. Nucleic Acids Research 43:D222-226.

30. Marco ML, Legac J, Lindow SE. 2005. Pseudomonas syringae genes induced during colonization of leaf surfaces. Environmental Microbiology 7:1379-1391.

31. Maynard D, Stewart J, Bettany J. 1984. Sulfur cycling in grassland and parkland soils. Biogeochemistry 1:97-111.

32. Nau-Wagner G, Boch J, Le Good JA, Bremer E. 1999. High-affinity transport of choline-O-sulfate and its use as a compatible solute in Bacillus subtilis. Applied and Environmental Microbiology 65:560-568.

33. ORSI BA, SPENCER B. 1964. Choline sulphokinase (sulphotransferase). The Journal of Biochemistry 56:81-91.

34. Østerås M, Boncompagni E, Vincent N, Poggi M-C, Le Rudulier D. 1998. Presence of a gene encoding choline sulfatase in Sinorhizobium meliloti bet operon: choline-O-sulfate is metabolized into glycine betaine. Proceedings of the National Academy of Sciences 95:11394-11399.

35. Rein U, Gueta R, Denger K, Ruff J, Hollemeyer K, Cook AM. 2005. Dissimilation of cysteate via 3-sulfolactate sulfo-lyase and a sulfate exporter in Paracoccus pantotrophus NKNCYSA. Microbiology 151:737-747.

36. Rivoal J, Hanson AD. 1994. Choline-O-sulfate biosynthesis in plants (identification and partial characterization of a salinity-inducible choline sulfotransferase from species of Limonium (Plumbaginaceae). Plant Physiology 106:1187-1193.

37. Rouached H, Berthomieu P, El Kassis E, Cathala N, Catherinot V, Labesse G, Davidian J-C, Fourcroy P. 2005. Structural and functional analysis of the C- terminal STAS (sulfate transporter and anti-sigma antagonist) domain of the Arabidopsis thaliana sulfate transporter SULTR1. 2. Journal of Biological Chemistry 280:15976-15983.

38. Sardiello M, Annunziata I, Roma G, Ballabio A. 2005. Sulfatases and sulfatase modifying factors: an exclusive and promiscuous relationship. Human Molecular Genetics 14:3203-3217.

39. Shibagaki N, Grossman AR. 2006. The role of the STAS domain in the function and biogenesis of a sulfate transporter as probed by random mutagenesis. Journal of Biological Chemistry 281:22964-22973.

40. Szameit C, Miech C, Balleininger M, Schmidt B, von Figura K, Dierks T. 1999. The iron sulfur protein AtsB is required for posttranslational formation of formylglycine in the Klebsiella sulfatase. Journal of Biological Chemistry 274:15375- 15381.

41. Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J, Simonovic M, Roth A, Santos A, Tsafou KP, Kuhn M, Bork P, Jensen LJ, von Mering C. 2015. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Research 43:D447-452.

42. Tralau T, Vuilleumier S, Thibault C, Campbell BJ, Hart CA, Kertesz MA. 2007. Transcriptomic analysis of the sulfate starvation response of Pseudomonas aeruginosa. Journal of Bacteriology 189:6743-6750.

43. Weinitschke S, Denger K, Cook AM, Smits TH. 2007. The DUF81 protein TauE in Cupriavidus necator H16, a sulfite exporter in the metabolism of C2 sulfonates. Microbiology 153:3055-3060.

44. Winsor G, Lam D, Fleming L, Lo R, Whiteside M, Yu N, Hancock R. 524 Brinkman FS. 2011. Pseudomonas Genome Database: improved comparative analysis 525 and population genomics capability for Pseudomonas genomes. Nucleic Acids Research 526:D596-600.

45. Wübbeler JH, Hiessl S, Meinert C, Poehlein A, Schuldes J, Daniel R, Steinbüchel A. 2015. The genome of Variovorax paradoxus strain TBEA6 provides new understandings for the catabolism of 3, 3′-thiodipropionic acid and hence the production of polythioesters. Journal of Biotechnology 209:85-95.

46. Wübbeler JH, Hiessl S, Schuldes J, Thürmer A, Daniel R, Steinbüchel A. 2014. Unravelling the complete genome sequence of Advenella mimigardefordensis strain DPN7T and novel insights in the catabolism of the xenobiotic polythioester precursor 3, 3′-dithiodipropionate. Microbiology 160:1401-1416.

47. Wübbeler JH, Steinbüchel A. 2014. New pathways for bacterial polythioesters. Current Opinion in Biotechnology 29:85-92.

48. Yu X, Lund SP, Scott RA, Greenwald JW, Records AH, Nettleton D, Lindow SE, Gross DC, Beattie GA. 2013. Transcriptional responses of Pseudomonas syringae to growth in epiphytic versus apoplastic leaf sites. Proceedings of the National Academy of Sciences 110:E425-E434.

CHAPTER 4. PLANT TISSUE-SPECIFIC BENEFITS OF QUATERNARY

AMMONIUM COMPOUNDS TO PSEUDOMONAS SYRINGAE DURING LEAF

AND SEED COLONIZATION.

Michael D. Millican and Gwyn A. Beattie

Abstract

Pseudomonas syringae must adapt to several abiotic stress factors while colonizing Phaseolus vulgaris, the most detrimental stresses being limitations in nutrient and water availability. Through a comparison between P. syringae B728a and mutants deficient in catabolism of choline, carnitine, choline-O-sulfate, glycine betaine, and phosphorylcholine, we evaluated the impact of each QAC on epiphytic and spermospheric fitness. While colonizing the leaf surface, P. syringae derives osmotolerance from host-derived choline to sustain populations during drying events. P. syringae will accumulate glycine betaine derived from choline and as water limitation increases will preferentially degrade glycine betaine and produce trehalose and NAGGN as compatible solutes. Choline derivatives, such as carnitine, choline-O-sulfate, and phosphorylcholine do not detectably impact epiphytic population of P. syringae.

Although water limitation was not a debilitating stress factor for P. syringae while colonizing the germinating P. vulgaris seeds, choline was degraded as a nutrient source throughout the germination process. Interestingly, P. syringae catabolized carnitine that was transiently released during radicle emergence and elongation to support population growth. Collectively, we demonstrated that plant-derived QACs provide osmotolerance

90 under water limiting conditions, and additionally functioned as nutrient sources supporting P. syringae B728a populations while colonizing distinct developmental stages of P. vulgaris.

Introduction

Pseudomonas syringae, a widespread and common plant pathogen, is a model organism for plant-bacterial interactions that has been extensively studied for its pathogenicity and ecology (18, 21, 22, 38). Among P. syringae strains that are foliar pathogens, P. syringae pv. syringae strain B728a is particularly adept at colonizing leaf surfaces, where it is subjected to frequent fluctuations in UV radiation, temperature, and water availability. Of the abiotic stress factors impacting epiphytic colonization, water availability is the most critical (1, 3, 16, 29). P. syringae tolerates increasing osmotic stress by accumulating compatible solutes either through de novo synthesis or through the import of osmoprotectants (17, 49). Compatible solutes are chemicals that accumulate to high concentration in the cytosol without interfering with cellular function, whereas osmoprotectants are chemicals that, when taken up from the environment, provide tolerance to hyperosmotic conditions by serving as, or being converted to, compatible solutes (11, 48, 49).

Osmoprotectants, such as glycine betaine and related quaternary ammonium compounds (QACs), have been described in all domains of life (20, 32, 41, 46, 50).

Here, we focus on QACs characterized as osmoprotectants, compatible solutes and nutrient sources in P. syringae (28), including choline, phosphorylcholine (PCH), choline-O-sulfate (COS), L-carnitine, and glycine betaine (GB) (6, 8). Three transporters

91 in P. syringae control QAC import for both osmoprotection and nutrient acquisition. The first QAC transporter is OpuC, an ABC-type transporter with broad substrate specificity.

OpuC is osmotically regulated and functions predominantly under hyperosmotic conditions to transport choline, PCH, COS, carnitine, and GB for osmoprotection (6).

BetT, a BCCT-type secondary transporter, also contributes to osmoprotection by transporting choline and acetylcholine (7). Unlike OpuC, BetT transports choline at low as well as high osmolarity, indicating that BetT can additionally function in nutrient acquisition. BetT has an unusually high capacity for choline transport, which results in better osmoprotection from choline than GB; this high capacity choline transport suggests that P. syringae is well suited for environments rich in choline (7).

Following uptake into P. syringae, QACs are converted to GB, and GB accumulates as the primary compatible solute (28). The conversion of QACs to GB is not universal in bacteria, as Bacillus subtilis and Listeria monocytogenes import and accumulate L-carnitine and COS as compatible solutes (24, 39, 45). Although GB accumulation is a critical step for effective osmoadaptation of P. syringae, high concentrations of cytoplasmic GB inhibit the de novo synthesis of the compatible solutes trehalose, a disaccharide, and N-acetylglutaminyl-glutamine amide (NAGGN), a dipeptide, in P. syringae strain B728a (28). Under moderate osmotic stress, the GB catabolism genes gbcAB are not expressed (28). Following an upshift to high osmotic stress, however, gbcAB expression increases, GB is catabolized, the GB-repressed genes treX and ggnA are de-repressed, and trehalose and NAGGN are synthesized and accumulated; the replacement of GB with compatible solutes that are synthesized de novo

92 suggests that high concentrations of GB may be toxic to the cell, consistent with findings in other bacterial species (23).

Some bacterial species can only use QACs for osmoprotection (24, 26, 33, 39), whereas others, like P. syringae, can use them for both osmoprotection and nutrition (28,

46, 47). P. syringae imports QACs for catabolism through the ABC-type transporter

CbcWV; cbcXWV expression and functional activity of the transporter is greatest at low osmolarity (9). CbcWV interacts with multiple substrate binding proteins (SBPs), each specific to one or more distinct QACs. The SBP encoded within the cbcXWV operon,

CbcX, binds both choline and GB. The other SBPs are encoded by genes in distinct locations throughout the genome: the SBP that binds GB, BetX, is encoded by a gene that is not in an operon, whereas the SBP that binds L-carnitine, CdhX, is encoded in the carnitine degradation operon ((9), chapter II) and the SBP for COS, CosX, is encoded in the COS degradation operon ((9), chapter III). The association of some SBP genes with catabolic operons, coupled with the negative impact of osmotic stress on CbcWV activity, is consistent with a role for the CbcWV transporter in nutrient acquisition rather than osmoadaptation (13, 19, 34, 35).

The metabolic network for the utilization of QACs as nutrients in P. syringae has two central intermediates, one at choline and one further downstream at GB (Figure 1).

PlcA, which is an ortholog of phospholipase C in P. aeruginosa (15), is hypothesized to cleave phosphatidylcholine to PCH. PCH can be cleaved by PchP, a phosphatase in the periplasm, producing choline and phosphate (Li and Beattie, unpublished data). Further catabolism in this metabolic network occurs in the cytoplasm. Following import, COS can be cleaved into choline and sulfate by the COS-specific sulfatase CosC (see Chapter

III). Choline, derived from COS, PCH, or other sources, can be converted to GB by

BetA, a choline dehydrogenase, and BetB, a betaine aldehyde dehydrogenase. Unlike other QACs, L-carnitine is converted directly to GB by the sequential actions of CdhA, a carnitine dehydrogenase, CdhC, a β-ketoacid cleavage enzyme, and CdhB, a thioesterase

(see Chapter II). Finally, GB can be catabolized to glycine through the betaine degradation pathway: GbcAB converts GB to dimethyglycine, which is converted to sarcosine by DgcAB, followed by conversion to glycine by SoxBDAG (28).

PC PlcA PCH PCH CHO CAR GB COS

PCH CHO PchP

Periplasm Transport Cytoplasm

BetC CHO COS BetBA CdhACB CAR GB GbcAB

GLY SAR DMG SoxBDAG DgcAB

Figure 1. Schematic diagram of QAC metabolic network. CHO – choline, PC – phosphatidylcholine, PCH – phosphorylcholine, COS – choline-O-sulfate, CAR – carnitine, GB – glycine betaine, DMG – dimethylglycine, SAR – sarcosine, GLY – glycine.

Plants produce many of the QACs that are potentially beneficial to P. syringae, including choline and phosphorylcholine, which can be derived from the major plant membrane lipid phosphatidylcholine (14). Through the use of a QAC-specific bioreporter system, choline was demonstrated to be highly abundant and biologically available to P. syringae cells in bean (Phaseolus vulgaris) leaves (8). A P. syringae deletion mutant lacking gbcAB demonstrated reduced fitness compared to the wild type during epiphytic colonization, indicating that GB catabolism can contribute to fitness

(28). The presence of bioavailable COS in plants has not yet been demonstrated, although our recent work found that COS is not present at detectable levels in P. vulgaris

(Chapter III). COS can accumulate as a sulfate storage molecule and compatible solute in several halophytic plants, such as Limonium species (35), but its presence in non- halophytic plants has not been well characterized. L-carnitine, which is a cellular catalyst for fatty acid transport into the mitochondria for β-oxidation, is present and biologically available to P. syringae in the spermosphere (5, 27, 34) (Chapter II). Thus, P. syringae has evolved to exploit a specific set of plant-derived QACs, but the interplay between, and relative influence of, these QACs on the dynamics of plant colonization is unclear.

In this study, we employed a collection of P. syringae mutants deficient in QAC transport and metabolism to evaluate the relative roles of distinct QACs in P. syringae

B728a interactions with P. vulgaris. Specifically, our goals were to (i) identify which

QACs were available to P. syringae during growth on leaves and seeds, (ii) characterize the dynamics of QAC availability following immigration onto leaves or seeds, and (iii) understand the role of these QACs as osmoprotectants versus nutritional sources for P. syringae during plant colonization. Collectively, these goals will allow us to elucidate a

95 model of how QACs influence the colonization of plants by a foliar plant pathogen at various stages of the plant life cycle and under distinct environmental conditions.

Materials and Methods

Bacterial strains, media, and growth conditions.

The bacterial strains used in this study are listed in Table 1. P. syringae strains were routinely grown on solid King’s B medium (25) at 28°C preceding inoculation into

MinAS (28) for growth at 25°C with shaking. Antibiotics were added as needed at the following concentrations (µg ml-1): rifampin (Rif), 100 and cyclohexamide (Cyclo), 100.

Cell growth was monitored in test tubes based on the optical density (OD) at 600 nm with shaking for aeration.

Epiphytic fitness assay.

Bean seeds (P. vulgaris cultivar Bush Blue Lake 274; Park Seeds, Hodges, South

Carolina, USA) were surface sterilized as follows: seeds were submerged in 70% ethanol for 1 min and 10% bleach for 30 s, and rinsed five times with sterile nanopure H2O.

After rinsing, seeds placed in a sterile petri dish were sorted based on the appearance of the seed coat. All seeds with cracked, wrinkled, or discolored seed coats were discarded.

Bean seeds were planted in 5-inch pots filled with pasteurized peat moss (Sunshine mix

SL-1; Sungro, Agawam, Massachusetts, USA) at a density of 10 seeds/pot. Plants were grown until the primary leaves were fully expanded under the following conditions:

25°C, 12 h/12h light/dark cycle, and 75% relative humidity. Bacteria grown on solid

King’s B medium for 48 h were suspended in water to a density of 2.5x105 cells ml-1.

Table 1. Bacterial strains

P. syrignae strains

B728a Wild Type; RifR Loper et al. (1987)

B728a deletion mutant lacking betA (Psyr_4732) ΔbetA Chen et al. (2013) encoding a choline dehydrogenase; RifR

B728a deletion mutant lacking, Psyr_2486-2491, Δngg Δtre Li et al. (2013) and Psyr_2992-3001; RifR

B728a deletion mutant lacking plcA (Psyr_1823) S. Li and G. Beattie, ΔplcA phospholipase C; RifR unpublished data

B728a deletion mutant lacking pchP (Psyr_4738) S. Li and G. Beattie, ΔpchP phosphorylcholine phosphatase; RifR unpublished data

B728a deletion mutant lacking gbcAB ΔgbcAB (Psyr_4776-4775) encoding glycine betaine Li et al. (2013) catabolism enzyme subunit A and B; RifR B728a deletion mutant lacking cosC (Psyr_0029) ΔcosC This work encoding sulfatase; RifR

B728a deletion mutant lacking cdhA (Psyr_2918) ΔcdhA This work encoding a carnitine dehydrogenase; RifR

B728a triple deletion mutant lacking opuCA BT (Psyr_4248-4252), betT (Psyr_4827), and cbcXWV Chen et al. (2010) (Psyr_4709-4711) QAC transporters; RifR

The plants were dip inoculated by immersion in the bacterial suspension for 1 min. After inoculation, the plants were immediately enclosed in a plastic bag such that the leaves were not touching the plastic, and then left on the bench at 25°C for 24 h.

Plants were de-bagged and moved to an environmental chamber (Conviron; Winnipeg,

Canada) under the following conditions: 12 h/12 h light/dark cycle, 25°C, and 45%

97 relative humidity. The bacterial populations were enumerated as described in Chapter II.

Three replicate experiments were performed and treated as blocks for a repeated measures analysis of the log-normal values using a split-plot design, where the strain was the whole plot factor and time was the split factor, with subsampling within the split-plot using Proc Glimmix in SAS. Individual analyses of each replicate experiment are shown in Appendix A.

Spermosphere fitness assay.

Sand was prepared and the field capacity was determined as described in Chapter

II. Bean seeds were surface sterilized as described above, and submerged for 1 min in bacterial suspension. After 1 min, the inoculum was decanted. Seeds were planted in washed sterile sand at 60% field capacity and incubated at 25°C under constant light

(EnviroGro T5 High output fluorescent lighting system, Hydrofarm Inc.; Petaluma,

California, USA). Bacterial populations on the seeds were enumerated as described in

Chapter II. Repeated measures analysis was carried out as described above.

Chemical analysis of the seed exudates

Bean seeds were germinated in sterile nanopure H2O for up to 3 days, and the exudates released during water imbibition, radicle emergence, and seedling establishment were recovered and analyzed via HPLC-MS for their QAC composition as described in

Chapter II.

Results

Many QAC transport and metabolism genes are upregulated during epiphytic colonization.

A previous study that evaluated the expression of B728a genes found that the majority of genes related to QAC utilization in P. syringae B728a are upregulated during epiphytic colonization compared to growth in a basal medium (51). Table 2 shows raw florescence intensity representing transcript abundance in cells grown in a liquid minimal medium and the fold-induction of these genes in cells recovered from leaf surfaces of plants that had been incubated for 24 h under high humidity conditions followed by 48 h of dry conditions. The genes related to COS and PCH catabolism were down-regulated approximately 2-fold in P. syringae cells from epiphytic sites. Within the carnitine locus, cdhX and cdhC, the first gene in the catabolic operon, were significantly upregulated.

Interestingly, the genes for choline and GB catabolism, along with the genes for the transporter CbcWV and the first genes in the OpuC operon, were upregulated, whereas that for the transporter BetT was not, consistent with its regulation primarily at the transcriptional level (7).

QACs influence colonization of the leaf surface.

We investigated the relative impact of plant-derived QACs on epiphytic colonization over time (Figure 2). In the first 24 h after inoculation, the leaf surfaces remained moist and the strains all exhibited similar growth (Figure 2A). At 48 hours post-inoculation (hpi), after 24 h of drying stress, the ΔbetA and BT mutants had smaller populations than the wild type (Figures 2B and 3A), whereas the populations of B728a,

Table 2. Fold changes relative to fluorescent intensities of genes related to the transport and metabolism of QACs in P. syringae B728a1

Fluorescence 3 Intensity Basal Fold- Psyr # Gene Function Medium change2 Quaternary Ammonium Compound (QAC) Transport Psyr_4249 opuCA Choline/carnitine/GB/COS transport 177 1.7 Psyr_4250 opuCB Choline/carnitine/GB/COS transport 488 1.7 Psyr_4251 opuCC Choline/carnitine/GB/COS transport 363 1.5 Psyr_4252 opuCD Choline/carnitine/GB/COS transport 851 1.4 Psyr_4709 cbcX Choline transport 629 3.2 Psyr_4710 cbcW Choline/carnitine/GB/COS transport 461 2.7 Psyr_4711 cbcV Choline/carnitine/GB/COS transport 188 3.1 Psyr_4827 betT Choline transport 22 1.5 Psyr_2916 cdhX Carnitine transport 75 1.9 Psyr_3758 betX glycine betaine transport 60 7.5 Psyr_0028 cosX Choline-O-sulfate transport 870 -1.8 Quaternary Ammonium Compound (QAC) Metabolism Psyr_4734 betI Choline metabolism 640 4.0 Psyr_4733 betB Choline metabolism 498 3.2 Psyr_4732 betA Choline metabolism 204 2.8 Psyr_4708 gbdR Glycine betaine metabolism 277 11.1 Psyr_4776 gbcA Glycine betaine metabolism 234 9.6 Psyr_4775 gbcB Glycine betaine metabolism 187 2.1 Psyr_2915 cdhR Regulator carnitine metabolism 153 1.1 Psyr_2917 cdhC Carnitine metabolism 56 1.5 Psyr_2918 cdhA Carnitine metabolism 68 1.2 Psyr_2919 cdhB Carnitine metabolism 53 1.1 Psyr_3236 dhcR Regulator acetoacetate/carnitine metabolism 77 -1.1 Psyr_3237 dhcA Acetoacetate/carnitine metabolism 162 16.4 Psyr_3238 dhcB Acetoacetate/carnitine metabolism 211 13.8 Psyr_0029 cosC Choline-O-sulfate metabolism 987 -2.5 Psyr_0027 cosY Choline-O-sulfate metabolism 1133 -1.2 Psyr_0026 cosE Choline-O-sulfate metabolism 123 -1.6 Psyr_4738 pchP Phosphorylcholine/choline metabolism 1193 -2.0 Psyr_1823 plcA Phosphatidylcholine/choline metabolism 88 3.9

______1 Data are derived from Gene Expression Omnibus accession no. GSE42544. 2 The fold change values represent the change in transcript abundance in epiphytic and apoplastic environments as compared to abundance in basal medium HMM. Yu et al. 2013 3 Fluorescent intensities across the B728a transcriptome ranged from 37 to 54119

100

A 8.0

7.5

7.0 TimeB728a vs B728a 6.5 TimeΔcosC vs CosC /g leaf) leaf) /g TimeΔpchP vs pchP cfu 6.0 TimeΔplcA vs plcA 5.5 Log( TimeΔcdhA vs cdhA Time vs gbcAB 5.0 Wet Dry ΔgbcAB TimeΔbetA vs betA 4.5 0 12 24 36 48 60 72 84 TimeBT vs BT Time (h) B C 8.0 8.0

7.5 7.5

7.0 7.0

6.5 6.5 /g leaf) leaf) /g /g leaf) leaf) /g cfu cfu 6.0 6.0

5.5 5.5 Log( Log(

5.0 Wet Dry 5.0 Wet Dry

4.5 4.5 0 12 24 36 48 60 72 84 0 12 24 36 48 60 72 84 Time (h) Time (h)

D 8.0

7.5

7.0

/g leaf) leaf) /g 6.5 cfu 6.0

Log( 5.5

5.0 Wet Dry

4.5 0 12 24 36 48 60 72 84 Time (h)

Figure 2. Choline and choline derivatives contribute to osmotolerance during epiphytic colonization. (A) B728a, ΔcosC, ΔpchP, ΔplcA, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during epiphytic colonization under water replete conditions followed by drying stress of 45% RH at 25°C. Data replotted to display growth dynamics between (B) B728a, ΔbetA, ΔgbcAB, and BT to demonstrate influence of compatible solute accumulation on epiphytic colonization; (C) B728a ΔplcA, ΔpchP, ΔbetA, and BT to demonstrate influence of phosphatidylcholine, phosphocholine, and choline on epiphytic colonization; (D) B728a, ΔcosC, ΔcdhA, and BT to demonstrate not all QACs influence epiphytic colonization. Populations are averaged across three independent experiments (individually presented in Appendix C). Reported as mean ± Standard Error of the Mean (SEM).

101

A B

7.5 7.5 A AB AB 7.0 A 7.0 ABC BC AB BCD AB AB CD AB AB D /g leaf) leaf) /g BC 6.5 6.5 cfu C Log ( Log

6.0 6.0

5.5 5.5

BT BT BT BT betA plcA betA plcA cosC cdhA cosC pchP cdhA pchP Δ Δ Δ Δ gbcAB gbcAB B728a B728a Δ Δ Δ B728a B728a gbcAB gbcAB Δ Δ Δ Δ Δ

Figure 3. Statistical comparisons of epiphytic populations levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are averaged across three independent experiments and compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05).

ΔgbcAB, ΔpchP, ΔcdhA, ΔplcA, and ΔcosC did not significantly differ in size (Figure

3A). Thus, strains deficient in choline but not GB catabolism were affected in fitness.

Between 48 and 72 hpi, the populations of the ΔgbcAB mutant did not change, whereas those of the ΔbetA and BT mutants increased (Figure 2B), indicating GB catabolism contributed to maintaining bacterial populations during sustained adaptation to drying stress. At 72 hpi, several of the mutants exhibited smaller populations than the wild type; these included ΔplcA, ΔbetA, ΔgbcAB, and BT (Figure 2B, 2C and 3B), demonstrating that the ability to catabolize phosphatidylcholine, choline, and GB, as well as to import

QACs, each contributed to maximal fitness of B728a on bean leaves. The ΔcosC and

ΔcdhA mutants showed a similar growth to B728a throughout the 72 h assay (Figure 2D),

102 indicating the absence of a role for COS or carnitine on epiphytic fitness; support for a role for phosphorylcholine, as reflected in the growth of the ΔpchP mutant, remains equivocal (Figure 3B).

Carnitine and choline contribute to colonization of the spermosphere.

We evaluated the role of QACs in spermosphere colonization over time by inoculating seeds with B728a, ΔggnΔtre, which is a mutant deficient in the de novo synthesis of the compatible solutes NAGGN and trehalose, ΔcosC, ΔpchP, ΔcdhA,

ΔbetA, ΔgbcAB, and BT (Figure 4). Between 0 and 24 hpi, the seeds were imbibing water, and all the strains exhibited similar growth (Figure 4A). The reduced growth of

BT relative to B728a at the later time points (Figures 4A-4D) indicates that an inability to import QACs is detrimental to population maintenance. The similar growth of B728a and ΔggnΔtre (Figures 4C and 5) during all stages of seed germination suggests that production of compatible solutes is not required for colonization of seeds and young seedlings. Additionally, the reduced growth of ΔgbcAB at both 48 hpi and 72 hpi

(Figures 4C and 5) indicates that utilization of QACs as nutrients influences population growth. At 48 hpi, when most of the seeds showed visible emergence of the radicle, the

ΔcdhA, ΔbetA, ΔgbcAB, and BT mutants had established smaller populations than B728a

(Figure 5A), demonstrating that choline and carnitine influenced population development. Impact of carnitine and choline catabolism on population trends persisted from 48 hpi to 72 hpi (Figure 5B), implying P. syringae gains sustained benefits from utilizing these two QACs as nutrients.

103

10 Time vs B728a 9 B728a ΔTimeggnΔ trevs tre- 8 ΔTimecosC vs cosC ΔTimepchP vs pchP /g tissue) tissue) /g 7 ΔTimecdhA vs cdhA cfu 6 ΔTimebetA vs betA

Log( ΔTimegbcAB vs gbcAB 5 BTTime vs BT Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

B C

10 10

9 9

8 8

/g tissue) tissue) /g 7 tissue) /g 7 cfu cfu 6 6 Log( Log( 5 5

Stage 1 Stage 2 Stage 3 Stage 4 Stage 1 Stage 2 Stage 3 Stage 4 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days) Time (days) D

/g tissue) tissue) /g 7 cfu 6 Log( 5

Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure 4. QAC catabolism contributes to colonization of germinating bean seeds. (A) B728a, ΔcosC, ΔggnΔtre, ΔpchP, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during colonization of germinating bean seeds in sterile sand under 24 h light and 25°C. Data replotted to display growth dynamics between (B) B728a, ΔbetA, ΔcdhA, ΔgbcAB, and BT to demonstrate relative impact of carnitine and choline (C) B728a ΔgbcAB, ΔnggΔtre, and BT to demonstrate impact of compatible solute accumulation; (D) B728a, ΔbetA, ΔcdhA, ΔpchP, and BT to demonstrate the relative impact of QACs, choline, carnitine, and phosphorylcholine, detected in seed exudates. Stages correspond to seed developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. Populations are averaged across three independent experiments (individually presented in Appendix C). Reported as mean ± SEM.

104

A B 10.0 10.0

9.5 9.5 A A A A A AB 9.0 AB 9.0 AB B 8.5 8.5

/g tissue) tissue) /g B BC BC

cfu C 8.0 BC BC 8.0 C Log ( Log 7.5 7.5

7.0 7.0

BT BT BT BT ggn ggn betA betA cosC pchP cdhA cosC Δ pchP cdhA Δ Δ Δ B728a B728a Δ gbcAB B728a B728a Δ Δ Δ gbcAB Δ Δ tre tre Δ Δ Δ Δ

Figure 5. Statistical comparisons of spermosphere population levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are averaged across three independent experiments and compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05).

HPLC-MS analysis revealed that choline, carnitine and phosphorylcholine were present in seed exudates while GB and COS were not detected. Choline was the most abundant QAC detected (Figure 6) and although choline levels in the seed exudates decreased over time after an initial release during imbibition, choline remained highly abundant throughout germination (Figure 6). Carnitine displayed a release pattern that corresponded with the growth dynamics of the ΔcdhA mutant (Figure 4C and 6), namely a greater population increase during the period between imbibition and radicle emergence

(stages 1 and 2) than between radicle emergence and radicle elongation/seedling establishment (stages 2 and 3/4) (Figure 4). The low levels of PCH detected did not change in abundance throughout the germination process (Figure 6).

105

A Stage 1 Stage 2 Stage 3 Stage 4

B 7.0

6.5

6.0

5.5

5.0

4.5 Log (abundance) (abundance) Log 4.0

3.5

3.0 0 1 2 3 4 5 Germination Stage

Figure 6. Bean seeds release the QACs choline, carnitine, and phosphorylcholine throughout germination. (A) Germinating seeds were placed into one of four developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. (B) Bean seeds were germinated to developmental stages 1-4, exudates were removed and analyzed via HLPC-MS for abundance of choline (m/z 104.1, black line), carnitine (m/z 162.1, red line), and phosphorylcholine (m/z 184.1, green line). Values are mean ± SE (stage 0, n=4; stage 1, n=13; Stage 2, n=46; stage 3, n=9; and stage 4, n=17). Results are representative of 3 replicated experiments (presented individually in Appendix C).

Discussion

The mechanisms that foliar pathogens utilize to establish and maintain populations on the plant host under water limitation stress have not been comprehensively characterized. In this study, we elucidated how the model organism P. syringae utilizes host-derived QACs to maintain populations during colonization of both the leaf and the seed. Specifically, we showed that P. syringae tolerated drying stress

106 during epiphytic colonization by eventually catabolizing glycine betaine, even though it had likely accumulated GB as a compatible solute; this degradation has been shown to enable the accumulation of the endogenous solutes trehalose and NAGGN (28). We also established that P. syringae utilized QACs as nutrient sources during colonization of the spermosphere. Interestingly, choline supported populations throughout the germination process, whereas carnitine supported populations during specific stages of germination that corresponded to distinct physiological events. A deeper understanding of host- pathogen dynamics in the phyllosphere could enable us to develop strategies to promote beneficial bacteria and limit pathogen growth in these environments, leading to reduced disease occurrence.

When P. syringae colonized leaf surfaces under high humidity conditions, QAC transport and metabolism did not have a large influence on population growth in the first

24 h (Figure 2A). This was likely due to nutrients leaching to the leaf surface due to the presence of standing water (30, 31, 36). It is clear from Figure 2D that not all QACs impacted leaf colonization, as ΔcdhA and ΔcosC maintained populations comparable to

B728a throughout the assay. However, the exposure to 24 h of drying stress between 24 and 48 hpi was associated with a decrease in the populations of BT and ΔbetA but not the other six strains (Figure 3A). Due to the inability of ΔbetA to convert choline to GB, the reduction in ΔbetA populations indicated that B728a accessed choline, and benefited from its use under drying conditions on leaves. A previous study (8) also found that

B728a accessed choline on leaves, but did not resolve whether choline was used for osmoprotection or nutrition. Since GbcAB is required for the use of GB for nutrition, the lack of a significant decrease in the ΔgbcAB populations provided evidence that the

107 choline was used for osmoprotection. Moreover, our data were consistent with the accumulation of GB rather than trehalose and NAGGN as a compatible solute during the initial 24 h of drying on leaf surfaces based on two lines of evidence. First, the ΔgbcAB mutant is known to accumulate GB in the presence of choline (Figure 1) and to rely solely on this GB as a compatible solute due to the accumulated GB inhibiting trehalose and NAGGN production (28), but this mutant did not differ significantly from B728a in its populations at 48 hpi. Secondly, the reduction in the BT populations during epiphytic growth demonstrated that de novo synthesis of compatible solutes cannot compensate for the lack of QAC-derived osmoprotection. Although the decrease in BT populations could be due the inability of BT to use QACs for nutrition, this was not supported based on the larger populations of ΔgbcAB as compared to BT at 48 hpi (Figure 2B and 3A).

We conclude that the P. syringae cells on leaves imported choline and this enabled the accumulation of sufficient GB to maintain P. syringae populations in response to the drying stress imposed between 24 and 48 hpi.

As the cells on leaves were exposed to continued drying stress, they appeared to transition away from a dependence on GB. In particular, between 48 and 72 hpi, ΔgbcAB did not grow whereas ΔbetA and BT did, and by 72 hpi, the populations of ΔgbcAB as well as ΔbetA and BT were lower than those of B728a (Figure 3B). These results were consistent with GB accumulation becoming detrimental to P. syringae during a period of sustained drying on the leaf surface, with the continued growth of ΔbetA and BT likely reflecting a transition from depending on GB to depending on compatible solutes that were synthesized de novo.

108

We began to investigate the source of the choline accessed by B728a on leaves by examining the growth of B728a mutants that were deficient in the catabolism of choline-derivatives. In particular, phosphatidylcholine is the major membrane lipid in eukaryotes, and it can be degraded to PCH by plant or bacterial enzymes (2, 40). We included the ΔplcA and ΔpchP mutants specifically to evaluate whether bacterial- mediated release of phosphatidylcholine and/or PCH contributed to B728a colonization of bean leaves. Looking at ΔplcA, the smaller populations of ΔplcA than B728a at 72 hpi but not 48 hpi (Figure 3) suggested at least a limited role of the bacteria in obtaining choline from PC during sustained drying stress. Since PlcA is predicted to function upstream of PchP (Figure 1), we would expect to observe a similar or greater decrease in

ΔpchP population with respect to B728a, as with ΔplcA; since we do not, our data neither supports nor disproves the role of PlcA in choline acquisition from PC; that this protein must be characterized further before a mechanism can be teased apart.

These data support the model shown in Figure 7 describing the mechanisms by which P. syringae is exposed to and tolerates osmotic stress on leaves following a rain event or dew deposition (Figure 7). When free water is present on a leaf, nutrients leach onto the leaf surface (44); these nutrients are likely comprised of sugars and organic acids

(4, 30, 36, 43) and support bacterial growth. As the leaf surface water evaporates, the solutes in the water increase in concentration, imposing an osmotic stress on the resident bacteria. A major mechanism by which P. syringae pv. syringae B728a, a well-adapted epiphyte, responds to the increased osmolarity is by importing choline, which has leached to the surface, and converting it to GB for accumulation as a compatible solute.

Accumulation of GB is sufficient to circumvent the initial shock of the drying stress.

109

Choline

Water promotes Choline Sugars Osmolarity of leaf leaching of choline Organic surface water and nutrients to the acids increases leaf surface

Osmoadaptation via osmoprotectant uptake

OpuC BetAB Choline Glycine BetT betaine

Choline is imported Glycine betaine is and converted to accumulated as a EVAPORATION glycine betaine compatible solute OF LEAF SURFACE WATER

Osmolarilty continues Osmolarity continues to increase, with little to increase and remaining available available choline choline decreases

Osmoadaptation via de novo synthesis Transition to reduced dependence on choline

GgnABC NAGGN ggnABC (Choline) Glycine NAGGN Tre Trehalose betaine GbcAB tre Trehalose (Glycine betaine) Glycine

Glycine betaine NAGGN and trehalose Remaining glycine NAGGN and trehalose repression of accumulate and replace betaine is catabolized accumulate as the primary glycine betaine as compatible solutes; endogenous solute synthesis is relieved compatible solutes colonies accumulate EPS

Figure 7. Model of the availability and function of QACs during P. syringae B728a colonization of leaf surfaces.

However, as evaporation continues, the level of stress increases, and the bacteria replace

GB with compatible solutes that are synthesized de novo. This involves degrading GB and synthesizing trehalose and NAGGN, which support long-term adaptation at the

110 cellular level. As the populations increase further in size, they also secrete extracellular polysaccharides, which further support long-term adaptation at the population level (37).

In contrast to osmotic stress as a major driver of B728a adaptation to leaf surfaces, osmotic stress did not impact B728a colonization of the spermosphere. The uptake of QACs contributed to seed colonization based on the reduced population sizes of BT at 48 and 72 hpi (Figure 5A and 5B). The BT populations did not differ from

ΔgbcAB at any time point (Figure 4C, 5A, and 5B), however, indicating that loss of uptake and loss of GB catabolism had similar impacts and thus QACs were used primarily or exclusively for nutrition. Moreover, the population levels of ΔggnΔtre and

B728a were similar whereas those of ΔgbcAB and BT were much reduced (Figure 4C,

5A, and 5B). Collectively, these results indicated that trehalose, NAGGN synthesis, and

GB accumulation did not detectably favor bacterial growth on seeds and developing seedlings, and therefore B728a was likely not subjected to osmotic stress during this phase of its lifecycle.

Compositional data of seed extracts coupled with the growth data of the QAC transport and catabolic mutants of B728a indicated that choline and carnitine were the primary QACs available during colonization of germinating bean seeds (Figure 4B, 5A, and 5B). The compositional data indicated large increases in choline and carnitine levels when the seeds were imbibing water, with gradual decreases after radicle emergence

(Figure 6); glycine betaine was not detected. The reduced population sizes of the ΔbetA,

ΔgbcAB and ΔcdhA mutants with respect to B728a (Figures 4B and 5) indicated that

B728a benefited from the catabolism of carnitine and choline among the QACs detected

(Figure 6). As demonstrated earlier, carnitine catabolism supports population

111

Carnitine Choline

Carnitine levels increase Increased Seed begins to Choline levels increase due to increased fatty nutrient imbibe water due to membrane acid metabolism during release due biogenesis and role of seed germination and to fissures choline as a precursor role of carnitine as a in the seed to phosphatidylcholine SEED carrier in β-oxidation coat GERMINATION

BetAB GbcAB Choline BetT Glycine CdhABC betaine Carnitine Cbc

Choline (Carnitine) Choline

Carnitine levels fall below Choline levels Carnitine levels decrease the limit of detection when remain high due as the seed transitions greening of seedling to continued from fatty acid metabolism tissues begins membrane to photosynthesis following biogenesis greening of seedling tissues

Figure 8. Model of the availability and function of QACs during P. syringae colonization of developing seedlings.

development during radicle emergence and elongation, when seeds produce carnitine and are likely relying on fatty acid metabolism for energy production (see Chapter II).

Choline release from germinating seeds was likely due to membrane biogenesis, which requires choline to produce phosphatidylcholine (13, 35). The fact that the

112 populations of ΔbetA were lower, although not significantly, than ΔgbcAB could be due to choline accumulation in the cytosol, as choline is positively charged at physiological pH and could be detrimental to cellular function (12, 42). PCH was also detected in the seed exudates (Figure 6); the fact that the growth of the ΔpchP mutant was slightly, but not significantly, reduced (Figure 5) provided weak evidence that B728a utilized this

PCH. PCH is required for membrane biogenesis during chloroplast production (10), and therefore would be expected to increase during the greening of the seedlings; we did not see evidence for this based on the absence of a major increase in the exudates upon seedling entry into stage 4 (Figure 6).

P. syringae growth dynamics support the model shown in Figure 8 of QAC release and utilization by bacteria on germinating seeds and developing seedlings. After a seed imbibes water and the seed coat cracks, carnitine and choline are released along with other nutrients. The P. syringae cells in the soil may use these nutrients, including carnitine, as signals to help direct them to germinating seeds (see Chapter II). P. syringae cells utilize both choline and carnitine as nutrient sources, but do not use them for osmoprotection, presumably because the presence of sufficient water for imbibition coupled with the release of relatively low levels of solutes minimizes osmotic stress.

Whereas choline, a precursor to the major membrane lipid phosphatidylcholine, remains abundant during germination and seedling development due to the continued need for membrane biogenesis, carnitine levels eventually decrease as the seed exhausts its lipid stores, decreasing its need for carnitine to serve as a carrier in β-oxidation, and transitions to a dependence on photosynthesis, as reflected in the greening of seedling tissue.

113

This study presented data that fits a real-world scenario where P. syringae B728a colonizes and establishes populations upon P. vulgaris leaves and seeds, and elucidates the differential impact of various QACs on these events. The presence and bioavailability of plant-derived choline has been investigated previously, though our study is first to implement choline directly in dynamic adaptations during colonization of the leaf surface. Our data establishes adaptation to drying stress on the leaf surface in P. syringae requires a dynamic shift in preferred compatible solutes as drying stress progresses in intensity, from glycine betaine to endogenous solutes trehalose and

NAGGN. There has been limited research on P. syringae as a colonist of germinating seeds and as a soil resident and through our investigation we have expanded the body of literature on bacterial-seed interactions. Specifically we have established that spermosphere colonist populations can derive benefit from individual components of seed exudates that are transiently released during germination. QACs are a major driving factor in P. syringae colonization of plant tissue and may serve as a model to help explain how many plant-associated microorganisms thrive in the harsh, constantly fluctuating abiotic conditions of the phyllosphere.

References

1. Axtell CA, Beattie GA. 2002. Construction and characterization of a proU-gfp transcriptional fusion that measures water availability in a microbial habitat. Appl Environ Microbiol 68:4604-4612.

2. Barker AP, Vasil AI, Filloux A, Ball G, Wilderman PJ, Vasil ML. 2004. A novel extracellular phospholipase C of pseudomonas aeruginosa is required for phospholipid chemotaxis. Mol Microbiol 53:1089-1098.

114

3. Beattie GA. 2011. Water relations in the interaction of foliar bacterial pathogens with plants. Annu Rev Phytopathol 49:533-555.

4. Beattie GA, Lindow SE. 1995. The secret life of foliar bacterial pathogens on leaves. Annu Rev Phytopathol 33:145-172.

5. Bourdin B, Adenier H, Perrin Y. 2007. Carnitine is associated with fatty acid metabolism in plants. Plant Physiol Biochem 45:926-931.

6. Chen C, Beattie GA. 2007. Characterization of the osmoprotectant transporter OpuC from pseudomonas syringae and demonstration that cystathionine-β- synthase domains are required for its osmoregulatory function. J Bacteriol 189:6901-6912.

7. Chen C, Beattie GA. 2008. Pseudomonas syringae BetT is a low-affinity choline transporter that is responsible for superior osmoprotection by choline over glycine betaine. J Bacteriol 190:2717-2725.

8. Chen C, Li S, Mckeever DR, Beattie GA. 2013. The widespread plant‐colonizing bacterial species pseudomonas syringae detects and exploits an extracellular pool of choline in hosts. Plant J 75:891-902.

10. Cornell RB, Ridgway ND. 2015. Ctp: Phosphocholine cytidylyltransferase: function, regulation, and structure of an amphitropic enzyme required for membrane biogenesis. Prog Lipid Res 59:147-171.

11. Csonka LN. 1989. Physiological and genetic responses of bacteria to osmotic stress. Microbiol Rev 53:121-147.

12. Currier SF, Mautner HG. 1974. On the mechanism of action of choline acetyltransferase. Proc Nat Acad Sci U S A 71:3355-3358.

13. Datko AH, Mudd SH. 1988. Phosphatidylcholine synthesis differing patterns in soybean and carrot. Plant Physiol 88:854-861.

14. De Rudder KE, Sohlenkamp C, Geiger O. 1999. Plant-exuded choline is used for rhizobial membrane lipid biosynthesis by phosphatidylcholine synthase. J Biol Chem 274:20011-20016.

115

15. Fitzsimmons LF, Hampel KJ, Wargo MJ. 2012. Cellular choline and glycine betaine pools impact osmoprotection and phospholipase C production in pseudomonas aeruginosa. J Bacteriol 194:4718-4726.

16. Freeman BC, Beattie GA. 2009. Bacterial growth restriction during host resistance to pseudomonas syringae is associated with leaf water loss and localized cessation of vascular activity in arabidopsis thaliana. Mol Plant- Microbe Interact 22:857-867.

17. Freeman BC, Chen C, Beattie GA. 2010. Identification of the trehalose biosynthetic loci of pseudomonas syringae and their contribution to fitness in the phyllosphere. Environ Microbiol 12:1486-1497.

18. Gross DC. 1991. Molecular and genetic analysis of toxin production by pathovars of pseudomonas syringae. Annu Rev Phytopathol 29:247-278.

19. Hanson A, Gage D. 1991. Identification and determination by fast atom bombardment mass spectrometry of the compatible solute choline-O-sulfate in Limonium species and other halophytes. Funct Plant Biol 18:317-327.

20. Hasegawa PM, Bressan RA, Zhu J-K, Bohnert HJ. 2000. Plant cellular and molecular responses to high salinity. Annu Rev Plant Biol 51:463-499.

21. Hirano SS, Upper CD. 1990. Population biology and epidemiology of Pseudomonas syringae. Annu Rev Phytopathol 28:155-177.

22. Hirano SS, Upper CD. 2000. Bacteria in the leaf ecosystem with emphasis on Pseudomonas syringae—a pathogen, ice nucleus, and epiphyte. Microbiol Mol Biol Rev 64:624-653.

23. Jovanovich S, Martinell M, Record M, Burgess R. 1988. Rapid response to osmotic upshift by osmoregulated genes in Escherichia coli and Salmonella typhimurium. J Bacteriol 170:534-539.

24. Kappes RM, Bremer E. 1998. Response of Bacillus subtilis to high osmolarity: Uptake of carnitine, crotonobetaine and γ-butyrobetaine via the ABC transport system OpuC. Microbiol 144:83-90.

25. King EO, Ward MK, Raney DE. 1954. Two simple media for the demonstration of pyocyanin and fluorescin. The J Lab Clinic Med 44:301-307.

26. Landfald B, Strøm AR. 1986. Choline-glycine betaine pathway confers a high level of osmotic tolerance in Escherichia coli. J Bacteriol 165:849-855.

27. Lelandais‐Brière C, Jovanovic M, Torres GA, Perrin Y, Lemoine R, Corre‐Menguy F, Hartmann C. 2007. Disruption of AtoCt1, an organic cation

116

transporter gene, affects root development and carnitine‐related responses in Arabidopsis. Plant J 51:154-164.

28. Li S, Yu X, Beattie GA. 2013. Glycine betaine catabolism contributes to Pseudomonas syringae tolerance to hyperosmotic stress by relieving betaine- mediated suppression of compatible solute synthesis. J Bacteriol 195:2415-2423.

29. Lindow SE. 1991. Determinants of epiphytic fitness in bacteria, p 295-314, Microbial ecology of Leaves. Springer.

30. Lindow SE, Brandl MT. 2003. Microbiology of the phyllosphere. Appl Environ Microbiol 69:1875-1883.

31. Lindow SE, Leveau JH. 2002. Phyllosphere microbiology. Curr Opin Biotechnol 13:238-243.

32. Mancinelli RL. 2005. Microbial life in brines, evaporites and saline sediments: The search for life on mars, p 277-297, Water on mars and life. Springer.

33. May G, Faatz E, Villarejo M, Bremer E. 1986. Binding protein dependent transport of glycine betaine and its osmotic regulation in Escherichia coli k12. Mol Gen Genetic MGG 205:225-233.

34. Mcneil PH, Thomas DR. 1975. Carnitine content of pea seedling cotyledons. Phytochem 14:2335-2336.

35. Mcneil SD, Nuccio ML, Ziemak MJ, Hanson AD. 2001. Enhanced synthesis of choline and glycine betaine in transgenic tobacco plants that overexpress phosphoethanolamine n-methyltransferase. Proc Natl Acad Sci U S A 98:10001- 10005.

36. Mercier J, Lindow S. 2000. Role of leaf surface sugars in colonization of plants by bacterial epiphytes. Appl Environ Microbiol 66:369-374.

37. Monier J-M, Lindow S. 2005. Aggregates of resident bacteria facilitate survival of immigrant bacteria on leaf surfaces. Microb Ecol 49:343-352.

38. Morris CE, Sands DC, Vinatzer BA, Glaux C, Guilbaud C, Buffiere A, Yan S, Dominguez H, Thompson BM. 2008. The life history of the plant pathogen Pseudomonas syringae is linked to the water cycle. ISME J 2:321-334.

39. Nau-Wagner G, Boch J, Le Good JA, Bremer E. 1999. High-affinity transport of choline-O-sulfate and its use as a compatible solute in Bacillus subtilis. Appl Environ Microbiol 65:560-568.

117

40. Qin C, Wang X. 2002. The Arabidopsis phospholipase D family. Characterization of a calcium-independent and phosphatidylcholine-selective pld1 with distinct regulatory domains. Plant Physiol 128:1057-1068.

41. Roberts MF. 2000. Osmoadaptation and osmoregulation in archaea. Front Biosci 5:D796-812.

42. Roberts MF. 2005. Organic compatible solutes of halotolerant and halophilic microorganisms. Saline Sys 1:1-30.

43. Ryffel F, Helfrich EJN, Kiefer P, Peyriga L, Portais J-C, Piel J, Vorholt JA. 2015. Metabolic footprint of epiphytic bacteria on Arabidopsis thaliana leaves. ISME J doi:10.1038/ismej.2015.141.

44. Schönherr J. 2006. Characterization of aqueous pores in plant cuticles and permeation of ionic solutes. J Exp Bot 57:2471-2491.

45. Verheul A, Rombouts FM, Beumer RR, Abee T. 1995. An ATP-dependent L- carnitine transporter in Listeria monocytogenes Scott A is involved in osmoprotection. J Bacteriol 177:3205-3212.

46. Wargo MJ. 2013. Homeostasis and catabolism of choline and glycine betaine: Lessons from Pseudomonas aeruginosa. Appl Environ Microbiol 79:2112-2120.

47. Wargo MJ, Hogan DA. 2009. Identification of genes required for Pseudomonas aeruginosa carnitine catabolism. Microbiol 155:2411-2419.

48. Wood JM. 2011. Bacterial osmoregulation: A paradigm for the study of cellular homeostasis. Annu Rev Microbiol 65:215-238.

49. Wood JM, Bremer E, Csonka LN, Kraemer R, Poolman B, Van Der Heide T, Smith LT. 2001. Osmosensing and osmoregulatory compatible solute accumulation by bacteria. Compar Biochem Physiol A: Mol Int Physiol 130:437- 460.

50. Yancey PH. 2001. Water stress, osmolytes and proteins. Am Zool 41:699-709.

118

CHAPTER 5. CONCLUSIONS AND FUTURE DIRECTIONS

Epiphytic bacteria must constantly respond and adapt to fluctuations in UV radiation, temperature, and water availability. Adaptation to water limitation is critical to effective colonization of epiphytic sites by all microorganisms residing in this environment (2). Quaternary ammonium compounds (QACs) are of particular interest in this context because they function as osmoprotectants. Additionally, plants produce several QACs, allowing for the potential QAC-mediated interaction between the plant host and epiphytic bacteria such as Pseudomonas syringae.

DhcAB metabolizes Acetoacetate, an intermediate of L-carnitine catabolism, in P. syringae, thus clarifying the carnitine degradation pathway in Pseudomonads. This study was the first to investigate the role of plant-derived carnitine in plant-bacteria associations and demonstrated a clear role for it in seed colonization. Utilization of carnitine by P. syringae during seed colonization is consistent with both a physiological role for carnitine in fatty acid oxidation and a role for fatty acid oxidation in the germination of bean seeds. Importantly, we established that a single QAC released during germination could be a major driver of bacterial growth during seed colonization.

These results raise the question of the breadth of plant species that produce carnitine at sufficient levels to influence colonizing microbes, given that the carnitine degradation genes are conserved across several pathogenic species. Research into carnitine in plants has focused mainly on its role in fatty acid metabolism in Arabidopsis and peas (5), but carnitine has been found at detectable levels in the tissues of other plants as well (1, 7).

Further investigation into the role of carnitine in other plant-microbe interactions,

119 specifically in the seeds, would indicate the generalizability of the phenomenon we observed. Furthermore, investigation into the impact of P. syringae carnitine utilization during colonization of aphids would shed light on distinct stages of the P. syringae lifecycle.

P. syringae can catabolize COS as a carbon, nitrogen, and sulfur source, and during growth in hyperosmotic stress, as an osmoprotectant, through the action of the choline sulfatase CosC. Similar to sulfonate degradation (8), the degradation of the sulfate ester, COS, is coupled with a sulfate transporter, CosE, which is likely functioning as a sulfate exporter to relieve P. syringae of sulfate toxicity. Further investigation is required to determine the mechanisms behind the toxicity observed upon deletion of cosE. One possible avenue would be to evaluate accumulation of 35S sulfate when ΔcosE is provided radiolabeled COS under osmotic stress. A potential pitfall of this experiment is the cost and the difficulty in purchasing and/or synthesizing radiolabeled COS. An alternative is to overexpress the sulfur assimilation pathway (3), to decrease the amount of sulfate accumulated in the cell and to avoid toxic levels of sulfate build up.

Specifically, over-expressing cysDNC and cysE could theoretically create a constitutive sulfate assimilation pathway; if this decreased toxicity of ΔcosE, it would directly support the role of CosE as a sulfate exporter. Although COS catabolism does not influence P. syringae colonization of P. vulgaris as either a nutrient source or as an osmoprotectant,

P. syringae may encounter COS produced from other organisms on leaves, such as epiphytic fungi (4). The data presented here provides the potential for investigation of

COS catabolism in association with other plant species including ones known to produce

COS, such as Limonium sp. (6). Obtaining sulfur on the leaf surface is a complex

120 interaction involving multiple sulfur sources; the epiphyte sulfur acquisition and utilization network model proposed here indicates that investigation into the mechanism by which foliar pathogens meet sulfur requirements will involve the analysis of several organosulfur networks including the alkyl- and aryl- sulfonates as well as the cysteine and methionine biogenesis pathways. This could be teased apart through investigation of the impact of compound mutations on epiphytic fitness. Further research into the sources of sulfur accessed by the foliar pathogens during plant-pathogen interactions could provide valuable insights into epiphytic ecology, thus helping to understand how microbes obtain micronutrients on the leaf surface.

There has been limited research on P. syringae as a colonist of germinating seeds and as a soil resident. Here we demonstrated the catabolism of the QACs choline and carnitine influenced colonization of the spermosphere, demonstrating that P. syringae has adapted to utilize compounds transiently released from its host. Choline is a major driver of osmotolerance during P. syringae epiphytic colonization, followed by the accumulation of trehalose and NAGGN to sustain bacterial populations during drying stress. Data presented here is in line with our model that P. syringae has evolved to thrive in environments rich in choline, such as the plant host. QACs are a major driving factor in P. syringae colonization of plant tissue and may serve as a model to help explain how many plant-associated microorganisms thrive in the harsh, constantly fluctuating abiotic conditions of the phyllosphere.

Collectively, this study advances the body of literature on QAC-mediated plant- pathogen interactions, specifically showing how P. syringae derives osmoprotection and nutritional benefits from a range of host-derived QACs during colonization of P.

121 vulgaris. A myriad of bacteria reside upon the leaf surface and have the opportunity to encounter QACs produced by the host plants; this study could be expanded to investigate other plant-microbe combinations as well as a microbiome study profiling the complete ecology of the leaf or seed, specifically in terms of the impact QAC utilization on the competition between colonizing microbes, whether these microbes are colonizing P. vulgaris or countless other plant species. Models developed based on data gathered here could help to understand how many epiphytes interact with their host and lead to discoveries that could eventually help to promote beneficial bacteria or limit the growth of pathogens on plants.

References

1. Arima J, Uesumi A, Mitsuzumi H, Mori N. 2010. Biochemical characterization of L-carnitine dehydrogenases from Rhizobium sp. and Xanthomonas translucens. Biosci, Biotechnol, Biochem 74:1237-1242.

2. Axtell CA, Beattie GA. 2002. Construction and characterization of a proU-gfp transcriptional fusion that measures water availability in a microbial habitat. Appl Environ Microbiol 68:4604-4612.

3. Beil S, Kertesz MA, Leisinger T, Cook AM. 1996. The assimilation of sulfur from multiple sources and its correlation with expression of the sulfate-starvation- induced stimulon in Pseudomonas putida S-313. Microbiol 142:1989-1995.

4. Bellenger N, Nissen P, Wood TC, Segel IH. 1968. Specificity and control of Choline–O-Sulfate transport in filamentous fungi. J Bacteriol 96:1574-1585.

5. Bourdin B, Adenier H, Perrin Y. 2007. Carnitine is associated with fatty acid metabolism in plants. Plant Physiol Biochem 45:926-931.

6. Hanson A, Gage D. 1991. Identification and determination by fast atom bombardment mass spectrometry of the compatible solute choline-O-sulfate in Limonium species and other halophytes. Funct Plant Biol 18:317-327.

7. Panter R, Mudd J. 1969. Carnitine levels in some higher plants. FEBS Lett 5:169-170.

122

8. Wübbeler JH, Hiessl S, Meinert C, Poehlein A, Schuldes J, Daniel R, Steinbüchel A. 2015. The genome of Variovorax paradoxus strain TBEA6 provides new understandings for the catabolism of 3, 3-thiodipropionic acid and hence the production of polythioesters. J Biotech 209:85-95.

123

APPENDIX A. SUPPLEMENTARY TABLES AND FIGURES FOR CHAPTER II

Table A.1 Primers used to construct mutants and RT-PCR operon analysis

Primers Sequences * Primers for constructing mutants F1-2918 5'-TGATGCAGCCGCTACTGGATCACAA-3' R1-2918KnC 5'-AGCCTACACAATCGCTCAAGACGTAAGAGTGCGGTGATT -3' F2-2918knca 5'-AATATCCGGGTAGGCGCAATCACTTGAGCTTCACGCAAT-3' R2-2918 5'-TCTGCGTGGCAATGAAGTTA-3' 2918scrnF 5'-TGAAATTCTCAGCCGTCTGG-3' 2918scrnR 5'-TGCTTTGCAGGACGATGGAA-3' F1-2919 5'-TGCCTCCGAGACATAAATCACC-3' R1-2919Kn 5'-AGCCTACACAATCGCTCAAGACGTGTCGGGGGATTCC-3' F2-2919kn 5'-AATATCCGGGTAGGCGCAATCACTTCGAAAAAGCCCTC-3' R2-2919 5'-CCTCGGTCTTGCAGAAATGTT-3' F1-2917 5'-ACCATGAACCGACTGATTGGCT-3' R1-2917kn 5'-AGCCTACACAATCGCTCAAGACGTCGGTGAGTGCGCAGGTGATGATGA-3' F2-2917kn 5'-AATATCCGGGTAGGCGCAATCACTCCAGTGAAATTCTCAGCCGTCTG-3' R2-2917 5'-ACATCGTCGATCAGTTTGTCGGT-3' F1-dhcAB 5'-CTATCGACTGTTGCGAGCATTCTC-3' R1-dhcKn 5'-AGCCTACACAATCGCTCAAGACGTGTCCAGCCATGATGATCCGCC-3' F2-dhcKn 5'-AATATCCGGGTAGGCGCAATCACTTCGTGGCCAAAACCGCAG-3' R2-dhcAB 5'-GCTGTGATCACCGGTTTGCTCGAA-3' RT2915F 5'-AATCACTCAAATCGGCGCTCC-3' RT2915R 5'-AAAAACACCACGGCTTCCTTGC-3' Primers for RT-PCR RT2916F 5'-AGCAAATCATTTTCGCCGGC-3' A RT2916R 5'-CGTCGCTGCCATTGAGGTAA-3' B RT2917F 5'-TACCGGTCACGCCCAAGCAAAT-3' C RT2917R 5'-AACAGCGGGTTATCGAGCAG-3' D RT2918F 5'-TGGAACAACAGGGCCTGGTA-3' E RT2918R 5'-GCCATAGCGCTTCGAGCAA-3' F RT2919F 5'-GCGTTCTACCTGCTGATCTT-3' G RT2919R 5'-ATCACCCGGCCAAGGAATTT-3' H

* Letters denote primers used for RT-PCR analysis in determining operon structure in Chapter 2 Figure 1.

124

A B

1.2 0.8

0.7 1.0 0.6

0.8 0.5 B728a cdhAΔcdhA 0.6 0.4 600 nm 600 nm cdhBΔcdhB 0.3

OD 0.4 cdhCΔcdhC 0.2 dhcABΔdhcAB 0.2 0.1 BT

0.0 0.0 0 10 20 30 40 50 0 20 40 60 80

Time (h) Time (h) Figure A.1. Deletion of cdhA, cdhB, cdhC, or dhcAB blocked growth on L-carnitine without interfering with central metabolism. A) Strains grown in MinAS medium, where 20 mM succinate is the sole carbon source. B) Strains grown in MinA medium with 20 mM L-carnitine added as the sole carbon source. Strain BT is deficient in carnitine uptake. Values are mean ± standard error of the mean (SEM) (n=5). Replicate experiment 2.

A B

1.2 0.8

1.0 0.6 0.8 B728a cdhAΔcdhA 0.6 0.4 600 nm 600 nm cdhBΔcdhB

OD 0.4 cdhCΔcdhC 0.2 dhcABΔdhcAB 0.2 BT

0.0 0.0 0 10 20 30 40 0 10 20 30 40 50

Time (h) Time (h) Figure A.2. Deletion of cdhA, cdhB, cdhC, or dhcAB blocked growth on L-carnitine without interfering with central metabolism. A) Strains grown in MinAS medium, where 20 mM succinate is the sole carbon source. B) Strains grown in MinA medium with 20 mM L-carnitine added as the sole carbon source. Strain BT is deficient in carnitine uptake. Values are mean ± standard error of the mean (SEM) (n=5). Replicate experiment 3.

125

A B C

1.0 0.45 0.8

0.40 0.7 0.8 0.35 0.6

0.30 0.5 0.6 600 nm 600 nm 0.25 0.4

0.4

OD 0.20 0.3 B728a

0.15 0.2 cdhAΔcdhA 0.2 0.10 0.1 dhcABΔdhcAB

0.0 0.05 0.0 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Time (h) Time (h) Time (h)

Figure A.3. DhcAB is involved in acetoacetate metabolism a central intermediate in leucine, phenylalanine, and carnitine catabolic pathways. Growth of B728a and the ΔcdhA and ΔdhcAB mutants in MinA medium with either (A) 20 mM L-leucine, (B) 20 mM L-phenylalanine, or (C) 20 mM acetoacetate, provided as carbon sources. Values are the mean ± SEM (n=5). Replicate experiment 2.

A B C 0.45 0.26 0.40

0.24 0.40 0.35 0.22 0.35 0.30 0.20 0.30 0.18 0.25

600 nm 600 nm 0.25 0.16 0.20 0.20 OD 0.14 B728a 0.15 0.15 0.12 cdhAΔcdhA 0.10 0.10 0.10 dhcABΔdhcAB

0.05 0.08 0.05 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 Time (h) Time (h) Time (h)

Figure A.4. DhcAB is involved in acetoacetate metabolism a central intermediate in leucine, phenylalanine, and carnitine catabolic pathways. Growth of B728a and the ΔcdhA and ΔdhcAB mutants in MinA medium with either (A) 20 mM L-leucine, (B) 20 mM L-phenylalanine, or (C) 20 mM acetoacetate, provided as carbon sources. Values are the mean ± SEM (n=5). Replicate experiment 3.

126

A B

14000 50

12000 40

10000

30 8000

6000 20

4000 Total abundance abundance Total 10 2000

0 0 B728a ΔdhcAB B728a ΔdhcAB

Figure A.5. DhcAB functioned by metabolizing acetoacetate as an intermediate by-product of the central carnitine catabolic pathway. Cells were grown in MinAS medium, washed, and incubated with 40 mM L- carnitine for 8 h before lysis and quantification of the (A) carnitine and (B) acetoacetate in the cell lysates by ESI QTOF mass spectrometry. Total abundance of acetoacetate and carnitine was calculated as the sum of the individual abundances of the protonated, sodiated, and potassiated adducts. Replicate experiment 2.

A B C 0.22 0.20 0.22

0.20 0.18 0.20

0.18 0.16 0.18 0.16 0.16 0.14 0.14 0.14 0.12 0.12 0.12 B728a 0.10 B728a 405nm/630nm 405nm/630nm 0.10 0.10 B728a + 1 mM Carnitine 0.08 B728a + CAR 0.08 0.08

OD Δ!cdhAcdhA 0.06 0.06 0.06 !cdhA + 1 mM Carnitine 0.04 0.04 0.04 ΔcdhA + CAR

0.02 0.02 0.02 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50

Time (h) Time (h) Time (h)

Figure A.6. Carnitine catabolism was not required for carnitine to function in osmoprotection. B728a and ΔcdhA were grown in MinAS with and without 1 mM L-carnitine (CAR) and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on optical density (OD) ratio at 405 nm/630 nm in microtiter plates. Values are means ± SE (n=3). Replicate experiment 2.

127

A B C 0.22 0.22 0.22

0.20 0.20 0.20

0.18 0.18 0.18

0.16 0.16 0.16

0.14 0.14 0.14 0.12 0.12 0.12 B728aB728a

405 nm/630 nm nm 405 nm/630 0.10 0.10 0.10 B728aB728a + + CAR 1 mM Carnitine 0.08 0.08 0.08 !cdhA

OD ΔcdhA 0.06 0.06 0.06 !cdhA + 1 mM Carnitine 0.04 0.04 0.04 ΔcdhA + CAR

0.02 0.02 0.02 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Time (h) Time (h) Time (h)

Figure A.7. Carnitine catabolism was not required for carnitine to function in osmoprotection. B728a and ΔcdhA were grown in MinAS with and without 1 mM L-carnitine (CAR) and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored using OD430 nm/630 nm in microtiter plates. Values are means ± SE (n=3). Replicate experiment 3.

128

A 9 A A B 8 B C C 7 D

WTB728a

Log (CFU/sample) (CFU/sample) Log 6 cdhAΔcdhA pNcdhAΔcdhA (pNcdhA) BTBT 5

Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure A.8. Carnitine utilization contributed to B728a colonization of the spermosphere during imbibition and radical emergence. Bean seeds were inoculated with B728a, ΔcdhA, ΔcdhA (pNcdhA), and BT. Population sizes were estimated at 0, 1, 2, and 3 days post-inoculation based on viable plate counts. Values are mean ± SEM (n=6). At each time point, values indicated by the same letter do not differ significantly (p<0.05 based on an ANOVA). Replicate experiment 2.

129

A A 9 B B C 8 C 7 C D

WTB728a

Log (CFU/sample) (CFU/sample) Log 6 cdhAΔcdhA pNcdhAΔcdhA (pNcdhA) BTBT 5

Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure A.9. Carnitine utilization contributed to B728a colonization of the spermosphere during imbibition and radical emergence. Bean seeds were inoculated with B728a, ΔcdhA, ΔcdhA (pNcdhA), and BT. Population sizes were estimated at 0, 1, 2, and 3 days post-inoculation based on viable plate counts. Values are mean ± SEM (n=6). At each time point, values indicated by the same letter do not differ significantly (p<0.05 based on an ANOVA). Replicate experiment 3.

130

A B

8.0 8.2

8.0 7.5 7.8

7.6 7.0 7.4 /g leaf) leaf) /g 6.5 7.2 cfu B728aB728a 7.0

Log( 6.0 cdhAΔcdhA 6.8 dhcABΔdhcAB B728aB728a 6.6 5.5 cdhAΔcdhA wet dry 6.4 dhcABΔdhcAB

5.0 6.2 0 10 20 30 40 50 60 70 80 0 12 24 36 48 60 Time (h) Time (h) Figure A.10. Carnitine does not contribute to colonization of leaves by B728a. B728a, ΔcdhA, and ΔdhcAB populations were monitored during (A) epiphytic colonization during 24 h water replete conditions followed by 48 h of drying stress; and (B) apoplastic colonization following vacuum infiltration. Populations were enumerated based on viable plate counts. Values are mean ± SEM (n=6). Replicate experiment 2.

A B

8.0 8.0

7.8 7.5 7.6

7.0 7.4

7.2 /g leaf) leaf) /g 6.5

cfu 7.0 B728aB728a

Log( 6.0 cdhAΔcdhA 6.8 dhcABΔdhcAB B728a 6.6 B728a 5.5 cdhAΔcdhA wet dry 6.4 dhcABΔdhcAB

5.0 6.2 0 10 20 30 40 50 60 70 80 0 12 24 36 48 60

Time (h) Time (h)

Figure A.11. Carnitine does not contribute to colonization of leaves by B728a. B728a, ΔcdhA, and ΔdhcAB populations were monitored during (A) epiphytic colonization during 24 h water replete conditions followed by 48 h of drying stress; and (B) apoplastic colonization following vacuum infiltration. Populations were enumerated based on viable plate counts. Values are mean ± SEM (n=6). Replicate experiment 3.

131

APPENDIX B. SUPPLEMENTARY TABLES AND FIGURES FOR CHAPTER III

Table B.1 List of primers for construction of deletion mutants and analysis of operon structure.

Primer Sequence * Primers for constructing mutants F1-0029 5'-AAATCATCCTTCCCAGCGGTACGA-3' R1-0029 5'-ATCCATTGTGAAAAGTTATGCGAAAGTGCC-3' F2-0029 5'-TTTTCCGCAACCTTGACCCTGCC-3' R2-0029 5'-ACTTAGCGTCGAGGCCTACAACTTCGC-3' F1-0027 5'-AGTCAGCAGTACATGCGCAACCA-3' R1-0027knca 5'-AGCCTACACAATCGCTCAAGACGTACTGTTCTCCAGAGTT-3' F2-0027knca 5'-AATATCCGGGTAGGCGCAATCACTCAGAAACGGTCTGAC -3' R2-0027 5'-GTAACTCAATGCCGCCCGGGT -3' F1-0026 5'-ATTTGCTGACAAGTTCGACAAGAA-3' R1-0026knca 5'-AGCCTACACAATCGCTCAAGACGTATGTTGGATGGGTCACT -3' F2-0026knca 5'-AATATCCGGGTAGGCGCAATCACTTTGAGTTGTAACTCA -3' R2-0026 5'-ATGGACCGGTACGTCGTCTT-3' Primers for RT-PCR RT29F 5'-TGACCAGTGACATCTACCCG-3' A RT29R 5'-AACGTCCGTCCATTGGCAAA-3' B RT28F 5'-AGTGACATTGCGGCCACCAA-3' C RT28R 5'-GGAACACCACAAAGGCCTTT-3' D RT27F 5'-CACTTATACCCCTCTCGCTG-3' E RT27R 5'-GTGGCGACGTTTGTATCTGA-3' F RT26F 5'-TCTGTACGCTGCGATTGTTC-3' G RT26R 5'-TTCCTGATTGGCATTGAGCA-3' H

* Letters denote primers used for RT-PCR analysis of operon structure in Chapter 3 Figure 2.

132

A B C 0.24 0.24 0.24 0.22 0.22 0.22 0.20 0.20 0.20 0.18 0.18 0.18 0.16 0.16 0.16 0.14 0.14 0.14 0.12 0.12 0.12 405nm/630nm 405nm/630nm 0.10 0.10 0.10 0.08 0.08 0.08 OD 0.06 0.06 0.06 0.04 0.04 0.04 0.02 0.02 0.02 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

MinASB728a MinASB728a MinASB728a Choline-cdhA O-sulfate Choline-cdhA O-sulfate Choline-cdhA O-sulfate NodhcAB carbon NodhcAB nitrogen NodhcAB sulfur Figure B.1. COS can function as a sole carbon, nitrogen, and sulfur source. Growth of B728a in (A) MinAS, MinA-C with COS, and MinA-C; (B) MinAS, MinA-N with COS, and MinA-N; (C) MinAS, MinA-S with COS, and MinA-S. Values are mean ± SEM (n=3). Replicate experiment 2.

A B C 1.0 1.0 1.0

0.8 x" 0.8 x" 0.8 x"

0.6 0.6 0.6 600 nm 600 nm

0.4 0.4 0.4 OD

0.2 0.2 0.2

0.0 0.0 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (h)

MinASB728a MinASB728a MinASB728a Choline-cdhA O-sulfate Choline-cdhA O-sulfate Choline-cdhA O-sulfate NodhcAB carbon NodhcAB nitrogen NodhcAB sulfur

Figure B.2. COS can function as a sole carbon, nitrogen, and sulfur source. Growth of B728a in (A) MinAS, MinA-C with COS, and MinA-C; (B) MinAS, MinA-N with COS, and MinA-N; (C) MinAS, MinA-S with COS, and MinA-S. Values are mean ± SEM (n=3). Replicate experiment 3.

133

A B C

1.0 1.0 1.0

0.8 0.8 0.8 B728aTime vs B782a MinA-C 0.6 0.6 0.6

600 nm 600 nm 0.4 B728aTime vs + B728a 10 mM MinA-C COS + COS 0.4 0.4 0.2 ΔTimecosC vs betC MinA-C OD 0.2 0.2 0.0 ΔTimecosC vs + betC 10 mM MinA-C COS + COS

0.0 -0.2 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (h)

Figure B.3. Deletion of cosC blocked growth on COS as a sole carbon, nitrogen, and sulfur source. Comparison between B728a and ΔcosC in (A) MinA-C with COS and MinA-C; (B) MinA-N with COS and MinA-N; (C) MinA-S with COS and MinA-S. Values are mean ± SE (n=3). Replicate experiment 2.

A B C 0.24 0.24 0.24

0.22 0.22 0.22

0.20 0.20 0.20 1 mM Carnitine mM 1 + cdhA

0.18 0.18 0.18 !

0.16 0.16 0.16 B728a

cdhA 0.14 0.14 0.14 !

0.12 0.12 0.12 B728a + 10 mM COS 405nm/630nm 405nm/630nm 0.10 0.10 0.10 Carnitine mM 1 + B728a

0.08 0.08 0.08 ΔcosC 0.06 0.06 0.06 B728a OD 0.04 0.04 0.04 ΔcosC + 10 mM COS 0.02 0.02 0.02 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure B.4. Deletion of cosC blocked growth on COS as a sole carbon, nitrogen, and sulfur source. Comparison between B728a and ΔcosC in (A) MinA-C with COS and MinA-C; (B) MinA-N with COS and MinA-N; (C) MinA-S with COS and MinA-S. Values are mean ± SE (n=3). Replicate experiment 3.

134

A B C 1.0 1.0 1.0

0.8 0.8 0.8 B728a 0.6 0.6 0.6 Time vs B782a MinA-C

600 nm 600 nm B728a + 10 mM COS 0.4 0.4 0.4 Time vs B728a MinA-C + COS

OD ΔcosE 0.2 0.2 0.2 Time vs betC MinA-C ΔTimecosE vs+ 10 betC mM MinA-C COS + COS 0.0 0.0 0.0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Time (h)

Figure B.5. Deletion of cosE blocked growth on COS as a sole carbon, nitrogen, and sulfur source. Comparison between B728a and ΔcosE in (A) MinA-C with COS and MinA-C; (B) MinA-N with COS and MinA-N; (C) MinA-S with COS and MinA-S. Values are mean ± SE (n=3). Replicate experiment 2.

A B C 0.24 0.24 0.24 0.22 0.22 0.22 0.20 0.20 0.20 0.18 0.18 0.18 B728a 0.16 0.16 0.16 Time vs B782a MinA-C 0.14 0.14 0.14 0.12 0.12 0.12 B728aTime vs+ 10B728a mM COSMinA-C + COS 0.10 0.10 0.10 405nm/630nm 405nm/630nm 0.08 0.08 0.08 ΔTimecosE vs betC MinA-C 0.06 0.06 0.06

OD 0.04 0.04 0.04 ΔTimecosE vs+ 10 betC mM MinA-C COS + COS 0.02 0.02 0.02 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure B.6. Deletion of cosE blocked growth on COS as a sole carbon, nitrogen, and sulfur source. Comparison between B728a and ΔcosE in (A) MinA-C with COS and MinA-C; (B) MinA-N with COS and MinA-N; (C) MinA-S with COS and MinA-S. Values are mean ± SE (n=3). Replicate experiment 3.

135

A B C 1.0 1.0 1.0

0.8 0.8 0.8

0.6 0.6 0.6

405nm/630nm 405nm/630nm 0.4 0.4 0.4 OD 0.2 0.2 0.2

0.0 0.0 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure B.7. COS can function as an osmoprotectant in B728a. B728a grown in MinAS (black line) and MinAS with 1 mM COS (red line) added with hyperosmotic stress imposed (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on optical density (OD) ratio 405 nm/ 630 nm in microtiter plates. Values are mean ± SE (n=3). Replicate experiment 2.

A B C 0.24 0.24 0.24

0.22 0.22 0.22

0.20 0.20 0.20

0.18 0.18 0.18

0.16 0.16 0.16

0.14 0.14 0.14

405nm/630nm 405nm/630nm 0.12 0.12 0.12

0.10 0.10 0.10 OD 0.08 0.08 0.08

0.06 0.06 0.06

0.04 0.04 0.04 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 Time (h)

Figure B.8. COS can function as an osmoprotectant in B728a. B728a grown in MinAS (black line) and MinAS with 1 mM COS (red line) added with hyperosmotic stress imposed (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD405 nm/ 630 nm in microtiter plates. Values are mean ± SE (n=3). Replicate experiment 3.

136

A B C 0.24 0.24 0.24

0.22 0.22 0.22

0.20 0.20 0.20

0.18 0.18 0.18

0.16 0.16 1 mM Carnitine mM 1 0.16 + cdhA

0.14 0.14 0.14 ! B728a

cdhA

0.12 0.12 0.12 !

405nm/630nm 405nm/630nm B728a + 1 mM COS 0.10 0.10 0.10 Carnitine mM 1 + B728a

OD 0.08 0.08 0.08 ΔcosC 0.06 0.06 0.06 B728a ΔcosC + 1 mM COS 0.04 0.04 0.04 0 10 20 30 0 10 20 30 0 10 20 30 Time (h)

Figure B.9. COS catabolism was required for COS to function in osmoprotection. B728a and ΔcosC were grown in MinAS with and without 1 mM COS and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD405 nm/ 630 nm in microtiter plates. Values are mean ± SE (n=3). Replicate experiment 2.

A B C 1.0 1.0 1.0

0.8 0.8 0.8

0.6 0.6 0.6 B728a

600 nm 600 nm Time vs B728a MinAS 0M 0.4 0.4 0.4 B728aTime vs B728a + MinAS1 mM 0M +COSCOS OD ΔcosC 0.2 0.2 0.2 Time vs betC MinAS 0M ΔTimecosC vs betC + MinAS 1 mM 0M +COS COS

0.0 0.0 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure B.10. COS catabolism was required for COS to function in osmoprotection. B728a and ΔcosC were grown in MinAS with and without 1 mM COS and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD600 nm in microtiter plates. Values are mean ± SE (n=3). Replicate experiment 3.

137

A B C 0.24 0.24 0.24

0.22 0.22 0.22

0.20 0.20 0.20

0.18 0.18 0.18

0.16 0.16 0.16

1 mM Carnitine mM 1 + cdhA

0.14 0.14 0.14 ! B728a

cdhA 405nm/630nm 405nm/630nm 0.12 0.12 0.12 !

B728a + 1 mM COS

0.10 0.10 0.10 OD Carnitine mM 1 + B728a

0.08 0.08 0.08 ΔcosE 0.06 0.06 0.06 B728a ΔcosE + 1 mM COS 0.04 0.04 0.04 0 10 20 30 0 10 20 30 0 10 20 30 Time (h)

Figure B.11. COS catabolism was required for COS to function in osmoprotection. B728a and ΔcosE were grown in MinAS with and without 1 mM COS and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD405 nm/ 630 nm in microtiter plates. Values are mean ± SE (n=3). Replicate experiment 2.

A B C 1.0 1.0 1.0

0.8 0.8 0.8

0.6 0.6 0.6 B728aTime vs B728a MinAS 0M 0.4 600 nm 600 nm B728aTime vs B728a + MinAS1 mM 0M +COSCOS 0.4 0.4 ΔTimecosE vs betC MinAS 0M OD 0.2 ΔTimecosE vs betC + MinAS 1 mM 0M +COS COS 0.2 0.2 0.0

-0.2 0.0 0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h)

Figure B.12. COS catabolism was required for COS to function in osmoprotection. B728a and ΔcosE were grown in MinAS with and without 1 mM COS and with hyperosmotic stress imposed by (A) 0 M NaCl, (B) 0.3 M NaCl, and (C) 0.6 M NaCl. Growth was monitored based on OD600 nm in microtiter plates. Values are mean ± SE (n=3). Replicate experiment 3.

138

A B 8.0 10

7.5 9

7.0 8

6.5 7 6.0 /g leaf tissue) leaf /g /g leaf tissue) leaf /g 6 cfu 5.5 cfu B728a 5 5.0 Log ( Log ΔcosC ( Log B728a 4.5 wet dry 4 ΔcosC 4.0 3 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h) Time (h)

Figure B.13. COS does not contribute to colonization of leaves by B728a. B728a and Δ cosC populations were monitored during (A) epiphytic colonization during 24 h water replete conditions followed by 48 h of drying stress, and (B) apoplastic colonization following vacuum infiltration. Populations were enumerated based on viable plate counts. Values are mean ± SEM (n=6). Replicate experiment 2.

A B 8 10

9 7

8 6

7 /g leaf tissue) leaf /g tissue) leaf /g 5 cfu B728a cfu 6 ΔcosC Log ( Log ( Log 4 wet dry 5 B728a ΔcosC 3 4 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Time (h) Time (h)

Figure B.14. COS does not contribute to colonization of leaves by B728a. B728a and Δ cosC populations were monitored during (A) epiphytic colonization during 24 h water replete conditions followed by 48 h of drying stress, and (B) apoplastic colonization following vacuum infiltration. Populations were enumerated based on viable plate counts. Values are mean ± SEM (n=6). Replicate experiment 3.

139

A B

10 5.5

5.0 9

4.5 8

4.0 /sample) 7 /sample) cfu cfu 3.5

6 3.0 Log ( Log Log ( Log

5 B728a 2.5 B728a ΔcosC ΔcosC 4 2.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 1 2 3 4 Time (h) Time (h)

Figure B.15. COS did not contribute to B728a colonization of the spermosphere. Bean seeds were inoculated with B728a, and ΔcosC, and planted in (A) sand under laboratory conditions, or (B) agricultural soil under field conditions. Population sizes were estimated at 0, 1, 2, and 3 days post-inoculation based on viable plate counts. Values are mean ± SEM (n=6). Replicate experiment 2.

/sample) 7 cfu

Log ( Log B728a 5 ΔcosC

4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (h)

Figure B.16. COS did not contribute to B728a colonization of the spermosphere. Bean seeds were inoculated with B728a, and ΔcosC, and planted in sand under laboratory conditions. Population sizes were estimated at 0, 1, 2, and 3 days post-inoculation based on viable plate counts. Values are mean ± SEM (n=6). Replicate experiment 3.

140

APPENDIX C. SUPPLEMENTARY FIGURES FOR CHAPTER IV A 7.5

7.0 TimeB728a vs B728a 6.5 TimeΔcosC vs CosC /g leaf) leaf) /g TimeΔpchP vs pchP cfu 6.0 TimeΔplcA vs plcA

Log( TimeΔcdhA vs cdhA 5.5 Time vs gbcAB Wet Dry ΔgbcAB TimeΔbetA vs betA 5.0 0 12 24 36 48 60 72 84 TimeBT vs BT Time (h) B C 7.5 7.5

7.0 7.0

6.5 6.5 /g leaf) leaf) /g /g leaf) leaf) /g cfu cfu 6.0 6.0 Log( Log( 5.5 5.5 Wet Dry Wet Dry 5.0 5.0 0 12 24 36 48 60 72 84 0 12 24 36 48 60 72 84 Time (h) Time (h) D 7.5

7.0

6.5 /g leaf) leaf) /g cfu 6.0 Log(

5.5 Wet Dry

5.0 0 12 24 36 48 60 72 84 Time (h)

Figure C.1. Choline and choline derivatives contribute to osmotolerance during epiphytic colonization. (A) B728a, ΔcosC, ΔpchP, ΔplcA, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during epiphytic colonization under water replete conditions followed by drying stress of 45% RH at 25°C. Data replotted to display growth dynamics between (B) B728a, ΔbetA, ΔgbcAB, and BT to demonstrate influence of compatible solute accumulation on epiphytic colonization; B728a, ΔplcA, ΔpchP, ΔbetA, and BT to demonstrate influence of phosphatidylcholine, phosphocholine, and choline on epiphytic colonization; (D) B728a, ΔcosC, ΔcdhA, and BT to demonstrate not all QACs influence epiphytic colonization. Reported as mean ± Standard Error of the Mean (SEM). Replicate experiment 1.

141

A B

7.5 7.5 A A AB AB AB 7.0 A 7.0 AB A B A A B A AB

/g leaf) leaf) /g AB 6.5 B 6.5 cfu Log ( Log

6.0 6.0

5.5 5.5

BT BT BT BT plcA betA plcA betA cosC cdhA cdhA cosC pchP pchP Δ Δ Δ Δ B728a B728a gbcAB gbcAB gbcAB gbcAB Δ B728a B728a Δ Δ Δ Δ Δ Δ Δ

Figure C.2. Statistical comparisons of epiphytic populations levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05). Replicate experiment 1.

142

A 8.5

8.0

7.5 TimeB728a vs B728a 7.0 TimeΔcosC vs CosC /g leaf) leaf) /g 6.5 TimeΔpchP vs pchP cfu 6.0 TimeΔplcA vs plcA

Log( 5.5 TimeΔcdhA vs cdhA TimeΔgbcAB vs gbcAB 5.0 Wet Dry TimeΔbetA vs betA 4.5 0 12 24 36 48 60 72 84 TimeBT vs BT Time (h) B C 8.5 8.5

8.0 8.0

7.5 7.5

7.0 7.0 /g leaf) leaf) /g /g leaf) leaf) /g 6.5 6.5 cfu cfu 6.0 6.0 Log( Log( 5.5 5.5

5.0 Wet Dry 5.0 Wet Dry

4.5 4.5 0 12 24 36 48 60 72 84 0 12 24 36 48 60 72 84 Time (h) Time (h)

8.5 D

8.0

7.5

7.0 /g leaf) leaf) /g

6.5 cfu

6.0 Log( 5.5 5.0 Wet Dry 4.5 0 12 24 36 48 60 72 84 Time (h) Figure C.3. Choline and choline derivatives contribute to osmotolerance during epiphytic colonization. (A) B728a, ΔcosC, ΔpchP, ΔplcA, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during epiphytic colonization under water replete conditions followed by drying stress of 45% RH at 25°C. Data replotted to display growth dynamics between (b) B728a ΔbetA, ΔgbcAB, and BT to demonstrate influence of compatible solute accumulation on epiphytic colonization; (C) B728a ΔplcA, ΔpchP, ΔbetA, and BT to demonstrate influence of phosphatidylcholine, phosphocholine, and choline on epiphytic colonization; (D) B728a, ΔcosC, ΔcdhA, and BT to demonstrate not all QACs influence epiphytic colonization. Reported as mean ± SEM. Replicate experiment 2.

143

A B

7.5 7.5 A AB BC A CD D 7.0 AB 7.0 DE BC BC EF BC CD F D

/g leaf) leaf) /g E 6.5 6.5 cfu Log ( Log

6.0 6.0

5.5 5.5

BT BT BT BT betA plcA betA plcA cosC cdhA cosC pchP cdhA pchP Δ Δ Δ Δ gbcAB gbcAB B728a B728a Δ Δ Δ B728a B728a gbcAB gbcAB Δ Δ Δ Δ Δ

Figure C.4. Statistical comparisons of epiphytic populations levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05). Replicate experiment 2.

144

A 7.5

7.0

6.5 TimeB728a vs B728a TimeΔcosC vs CosC /g leaf) leaf) /g 6.0 TimeΔpchP vs pchP cfu TimeΔplcA vs plcA 5.5

Log( TimeΔcdhA vs cdhA 5.0 Time vs gbcAB Wet Dry ΔgbcAB TimeΔbetA vs betA 4.5 0 12 24 36 48 60 72 84 TimeBT vs BT Time (h) B C 7.5 7.5

7.0 7.0

6.5 6.5 /g leaf) leaf) /g /g leaf) leaf) /g 6.0 6.0 cfu cfu

5.5 5.5 Log( Log(

5.0 5.0 Wet Dry Wet Dry

4.5 4.5 0 12 24 36 48 60 72 84 0 12 24 36 48 60 72 84 Time (h) Time (h)

7.5 D

7.0

6.5 /g leaf) leaf) /g 6.0 cfu

5.5 Log(

5.0 Wet Dry 4.5 0 12 24 36 48 60 72 84 Time (h)

Figure C.5. Choline and choline derivatives contribute to osmotolerance during epiphytic colonization. (A) B728a, ΔcosC, ΔpchP, ΔplcA, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during epiphytic colonization under water replete conditions followed by drying stress of 45% RH at 25°C. Data is replotted to display growth dynamics between (B) B728a ΔbetA, ΔgbcAB, and BT to demonstrate influence of compatible solute accumulation on epiphytic colonization; (C) B728a ΔplcA, ΔpchP, ΔbetA, and BT to demonstrate influence of phosphatidylcholine, phosphocholine, and choline on epiphytic colonization; (D) B728a, ΔcosC, ΔcdhA, and BT to demonstrate not all QACs influence epiphytic colonization. Reported as mean ± SEM. Replicate experiment 3.

145

A B

7.5 7.5

7.0 7.0

/g leaf) leaf) /g A A 6.5 6.5 AB cfu AB AB A B AB AB

Log ( Log B C B B B C 6.0 6.0

5.5 5.5

BT BT BT BT betA plcA betA plcA cosC cdhA cosC pchP pchP cdhA Δ Δ Δ Δ gbcAB gbcAB B728a B728a Δ Δ Δ B728a B728a gbcAB gbcAB Δ Δ Δ Δ Δ

Figure C.6. Statistical comparisons of epiphytic populations levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05). Replicate experiment 3.

146

10 Time vs B728a 9 B728a ΔTimeggnΔ trevs tre- 8 ΔTimecosC vs cosC ΔTimepchP vs pchP 7 /g tissue) tissue) /g ΔTimecdhA vs cdhA cfu 6 ΔTimebetA vs betA ΔTimegbcAB vs gbcAB Log( 5 BTTime vs BT Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

B C

10 10

9 9

8 8

/g tissue) tissue) /g 7 tissue) /g 7 cfu cfu 6 6 Log( Log( 5 5

Stage 1 Stage 2 Stage 3 Stage 4 Stage 1 Stage 2 Stage 3 Stage 4 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days) Time (days) D

7 /g tissue) tissue) /g cfu 6 Log( 5

Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure C.7. QAC catabolism contributes to colonization of germinating bean seeds. (A) B728a, ΔcosC, ΔggnΔtre, ΔpchP, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during colonization of germinating bean seeds in sterile sand under 24 h light and 25°C. Data replotted to display growth dynamics between (B) B728a, ΔbetA, ΔcdhA, ΔgbcAB, and BT to demonstrate relative impact of carnitine and choline (C) B728a, ΔgbcAB, ΔnggΔtre, and BT to demonstrate impact of compatible solute accumulation; (D) B728a, ΔbetA, ΔcdhA, ΔpchP, and BT to demonstrate the relative impact of QACs, choline, carnitine, and phosphorylcholine, detected in seed exudates. Stages correspond to seed developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. Reported as mean ± SEM. Replicate experiment 1.

147

A B 10.0 10.0

9.5 9.5 A A AB AB AB 9.0 AB 9.0 AB AB 8.5 B 8.5

/g tissue) tissue) /g B B B B cfu 8.0 8.0 B BC

Log ( Log C 7.5 7.5

7.0 7.0

BT BT BT BT ggn ggn betA betA cosC cosC pchP cdhA pchP cdhA Δ Δ Δ Δ B728a B728a B728a B728a Δ Δ gbcAB gbcAB Δ Δ Δ Δ tre tre Δ Δ Δ Δ

Figure C.8. Statistical comparisons of spermosphere population levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05). Replicate experiment 1.

148

10 Time vs B728a 9 B728a ΔTimeggnΔ trevs tre- 8 ΔTimecosC vs cosC ΔTimepchP vs pchP /g tissue) tissue) /g 7 ΔTimecdhA vs cdhA cfu 6 ΔTimebetA vs betA

Log( ΔTimegbcAB vs gbcAB 5 BTTime vs BT Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

B C

10 10

9 9

8 8

/g tissue) tissue) /g 7 tissue) /g 7 cfu cfu 6 6 Log( Log( 5 5

Stage 1 Stage 2 Stage 3 Stage 4 Stage 1 Stage 2 Stage 3 Stage 4 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days) Time (days) D

/g tissue) tissue) /g 7 cfu 6 Log( 5

Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure C.9. QAC catabolism contributes to colonization of germinating bean seeds. (A) B728a, ΔcosC, ΔggnΔtre, ΔpchP, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during colonization of germinating bean seeds in sterile sand under 24 h light and 25°C. Data replotted to display growth dynamics between (B) B728a, ΔbetA, ΔcdhA, ΔgbcAB, and BT to demonstrate relative impact of carnitine and choline (C) B728a, ΔgbcAB, ΔnggΔtre, and BT to demonstrate impact of compatible solute accumulation; (D) B728a, ΔbetA, ΔcdhA, ΔpchP, and BT to demonstrate the relative impact of QACs, choline, carnitine, and phosphorylcholine, detected in seed exudates. Stages correspond to seed developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. Reported as mean ± SEM. Replicate experiment 2.

149

A B 10.0 10.0

9.5 9.5 AB A AB A AB AB 9.0 AB 9.0 AB B 8.5 B 8.5 /g tissue) tissue) /g B

cfu C BC 8.0 C 8.0 C C Log ( Log

7.5 7.5

7.0 7.0

BT BT ggn BT BT betA ggn cosC pchP cdhA Δ betA Δ cosC cdhA pchP B728a B728a Δ Δ gbcAB Δ Δ Δ B728a B728a Δ gbcAB tre Δ Δ Δ tre Δ Δ Δ

Figure C.10. Statistical comparisons of spermosphere population levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05). Replicate experiment 2.

150

10 Time vs B728a 9 B728a ΔTimeggnΔ trevs tre- 8 ΔTimecosC vs cosC ΔTimepchP vs pchP /g tissue) tissue) /g 7 ΔTimecdhA vs cdhA cfu 6 ΔTimebetA vs betA

Log( ΔTimegbcAB vs gbcAB 5 BTTime vs BT Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

B C

10 10

9 9

8 8

/g tissue) tissue) /g 7 tissue) /g 7 cfu cfu 6 6 Log( Log( 5 5 Stage 1 Stage 2 Stage 3 Stage 4 Stage 1 Stage 2 Stage 3 Stage 4 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days) Time (days) D

7 /g tissue) tissue) /g

cfu 6

Log( 5 Stage 1 Stage 2 Stage 3 Stage 4 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time (days)

Figure C.11. QAC catabolism contributes to colonization of germinating bean seeds. (A) B728a, ΔcosC, ΔggnΔtre, ΔpchP, ΔcdhA, ΔgbcAB, ΔbetA, and BT populations were tracked during colonization of germinating bean seeds in sterile sand under 24 h light and 25°C. Data replotted to display growth dynamics between (B) B728a, ΔbetA, ΔcdhA, ΔgbcAB, and BT to demonstrate relative impact of carnitine and choline (C) B728a, ΔgbcAB, ΔnggΔtre, and BT to demonstrate impact of compatible solute accumulation; (D) B728a, ΔbetA, ΔcdhA, ΔpchP, and BT to demonstrate the relative impact of QACs, choline, carnitine, and phosphorylcholine, detected in seed exudates. Stages correspond to seed developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. Reported as mean ± SEM. Replicate experiment 3.

151

A B 10.0 10.0

9.5 9.5 A A

A AB 9.0 A 9.0 AB A AB

8.5 8.5 B B /g tissue) tissue) /g B BC cfu 8.0 C 8.0 C C C Log ( Log 7.5 7.5

7.0 7.0

BT BT BT BT ggn ggn betA betA cosC cosC pchP cdhA pchP cdhA Δ Δ Δ Δ B728a B728a Δ B728a B728a Δ gbcAB gbcAB Δ Δ Δ Δ tre tre Δ Δ Δ Δ

Figure C.12. Statistical comparisons of spermosphere population levels at (A) 48 hpi and (B) 72 hpi. Letters denote statistical groupings based on repeated measures analysis. Populations are compared using a repeated measures analysis. Statistical groups signify significant difference (p>0.05). Replicate experiment 3.

152

A Stage 1 Stage 2 Stage 3 Stage 4

B 6.5

6.0

5.5

5.0

4.5 Log (abundance) (abundance) Log

4.0

3.5 0 1 2 3 4 5 Germination Stage

Figure C.13. Bean seeds release the QACs choline, carnitine, and phosphorylcholine throughout germination. (A) Germinating seeds were placed into one of four developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. (B) Bean seeds were germinated to developmental stages 1-4, exudates were removed and analyzed via HLPC- MS for abundance of choline (m/z 104.1, black line), carnitine (m/z 162.1, red line), and phosphorylcholine (m/z 184.1, green line). Values are mean ± SEM. Replicate experiment 1.

153

A Stage 1 Stage 2 Stage 3 Stage 4

B 6.5

6.0

5.5

5.0

4.5

4.0

Log (abundance) (abundance) Log 3.5

3.0

2.5 0 1 2 3 4 5 Germination Stage

Figure C.14. Bean seeds release the QACs choline, carnitine, and phosphorylcholine throughout germination. (A) Germinating seeds were placed into one of four developmental stages: stage 1, imbibition; stage 2, radical emergence; stage 3, radical elongation; and stage 4, lateral root formation/greening. (B) Bean seeds were germinated to developmental stages 1-4, exudates were removed and analyzed via HLPC- MS for abundance of choline (m/z 104.1, black line), carnitine (m/z 162.1, red line), and phosphorylcholine (m/z 184.1, green line). Values are mean ± SEM. Replicate experiment 2.