An Integrative Approach to Understanding Pyoluteorin Biosynthesis in the Biological Control Bacterium Pseudomonas Fluorescens Pf-5

AN ABSTRACT OF THE DISSERTATION OF

Brian Nowak-Thompson for the degree of Doctor of Philosophy in Biochemistry and Biophysics presented on March 14,1997. Title: An Integrative Approach to Understanding Pyoluteorin Biosynthesis in the Biological Control Bacterium Pseudomonas fluorescens Pf-5.

Redacted for Privacy Abstract Approved:

Pyoluteorin is a chlorinated, mixed polyketide metabolite produced by Pseudomonas fluorescens Pf-5. Pyoluteorin exhibits antifungal activity against Pythium ultimum, the causal agent of preemergence damping-off disease. Biosynthesis of this compound was studied using an approach that integrated nucleotide sequence analysis of a region required for pyoluteorin production and isotopic labelling studies of proposed pathway intermediates. ['COOK- L-Proline was demonstrated to be the primary precursor of the chlorinated pyrrole moiety within pyoluteorin. Hypothetical pathway intermediates were synthesized and tested in vivo for their incorporation into pyoluteorin. The results of these studies, however, were equivocal and definitive conclusions regarding the biochemical pathway leading to pyoluteorin could not be made. Nucleotide sequencing of the region required for pyoluteorin production identified several clustered biosynthetic genes. pltB and pltC encode a Type I polyketide synthase required for pyoluteorin biosynthesis. A pathway for resorcinol ring formation is proposed based on the catalytic domain organization of the polyketide synthase. The deduced peptide sequences of pltA, pltD, and a partial open reading frame, orf-S, exhibit similarity to the halogenases that are involved in tetracycline and pyrrolnitrin biosynthesis. Although halogenases are a poorly characterized class of enzymes, some aspects of their catalytic mechanism regarding pyoluteorin biosynthesis are discussed based on peptide sequence analysis. The regulatory gene pltR was also identified and appears to encode a transcriptional activator protein with similarity to the LysR family of transcriptional regulators. An Integrative Approach to Understanding Pyoluteorin Biosynthesis in the Biological Control Bacterium Pseudoinonas fluorescens Pf-5.

by Brian Nowak-Thompson

A DISSERTATION

submitted to

Oregon State University

in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Complete March 14, 1997 Commencement June 1997 Doctor of Philosophy dissertation of Brian Nowak-Thompson presented on March 14, 1997

APPROVED: Redacted for Privacy

Major Professor, representing Biochemistry and Biophysics Redacted for Privacy Head of Department of Biochemistry and Biophysics

Redacted for Privacy

Dean of GradweXe School

I understand that my thesis will become part of the permanent collection of Oregon State University libraries. My signature below authorizes release of my thesis to any reader upon request. Redacted for Privacy Brian Nowak-Thompson ACKNOWLEDGMENTS

For their non-scientific contributions: I gratefully acknowledge the undying patience and loving support of my life-long companion (who is, fortunately, also my wife), Jo. Her willingness to endure my lengthy stay in graduate school is not to be forgotten. Good baby Everett was also a major motivating force behind the completion of this writing and is the newest person I will probably ever see. I also thank Dr. Glenn Tinseth for his insightful discussions on brewing and fermentations and damn him for finishing first!

For their scientific contributions: I would like to express my gratitude to Nathan Corbell, Nancy Chaney, Marcella Henkles, Cheryl Whistler, Dr. Mary Hagen, and Dr. Virginia Stockwell for their research contributions that have provided the biological context that is such an integral part of this project. I especially thank Dr. Jennifer Kraus for her initial work that has made this project so successful. Dr. Chris Melville is also recognized for initiating many stimulating discussions concerning chemical synthetic methods. Finally, I am forever indebted to Drs. Joyce Loper and Steven Gould who have taught me as much about being a scientist as they have about science. Their confidence and guidance have been very much appreciated. Dr. Loper is gratefully recognized for fostering the collaborations I have taken part in and the many professional contacts I have had the opportunity to make during my tenure. TABLE OF CONTENTS

rag Chapter 1. Biological Control of Plant Pathogens: A Paradigm for Understanding Bacterial Physiology and the Ecological Role of Secondary Metabolites 1

1.1 Introduction 1

1.2 Antibiosis as a mechanism in biological control 2

1.3 Statement of research objectives 9

Chapter 2. Biochemical Analysis of the Pyoluteorin Pathway 11

2.1 Introduction 11

2.1.1 Rudimentary polyketide pathways: the tetraketides 11 2.1.2 Unusual polyketide starter units 15 2.1.3 Post-assembly modifications of polyketide metabolites 20

2.2 Biosynthetic Analysis of Pyoluteorin 22

2.3 Results and Discussion 25 2.3.1 Preliminary studies: pyoluteorin production and isolation 25 2.3.2 Incorporation of proline into pyoluteorin 26 2.3.3 Involvement of deschloropyoluteorin 28 2.3.4 Involvement of pyrrole-2-carboxylic acid 32 2.3.5 Involvement of tetrahydrodeschloropyoluteorin 34

2.4 Conclusion 37

2.5 Experimental methods 37

Chapter 3. Nucleotide Sequence Analysis of Loci Required for Pyoluteorin Biosynthesis 52

3.1 Introduction 52 TABLE OF CONTENTS (Continued)

Page

3.1.1 Genetic organization of polyketide synthases 52 3.1.2 Halogenating enzymes in microorganisms 55

3.2 Results and Discussion 56

3.2.1 Nucleotide sequencing of the pyoluteorin region 56 3.2.2 Identification and sequence analysis of the pyoluteorin PKS 58 3.2.3 Identification and sequence analysis of the pyoluteorin halogenases 67

3.3 Conclusions 70

3.4 Experimental methods 71

Chapter 4. Nucleotide Sequence Analysis of a Putative Transcriptional Activator Regulating Pyoluteorin Biosynthesis 73

4.1 Introduction 73

4.2 Results and Discussion 75

4.2.1 Nucleotide sequence analysis of pltR 75 4.2.2 Similarity of PltR to LysR-type transcriptional activators 76 4.2.3 A putative P1tR binding site within the pyoluteorin gene cluster 79 4.2.4 Models for PltR-dependent regulation of pyoluteorin biosynthesis 80

4.3 Conclusions 81

4.4 Experimental Methods 81

Chapter 5. Concluding Summary 83

5.1 Pyoluteorin biosynthesis 83

5.2 Regulation of pyoluteorin biosynthesis 85 TABLE OF CONTENTS (Continued)

Page

Bibliography 86

Appendices 110

Appendix 1. Production of 2,4-Diacetylphloroglucinol by the Biocontrol Agent Pseudomonas fluorescens Pf-5 111 Appendix 2. Nucleotide Sequence and Translated Product of p1tB 116 Appendix 3. Nucleotide Sequence and Translated Product of pltC 126 Appendix 4. Nucleotide Sequence and Translated Product of pltA 133 Appendix 5. Nucleotide Sequence and Translated Product of pltD 135 Appendix 6. Nucleotide Sequence and Translated Product of orfS 138 Appendix 7. Nucleotide Sequence and Translated Product of pltR 139 LIST OF FIGURES

Figure raga

1-1. Antibiotics with inhibitory activity against plant pathogens 4

3-1. Genetic organization and part of the reaction pathway catalyzed by a Type I PKS 53 3-2. Genetic organization and part of the reaction pathway catalyzed by a Type II PKS 54 3-3. Genomic region required for pyoluteorin production and the biosynthetic genes identified by nucleotide analysis 57 3-4. Modular organization of the functional domains identified in P1tB and P1tC 59 3-5. Deduced peptide sequence alignment for the three PKS modules encoded by pltB and pltC. 61 3-6. Proposed route for formation of the pyoluteorin resorcinol moiety. 66 3-7. The deduced peptide sequence alignments for the putative halogenases PltA, PltD, and Orf-S with PrnC and Cts4. 68 4-1. Genomic region required for pyoluteorin production and the regulatory gene identified by nucleotide analysis. 75

4-2. Gap alignment of P1tR and a LysR profile sequence 77

4-3. The putative promoter region of pltR 79 An Integrative Approach to Understanding Pyoluteorin Biosynthesis in the Biological Control Bacterium Pseudomonas fluorescens Pf-5.

Chapter 1. Biological Control of Plant Pathogens: A Paradigm for Understanding Bacterial Physiology and the Ecological Role of Secondary Metabolites

1.1 INTRODUCTION

Traditionally, research in plant pathology and plant physiology has focused on two types of microbial interactions, namely parasitism/predation and mutualism (symbiosis). Because both types of interactions generally involve a plant host, and since the respective disciplines often emphasize agricultural applications, the importance of microbe-microbe interactions within agricultural systems is often overlooked. Nevertheless, an astounding number of interactions, both direct and indirect, inevitably occur among the individual populations within any microbial community and can affect microbe-plant interactions. It is possible within certain agricultural systems to exploit the microbial interactions in order to biologically control plant pathogenic organisms that are not managed adequately by current control practices. Biological control is defined as the reduction in the amount of inoculum or the virulence of a pathogen accomplished by one or more organisms, excluding humans (Cook and Baker, 1983). Biological control of plant pathogens can be accomplished through modifying cultural practices to increase or maintain native antagonistic populations or by artificially introducing antagonists or hypovirulent strains of the pathogen either into the environment or onto the plant host. Therefore, successfully implementing a biological control program requires an understanding of not only the agonist- antagonist relationship, but also the environmental influences affecting the establishment of antagonist populations. Many ecologists advocate that homeostasis of very diverse communities is generally maintained via negative feedback, wherein an environmentally induced change in one population 2

elicits a response within other communal populations so as to reduce the severity of the initial environmental stimulus (Bull and Slater, 1982). Because environmentally induced changes within microbial populations are manifested as observable phenotypes, biological control systems serve as useful paradigms for physiological processes that occur in response to environmental conditions. As a result, biological control research provides a forum for understanding biochemical processes as well as the ecological mechanisms regulating these processes. Biological control organisms utilize a variety of mechanisms to inhibit their target pathogens. Competitive exclusion of a pathogen by a biocontrol antagonist results when the ecological niches occupied by both organisms overlap. The extent of ecological overlap may be nearly complete as in the competition between a virulent and a hypovirulent strain of a pathogen (Paulitz et al, 1987; MacDonald and Fulbright, 1991; Chen et al, 1994). Alternatively, two unrelated organisms may compete for limiting resources such as available iron (Loper and Buyer, 1991) or carbon sources (Nelson and Maloney, 1992; Pau litz, 1990). Other mechanisms that have also been implicated in biological control include parasitism (Adams, 1990), cross protection (Fulton, 1986), and antibiosis (Fravel, 1988). Details of biocontrol systems utilizing these mechanisms are described within the many comprehensive reviews that have been published (Weller, 1988; DeFago and Haas, 1990; Gould, 1990; Gutterson, 1990; Nelson and Maloney, 1992; O'Sullivan and O'Gara, 1992; Loper et al, 1994; Thomashow and Weller, 1995). It should be emphasized that in many systems, several biocontrol mechanisms concertedly contribute to the total biological control activity observed for a single antagonist.

1.2 ANTIBIOSIS AS A MECHANISM IN BIOLOGICAL CONTROL

Antibiotics are relatively small organic compounds produced by microorganisms via secondary metabolic pathways. Their production generally is not considered essential for survival of the producing organism, and the ecological importance of secondary metabolites has been debated for many years (Gottlieb, 1976; de Lorenzo and Aguilar, 1984; Williams and Vickers, 1986; Maplestone et al, 1992; Stone and Williams, 1992; Vining, 1992). Research 3 with biological control systems has demonstrated that, in some instances, antibiotics contribute to the ecological competence of the producing organism (Mazzola et al, 1992). Nevertheless, their production is not an absolute requirement for survival of bacteria in natural habitats. Historically, it has been difficult to establish a role for antibiotic production in the biological control of plant diseases. Although antibiotics produced by biocontrol agents in vitro clearly supressed growth of target pathogens, there was little evidence for in situ antibiotic production by biocontrol organisms inhabiting soil or plant surfaces. Recently, molecular genetic techniques have been used to construct mutants that are deficient in antibiotic production. Comparison of the biocontrol activities of the isogenic mutant and the parental strains has provided further correlations between antibiotic production and the biocontrol activity of a microorganism in natural habitats. Unfortunately, selection of the mutants generally has relied upon in vitro screening for pathogen inhibition without biochemical characterization. Mutations affecting antibiotic production are not limited to lesions within biosynthetic loci; antibiotic production is mediated by global regulatory loci as well as genes closely linked to the biosynthetic gene cluster. Furthermore, lesions within genes affecting antibiotic precursor biosynthesis can also result in a non- producing phenotype. Mutants selected for loss of antibiotic production commonly have pleiotropic phenotypes (Kraus and Loper, 1992; Laville et al, 1992; Haas et al, 1993; Lam et al, 1993; Willis et al, 1994). Therefore, demonstrating the dependence of biocontrol activity through a molecular genetic approach requires thoroughly characterized mutants with genetic lesions in structural genes for antibiotic biosynthesis. The following examples highlight several secondary metabolites (Figure 1-1) that are produced by organisms exhibiting biological control activity. Although some of these compounds have not been definitively established as contributing factors to biological control, they nevertheless serve to illustrate the importance of molecular genetics in biological control research as well as the requirement for rigorously characterizing the mutants used in these studies. 4

Agrocin 84. Agrocin 84 is a bacteriostatic antibiotic produced by the biological control organism Agrobacterium radiobacter K84 that inhibits DNA synthesis in A. tumefaciens, the causal agent of crown gall disease (Moore and Warren, 1979; Kerr, 1980; Farrand, 1990). Agrocin 84 production is a plasmid determined phenotype of A. radiobacter K84. Curing A. radiobacter of the plasmid required for agrocin 84 production significantly reduces the biocontrol activity of this bacterium, but non-producing strains partially suppress crown gall formation when introduced onto the plant host prior to infection by A. tumefaciens. Although agrocin 84 production is a major determinant of biocontrol activity, other factors also contribute to disease suppression. Increasing the agrocin 84 biosynthetic gene dosage, however, does not further improve biocontrol activity (Shim et al, 1987).

Agrocin 84 000H

N H Phenazine-1-carboxylic Pyrrolnitrin acid

o CH 0 OH

HD CH N a OH 0 H 2,4-Diacetylphloro glucinol Pyoluteorin

Figure 1-1. Antibiotics with inhibitory activity against plant pathogens. The antifungal metabolite oomycin A has been omitted because its molecular structure is unknown. 5

Gliovirin. Gliocladium virens is a soil fungus with a demonstrated ability to parasitize Rhizoctonia solani (Tu and Vaartaja, 1980; Tu, 1980; Howell, 1982). Although G. virens also is able to inhibit the plant pathogenic fungus Pythium ultimum, initial evidence suggested that a non-parasitic mechanism was responsible for this activity (Howell, 1982). Subsequently, the antifungal metabolite gliovirin was isolated and structurally characterized from G. virens culture extracts (Howell and Stipanovic, 1982). Gliovirin inhibits only oomycete fungi, such as P. ultimum and Phytopthora species (Stipanovic and Howell, 1983). Mutagenesis of G. virens has allowed the isolation of gliovirin non-producing and over-producing strains. Gliovirin non-producing mutants were significantly less effective at controlling preemergence damping-off in Pythium infested soils, but the over-producing mutants did not increase disease suppression as compared to the parental strain. Although it is tempting to conclude that gliovirin is a contributing factor to the biocontrol activity of G. virens against P. ultimum, these studies have not demonstrated in situ gliovirin production. Furthermore, the non-producing and over-producing mutants have not been thoroughly characterized to ascertain if the genetic lesions affect only gliovirin production. Therefore, further studies are necessary before the role of gliovirin in the biological control of P. ultimum can be unequivocally demonstrated.

Oomycin A. Pseudcmonas fluorescens Hv37a exhibits in vitro antifungal activity against P. ultimum (Gutterson et al, 1986) and produces several antifungal metabolites (James and Gutterson, 1986). Although none of the compounds produced by this strain have been structurally characterized, one metabolite, oomycin A, is the major contributing factor for biocontrol activity (Howie and Sus low, 1991). Genetic analysis within Hv37a has identified at least four unlinked loci required for oomycin A production (Gutterson et al, 1988; Gutterson, 1990). Several of these loci are responsible for induction of the biosynthetic genes by the oxidation of glucose via glucose dehydrogenase. The in situ expression of an oomycin A biosynthetic gene was demonstrated using a lacZ transcriptional fusion in both an oomycin A producing and non-producing strain (Howie and Sus low, 1991). Additional evidence for the ecological role of oomycin A production within Hv37a was obtained by placing 6

a constitutive tac promoter upstream of the oomycin A biosynthetic gene cluster (Howie et al, 1989; Gutterson, 1990). By introducing the unregulated loci into the Hv37a chromosome, an increase in oomycin A production correlated with a significant increase in biological control of P. ultimum as compared to the wild-type strain. Furthermore, transcriptional fusions of the tac-derived promoter to a lux reporter gene revealed an increase of biosynthetic gene expression in situ (Gutterson, 1990). Demonstrating biosynthetic gene expression in situ, however, does not necessarily establish antibiotic production because other regulatory mechanisms may ultimately control antibiotic biosynthesis.

Phenazine-1-carboxylic acid. Suppression of the phytopathogenic fungi Gaeumannomyces graminis var. tritici has been correlated to the production of phenazine-1-carboxylic acid by the biocontrol organisms Ps. fluorescens 2-79 (Thomashow and Weller, 1988) and Ps. aureofaciens 30-84 (Pierson and Thomashow, 1993). The production of phenazine-1-carboxylic acid is one of the few cases in which a direct correlation has been demonstrated between in situ antibiotic production and biological control activity (Thomashow et al, 1990). Phenazine-1-carboxylic acid biosynthesis also has a long term, positive influence on the ecological fitness of producing strains (Mazzola et al, 1992). Furthermore, the efficacy of Ps. fluorescens 2-79 as a biocontrol agent is a function of the target pathogen sensitivity to phenazine-l-carboxylic acid (Mazzola et al, 1995). Antibiotic production, therefore, is not the sole determinant of biocontrol activity since non-producing mutants partially suppress disease (Thomashow and Weller, 1988). The in situ production of phenazine-1-carboxylic acid is influenced by several edaphic factors such as the proportions of sand, organic matter, silt, and clay, soil pH, metal ions, and total carbon (Thomashow and Weller, 1995). Although these factors affect antibiotic production, it is likely that their influence is indirect, exerted through complex interactions between the host, pathogen, biocontrol agent, and soil. Phenazine-l-carboxylic acid production also is regulated directly in response to cell density (reviewed in Pierson and Pierson, 1996). Two loci, phzR and phzl, have been identified via nucleotide sequence analysis and are linked to genes encoding the biosynthetic enzymes 7

of the phenazine pathway (Pierson et al, 1995). PhzR and PhzI belong to the LuxR/LuxI family of cell-density dependent regulatory proteins and are required for antibiotic production in Ps. aureofaciens 30-84. PhzI shares sequence similarity to several N-acyl-L-homoserine lactone synthases. PhzR is presumably a transcriptional activator that induces phenazine-1-carboxylic acid production in response to the N-acyl-L-homoserine lactone derivative produced by PhzI. It is hypothesized that fungal infection is actually required for the Ps. aureofaciens population to reach the threshold density required for phenazine production as a result of an increase in root exudates released from root lesions.

Pyrrolnitrin. Pyrrolnitrin is produced by many Pseudomonas spp. (Leisinger and Margraff, 1979) and was identified initially as a contributing factor to biological control in Ps. fluorescens Pf-5 (Howell and Stipanovic, 1979). Pyrrolnitrin inhibits several phytopathogenic fungi; however, it is most effective in situ against Rhizoctonia solani. Although the severity of R. solani infection as measured by seedling emergence was reduced by treating seeds with purified pyrrolnitrin, greater protection was afforded when Pf-5 itself was used as a seed treatment (Howell and Stipanovic, 1979). These results suggest that other biocontrol mechanisms may contribute to the inhibition of R. solani by Pf-5. Further evidence for the role of pyrrolnitrin production in biological control has been obtained in studies of Ps. fluorescens BL915 (Ligon et al, 1996). A 6.2-kb genomic region within BL915 is required both for pyrrolnitrin production and for biological control activity (Hill et al, 1995). Nucleotide sequence analysis and pyrrolnitrin production in heterologous strains containing the cloned 6.2-kb region is compelling evidence for the existence of pyrrolnitrin biosynthetic genes within this region (Ligon et al, 1996; GenBank accession no. U74493). These results suggest that pyrrolnitrin is a major contributing factor to the biological control activity of BL915. Pyrrolnitrin production does not appear to increase the ecological competence of BL915 (Hill et al, 1994); however, the duration of this competition study was relatively short (12 days) and may not have been sufficient to detect slight differences between the ecological competence of non-producing and wild-type strains. 8

2,4-Diacetylphloroglucinol. 2,4-Diacetylphloroglucinol is produced by many isolates of fluorescent pseudomonads and exhibits activity against a variety of plant pathogenic fungi (Thomashow and Weller, 1995). Production of this antifungal metabolite is globally regulated by the GacA/ApdA two component regulatory system within Ps. fluorescens (Laville, et al, 1992; Corbell and Loper, 1995) as well as by the sigma factors RpoS and RpoD (Schnider et al, 1995; Sarniguet et al, 1995). In independent studies using two Ps. fluorescens strains, antibiotic non-producing mutants were obtained by transposon mutagenesis (Vincent et al, 1991; Fenton et al, 1992). In both cases, the isolated mutants were less suppressive against P. ultimum or G. grorninis var. tritici than were the wild-type strains. Subsequently, a genomic DNA region conferring 2,4-diacetylphloroglucinol production was cloned from both strains and when transferred into heterologous hosts, significantly increased their biocontrol activity against the target pathogens (Fenton et al, 1992; Bangera and Thomashow, 1996). Four open reading frames that encode biosynthetic enzymes required for 2,4-diacetylphloroglucinol production have been identified within a 6.5-kb fragment from Ps. fluorescens Q2-87 (Thomashow et al, 1996; GenBank accession no. U41818). Since the genes expressed heterologously in the previous studies were specifically involved in 2,4 diacetylphloroglucinol biosynthesis, it is very likely that this antibiotic is a major contributing factor of biological control.

Pyoluteorin. Pyoluteorin is produced by the biological control agent Ps. fluorescens Pf-5 and is active against the plant pathogenic fungus P. ultimum (Howell and Stipanovic, 1980). Treatment of cotton seeds with either pyoluteorin or Pf-5 cultures prior to planting in Pythium infested soils results in similar levels of disease suppression and was initially the basis for demonstrating the contribution of pyoluteorin towards the biocontrol activity of Pf-5 (Howell and Stipanovic, 1980). In a later study, Tn5 mutagenesis of Pf-5 was used in an attempt to further correlate pyoluteorin production in the biocontrol of P. ultimum on cucumber (Kraus and Loper, 1992). Seven pyoluteorin non-producing mutants were isolated; however, they were just as effective in disease suppression as the wild-type strain. More recently a study using Ps. fluorescens CHAO has demonstrated that pyoluteorin 9

contributes to biological control on cress but not on cucumbi, leading to the conclusion that the plant host appears to affects pyoluteorin's contribution to biological control (Maurhofer et al, 1994). Subsequent characterization of the Pf-5-derived Tn5 mutants localized all seven of the transposon insertions to within a 21-kb genomic region (Kraus and Loper, 1995). Since it is generally accepted that biosynthetic genes are closely linked, it was assumed that this region encoded pyoluteorin biosynthetic enzymes. The expression of this region in situ was assessed on cotton and cucumber using an ice nucleation transcriptional reporter gene fusion. Although gene expression occurred on both plant hosts, expression on cucumber occurred after the infection period required by P. ultimum (Nelson et al, 1986). In light of the carbon source effects on pyoluteorin production within Pf-5 (Kraus and Loper, 1992; Nowak- Thompson et al, 1994), differences in seed exudates could account for the plant host-dependent pattern of gene expression.

1.3 STATEMENT OF RESEARCH OBJECTIVES

The research described within this treatise represents an integrated approach to understanding the biosynthesis of pyoluteorin within Ps. fluorescens Pf-5. Pf-5, originally isolated from a cotton rhizosphere, is effective at controlling Pythium damping-off of cotton (Howell and Stipanovic, 1980) and cucumber (Kraus and Loper, 1992), Rhizoctonia damping-off of cotton (Howell and Stipanovic, 1979), and ascocarp formation by Pyrenophora tritici repentis on wheat straw residue (Pfender et al, 1993). In addition to producing pyoluteorin, Pf-5 also produces pyrrolnitrin (Howell and Stipanovic, 1979), 2,4-diacetylphloroglucinol (Nowak-Thompson et al, 1994; See Appendix 1), hydrogen cyanide (Kraus and Loper, 1992), and a pyoverdine siderophore (Kraus and Loper, 1992). Because of the importance of pyoluteorin in biological control (Section 1.2), characterization of the genomic region required for pyoluteorin production has been undertaken. Furthermore, pyoluteorin is a structurally interesting metabolite due to the halogenated pyrrole moiety, a relatively rare functionality found in terrestrial natural products. Therefore, two principle objectives guided the research described within this treatise: the 10 elucidation of the pyoluteorin biosynthetic pathway using an isotope feeding methodology, and the nucleotide sequence analysis of the genomic region required for pyoluteorin production. Although the results herein do not provide a complete understanding of pyoluteorin biosynthesis, they nevertheless have revealed several interesting aspects of this pathway and have established a solid framework for future studies concerning both the biochemical pathway and the regulation of pyoluteorin biosynthesis. 11

Chapter 2. Biochemical Analysis of the Pyoluteorin Pathway

2.1 INTRODUCTION

The polyketide origin of natural products was originally proposed by Collie (1907) based on the observation that successive condensations and dehydrations of a few simple aldehydes or ketones can give rise to more complex structures. However, the biological significance of this proposal was not demonstrated until the introduction of radioisotopes in the 1950's. 14C_ labelling studies provided the necessary evidence for Collie's original proposal, and also lead to an expansion of the biosynthetic hypothesis as an aid in the structural elucidation of natural products (Birch, 1957). Many polyketide biosynthetic pathways have since been probed using a variety of techniques, and many reviews have highlighted the recent developments within the realm of polyketide natural products (Robinson, 1991; Fenical, 1993; Plater and Strohl, 1994; Rohr, 1995; Simpson, 1995; O'Hagan, 1995; Khos la and Zawada, 1996). The structural diversity of the polyketides makes it necessary to limit this discussion to a few thoroughly characterized metabolic pathways, some of which represent "mixed" polyketide pathways (i.e. not solely acetate-derived). Although the polyketides described herein possess relatively simple molecular structures when compared to higher molecular weight polyketides, they nevertheless serve to illustrate the biochemical transformations often encountered in polyketide biosynthesis. The biochemical reactions required for polyketide chain assembly and their analogy to fatty acid synthesis have been reviewed (Hopwood and Sherman, 1990), and will not be presented here.

2.1.1 Rudimentary polyketide pathways: the tetraketides

One of the most thoroughly studied polyketide systems in which both the genetic and biochemical aspects are understood, is the 6-methylsalicylic acid (6MSA) pathway found within various Penicillium species. Several 12

independent isotope labelling studies have firmly established the polyketide origin of 6MSA, 2.1 (Scheme 2-1; Birch 1967; Spencer and Jordan, 1990). 6MSA biosynthesis is processive and requires an NADPH-dependent reduction before addition of the final malonyl-CoA unit (Dimroth et al, 1970; Scott et al, 1974). Furthermore, elimination of the C-2 malonyl-CoA methylene hydrogens required for aromatization of the carbocyclic ring occurs stereospecifically (Spencer and Jordan, 1990; Jordan and Spencer, 1990). 6MSA synthase has been isolated and characterized (Dimroth et al, 1970; Scott et al, 1974; Spencer and Jordan, 1992a). The gene encoding the 6MSA synthase has been cloned and sequenced (Beck and Ripka, 1990). The deduced polypeptide sequence exhibits similarity to several bacterial polyketide synthases and contains discrete catalytic domains required for polyketide assembly (see Chapter 3).

Scheme 2-1

0 0

Ha Dor'SCoA Hs soiegLLSCoA O coi c 02. 0c1,rC H3 NADPH -----3111111. EnZSNrrolTh.rC H3 ....--....00.. CoAS Ac H3 0 0 0 SEnz O 0

HO H3 ?I' CoA Ha - H0H2O C H3 C 01 C H3

SEnz SEnz 014 0 HR. 0 0 HO 0 2.1

Orsellinic acid (OA; see Scheme 2-4, pg. 20) is a metabolite produced by Penicillium spp. that is structurally related to 6MSA, but contains an additional C4 hydroxyl moiety. OA synthase has been purified, and although it does not require exogenous NADPH for catalysis, it exhibits properties similar to 6MSA synthase (Gaucher and Shepherd, 1968). The OA synthase gene sequence 13 has not been reported. With the exception of the NADPH-dependent reduction at C4 of the aromatic ring, OA and 6MSA biosynthesis appear to occur via identical reaction pathways (Gatenbeck and Mosbach, 1959; Spencer and Jordan, 1992b). The bacterial metabolite 2,4-diacetylphloroglucinol (DAPG), 2.2, is produced by many fluorescent Pseudomonas species (Nowak-Thompson et al, 1994 and references therein; see Appendix 1). Although the DAPG structure certainly suggests that it is acetate derived (Birch, 1967; Mann, 1987), this has never been experimentally proven. Nevertheless, two pieces of experimental evidence support this hypothesis. The identification of monoacetyl phloroglucinol (MAPG), 2.3, and the detection of enzyme activity converting MAPG to DAPG in a Pseudomonas species argues in favor of DAPG originating as a simple tetraketide (Shanahan et al, 1993). Furthermore, the biosynthetic gene required for DAPG production exhibits sequence similarity to several chalcone synthases (Thomashow et al, 1996), plant enzymes that utilize three malonyl-CoA equivalents in precursor formation for flavonoid phytoalexins (Kreuzaler and Halbrock, 1975; Hrazdina et al, 1976). Therefore, DAPG biosynthesis most likely occurs as depicted in Scheme 2-2.

Scheme 2-2

0 0 C H3

CO 2 SEnz CH HO OH

EnzS.C H3 C H3 JP" C H3 IT C H3 0 0 0 OH 0 OH 0 2.3 2.2

DAPG represents a fundamentally different ring structure from 6MSA due to the folding of the extended polyketide chain prior to cyclization (Scheme 2-3). Cyclization of 6MSA requires C-C bond formation between C2 and C7 of the polyketide chain, whereas Cl and C6 are involved in bond formation 14

within DAPG. Differences in cyclization regiospecificity are also apparent in naringenin chalcone, 2.4, and resveratrol, 2.5, plant-specific polyketide metabolites required for flavonoid and stilbene production, respectively (Mann, 1987; Kindl, 1985). Oxygenation patterns on the phenol ring highlight the similarity of naringenin chalcone to DAPG; however, the relationship of resveratrol to 6MSA is somewhat obscured by an apparent decarboxylation (Scheme 2-3). Although both chalcone and resveratrol synthase catalyze polyketide biosynthesis via successive condensations of malonyl-CoA, these enzymes are distinct from the polyketide synthases (Schroder et al, 1988; Lanz et al, 1991).

Scheme 2-3

I SEnz

----Imp. .7.7".=.: C H3 0 0 OH 0 OH 0 2.4

0 0 OH 1 OH 0 SEnz -Ow. 7 C H3 (F10)"

Although only two different ring systems are possible for any of the tetraketides, applying this same principle to larger polyketides greatly increases the structural variability within this class of natural products. Structural variation is an important consideration in biocombinatiorial design of novel compounds (McDaniel et al, 1995). It has also been suggested that folding patterns may represent biosynthetic and evolutionary relationships within the multicyclic polyketides (Rohr, 1992). 15

2.1.2 Unusual polyketide starter units

For many years, the contribution of starter units to polyketide structural variability has been an understated aspect of polyketide biosynthesis. However, recent proposals regarding biocombinatorial engineering of polyketide natural products have suggested manipulating the starter unit specificity of the polyketide synthase as a means of introducing novel structural variations (Katz and Donadio, 1993; Hutchinson and Fugii, 1995; Rohr, 1995; Khosla and Zawada, 1995). Alternative starter units can sometimes be introduced into polyketide structures through precursor-directed biosynthesis (reviewed by Thiericke and Rohr, 1993). Of the naturally occurring polyketide starter units, acetate and propionate are, by far, the most widely used, but some structures exhibit alternative starter units (Table 2.1; also see Scheme 2-3 for starter units within naringenin chalcone and resveratrol.). Nevertheless, polyketide natural products possessing a non-acetate starter unit are encountered relatively infrequently. Starter units generally can be identified within the polyketide structure except in cases of extensive post-assembly modifications or structural rearrangements. Only precursors that have been experimentally demonstrated as polyketide starter units are listed in Table 2.1. Although this list is extensive, it should not be considered complete. The existence of amino acid polyketide starter units is especially interesting in light of the similar thiotemplate mechanisms used in polyketide and non-ribosomal peptide biosynthesis (Kleinkauf and von Dohren, 1995 and 1996). The substrates recognized by polyketide synthases and peptide synthetases are the activated CoA thioester and acyl adenylated amino acids, respectively. Since the activated amino acids eventually form a thioester linkage with a 4'-phosphopantetheine prosthetic group in both the synthases and synthetases, it is tempting to speculate on the possibility of substrate competition within an organism possessing both biosynthetic pathways. It is likely that these activation strategies serve to channel substrates to the individual pathways and avoid direct competition between both pathways. Nevertheless, the activating enzymes offer a potential point for the coordinated regulation of polyketide and non-ribosomal peptide biosynthesis. 16

Table 2.1 Non-acetate compounds that incorporate into the starter units of polyketide natural products. The location of each starter unit is indicated by darkened bonds within each structure.

Amino Acids

proline (B.N-T., J. Loper, and S. Gould, OH 0 H unpublished results; Cuppels et al, 1986) Pyoluteorin

a if. 0 (Kawamura et al, 1996) C H3 Pyralomicin a 02N

(Carter et al, 1989) OHO H Dioxapyrrolomycin

o-acetyl-L-serine 3C.pN _ O% (Kobayashi et al, 1986) Rhizoxin

glycine 0 H3C H3C H3C) rk''')Y1 C F13 C H3 (Carmeli et al, 1993) Tolytoxin

13phenylalanine

(Omura et al, 1982)

Hitachimycin 17

Table 2.1 (cont'd)

Glycerol 000 (Lee et al, 1987) 0 Aplasmomycin

Linear Carboxylic Acids

malonate

(Jeffs and McWilliams, 1981)

H Cycloheximide

OH NK4e2 , 7 al Nit OH 0H0 0 0 (Thomas and Williams, 1983) Oxytetracycline

(Bockholt et al, 1994) F1300 N 0

H3 Lysolipin

hexanoate OH 0 OH 110111 (Townsend et al, 1984; C H3 Brobst and Townsend, 1994) 0 Averfulvin

hexadecanoate

(Walters et al, 1990)

Anacardic Acid 18

Table 2.1 (cont'd)

Branched and Cyclic Carboxylic Acids

2-methylpropionatel

(Tsou et al, 1989)

2-methylbutanoatel

1-0 0 (Al-Douri and Dewick, 1988) Lathodoratin

3-hydroxycyclo hexanoate2 (Thiericke et al, 1990) Asukamycin 3,4-dihydroxy cyclohexanoate C H3 CH (Lowden et al, 1996) Rapamycin

Aromatic Carboxylic Acids

benzoate3

(Burns et al, 1979) OH 0

(Jones et al, 1992) C H3 Squalistatins 19

Table 2.1 (cont'd)

Aromatic Carboxylic Acids (cont'd)

o-hydroxybenzoate

(Aragozzini et al, 1988) cD2H Thermorubin

m-hydroxybenzoate

R= yy (Bennett and Lee, 1988)

Mangostin

3-amino-5-hydroxy benzoate4

(Kibby et al, 1980; Lee et al, 1993; Ghisalba and Nuesch, 1981)

1 Identical starter units are incorporated into several avermectin antibiotics (Chen et al, 1989; Schulman et al, 1986). 2 Ansatrienin A is a related antibiotic utilizing this precursor as the polyketide starter unit (Moore et al, 1993). 3 Benzoate also serves as the starter unit for vulgamycin biosynthesis (Seto et al, 1976). 4 The me ta-C7N unit is also used in other ansamycin antibiotics; see Staley and Rinehart, 1991. 20

2.1.3 Post-assembly modifications of polyketide metabolites

Post-assembly modification of polyketide structures greatly expands the structural variability of simple polyketides such as 6MSA or OA. The extent of these modifications range from simple transformations of functional groups (i.e. methylations or oxidations/reductions) to complex oxidative cleavages and carbon skeleton rearrangements. Methylation by S-adenosyl methionine and oxidative decarboxylation of the OA sub-structure are required for biosynthesis of the fungal benzoquinone, shanorellin, 2.6, (Scheme 2-4; Wat et al, 1971). Because the carbon skeleton of OA is found within a variety of fungal metabolites (Simpson, 1991), OA is a likely precursor to shanorellin. Nevertheless, this possibility has not been tested.

Scheme 2-4

[Ox] *CH3 -AdoMet OH C H3 OH C 02H C 02H C 02H OA

*C H3 *C H3-AdoMet H3C -Nap.-C 02 OH HO C 02H

Two well known examples of an oxidative C-C bond cleavage followed by a structural rearrangement are found in the pathways leading to patulin, 2.7, and penicillic acid, 2.8 (reviewed by Zamir, 1980 and Mann, 1987). Both patulin and penicillic acid are carcinogenic Penicillium metabolites derived from 6MSA and OA, respectively. Patulin biosynthesis requires ring cleavage at the C2-C3 bond of the aromatic ring. The epoxide intermediate was isolated from a patulin non-producing mutant and shown to efficiently incorporate 21 into patulin (Scheme 2-5). In contrast, ring cleavage required for penicillic acid biosynthesis occurs between the C4-05 bond of the aromatic ring (Scheme 2-6). In light of the reaction pathway leading to patulin, it has been suggested that an epoxyquinone may serve as a penicillic acid precursor (Zamir, 1980).

Scheme 2-5

-C 02 [Ox] OH

C H2CH C HO

r°F1 0 HO C.3 -110. 0 F/32CC HO

2.7

Scheme 2-6

CH *O H3-AdoMet [Ox]

C H3 *C H30 C H3 -CO2 *C H30 a-13 C 02H C 02H O

Oxidative coupling of phenolic compounds is another post-assembly modification found in many natural products and is not limited to polyketide derived structures. Because oxidative coupling has been discussed extensively elsewhere (Brown, 1967; Geismann and Crout, 1969), it will only be mentioned 22

briefly as it applies to usnic acid biosynthesis. Usnic acid, 2.9, is produced by several lichen species and labelling studies have established the biosynthetic pathway (Scheme 2-7; Taguchi et al, 1969). Although 4-methyl-MAPG, 2.10, is a precursor to 2.9, MAPG itself does not incorporate into usnic acid. It was therefore concluded that the aromatic methyl group is introduced prior to cyclization. It is possible that the 4-methyl moiety actually originates via a propionate extender unit incorporated by the bacterial symbiont of the lichen.

Scheme 2-7

FI3 0

"4111E-110. H3C OH 2.10

2.2 BIOSYNTHETIC ANALYSIS OF PYOLUTEORIN

It is apparent from the oxygenation pattern of pyoluteorin, 2.11, that the resorcinol moiety most likely results from the condensation of three malonyl- CoA equivalents with proline, or a derivative thereof, serving as the starter unit for polyketide assembly. The specific incorporation of [1,2-13C2] acetate into pyoluteorin has demonstrated that pyoluteorin is indeed a polyketide metabolite (Cuppels et al, 1986) as evidenced by the labelling pattern found within the resorcinol moiety (Scheme 2-8). Furthermore, the non-uniform 23

labelling of the dichloropyrrole suggested that this moiety originated from a tricarboxylic acid cycle intermediate, a conclusion that is consistent with a proline derived starter unit (Scheme 2-9; Cuppels et al, 1986).

Scheme 2-8

[rAscoA1 SEnz cot OH CI EnzS N CI 0 0 H OH 0 H 2.11

While the assembly of the resorcinol ring is by no means atypical, the use of an unusual starter unit by the presumed pyoluteorin polyketide synthase (PKS) justified further study of this system. The rapamycin synthase is the only PKS utilizing an unusual starter unit that has been characterized (Aparicio et al, 1996). Considering the current advances in PKS molecular genetics, a combinatorial approach to drug design is possible, and the ability to introduce diverse starter units into a polyketide offers an additional variable for drug design. The basis for starter unit selection by a PKS, however, is unknown. Therefore, the concurrent genetic characterization of the pyoluteorin PKS and identification of the required starter unit substrate may reveal factors influencing starter unit fidelity. Further justification for this study focuses on formation of the pyoluteorin dichloropyrrole ring. Over 2000 halogenated natural products have been reported (Gribble, 1996), but despite the proliferation of these compounds, little is known regarding the halogenation reactions involved in their synthesis. Because halogenation is known to increase the pharmacological effects of many compounds (Neidelman and Geigert, 1987), it is conceivable that a mechanistic understanding of biohalogenations could be exploited for the synthesis of new or otherwise costly antibiotics. 24

Scheme 2-9

Initial labelling of TCA cycle inter mediates and proline by 13C-acetate: 0 A COiH3C _v..Cr '02ey 0 C 02 ...410,.IIP --110. -O. in C 02- _. -02,C).C/%C Oi --ON- N 0 --Do- 0 H a-ketoglutarate proline Further incorporation of 13C-acetate into TCA cycle intermediates and proline: / .02CN/4440 02.

.02e)fC 02 CO2

o O \[0H3CAO [H. 01

-o2c')(co2- -02nC0 02- oey."coi -o2e)1/4"co2. HO C 02 HO 0 02- Ho COI Ho coi 4 4 4 4 cj ,Lco %cc02f 4 4 25

A considerable amount of effort has been directed, therefore, towards the isolation and characterization of haloperoxidases, enzymes that are capable of forming carbon-halogen bonds in the presence of halide ions and hydrogen peroxide (van Pee, 1996). However, it has yet to be demonstrated that any of the haloperoxidases so far characterized are responsible for the in vivo halogenation of known natural products. It is quite likely that elucidating the pyoluteorin biosynthetic pathway will identify a substrate for the halogenating enzyme(s) involved. Such a discovery would allow the development of a specific assay required for the isolation and subsequent characterization of this unknown class of enzymes.

2.3 RESULTS AND DISCUSSION

2.3.1 Preliminary studies: pyoluteorin production and isolation

Pyoluteorin production was evaluated in a variety of liquid culture media. HPLC and TLC analysis of culture extracts indicated that pyoluteorin production was greatest in cultures grown on either modified King's medium B (King et al, 1954) or Bacto Peptone medium (Difco, 1984). Pyoluteorin production did not differ between these two media, and modified King's medium B was chosen for further studies. By varying the peptone source within King's medium B, it was discovered that phytone resulted in approximately a two-fold increase in pyoluteorin production as compared to proteose peptone #3. Several carbon sources were also used in an attempt to further improve production. The addition of glucose to the production medium appeared to increase DAPG production at the expense of pyoluteorin. The differential production of these antibiotics in response to carbon source parallels previous results demonstrating that transcriptional activity of a promoter within the pyoluteorin gene cluster is repressed in cultures containing glucose, but induced in cultures supplemented with glycerol (Loper and Kraus, 1995). Aeration effects were examined by varying the flask size to medium volume ratio. In general, flasks containing a smaller volume of medium 26

result in elevated pyoluteorin production, presumably due to an increase in aeration rate. This result contrasts with an earlier report that pyoluteorin is produced only in non-aerated flask cultures (Bencini et al, 1983). Lower temperatures (18 °C vs 27 °C) also supported greater metabolite production. Although it is not obvious why temperature would serve to increase production, it is possible that the lower temperature more closely approximates the native habitat of Pf-5. A previously developed purification method for pyoluteorin required the fractionation of a crude extract by both column and thin layer chromatography followed by derivatization and further purification (Cuppels et al, 1986). Several alternative solvent systems (see Section 2.5) therefore were developed that provided pyoluteorin of sufficient purity after a single silica column chomatography for NMR spectroscopic analysis.

2.3.2 Incorporation of proline into pyoluteorin

In light of the previous biosynthetic study (Cuppels et al, 1986), the initial objective was to establish if proline is the primary precursor of the dichloropyrrole moiety. Therefore, an initial study was conducted in which R L-proline was administered to growing cultures of Pf-5. After isolation and recrystallization to a constant specific activity, 19% of the total radioactivity added to the culture was detected in pyoluteorin. Although it was possible that proline may have been catabolized and the labelled products subsequently incorporated into pyoluteorin, the high level of radioactivity in the final product argued against this possibility. Nevertheless, [1-13C] -L-proline was administered to Pf-5 cultures in order to demonstrate the specific incorporation of proline into pyoluteorin. Purification and 13C-NMR analysis of pyoluteorin revealed a specific enrichment of 5.7% over and above the natural abundance carbonyl resonance for the 13C spectrum and provided definitive evidence that the dichloropyrrole is derived from proline (Scheme 2-10). 27

Scheme 2-10

0 H CH 0 H 2.11

Based on the results of these simple feeding experiments, several biosynthetic pathways could be envisioned in which proline serves as the primary precursor to pyoluteorin (Scheme 2-11). The proposed pathways all involve a pyrrolidine to pyrrole oxidation, the subsequent chlorination of the pyrrole, and polyketide assembly of the resorcinol moiety. Differences among the proposed pathways reside in the temporal order of the transformations. It was expected that biosynthetic feeding studies would demonstrate involvement of deschloropyoluteorin, 2.12, and either pyrrole 2-carboxylic acid, 2.13 or tetrahydrodeschloropyoluteorin, 2.14, and provide evidence for the intermediates in the metabolic pathway leading to pyoluteorin.

Scheme 2-11

Ho H O 0 H proline 2.13

a OH 0 H 0H n H 2.14 2.12 2.11 28

2.3.3 Involvement of deschloropyoluteorin

Deschloropyoluteorin (DCP) was a logical target compound for the subsequent feeding studies because two of the three proposed pathways converge at this putative intermediate. Furthermore, DCP was readily synthesized (Scheme 2-12) by Friedel-Crafts aroylation of pyrrole with 2,6 dimethoxybenzoyl chloride, 2.15, (Rao and Reddy, 1990) followed by deprotection of 2.16 with aluminum bromide (Cue et al, 1981). The intermediacy of DCP in the pyoluteorin pathway was first examined using a radioisotope trapping method. [1-14C]-Acetate was added to cultures of Pf-5 with the expectation that pyoluteorin pathway intermediates would be radiolabelled. Shortly after adding the labelled precursor, the cells were collected, lysed, and unlabelled synthetic DCP was added to the cell lysate. DCP was subsequently re-isolated from the lysate, but no radioactivity could be detected in the purified sample.

Scheme 2-12

AlC13; 5 °C OC H 3 AIBr3 0 O. --Hp. C H30 0 N CH30 0 H H 2.15 2.16 2.12 (72% yield; (40% yield) 30% overall)

This unexpected result unlikely was due to insufficient labelling of pathway intermediates since approximately 3% of the total radioactivity was detected in pyoluteorin. Other explanations, however, could account for the lack of evidence from this experiment. The metabolic flux through the pyoluteorin pathway may have been rapid enough to prevent accumulation of pathway intermediates. Addition of unlabelled carrier DCP could sufficiently dilute the radiolabel to render it undetectable. Alternatively, it is also possible that DCP was not a pyoluteorin pathway intermediate. Nevertheless, the outcome of this experiment does not allow for any definitive conclusions regarding the role of DCP in the pyoluteorin pathway. 29

In light of the unsuccessful result from the isotope trapping experiment, the involvement of DCP in pyoluteorin biosynthesis was further investigated by isotope labelling studies. Treatment of DCP with 2H-trifluoroacetic acid afforded the deuterated compound, 2.17 (Scheme 2-13). Greater than 90% proton-deuterium exchange occurred on the pyrrole ring and approximately 70% proton-deuterium exchange occurred at two of the three positions on the resorcinol moiety.

Scheme 2-13

2H-TFA

112; 75 °C

2.17 (76% yield)

Higher reaction temperatures resulted in low yields of the desired product, and this was likely the result of an acid-catalyzed rearrangement of the acylpyrrole (Carson and Davis, 1981). Attempts to deuterate DCP using 2H20 and BF3-Et20 (McGeady and Croteau, 1993) were unsuccessful. Compound 2.17 was subsequently fed to cultures of Pf-5 and pyoluteorin analyzed for deuterium enrichment by 2H NMR spectroscopy. Based on the amount of 2.17 used in this feeding experiment (50 mg; 240 gmole) and a minimal NMR detection limit of 1 gmole 2H-pyoluteorin, the detectable limit for incorporation of DCP was approximately 0.5%. However, no deuterium was detected in the pyoluteorin sample isolated. Therefore, no valid conclusions regarding the intermediacy of DCP in the pyoluteorin pathway could be made on the basis of this experiment. Two observations made during the course of the deuterium feeding experiment offer some insight to the limitations occasionally encountered in biosynthetic labelling studies. Since DCP is only sparingly soluble in water and the majority of 2.17 administered to the cell culture was recovered (98% 30 with no loss of the label), it is possible that an insufficient amount of DCP was able to cross the cell membrane. The outer membrane structure of Gram- negative bacteria is an effective barrier to macromolecules and allows limited diffusion of hydrophobic compounds (see references within Vaara, 1992). Several polycationic agents and divalent cation chelators are reported to decrease antibiotic resistance in Gram-negative bacteria by increasing the permeability of the outer membrane (Brown and Richards, 1965; Leive, 1967; Viljanen et al, 1986; Vaara, 1990; Vaara, 1992). Therefore, several cationic detergents, polylysine, and polymyxin B nonapeptide (PMBN) were tested for their effects on pyoluteorin production and DCP utilization by Pf-5 cultures. In addition, the organic solvents toluene and dimethylsulfoxide (DMSO) were also examined for their peameabilizing properties. It was expected that by increasing cell permeability to DCP, an increase in pyoluteorin production would be accompanied by a corresponding decrease in the concentration of exogenous DCP. Cell pellets from the detergent- and toluene-amended Pf-5 cultures were visibly different from the control culture, and no pyoluteorin was detected in the corresponding culture extracts. Both PMBN and polylysine negatively affected pyoluteorin production in MgC12-containing cultures. However, the suppressive effect of PMBN was not apparent in a culture medium lacking MgC12 and amended with ethylenediaminetetraacetic acid (EDTA). Interestingly enough, DMSO did not affect pyoluteorin production when compared to the control culture. Nevertheless, no changes in pyoluteorin production were observed when DCP was added to cultures. In spite of the negative evidence resulting from this study, it was reasoned that including either PMBN or DMSO in the culture medium may increase the chances for a obtaining a positive result. Since there was no apparent change in pyoluteorin production when DCP was added to the cultures however, the use of radioisotopes to increase sensitivity in future analysis was considered prudent. A semi-synthetic strategy was initially considered to obtain radiolabelled DCP by reductive dechlorination of labelled pyoluteorin obtained from "C-acetate fed cell cultures (Takeda, 1958; Birch et al, 1964). However, catalytic hydrogenation of pyoluteorin on platinum oxide or palladium metal was unsuccessful. A total synthesis of 14C -DCP was therefore undertaken based on the Friedel-Crafts aroylation shown in Scheme 2-12. Labelled [14C0] -2,6- dimethoxybenzoic acid, 2.18, was readily available (Scheme 31

2-14) by direct metalation of 1,3-dimethoxybenzene, 2.19, followed by treatment of the organolithium species with 14CO2 generated from Na214CO3 (Gilman and Ess, 1933; Gilman et al, 1940). Despite reports that addition of tetramethylenediamine (TMEDA) to the reaction mixture increases the overall yield and affects the regiospecificity of metalation (Slocum and Jennings, 1976; Cabiddu et al, 1979; Cabiddu et al, 1981), use of this reagent resulted in yields of <5%. However, reaction yields were increased ten-fold when TMEDA was omitted and THF was used as the reaction solvent instead of diethyl ether; a similar observation has been reported previously (Meyers and Avila, 1980). Conversion of 2.18 to the acid chloride, 2.15, was achieved with thionyl chloride (Bosshard et al, 1959), and the labelled product was used directly for the Friedel-Crafts aroylation without purification.

Scheme 2-14

H3 CCH3 1) n-BuLi, THF SOCl2 2) 4C 02 C H30 C H30 0 C H30 0 2.19 2.18 2.15 (78% yield)

The labelled DCP thus synthesized was administered to Pf-5 cultures grown in medium lacking MgC12 and amended with EDTA and PMBN. Isolation of pyoluteorin from the culture supernatant followed by recrystallization to constant specific activity revealed that none of the labelled DCP was incorporated. The lowest detectable limit of incorporation was calculated as 0.005% based on the amount of pyoluteorin produced (17.1 mg; 62.6 pimole), a weighing accuracy of ± 0.10 mg, and a specific activity of at least 103 dpm mg-1 for the recovered pyoluteorin. Despite the considerable amount of effort focused on the possible involvement of DCP in the pyoluteorin biosynthetic pathway, no evidence was obtained suggesting that DCP is a pyoluteorin precursor. Although it is tempting to refute the intermediacy of DCP, such a conclusion based solely 32

on a lack of evidence would be unsound. Therefore, none of the proposed pathways leading to pyoluteorin could be discounted from these studies and other putative intermediates were considered.

2.3.4 Involvement of pyrrole-2-carboxylic acid

The putative intermediate pyrrole-2-carboxylic acid (PCA) lies at a branch point in the hypothetical biosynthetic routes leading to pyoluteorin (Scheme 2-11) and therefore was chosen as the next target for further study. Although synthesis of PCA is feasible by reacting pyrrole with methyl chloroformate (Hodge and Rickards, 1963) or with trifluoroacetic anhydride in the presence of an acid scavenger followed by base hydrolysis (Sonnet, 1971), isotopic labelling of PCA is not convenient using either of these strategies. [13CO2H] PCA, 2.20, was synthesized instead (Scheme 2-15) from pyrrole and 13CO2 via pyrrylmagnesium bromide (Letellier and Bouthillier, 1957). Subsequently, 2.20 was added to Pf-5 cultures as an aqueous solution, and pyoluteorin was isolated and analyzed for isotopic enrichment. However, no enrichment of the carbonyl resonance in the 13C NMR spectrum was observed. The lowest detectable limit for incorporation of 2.20 into pyoluteorin was calculated as 0.3% based on the amount of pyoluteorin isolated (46.1 mg; 171 limole), a 13C NMR enrichment of at least 1.1% over natural abundance, and the amount of 2.20 administered to the culture (75 mg; 681 1.tmole).

Scheme 2-15

EtMgBr 13CO2

H 0 MgBr 2.20 (28% yield) 33

Failure to detect incorporation of the free acid into pyoluteorin provoked consideration of the initial PKS substrate required for biosynthesis. If the pyrrole is the actual starter unit for polyketide biosynthesis, it may be recognized by the PKS only as the ACP-CoA thioester. The use of N acetylcysteamine (NAC) thioester derivatives to mimic the ACP-bound substrate allows recognition of an advanced intermediate by active-sites within the PKS. This method has been used on occassion to incorporate postulated polyketide intermediates into several natural products (Yue et al, 1987; Cane and Ott, 1988; Brobst and Townsend, 1994; Dutton et al, 1994; Hai les et al, 1994a; Hai les et al, 1994b; Cane et al, 1995 and references therein). Therefore, 2.20 was converted to the NAC thioester 2.21 (Scheme 2-16) by treatment with dicyclohexaneylcarbodiimide (Brobst and Townsend, 1994). The labelled product was dissolved in DMSO and administered to a 1 L culture of Pf-5. Generally, 40-55 mg of pyoluteorin can be recovered from a 1 L culture, however in this instance, only 15 mg was isolated. Nevertheless, based on the amount of pyoluteorin produced (15 mg; 54.9 umole), a 13C NMR enrichment of at least 1.1% over natural abundance, and the amount of 2.21 administered to the culture (93 mg; 440.8 umole), the lowest detectable limit for incorporation of 2.21 into pyoluteorin was calculated to be 0.1%. The isolated pyoluteorin, however, was not labelled.

Scheme 2-16

0 N-acetylcysteamine NJrSN..1 NAc H 0 DC C; DMAP; 0 °C H 0 H 2.20 2.21 (41% yield)

The absence of any positive results from these feeding experiments again obviates any definitive conclusion regarding the role of PCA in the pyoluteorin pathway. As was the case for results described in Section 2.3.3, none of the proposed pathways leading to pyoluteorin could be discounted. 34

2.3.5 Involvement of tetrahydrodeschloropyoluteorin

The involvement of the proposed biosynthetic intermediate tetrahydro deschloropyoluteorin (THDCP, 2.14 in Scheme 2-11) was examined via isotopic feeding studies in a final attempt to elucidate the biosynthetic route leading to pyoluteorin. While the feeding study itself did not present any particular difficulty, synthesis of labelled THDCP was not as readily accomplished as was the syntheses of either DCP or PCA. Direct hydrogenation of 2.16 over palladium or platinum catalysts was unsuccessful, presumably due to the aromatic resonance stabilization of the pyrrole. However, N protection of the pyrrole as the tert-butoxycarbonyl derivative 2.22 (Scheme 2-17) provided sufficient delocalization of the nitrogen electron pair for reduction to occur (Kaiser and Muchowski, 1984). Although this methodology looked promising, complete deprotection of 2.23 was unsuccessful despite the variety of conditions examined (Greene and Wuts, 1991).

Scheme 2.17

OCH3 0 4'LLIDIBu OCH3 )2 -Do-H2 K*OtBu PcVC N C H30 0 H C H30 0 tBoc C H30 0 tBoc 2.16 2.22 (75% yield) (83% yield)

OCH3 TFA BBr3

0°C - 70°C N C H30 0 H2 OOCCF C H30 0 H 2.23 (57% yield; (54% yield) 19% overall) 35

An alternative protection strategy was attempted via catalytic hydrogenation of an N,0,0-tri-tert-butoxycarbonyl derivative of 2.12. Although complete characterization of the product was not undertaken, resonances in the 1H NMR spectrum of the isolated product suggested only a partial reduction of the pyrrole. It is very likely that steric constraints of the tert-butoxycarbonyl protecting groups were responsible for incomplete reduction of the pyrrole. Therefore, protection of the resorcinol ring with a less sterically demanding protecting group was examined within this synthetic strategy. Treatment of 2.12 with benzyl bromide resulted in protection of a single phenolic group as the benzyl aryl ether 2.24 (Scheme 2-18). Conversion of this ether to 2.25 was afforded by treatment with base and di-tert-butyl dicarbonate. Catalytic hydrogenation allowed full reduction of the pyrrole to the pyrrolidine, presumably due to the loss of the benzyl protecting group, thereby relieving the steric effects limiting pyrrole reduction. Removal of the remaining tert-butoxycarbonyl groups with trifluoroacetic acid resulted in THDCP as the desired product. The reaction scheme was then repeated using a mixture of [13CO]- and [14CO] -DCP.

Scheme 2-18

Cs H5 CH2Br OtBoc

I I Na2CO3 N OtBu BzO O H BzO 0 tBoc 2.12 2.24 2.25 (74% yield from DCP)

H2 TFA

Pd/C 0 °C HD 0 H

(92% yield) 2.14 (>98% yield; 67% overall) 36

Extremely low levels of radioactivity were found in the pyoluteorin purified from Pf-5 cultures that had been fed [1300/140-ju THDCP. Within the parameters generally accepted for biosynthetic feeding studies, the permissible error in determining radiochemical purity of the final product is + 3% of the specific molar radioactivity over the course of three recrystallizations. The total disintegrations per minute (dpm) that were counted for the purified samples were in the range of 150-400 dpm. The specific activity of each sample from the last three recrystallizations fell within the permissible error range (969.39 ± 15.63 dpm mg-1 = 1.6% error) and corresponded to an observed incorporation of 0.1%. However, due to the low levels of radioactivity for each counted sample, the specific activities were recalculated taking into account background radioactivity of -30 dpm. The standard deviation of the adjusted specific activities was more than twice the acceptable range (768.01 ± 59.52 dpm mg-1 = 7.5% error). It should also be noted that approximately 90% of the total radioactivity added to the culture remained in the supernatant following extraction of the pyoluteorin and could not be recovered. Despite the presence of a 13C label in THDCP administered to the Pf-5 culture, no evidence of enrichment was detected in the isolated pyoluteorin sample. However, in light of the low incorporation of radioactivity, this result was expected since the minimum detectable incorporation of ['CO] THDCP for this experiment was approximately 0.8%. Developing definitive conclusions from these results is difficult due to the inconsistency of the two interpretations of the radioisotope incorporation data. On one hand, it appears that THDCP is, in fact, incorporated into pyoluteorin, suggesting that it is a pathway intermediate. Furthermore, since 90% of the radiolabel remained in the culture supernatant, the cell membrane may not be permeable to THDCP, thus preventing efficient utilization of this putative intermediate. If this is the case, the actual incorporation of THDCP into pyoluteorin was approximately 1%. Without a complete characterization of the labelled compound remaining in the supernatant, however, this point is merely speculation. The other possible interpretation of the described result is that THDCP is not a pathway intermediate. While this conclusion is supported by the lack of evidence implicating DCP as a pathway intermediate, both of these statements are based on negative evidence and, therefore, are suspect. 37

2.4 CONCLUSION

Although this study has demonstrated that proline is the primary precursor leading to formation of the dichloropyrrole moiety of pyoluteorin, the results regarding postulated advanced intermediates within the pyoluteorin biosynthetic pathway are inconclusive. No evidence was obtained for the involvement of either DCP or PCA within the pathway despite the various approaches used. Although isotope labelling data suggests that THDCP is a pathway intermediate, this evidence is tenuous at best since the extent of incorporation is low, given the level of background radioactivity. Furthermore, if THDCP is a pathway intermediate, DCP should have been incorporated into pyoluteorin. The lack of evidence resulting from this study has forestalled identification of the starter unit required for polyketide assembly as well as the intermediate preceding halogenation. This is rather unfortunate since these points were the basis of justification for this approach. The limited success of these isotope feeding experiments brings into question the usefulness of this approach for future work with this organism. An isotopic labelling methodology will most likely be successful only if used in conjunction with a cell free extract or purified recombinant proteins. Because several genes within the pyoluteorin biosynthetic gene cluster have been cloned and sequenced (see Chapter 3), purified recombinant proteins can be obtained for use in future studies.

2.5 EXPERIMENTAL METHODS

Chromatographic analysis of pyoluteorin. Production of pyoluteorin was routinely assayed from a crude extract preparation of bacterial cultures. A 5-10 mL aliquot was removed from a bacterial culture and cells were removed by centrifugation. The supernatant was then acidified to pH < 2 with 1 N HC1. The solution was subsequently extracted with ethyl acetate (10% vol, 3x), the pooled organic layers back extracted with water (5% vol, 3x), and the solvent removed by rotary evaporation. The pyoluteorin contained within the crude extract was assayed using either thin layer or 38 high performance liquid chromatography (TLC and HPLC, respectively). TLC analysis: Efficient resolution of pyoluteorin from the culture extract dissolved in Me0H was observed on aluminum backed, normal phase silica gel plates (Merck Kieselgel 60 F254) developed using any one of the solvent systems listed below. Pyoluteorin was visualized as an orange-brown derivative upon treatment with aqueous diazosulfanilic acid (1 vol. of 5% NaNO2, 2 vol. of 0.9% sulfanilate in 1M HC1, and 3 vol. of 20% K2CO3). Solvent Mixture Rf Value CHC13 / Me0H 0.32 CHC13 / EtOAc 0.34 CHC13 / Acetone 0.34 Toluene/ Acetone 0.38

It was also possible to resolve pyoluteorin (Rf 0.32) on reverse phase C18 silica TLC plates (Whatman KC18F) developed with a mixture of 50% 0.5 M NaCl and 50% Me0H. HPLC analysis: Pyoluteorin was analyzed using a Waters C18 Nova-pak column (8 x 100 mm, 4p) eluted with 45% H20/25% Me0H/30% ACN at 1.5 ml miri 1. Samples were dissolved in Me0H for injection onto the column and the eluant was monitored by UV absorbance at 310 nm. The retention time for pyoluteorin was 3.25 min. Pyoluteorin could also be assayed by normal phase HPLC (Nova-pak, 8 x 100 mm, 414 with 75% hexane/25% EtOAc at 1.5 ml min-1 (retention time 6.3 min). This method was not routinely used, however.

Optimization of pyoluteorin production. Duplicate 5 mL cultures of Pseudomonas fluorescens Pf-5 (JL4092) were grown in 523 media (Kado and Heskett, 1970), King's medium B modified to contain 0.5% wt/vol glycerol (King et al, 1954), B10 medium (Nowak-Thompson and Gould, 1994), Nutrient Broth supplemented with 1% glycerol (Difco Laboratories, Detroit, MI), and Bacto Peptone (Difco). Samples taken from each fermentation at 24 h, 48 h, and 70 h, were assayed by TLC and HPLC for relative pyoluteorin production. Peptone source: To investigate the effect of peptone sources on pyoluteorin production, equal weights of Bacto Peptone (Difco Laboratories, 39

Detroit, MI), phytone peptone (Becton-Dickinson, Cockeysville, MD), soytone peptone (Difco), neopeptone (Difco), and proteose peptone #3 (Difco) were substituted in the modified King's medium B formulation. In addition, two flasks containing proteose peptone #3 were supplemented with either Bacto yeast extract (Difco) or Amberex yeast extract (Red Star Specialties, Milwaukee, WI). Extractions and analysis for metabolites were performed as above. Carbon source: Equal weights of glycerol, sucrose, glucose, fructose, and gluconate were substituted as carbon sources into modified King's medium B formulation containing phytone peptone. Filter-sterilized solutions of glucose and fructose were added to the sterile medium aseptically. Extractions and analysis for metabolites were performed as above. Flask size: Medium to flask volume ratio effects on pyoluteorin production were examined using cultures of a pyoluteorin over-producing Tn5 mutant of Ps. fluorescens Pf-5 (JL4239). Cultures were grown in 2-500 mL flasks containing either 160 or 80 mL of modified King's medium B containing phytone peptone, and in 2-250 mL flasks containing either 80 or 40 mL of the same medium at 20 °C for 48 h. Pyoluteorin production was also examined under oxygen-limiting conditions. Cultures in 2-250 mL flasks containing either 80 or 40 mL of the modified medium were incubated without aeration at 20 °C for 48 h. Cultures maintained without shaking grew poorly and were not analyzed. Temperature effect: The temperature effect on pyoluteorin production was investigated for both strains JL4092 and JL4239. Strains were grown in modified King's medium B containing phytone peptone and either fructose or glycerol at 18 °C, 22 °C, and 28 °C for 48 h. Extractions and analysis for metabolites were performed as above.

Production of pyoluteorin by Pf-5 in culture. For feeding studies and routine isolation of pyoluteorin, cultures of Ps. fluorescens Pf-5 were grown in a modified King's medium B (referred to hereafter as phytone medium) formulated as follows: 40

phytone peptone 2.0% Glycerol 0.5% K2HPO4 0.15% MgSO4 7 H2O 0.15%

All percentages are mass/volume. Medium was made to volume with deionized water and adjusted to between pH 7.0-7.2. Production was optimal in 40 mL cultures contained in 250 mL Erlenmeyer flasks; however, large scale cultures were grown in 100 mL of phytone medium contained in 1 L Erlenmeyer flasks. All cultures were incubated at 19-22 °C at 150 RPM for 42-48 h.

Isolation of pyoluteorin. Cells were removed from a 1 L culture of Ps. fluorescens Pf-5 (10-1 L flasks containing 100 mL phytone medium) by centrifugation and the supernatant was acidified to pH < 2 with 6N HC1. The solution was extracted with ethyl acetate (10% vol, 3x) and the pooled organic layers back extracted with water (5% vol, 3x). The organic layer was dried over anhydrous MgSO4 and the solvent was subsequently removed by rotary evaporation. The crude extract was adsorbed onto a small amount of silica gel and applied to a 2.5 x 15 cm silica gel (Silica Gel 60, 40-63 EM Science, Gibbstown, NJ) column equilibrated in either 3:1 toluene/acetone or 4:1 CHC13/acetone and fractionated by flash chromatography (Still et al, 1978) with the equilibrating solvent. The material thus isolated was further purified by recrystallization from hot CHC13 and was spectroscopically identical to pyoluteorin. 1H NMR (300 MHz; d6-acetone) 8 9.02 (br s, exchangeable), 7.15 (t, 1H, J = 8.2 Hz), 6.80 (s, 1H), 6.47 (d, 2H, J = 8.2 Hz), and 3.00 (br s, exchangeable). 13C NMR (75 MHz; d6-acetone) 8 183.1, 157.2, 132.5, 130.9, 119.5, 117.3, 113.0, 110.5, and 107.5.

L-14C1 feeding study. Phytone medium (40 mL in 250 mL shake flask) was inoculated with Ps. fluorescens Pf-5 (over-producing mutant JL4239). The culture was grown under standard conditions (see above) for 24 41 h. One milliliter of this culture was used to inoculate 100 mL phytone medium in each of two, 1-L shake flasks. After 16 h, 9.99 ptCi (2.20 E 7 dpm) of [U -14C] L-proline (ICN, Irvine, CA) dissolved in 18 mL H2O was divided into two equal portions and added aseptically to each 100 mL culture. Both cultures were combined 44 h after inoculation and pyoluteorin was extracted from the culture supernatant. Authentic, unlabelled pyoluteorin (20 mg) was added to the crude extract, and the combined samples purified by flash chromatography yielding 56 mg (207 1.unol) of pyoluteorin. The sample was recrystallized to within 1% of constant specific radioactivity (1.18 E 5 ± 1.05 E 3), indicating a 19 % incorporation of proline into pyoluteorin.

[1-"a-L-Proline feeding study. A 40 mL seed culture of Ps. fluorescens Pf-5 (wild-type strain JL4239) was grown under standard conditions (see above) for 24 h. One milliliter of this culture was used to inoculate 100 mL phytone medium in each of ten 1-L shake flasks. Sixteen milligrams of [1-13C] -L-proline (CIL, Cambridge, MA) was dissolved in 10 mL H2O and equal portions added aseptically to each of the culture flasks 13 h after inoculation. 48 h after culture inoculation, pyoluteorin was isolated (47 mg; 173 innol) from the combined cultures and recrystallized from hot CHC13. An enrichment of 5.7% over natural abundance of the pyoluteorin carbonyl resonance (8 183.0) was observed in the 13C NMR spectrum (75 MHz; d6-acetone).

Synthesis of deschloropyoluteorin (2-(2',6'-dihydroxybenzoy1)-1H pyrrole; DCP; Scheme 2-12). 2,6-Dimethoxybenzoyl chloride (2.0 g; 9.97 mmol) and pyrrole (1 mL; 14.4 mmol) were combined in 20 mL dichloroethane and cooled to 5 °C in an ice/water bath. A slurry of A1C13 (1.5 g; 11.25 mmol) suspended in 20 mL anhydrous CH2C12 was added to the mixture over 30 min with stirring under an inert atmosphere. The reaction was allowed to warm to room temperature and stirred for 1 h. Work up of the reaction and purification of the desired compound was as described (Rao and Reddy, 1990). Purification yielded 820 mg (40% yield) of 2,6-dimethoxy-DCP: mp 197-198 °C [lit. (Cue et al, 1981) mp 192-195 °C[;1H NMR (300 MHz; d6 -DMSO) 8 11.85 (br s, exchangeable), 7.34 (t, 1H, J = 8.5 Hz), 7.08 (m,41-1), 6.70 (d, 2H, J = 42

8.5 Hz), 6.29 (m, 1H), 6.10 (m, 1H), 3.65 (s, 6H); 13C NMR (75 MHz; d6-acetone) 8 184.5, 158.7, 136.9, 131.3, 125.6, 119.4, 118.6, 110.8, 105.1, and 56.2; EI HRMS m/z 231.0896 (calc. exact mass for C13H13NO3 231.0895). The 2,6-dimethoxy derivative (800 mg, 3.46 mmol) was subsequently dissolved in 115 mL of benzene and added to a vigorously stirred solution of A1Br3 (3.0 g; 11.28 mmol in 30 mL benzene). The bright orange reaction mixture was stirred for 5 h at room temperature. The reaction was then added to a 1 L flask containing 130 mL of 3 N HCl and 30 mL EtOAc and stirred for 30 min. The aqueous layer was extracted with EtOAc (3 x 30 mL) and the combined organic layers dried over anhydrous MgSO4. The crude extract was concentrated to dryness and fractionated on a flash silica gel column (5 x 15 cm) eluted in CHC13 and Me0H (9:1) yielding 510 mg of DCP (72% yield): mp 143-145 °C [lit. (Takeda, 1958) mp 142-145 °C]; UV. (HPLC solvent) 298 nm; 11-1 NMR (300 MHz; d6-DMS0) 8 11.76 (br s, 1H), 9.34 (s, 2H), 7.06 (m, 1H), 6.98 (t, 1H, J = 8.1 Hz), 6.39 (m, 1H), 6.35 (d, 2H, J= 8.2 Hz), and 6.11 (m, 1H); 13C NMR (100 MHz; d6 -DMSO) 8 183.6, 155.6, 133.3, 129.6, 125.1, 118.1, 116.1, 109.6, and 106.4; EI HRMS m/z 203.0583 (calc. exact mass for CI IH9N03203.0582).

Deschloropyoluteorin "C-isotope trapping experiment. Two sets of duplicate production cultures (a total of 4-40 mL cultures in 250 mL flasks) were used in this experiment. One set was comprised of two flasks containing phytone medium and the two flasks within the other set contained phytone medium in which glucose was substituted for glycerol (phytone / glucose medium). Each of the production cultures were inoculated with 400 JAL of a Ps. fluorescens Pf-5 (JL4092) seed culture grown for 24 h. After 9.5 h, an aqueous solution of [1-14C1-sodium acetate (10 liCi) was added to each of the production cultures. The cells were collected by centrifugation from one phytone and one phytone/glucose culture at 14.5 h and 18.5 h following inoculation. The cells were then resuspended in 3-5 mL of the culture supernatant and an equal weight of micro glass beads. The suspension was sonicated on ice using a probe-type sonicator for 15 min at 90% full power. The sample was then centrifuged to remove cell debris and glass beads. Approximately 60 mg of DCP (in Me0H) was added to each lysate solution. 43

The DCP was isolated from the cell lysate using the methodology described for pyoluteorin with one important addition. Since DCP and pyoluteorin co-elute from the flash silica gel column, it was necessary to further purify the sample by reverse phase chromatography. The DCP/pyoluteorin mixture was applied to a 1.5 x 15 cm C18 reverse-phase silica gel column (Bakerbond; 40 gm; JT Baker, Phillipsburg, NJ) and eluted with 50% aqueous Me0H. Repeated recrystallization from benzene of the recovered DCP resulted in complete loss of detectable radioactivity. The combined fractions containing pyoluteorin recovered from the C18 silica column were determined to contain approximately 2.8% of the total radioactivity that had been added to the cultures. Unlabelled pyoluteorin was added to these fractions, but it was not possible to recrystallize pyoluteorin to determine accurate levels of incorporation. Therefore, a sample of the recovered pyoluteorin was resolved from impurities by 2D-TLC (silica gel; 4:1 toluene-acetone; 9:1 CHC13-Me0H) and subjected to scintillation counting.

Deuterium exchange of deschloropyoluteorin with 21-1-trifluoroacetic acid (Scheme 2-13). DCP (64 mg; 315 gmol) was dissolved in 1 mL of 2H trifluoroacetic acid and 700 gL of 21120. The solution was placed in a thick walled, glass tube fitted with a telfon stopcock and frozen in a dry ice bath. The tube was repeatedly evacuated and flushed with N2 (4 x). After the final flush, the tube was sealed and heated in an oil bath at 75 °C overnight. Solvent was removed under high vacuum and the crude product was eluted on a flash silica gel column with 9:1 CHC13 -MeOH yielding 50 mg of purified product (76% yield). The extent of labelling was estimated as >90% for the pyrrole ring and -70% for the m eta-positions of the resorcinol ring, based on the integration of resonances within the 1H NMR spectrum. Since the carbon resonance corresponding to the pa ra-position within the resorcinol ring exhibited limited isotopic splitting, it was presumed that limited deuterium exchange occurred at this position. Therefore proton integrations were normalized to the para-proton resonance. 44

2H- Deschloropyoluteorin feeding study. Phytone production medium (5 x 100 mL in 1 L shake flasks) was inoculated with 800 'IL of an overnight culture of Ps. fluorescens Pf-5 (JL4092). 400 piL of a 2H-DCP solution (50 mg in 2 mL EtOH) was added aseptically to each of the production cultures at 20 h after inoculation. Cultures were incubated until 48 h after inoculation when the supernatant was extracted as described above. The crude extract was fractionated on a 2 x 20 cm C 18 reverse-phase silica column (Bakerbond; 40 gm; JT Baker, Phillipsburg, NJ) and eluted with a 3:2 Me0H-water mixture. Both DCP and pyoluteorin were recovered (52.2 mg and 15.5 mg, respectively). 2H NMR analysis of pyoluteorin showed no evidence of 2H-DCP incorporation. The 1H NMR spectrum for the recovered DCP was identical to authentic material and indicated the exchange of the label had not occurred during the course of the experiment.

Cell permeability study. Cultures (5 mL) of Ps. fluorescens Pf-5 (JL4092) were grown in phytone medium for 15 h at which time individual cultures received one of the following: 100 pit toluene, 100 pil DMSO, 50 mg mixed alkyltrimethylammonium bromide (primarily C14), 50 mg cetylpyridinium bromide, 50 mg methyltrioctylammonium chloride, PMBN to 1 pig mL-1 and 3 pig mL-1, or poly-L-lysine (MW 4000-15000) to 8 pig mL-1 and 16 lig mL-1. Additional 5 mL cultures were prepared in which MgC127H20 was omitted from the phytone medium and replaced with EDTA to a final concentration of 6.5 pig mL-1. The phytone/EDTA cultures were amended with either PMBN to 1 pig mL-1 and 3 pig mL-1 or poly-L-lysine to 8 pig mL-1 and 16 pig mL-1, 15 h after inoculation. All of the cultures also received 10 pit of a stock DCP solution (24 mg dissolved in 1.2 mL) at the same time point. Cultures were allowed to grow for an additional 26 h. Pyoluteorin production and changes in DCP concentration were analyzed qualitatively by HPLC as described above.

Synthesis of 2,6-dimethoxy-['4C0]-benzoyl chloride (Scheme 2-14). 1,3 Dimethoxybenzene (261 piL; 276 mg; 2 mmol) and n-butyl lithium (1.25 mL of a 1.6 M hexaneane solution; 2 mmol) were combined in a round bottom flask (25 mL) containing a stir bar and 5 mL anhydrous THF. The solution 45 was refluxed for 4 h under Ar, after which time the condenser was removed and the flask fitted with a rubber septum. A mixture of 220 mg Na213CO3 (CIL, Cambridge, MA) and 250 gCi Na214CO3 (ARC, St. Louis, MO) was placed in a separate round bottom flask (10 mL) containing a stir bar and fitted with a rubber septum. Both flasks were frozen in liquid N2, evacuated, and flushed with Ar (3x). Upon final evacuation, the two flasks were connected via candula needle and allowed to warm to room temperature. Concentrated H2SO4 (5 mL) was added to the Na2CO3 mixture and the organolithium reagent was again frozen in liquid N2 until CO2 evolution ceased. The candula needle was removed, and the reaction was warmed to room temperature and stirred for 24 h. The residue remaining in the reaction flask was dissolved in 10% NaHCO3 and extracted with Et20. The aqueous phase was subsequently acidified to pH 2 with 6 N HC1 and the reaction product extracted into Et20. The organic layer was dried over anhydrous MgSO4 and evaporated, yielding 209 mg of [14C0] -2,6- dimethoxybenzoic acid (0.055 gCi mg-1; 78% yield): mp 183-185 °C [lit (Aldrich, 1997) mp 186-187 °C]; 1H NMR (300 MHz; CDC13) 7.33 (t, 1H, J = 8.4 Hz), 6.60 (d, 2H, J = 8.4 Hz), and 3.88 (s, 6H); 13C NMR (75 MHz; CDC13) 8 170.6, 157.8, 131.8, 111.5, 104.0, and 56.1; EI HRMS m/z 182.0579 (calc. exact mass for C9H1003 182.0579). The entire sample of the labelled benzoic acid derivative was dissolved in a minimal volume of SOC12 and 1 drop of dimethylformamide was then added. The reaction was stirred overnight at room temperature under Ar. The solvent was subsequently removed under high vacuum and the material used directly in the Friedel-Crafts reaction as described above yielding 95.5 mg of labelled 2,6-dimethoxy-DCP, approximately half (48 mg) of which was subsequently converted to labelled DCP (38 mg, sp. act. 0.982 tiCi

P4C01-deschloropyoluteorin feeding study. Ps. fluorescens Pf-5 (JL4092) cultures (3 x 90 mL in 1 L shake flasks) grown in phytone medium lacking MgC127H20 under standard conditions were started from one milliliter of an overnight culture. Cultures were amended with PMBN (3 pg mL-1) and EDTA (6.5 pg mL-1) 14 h after inoculation. The labelled DCP (14.73 mg in 3 mL DMSO) was simultaneously added to the cultures and growth continued 46

until 50 h after initial inoculation. Pyoluteorin was isolated using the method described for the 2H-DCP feeding study. Following three recrystallizations of the recovered pyoluteorin from hot CHC13, no radioactivity was detected.

rOD21-1)-Pyrrole-2-carboxylic acid synthesis (PCA; Scheme 2 -15). In a round bottom flask fitted with a rubber septum and stir bar, 20 mL of Et20 was cooled to -10 °C followed by addition of 10 mL of EtMgBi (3.0 M ethereal solution). A solution of 2.08 mL pyrrole (2.0 g; 29.9 mmol) in 15 mL Et20 was added drop-wise and the reaction stirred for 15 min. The reaction mixture was frozen in liquid N2, evacuated, and allowed to thaw in an ice bath (3x). Upon final thawing, the reaction flask was connected via a candula needle to a second flask containing 2.05 g Na213CO3 (CIL, Cambridge, MA) and a stir bar under Ar. Concentrated H2SO4 (10 mL) was added to the Na213CO3 and the Grignard reagent was immediately immersed in liquid N2 until CO2 evolution ceased. The candula needle was removed, and the reaction warmed to room temperature over 3 h. The mixture was then poured into an ice slurry of 3 M H2SO4 and the reaction product partitioned into Et20. The organic layer was dried over anhydrous MgSO4, treated with activated charcoal, and filtered. Crystallization was induced by partial evaporation of the solvent and refrigeration, yielding 975 mg of while crystals: mp 204-207 °C (dec) [lit. (Letellier and Bouthillier, 1957) mp 204-207 °C]; 1H NMR (400 MHz; d6-acetone) 8 10.89 (br s), 9.25 (br s), 7.06 (m, 1H), 6.87 (m, 1H), and 6.23 (m, 1H); 13C NMR (100 MHz; d6-acetone) 8 162.9, 124.5, 123.5, 116.2, and 110.6; EI HRMS for [12CO2H] -PCA m/z 111.0320 (calc. exact mass for C5H5NO2 111.0320); EI LRMS for [13CO2F1] -PCA m/z 112.1. Calculated 13C-enrichment of the synthetic sample was 96% based on LRMS data.

FI313211]-Pyrrole-2-carboxylic acid feeding study. Phytone medium (3 x 100 mL in 1 L shake flasks) was inoculated with one milliliter of an overnight culture ofPs. fluorescens Pf-5 (over-producing mutant JL4239). Cultures were grown under standard conditions. Labelled PCA (75 mg) was dissolved in 18 mL of 0.1 M NaHCO3 and administered in equal portions to the cultures at 14 h and 21 h after inoculation. After 46 h of growth, pyoluteorin was isolated 47

from the culture supernatant as described above. The sample (47 mg) was recrystallized from hot benzene and analyzed by 13C NMR. Comparison of 13C line intensities for the isolated sample to those of an identical, unlabelled pyoluteorin sample did not reveal any evidence of isotopic enrichment.

(13C0]-Pyrrole-2-carboxylic acid N-acetylcysteamine thioester (NAC PC0]-PCA; Scheme 2-16). N,S-Diacetylcysteamine (2.4 g; 14.8 mmol) was hydrolyzed in 80 mL aqueous KOH (2.8 g, 49.9 mmol) for 2 h. The reaction pH was adjusted to 5.0 with 2 N HCl and the mixture was saturated with NaCl. The brine solution was subsequently extracted with dichloromethane yielding N-acetylcysteamine as a clear oily residue: 1I-1 NMR (300 MHz; CDC13) 8 7.03 (br s, 1H), 3.28 (q, 2H, J = 6.6 Hz), 2.53 (dt, 2H, J= 6.6 Hz and 8.4 Hz), 1.89 (s, 3H), and 1.36 (t, 1H, J = 8.4 Hz). An aliquot of the N-acetylcysteamine preparation (800 A) was combined with [13CO2H] -PCA (250 mg, 2.23 mmol) and 4-dimethylaminopyridine (27.5 mg, 0.23 mmol) in 15 mL of anhydrous dichloromethane. The mixture was cooled to 0 °C and the flask flushed with Ar. 1,3-Dicyclohexaneylcarbodiimide (460 mg, 2.71 mmol) was added and the reaction was maintained at 0 °C with stirring. The reaction flask was subsequently removed from the ice bath, allowed to warm to room temperature, and stirring continued for 4 h. The reaction mixture was then filtered through celite and placed under high vacuum overnight. Purification of the NAC-thioester was afforded by flash silica chromatography (2.5 x 15 cm column, 4:1 EtOAc- hexane) followed by recrystallization from 1,1,1-trichloroethane. A total of 195 mg of pure thioester was obtained: mp 110-111 °C ; UVmax (Et0H) 296 nm; IR (neat) 3304, 2930, 2353, 1628, and 1598 cm-1;11 NMR (300 MHz; d6-acetone) 8 7.31 (br s), 7.11 (m, 1H), 6.91 (m, 1H), 6.19 (m, 1H), 3.32 (q, 2H J= 6.5 Hz), 3.06 (t, 2H, J = 6.5 Hz), 2.82 (s, 1H), and 1.82 (s, 3H); 13C NMR (75 MHz; d6-acetone) 8 180.1, 169.5, 130.2, 124.8, 115.2, 110.3, 39.5, 27.6, and 22.2; EI HRMS for NAC-P2CORCA m/z 212.0619 (calc. exact mass for C9H12N202S 212.0619); EI LRMS for [13CO21-1] PCA m/z 213.1. Calculated 13C-enrichment of the synthetic sample was 96% based on LRMS data. 48

(13C0]-Pyrrole-2-carboxylic acid N-acetylcysteamine thioester feeding study. Ps. fluorescens Pf-5 (JL4092) cultures (10 x 100 mL in 1 L shake flasks) grown in phytone medium under standard conditions were started from 800 !IL of an overnight culture. NAC-[13C0]-PCA (92.3 mg) was dissolved in 4 mL DMSO and equal portions were administered to the cultures at 13.5, 15.5, 18, and 20 h after inoculation. Pyoluteorin (15 mg) was isolated from the combined culture supernatants 48 h after inoculation as described above. No enrichment was observed for the 13C carbonyl resonance in the NMR spectrum of the isolated pyoluteorin sample.

Synthesis of protected tetrahydrodeschloropyoluteorin (242-hydroxy 6'-methoxybenzoy1)-1H-pyrrolidine; Scheme 3-17). 2,6-dimethoxy-DCP (203 mg, 0.88 mmol) potassium-tert-butoxide (100 mg, 0.89 mmol), and di-tert-butyl dicarbonate (225 210.9 mg, 0.97 mmol) were refluxed in 30 mL anhydrous THE under Ar for 2 h. The reaction was washed twice with saturated NaCI. The reaction was fractionated on a silica gel flash column (1 x 15 cm) eluted with 3:7 EtOAc- hexane, yielding 218 mg of the N-protected compound: mp 66-68 °C; UVmax (MeOH) 212, 248, and 286 nm; IR (CHC13) 2975, 2944, 1753, 1673, 1605, and 1475 cm-1; 1H NMR (300 MHz; CDC13) 8 7.40 (dd, 1H, J = 3.4 and 1.5 Hz), 7.29 (t, 1H, J = 8.4 Hz), 6.58 (dd, 1H, J = 3.4 and 1.5 Hz), 6.54 (d, 2H, J = 8.4 Hz), 6.10 (dd, 1H J = 3.4 and 1.5 Hz), 3.74 (s, 6H), and 1.60 (s, 9H); 13C NMR (75 MHz; CDC13) 6 182.6, 157.9, 149.3, 134.9, 130.7, 128.9, 124.3, 118.8, 109.8, 104.0, 84.5, 55.9, and 27.4; EI HRMS m/z 331.1418 (calc. exact mass for C18H21N05 331.14197). The N-protected pyrrole (164 mg, 0.495 mmol) was catalytically hydrogenated over 10% Pd/C (77 mg) in MeOH (2 mL) containing 2 drops conc. HC1 under 1 atm H2 for 36 h with vigorous stirring. The suspension was then filtered through glass wool and the filtrate triturated with CHC13 forming a white solid. The precipitate was chromatographed on silica gel with 1:1 EtOAc- hexane yielding 138 mg of the 2,6-dimethoxy-tBoc-pyrrolidine as an oily residue: UV. (MeOH) 257 (shoulder) and 276 nm; IR (CHC13) 2970, 2933, 2884, 1726, 1707, 1605, and 1473 cm-1;1H NMR (300 MHz; d4 -MeOH) 8 7.37 (t, 1H, J = 8.4 Hz), 6.71 (d, 2H, J = 8.4 Hz), 4.94 (t, 1H, J= 5.7 Hz), 3.80 (s, 6H), 3.43 (m, 2H), 2.04 (m, 1H), 1.89 (m, 2H), and 1.41 (s, 9H); 13C NMR (75 49

MHz; CDC13) 8 200.8, 157.5, 154.5, 131.0, 117.9, 103.9, 79.1, 66.2, 55.7, 46.2, 28.5, 28.1, and 22.9; EI FIRMS m/z 336.1786 (calc. exact mass for Ci8H25N05336.18110). The t-Boc protecting group was removed by treatment of the pyrrolidine (138 mg, 0.412 mmol) with 1.5 mL trifluoroacetic acid under an inert atmosphere at 0 °C for 40 min. Evaporation of the solvent under high vacuum yielded the 2,6-dimethoxy-pyrrolidine derivative as an oily residue: UVmax (MeOH) 214 and 274 nm; IR (CHC13) 2966, 2842, 2768, 1790, 1722, 1673, and 1592 cm-1; 11-1 NMR (300 MHz; CDC13) 8 9.59 (br s), 7.40 (t, 1H, J = 8.4 Hz), 6.60 (d, 2H, J = 8.4 Hz), 5.21 (br s, 1H), 3.80 (s, 6H), 3.53 (m, 2H), 2.13 (m, 1H), 2.11 (m, 2H), and 2.07 (m, 1H); 13C NMR (75 MHz; CDC13) 8 197.1, 158.1, 133.8, 113.4, 104.2, 66.9, 56.0, 47.0, 27.9, and 23.9; EI FIRMS m/z 236.1287 (calc. exact mass for C13H18NO3 236.1287). The 2,6-dimethoxy pyrrolidine derivative (18 mg, 0.52 mmol) was dissolved in 2 mL anhydrous dichloromethane, was cooled to -70 °C and BBr3 (510 !IL of 1.0 M solution) was added drop-wise. The reaction was removed from the dry ice bath and allowed to warm to room temperature and stirred for 2 h. The reaction mixture was poured into an ice slurry (20 mL), stirred for an additional 20 min, saturated with NaC1, and subsequently extracted into dichloromethane. Reaction work up yielded 6.5 mg of product. 1H NMR analysis revealed that only a single methyl group was removed and therefore, complete characterization of the compound was not undertaken. 1F1 NMR (300 MHz; CDC13) 8 8.36 (br s), 7.45 (t, 1H, J = 8.4 Hz), 6.61 (d, 1H, J = 8.4 Hz), 6.42 (d, 1H, J = 8.4 Hz), 5.50 (br s, 1H), 3.95 (s, 3H), 3.63 (m, 2H), 2.64 (m, 1H), 2.12 (m, 2H), and 1.91 (m, 1H).

Synthesis of labelled tetrahydrodeschloropyoluteorin (2-(2',6'-dihydroxy 13CO/4C01-benzoy1)-1H-pyrrolidine; [13C0114C0]-THDCP; Scheme 2-18). Labelled DCP (63 mg, 0.310 mmol) was dissolved in 7 mL dimethylformamide containing Na2CO3 (72 mg, 0.683 mmol) and benzyl bromide (81 III., 116 mg, 0.683 mmol). The reaction flask was charged with Ar and heated for 3 h at 60-65 °C. Solvent was removed overnight under vacuum and the residue dissolved in 5-10 mL of EtOAc. The solution was subsequently washed with 1N HCl (2x) followed by saturated NaCl (2x). Benzylated product was purified via flash silica chromatography (1 x 15 cm, 7:13 EtOAc- hexane) as an oily 50

residue: UV,.. (EtOH) 295 nm; IR (CHC13) 3393, 2965, 2365, 1604, and 1551 cm-1; 1H NMR (300 MHz; CDC13) 5 9.98 (br s), 7.28 (m, 4H), 7.17 (m, 2H), 7.01 (m, 2H), 6.67 (d, 1H, J = 8.0 Hz), 6.57 (d, 1H, J = 8.1 Hz), 6.24 (m, 1H), and 5.06 (s, 2H);13C NMR (75 MHz; CDC13) 8 184.7, 159.4, 157.6, 136.2, 133.4, 132.6, 128.3, 127.7, 126.9, 125.6, 120.7, 113.8, 111.0, 110.5, 104.3, and 70.6; EI HRMS m/z 293.1045 (calc. exact mass for C18H15NO3 293.1052). O-Benzyl-DCP (98 mg, 0.334 mmol) was combined with potassium tert butoxide (93.8 mg, 0.836 mmol) and di- tert-butyl dicarbonate (192 182.4 mg, 0.836 mmol) and the mixture refluxed in 25 mL anhydrous THE under Ar for 2.5 h. The reaction products were fractionated on a silica gel column (2 x 15 cm, 1:4 EtOAc- hexane) and the fully protected product was recrystallized from 1,1,1-trichloroethane and hexane yielding a white crystalline compound: mp 119-120 °C; UVmax (EtOH) 254 and 287 nm; IR (CHC13) 2977, 2936, 2876, 2359, 1759, and 1720 cm-1; 1H NMR (300 MHz; CDC13) 8 7.42 (m, 1H), 7.35 (t, 1H, J = 8.2 Hz), 7.26 (m, 3H), 7.26 (m, 2H), 6.86 (d, 1H, J = 8.2 Hz), 6.84 (d, 1H, J = 8.2 Hz), 6.58 (m, 1H), 6.11 (m, 1H), 5.03 (s, 2H), 1.49 (s, 9H), and 1.40 (s, 9H); 13C NMR (75 MHz; CDC13) 8180.5, 157.2, 151.4, 149.3, 149.0, 136.2, 134.6, 130.8, 128.4, 128.3, 127.7, 126.9, 124.1, 123.4, 115.5, 110.2, 110.0, 84.6, 83.4, 70.6, 27.4, and 27.3; EI HRMS m/z 493.2098 (calc. exact mass for C28H31N07 493.2101); Elemental analysis: found C 68.10, H 6.42, N 2.78, 0 22.70 (anal. calc. for C28H31N07: C 68.18, H 6.34, N 2.84, 0 22.64). Hydrogenation of the protected pyrrole derivative (114 mg, 0.231 mmol) was accomplished over 10% Pd/C (20 mg) in 3 mL EtOH under 1 atm 112 for 5 h. The catalyst was removed by filtering the reaction through glass wool and the products fractionated by flash silica chromatography (1.5 x 15 cm, 3:17 EtOAc- hexane) yielding 85 mg of N,O- t- Boc- pyrrolidine as an oily residue: U Vmax (EtOH) 210 (shoulder), 254, and 312 nm; IR (CHC13) 3150, 2977, 2936, 2882, 2246, 1770, 1711, and 1693 cm-1; 1H NMR (300 MHz; CDC13) 8 12.44 (s, 1H), 11.80 (br s, 1H), 7.43 (t, 1H, J = 8.3 Hz), 7.36 (t, 1H, J= 8.3 Hz), 6.88 (dd, 1H, J = 1.0 and 8.3 Hz), 6.85 (dd, 1H, J = 1.0 and 8.3 Hz), 6.73 (dd, 1H, J = 1.0 and 8.3 Hz), 6.70 (dd, 1H, J = 1.0 and 8.3 Hz), 5.13 (m, 2H), 3.67 (m, 2H), 3.47 (m, 2H), 2.38 (m, 1H), 2.24 (m, 1H), 1.93 (m, 6 H), 1.56 (s, 9H), 1.53 (s, 9H), 1.45 (s, 9H), and 1.23 (s, 9H); 13C NMR (75 MHz; CDC13) 8 205.6, 203.6, 163.8, 161.5, 154.9, 153.6, 150.9, 150.8, 150.6, 150.0, 135.4, 134.2, 116.1, 116.0, 114.9, 113.5, 113.3, 112.4, 84.6, 84.1, 80.4, 79.7, 64.9, 64.7, 46.9, 46.6, 30.3, 29.7, 29.5, 28.3, 27.9, 27.6, 51

23.8, and 22.9; EI HRMS m/z 407.1944 (calc. exact mass for C211-129N07 407.1944). Full deprotection of the N,O- t- Boc- pyrrolidine intermediate (84 mg, 0.206 mmol) occurred upon treatment with trifluoroacetic acid at 0 °C for 20 min. Solvent was removed under high vacuum and the residue recrystallized from EtOAc and pentane, yielding 35 mg of [13C0/14C0] -THDCP (sp. act. 0.982 m.Ci mg-1): mp 186-190 °C (dec); UVmax (EtOH) 278 and 356 nm; IR (CHC13) 2995, 2757, 2365, 1670, 1634, and 1610 cm-1; 111 NMR (300 MHz; d4-Me0H) 8 7.35 (t, 1H, J = 8.2 Hz), 6.43 (d, 2H, J = 8.2 Hz), 5.50 (dd, 1H, J= 5.5 and 9.3 Hz), 3.42 (m, 2H), 2.67 (m, 1H), 2.09 (m, 2H), and 1.98 (m, 1H); 13C NMR (75 MHz; d4 -MeOH) 8 202.6, 166.3, 142.1, 111.2, 110.9, 70.1, 50.0, 33.2, and 26.9; EI HRMS m/z 207.08950 (calc. exact mass for C11H13NO3 207.08954).

[13COP4C0]-Tetrahydrodeschloropyoluteorin feeding study. Ps. fluorescens Pf-5 (JL4092) cultures (7 x 100 mL in 1 L shake flasks) inoculated from a common seed culture were grown in phytone medium under standard conditions. Labelled THDCP (35 mg, 12.9 liCi) was dissolved in 14 mL H20, 1 mL of which was added to each culture 13 h and 16 h after inoculation. Cultures were grown until 42 h after inoculation. Pyoluteorin was isolated using standard methods and analyzed by 13C NMR. No enrichment above natural abundance was detected; therefore, the sample was recrystallized to constant specific radioactivity from hot CHC13. Specific activity was calculated with and without background correction (31.6 dpm). 52

Chapter 3. Nucleotide Sequence Analysis of Loci Required for Pyoluteorin Biosynthesis

3.1 INTRODUCTION

A genomic region was identified previously within Ps. fluorescens Pf-5 that is required for pyoluteorin production (Kraus and Loper, 1995; see Chapter 1). Based on the fact that this region spanned approximately 21-kb and that transposon insertions within this region appeared to only affect pyoluteorin production, it was suggested that the structural genes encoding the pyoluteorin biosynthetic enzymes were localized within this 21-kb fragment. Because pyoluteorin is a polyketide-derived metabolite (see Chapter 2), it was likely that genes encoding a polyketide synthase could be identified via nucleotide sequence analysis of the 21-kb genomic region. Although the principal objective of this study was to identify the pyoluteorin polyketide synthase, other loci also were identified that are likely responsible for chlorination of a pyoluteorin biosynthetic intermediate(s). The identification of genes encoding the polyketide synthase and the halogenases required for pyoluteorin biosynthesis is intriguing in that the polyketide synthases have been characterized extensively whereas halogenases are an almost unknown class of enzymes.

3.1.1 Genetic organization of polyketide synthases

The biochemical pathways as well as the biosynthetic enzymes required for fatty acid and polyketide biosynthesis share many similarities (Hopwood and Sherman, 1990; Katz and Donadio, 1993). Both processes involve the serial condensation of short chain CoA-thioesters (usually acetate-derived) and the subsequent reductive modification of the extended carbon chain. These reductive modifications involve the same types of transformations, namely, carbonyl reduction, dehydration, and enoyl reduction. Whereas fatty acid synthases (FASs) redundantly catalyze each one of these reactions and 53 generally recognize only malonyl-CoA as a substrate, PKSs exhibit remarkable control over the extent to which each newly added carbon unit is modified. Furthermore, PKSs can utilize a variety of carboxylic acid., (e.g. acetate, propionate, or butyrate) as starter units (see Table 2.1) or for extending the carbon chain. Therefore, polyketides generally possess an elaborate carbon skeleton containing several oxidized functionalities. The protein structures of fatty acid and polyketide synthases are also similar, and these enzymes are classified as either Type I or Type II proteins. Like the Type I FASs present in vertebrates, Type I PKSs are large multifunctional proteins containing several catalytic domains required for polyketide assembly. These domains are organized into one or more enzymatic modules, each containing a minimal core set of domains: a ketosynthase (KS), an acyltransferase (AT), and an acyl carrier protein (ACP). Additional enzymatic domains, such as a ketoreductase (KR), may also be present within an individual Type I PKS module and are located between the AT and the ACP domains. In Type I PKSs possessing several modules, each module is responsible for the addition and modification of a single extender unit. In addition, these enzymes are processive and typically require f3-carbonyl modification prior to addition of the next carbon unit.

module

6-Methylsalicylate Synthase Penicillium patulutn I KS AT DH KR ACF> (5.2 kb)

0 0 KR DH KS/AT SE SEnz o 0 -1111,'

Figure 3-1. Genetic organization and part of the reaction pathway catalyzed by a Type I PKS. For this example, a single enzymatic module is encoded by a single ORF. Only the reduction, dehydration, and the final condensation within the pathway are presented to illustrate the processive mechanism of the Type I PKSs. Abbreviations are as follow: KS, ketosynthase; AT, acyltransferase; DH, dehydrase; KR, ketoreductase; ACP, acyl carrier protein; SEnz, thioester linkage to the enzyme. 54

The genes encoding several Type I PKSs have been characterized from both bacterial and fungal sources. Fungal Type I PKSs generally are encoded by a single ORF containing one module and are responsible for biosynthesis of an aromatic metabolite, as exemplified by the 6-methyl salicylate synthase (6MSAS) from Penicillium patulum (Figure 3-1; Beck et al, 1990). Bacterial Type I PKSs, however, often contain multiple modules distributed among several proteins and generally synthesize non-aromatic, macrolide antibiotics. Examples of bacterial Type I PKSs exist in which 2, 4, 5, or 6 modules are present within a single peptide (Cortes et al, 1990; Donadio et al, 1991; Aparicio et al, 1996).

Actinorhodin PKS Streptomyces coelicolor A3

KR KS AT I CLF ACP Cyclase Cyclase

minimal PKS

PKS 0 0 0 PKS/KR 10x Acetate --Yo 1111" 0 0 0 OH sEnz HO 2 Actinorhodin

Figure 3-2. Genetic organization and part of the reaction pathway catalyzed by a Type II PKS. Only the first reduction and cyclization within the pathway are presented to illustrate the non-processive mechanism of the Type II PKSs. ORFs are indicated by open boxes. Abbreviations are as follow: KS, ketosynthase; AT, acyltransferase; CLF, chain length determining factor; KR, ketoreductase; ACP, acyl carrier protein; SEnz, thioester linkage to the enzyme.

In contrast to the multifunctional protein Type I systems, the Type II PKSs are multienzyme complexes comprised of a few, principally mono- functional proteins (Figure 3-2). Three essential proteins (a bifunctional KS/AT, a chain length-determining factor (CLF), and an ACP) are present in 55 all of the Type II PKSs described to date (Hutchinson and Fujii, 1995). The KA/AT, CLF, and ACP comprise a minimal PKS, which is necessary and sufficient for polyketide biosynthesis (McDaniel et al, 1994). Unlike the processive mechanism of the Type 1 PKSs, active sites within the minimal Type II PKS are used iteratively to form an extended poly -(3- ketone chain prior to any reductive modifications of the carbonyls (i.e. a non-processive mechanism). The regiospecificity of the cyclization process is dependent upon several cyclases and, in some instances a ketoreductase, that function in concert with the minimal PKS (Hutchinson and Fujii, 1995). Other proteins within the Type II PKSs are thought to be required for starter unit specificity or post assembly modifications of functional groups present within the polyketide structure. Although the resveratrol and chalcone synthases are distinct from the polyketide synthases, they nevertheless are responsible for the formation of polyketide natural products. These synthases are single, homodimeric proteins that do not require the 4'-phosphopantetheinyl prosthetic group necessary for the PKSs (Kreuzaler et al, 1979; Schroder et al, 1988), and they possess a single cysteine residue that is required for activity (Lanz et al, 1991). Furthermore, the dissimilarity between the peptide sequences of the resveratrol/chalcone synthases and the PKSs suggests that these two classes of enzymes have convergently evolved to utilize a similar catalytic mechanism (J. Schroder, personal communication). Within the fluorescent pseudomonads, only two PKSs have been previously reported. The PKS required for 2,4-diacetylphloroglucinol biosynthesis in Ps. fluorescens Q2-87 exhibits similarity to the chalcone synthases (Thomashow et al, 1996). A Type II PKS has been identified within Ps. syringae and is responsible for biosynthesis of the coronafacic acid moiety of the phytotoxin coronatine (Penfold et al, 1996).

3.1.2 Halogenating enzymes in microorganisms

Despite the existence of a few thousand halogenated secondary metabolites (Gribble, 1996), practically nothing is known about the enzymes responsible for the halogenation of these compounds. Historically, a non-specific enzyme 56

assay based on the halogenation of monochlorodimedone in the presence of hydrogen peroxide and halide ion (Hewson and Hager, 1980) has been the principal method used in the isolation of known haloperoxidases. Recently, however, gene disruption experiments have demonstrated that several of the proteins thus isolated are not required for halo-metabolite production in the pathways to which they have been assigned (reviewed by van Pee, 1996). Furthermore, halogenation by many of these halogenating proteins is non regiospecific, suggesting the protein has a negligible role in directing the reaction. Sequence comparison of these haloperoxidases has revealed structural similarities to several serine-hydrolases (Pelletier and Altenbuchner, 1995). The halogenations catalyzed by these proteins are presumed to involve hydrogen peroxide hydrolysis of the serine ester within the hydrolase catalytic triad. The peracid thus formed reacts with halide ion generating the reactive halogenating species (van Pee, 1996). In light of these observations, it is quite possible that the halogenating activity of these proteins is an experimental artifact and not representative of their biological function. Genes encoding halogenases participating in secondary metabolic pathways have been identified only recently. The encoded enzymatic activities have been experimentally verified by either complementation or gene disruption. Therefore, the discovery of several loci within the pyoluteorin gene cluster that may encode halogenases is significant. Although data concerning the catalytic mechanism of these enzymes is presently not available, it is possible that alignment of the deduced amino acid sequences will identify invariant residues that may be required for catalytic activity.

3.2 RESULTS AND DISCUSSION

3.2.1 Nucleotide sequencing of the pyoluteorin region

Nucleotide sequence analysis was initiated on genomic DNA of Pf-5 flanking each of the Tn5 insertions. One of the Tn5 inverted repeats and the flanking Pf-5 genomic DNA was cloned from each of five Tn5 mutants (Figure 3-3, insertions 4236, 4296, 4128, 4274, and 4175). The nucleotide sequence 57

of the genomic DNA flanking each Tn5 insertion was obtained via an oligonucleotide primer complementary to nucleotides 18-37 of the Tn5 inverted repeat sequence (Gen Bank accession no. L19385). One of these sequences (insertion 4274) showed significant similarity to genes encoding the erythromycin polyketide synthase (PKS) of Saccharopolyspora erythraea. Because pyoluteorin is a polyketide-derived metabolite (Cuppels et al, 1986; see Chapter 2), a PKS should be required for pyoluteorin biosynthesis. Therefore, initial sequence data was consistent with the hypothesis that insertion 4274 was within a pyoluteorin biosynthesis gene. Sequence analysis was continued with wild-type DNA of Pf-5 subdoned from cosmid pJEL1938, which contained the 18.4 kb region depicted in Figure 3-3.

Hind111 4236 4296 4128 4274 4175 4366 EcoRl EcoRl Hind!!! EcoR1 EcoRlt I EcoRl I t 1 orf-S pitA p/tB pltC pltD

1.3 kb

Figure 3-3. Genomic region required for pyoluteorin production and the biosynthetic genes identified by nucleotide analysis. All transposon insertions are numerically labelled. Tn5 insertion locations are designated by black diamonds. The orientation of a Tn3nice reporter insertion (4366) is is indicated by the open flag. Loci names are below the corresponding ORFs.

Four complete ORFs in addition to a partial ORF were identified by codon preference analysis (West and Iglewski, 1988). The inferred peptide sequences of the two largest ORFs (pltB and pitC) exhibit extensive similarity to several PKSs. The identification of several functional domains within P1tB and P1tC that are required for polyketide assembly is convincing evidence for the assigned function of these proteins. The inferred gene products of pltA, pltD, and orf-S share sequence similarity to other halogenases. Because halogenases are a poorly characterized class of enzymes, however, enzyme function cannot yet be predicted from sequence similarity alone. 58

3.2.2 Identification and sequence analysis of the pyoluteorin PKS

Nucleotide sequence analysis of pltB and pltC. Two ORFs, pltB (7.4 kb) and pltC (5.3 kb), which encode proteins 2458 and 1774 amino acids in length, respectively, were identified within the pyoluteorin biosynthesis gene cluster and share sequence similarity to known PKSs (nucleotide sequence data for pltB and pltC are in Appendices 2 and 3, respectively). The translation initiation sites for both pltB and pltC were assigned based on the existence of postulated Shine-Dalgarno sequences 8 and 6 bp, respectively, upstream from the identified start codons. It is unlikely that an alternative translation initiation site exists upstream of the identified pltB start codon since it is immediately preceded by an in-frame stop codon. Although it appears that pltB and pltC are separated by 48 nucleotides (nt), a possible alternative pltC start codon exists 12 by downstream of pltB. A pltC gene initiated from this alternative start codon would encode 12 additional amino acid residues on the N-terminus of P1tC. However, no convincing Shine-Dalgarno sequence exists immediately preceding this alternative start site. Nevertheless it is possible that the two ORFs may be translationally coupled, given the short distance between the end of pltB and this possible pltC initiator codon (McCarthy and Gualerzi, 1990). Comparison of pltB and pltC nucleotide sequences with entries in either the EMBL or Gen Bank databases revealed that these ORFs exhibited similarities to several known Type I PKSs. At both the nucleotide and amino acid level, the most significant alignments of pltB and pltC were found with the eryA gene of S. erythraea, which encodes the PKS required for erythromycin production (Cortes et al, 1990; Donadio et al, 1991). Conserved amino acids found in the catalytic sites of several PKSs are also present in the deduced amino acid sequences of PltB and PltC and define catalytic domains within these proteins. These results suggest that both PltB and PltC possess enzymatic functions necessary for polyketide assembly. 59

Modular organization of the pyoluteorin PKS catalytic domains. The genetic organization of the Pf-5 pyoluteorin PKS exhibits features common to both bacterial and fungal Type I PKSs. Three KS, AT, and ACP catalytic domains and a single functional KR domain are present in P1tB and PDC (Figure 3-4). Two modules are present within P1tB. Of these, module 1 (1013 aa) is located at the N-terminus of P1tB and contains only a core set of domains. Module 2 (1424 aa), located in the C-terminus of P1tB, contains an apparently non-functional KR domain in addition to a core set of domains. A single module (1774 aa) is present within PDC and contains the only functional KR domain identified within the this PKS as well as a presumed non-functional dehydrase (DH) domain. The simplified domain organization of module 1 and the presence of inactive catalytic domains (i.e. KR in module 2 and DH in module 3) have precedence in bacterial systems (Aparicio et al, 1996; Donadio et al, 1991). Nevertheless, several features of the pyoluteorin PKS such as the absence of a loading or thioesterase (TE) domain and a single module within PltC, are attributes normally associated with fungal PKSs.

ACP-1 ACP-2 P1tB KS-1 AT-1 KS-2 AT-2 a 1 module 1 module 2

ACP-3 pltc Ili KS-3 Naval AT-3

I 1 400 aa module 3

Figure 3-4. Modular organization of the functional domains identified in P1tB and PltC. The individual domains within each of the three modules are designated by open boxes and are labelled according to their assigned function. Stippled boxes represent the non-functional KR and DH domains. Interdomain regions are represented by solid lines. Abbreviations are as follow: KS, ketosynthase; AT, acyltransferase; CLF, chain length determining factor; KR, ketoreductase; ACP, acyl carrier protein. 60

Catalytic domains within the pyoluteorin PKS. Catalytic domains were identified within the pyoluteorin PKS modules using ProfileGap analysis as well as multiple sequence alignments for identified domains within known PKS and FAS sequences (Figure 3-5). Well-defined modules such as KSs could be identified by either method. However, poorly defined regions or regions in which diagnostic sequence motifs were not found could be identified only by using a statistical analysis of the Profile Gap alignments. Active-site motifs and the invariant residues within the protein sequence could be identified only by using multiple sequence alignments. The C-terminal and N-terminal limits to any given domain also differed depending on whether multiple sequence alignment or Profile Gap analysis was used to define the domain (data not shown). It was not possible, therefore, to define the precise length of each domain within the pyoluteorin PKS. Although the statistical capability of Profile Gap analysis does not take into account all statistical properties of biological sequences (Lipman et al, 1984), when used in conjunction with multiple sequence alignments, this alignment tool was invaluable in identifying regions of limited similarity within poorly characterized domains. A KS domain of approximately 415 as is present within each of the three modules of the pyoluteorin PKS. An active-site Cys residue within the sequence motif GPxxxxxxxCxSxL (Cortes et al, 1990) is present roughly 170 residues from the N-terminal start of each KS domain. Two His residues found to be invariant in both Type I and Type II PKSs (Aparicio et al, 1996) are also present within the KS of modules 1, 2, and 3. It has been suggested that one of these His residues may increase the reactivity of the active-site Cys by acting as a general base (Aparicio et al, 1996). Active-site serine residues exist within the PxxxxGHSxGxxxA sequence motif (Cortes et al, 1990) for each of the pyoluteorin PKS AT domains. The AT domain of module 3 exhibits a Pro)Ala substitution within this sequence motif. Also present within the AT domains of modules 1 and 3 are specific His, Arg, and Gln residues that have been identified in all Type I PKSs and FASs (Aparicio et al, 1996) and that are involved in the active-site structure of the malonyl-CoA:acyl carrier protein transacylase of E. coli (Serre et al, 1995). In contrast to modules 1 and 3, the assigned AT region of module 2 exhibits limited similarity to the sequences of known AT domains and 61

Figure 3-5. Deduced peptide sequence alignment for the three PKS modules encoded by pltB and pltC. Each functional domain is delimited by a solid black bar above the aligned sequences. The dashed regions of this bar designate the approximate boundaries for each domain. Profile sequences surrounding active-site regions within each identified domain are also located above the aligned sequences. The consensus sequence located below the aligned pyoluteorin PKS modules was generated from conserved residues (identified by uppercase text) within each domain. A 464 aa region from P1tC has been eliminated for clarity and is designated by the double headed arrow. Individual aa residues discussed in the text are in bold type. " " R". ; ° nma nom i/ a -,;;; I g .6: RN, 4 iRili,'.r. iRiit gilEir; ON: i.,:iiit 1 I 1111 if.,:lic IRiit IR17,1; IR,,:ii: 1,4q!!!!!!!!!!! i!!!! I!!! OM! 1

3 St4 4 1. 01 444 , fi. 4., in i t.: ill!t ii k i.1 airl.. t ...1gz. au: r {za : rar rri: E. r. ut: ''..q.1 g i 11! 1 11.1 t IV iii 1r 11 0:4 ill ill 4 1.1 I riiE a ao.,iii 1, faOBI I ill Iiil i ii ii evHit ..-41 1 !HI rr. A.!' a' Iiigt la° 1H.1 :.-/ lii fi?.. 4 -I ?:.ii! 31: r..r I P iii Ull11 i ill rt111, :7 a iii Ili ! ili .e "1 'ai.rz 1 III HI :OP. i 1 + iii if,:i ...1, i 1 if elfi : r r. ..: ',- ip 0.1 i a:.:41 :17 r ir:E: i II': 7: 4 II: 121 IV .49.4a 8'2.: SS: oo I !... 2r.: :7 ;g; VI 1 ;II t in i .?. ni 11 111 1 in I ; .....itl Fri + Et I 2 41.'111 i :7 .ii ??.i al ill 4 :4;< i :v,:i. iii i ill i 6-*: 4- f i 1 r; 1 II! '1111 i i .91 1: !it it4 s *1 ilti ii/ 11 1, ja- iii r a a ltk lit 1 ail 141 VS 1.1 1.11. 1141:1. 0 4 1:::: I i Fli 1 R & 3 : 332 i -111 SE4 F #ii Si: II; Di . V.-2..ii 44i a4 r illE 7v:a iF,A 5.:.F--1 44,' ill 1 i V.-affi i 155 1 I a gig, la I- fpI I Aii Ao PE Ir :`ii: + Hi: 51 Fit° W.0Wes 0..a "I 311 i .. Cil I. ;1! trg 1 ! f al f 111 ' :11; in I i f iii 4 i i Eli 54'.§ ? 2 c 9 di thli Hi 4 Etl. .,: i,11 "1 iil. 7r.-E: til i. ill itiii Y 04 A r4 Io iii tel its ? l' ii:'' 1 iii P : gii Lia .1.91. i la i r Id !. I 1 HI I all: W. i alr 11 iii t ,,P. tip. Fil l in ..r,.. 1 1Alf gg- ilii : az ifi z Fliiel: : .1; i51 i irA ae s..1 'Oil HI gzA EEE : k&g I I ill 1 eiti 1" ?Ili in iis ,1.31 t .1: ? -41,-. .:.".2i 13: ii :$S r': iF,E di :..4.7a r. fri l otiE z r. :71 irt 1 E 7: 2 ;-.,' 1 1:, E ?..: i' F .: :. .7: 0 ...1 Vii ?...1 3 iii 63

possesses several anomalous features. The Pro and His residues found in the motif surrounding the active-site serine have been substituted by Ala and Phe residues, respectively. Furthermore, of the His, Arg, and Gln residues mentioned above, only the Arg is present in the deduced protein sequence. Two sequence insertions of 8 and 64 amino acids C-terminal to the active-site Ser are also apparent from the multiple sequence alignment with known AT domains. The enzymatic function of all ACP domains requires covalent attachment of 4'-phosphopantetheine to the Ser residue within an LGxDS sequence motif (Cortes et al, 1990). The Gly and Ser residues of this motif are present within the C-terminal regions of modules 1, 2, and 3, but substitutions for the Leu and Asp residues are present within the assigned ACP domains. In both modules 1 and 2, an Asp -*Ser substitution has occurred. Additionally, a Leu>Met substitution has occurred in module 2 as well as a Leu -*Tyr substitution in module 3. Apparently these substitutions do not compromise enzymatic function, because based on the model for polyketide biosynthesis, functional ACP domains are an absolute requirement of the PKS. Consistent with this conclusion, it is generally recognized that sequences flanking the ACP active site motif do not exhibit noticeable sequence similarity among PKSs. A KR domain was identified within a 178 amino acid region between the identified AT and ACP domains of module 3. The invariant GxGxxGxxxA motif, which is found in all functional KR domains (Aparicio et al, 1996) and is believed to form part of the binding site for the cofactor NADP(H) (Scrutton et al, 1990), is present in the module 3 KR domain. In contrast, an analogous region within module 2 of PltB exhibited very limited similarity to the C-terminal region of a KR profile sequence. Extensive substitutions of conserved KR amino acid residues within this module 2 region make it unlikely that this domain is functional within the pyoluteorin PKS. Therefore, the pyoluteorin PKS appears to contain only one functional KR domain residing in module 3. Within module 3 of the pyoluteorin PKS, a 558 as region between the AT and KR domains was examined for the presence of DH domain. Statistical analysis of the alignments between this region and a DH profile sequence suggests that limited sequence similarity does exist. Although DH catalytic 64

domains are located typically between AT and KR catalytic domains in Type I FASs (and some PKSs), DH domains are not well characterized, and it is difficult to identify a DH domain on the basis of sequence homology alone. No TE domain was identified within either P1tB or PltC. Although TE domains are not strictly required within Type I PKSs, those that have been identified are located at the C-terminal end of the last ORF (Donadio and Katz, 1992; Swan et al, 1994; Yu and Leonard, 1995). It is unlikely that a TE domain exists at the end of PltC since only 30 nt exist between the C-terminus of the ACP and the translational stop codon.

Non-catalytic regions of the pyoluteorin PKS. The differences in lengths of the interdomain regions (IDRs) within P1tB and P1tC accounts for the variable module sizes within the pyoluteorin PKS. Whereas the IDR between the KS and AT within modules 1 and 3 are approximately 100 residues in length, the same region within module 2 contains 248 aa residues. The KS/AT IDR of modules 1 and 3 align with the N-terminal portion of the KS/AT IDR of module 2. The AT and KR domains of module 3 are separated by 558 aa and the AT to ACP distances in modules 1 and 2 are 90 and 260 residues, respectively. A qualitative, composite alignment derived from the individual pair-wise alignments for the C-terminal IDR of each module suggests limited sequence similarity immediately preceding the ACP domains. However, the disparity among the C-terminal IDR for modules 1, 2 and 3 precludes further comparison.

The pyoluteorin PKS does not possess a loading domain. Initiation of polyketide assembly requires formation of an enzyme-substrate complex between an activated starter unit (e.g. acetyl CoA) and the first KS domain. Within bacteria, enzyme-substrate complex formation was attributed initially to a loading module comprised of an extra AT and ACP domain located between the N-terminus of module 1 and the first KS domain (Donadio et al, 1991). Characterization of the rapamycin PKS indicates, however, that N-terminal loading modules are not restricted to AT and ACP catalytic domains (Aparicio et al, 1996). In the rapamycin PKS, three catalytic domains, 65

a carboxylic acid:CoA ligase, an enoyl reductase, and an ACP, are located at the N-terminus of the first PKS module and are likely involved with activation and reduction of the initial substrate (Lowden et al, 1996). The activated starter unit is then transferred directly to the first KS domain for polyketide assembly. In contrast to the bacterial PKSs, loading modules have not been reported within fungal Type I enzymes. Furthermore, the heterologous expression of 6MSAS in the bacterium Streptomyces coelicolor A3(2) (Bedford et al, 1995) implies that polyketide initiation can occur through an alternative mechanism within fungal Type I PKSs. Although no evidence for a loading module within the pyoluteorin PKS was found, this enzyme may possess an inherent, but as of yet unidentified, transferase activity required to initiate polyketide biosynthesis. It is also possible that the initial substrate is directly transferred from another protein functioning earlier in the pathway. A gene encoding a CoA ligase within the pyoluteorin locus has tentatively been identified (N.L. Chaney and J.E. Loper, unpublished) and may be involved in the activation and subsequent transfer of the starter unit onto the PKS. Alternatively, the pyoluteorin PKS may utilize the transferase activity from a distinct, yet related pathway. The suggestion that a malonyltransferase is required for both fatty acid and polyketide biosynthesis within two Streptomyces species (Reville et al, 1995; Summers et al, 1995) are of particular relevance to this possibility.

Polyketide assembly of the pyoluteorin resorcinol moiety. The presence and organization of catalytic domains identified within the pyoluteorin PKS correspond to resorcinol ring formation within pyoluteorin. The addition of each C2 extender unit and modification of the resulting 13-carbonyl structure could be catalyzed by the combined activities within the individual modules of the pyoluteorin PKS (Figure 3-6). Although resorcinol ring synthesis within pyoluteorin requires dehydration at the 3-position, no functional DH domain was clearly identified. Therefore, dehydration may occur after release of the fully extended polyketide from the PKS. Tautomerization and the subsequent aromatization of I could very well be non-enzymatic processes due to the resonance stabilization energy resulting from this structural rearrangement (Armstrong and Patel, 1993). Release of the fully extended substrate from the 66

PKS apparently does not require a dedicated TE domain, as none was found at the C-terminus of P1tC. However, a separate gene with similarity to several TEs has been identified downstream of pltB and pltC and may terminate polyketide biosynthesis (N.L. Chaney and J.E. Loper, unpublished). Alternatively the PKS alone may terminate polyketide biosynthesis by cyclization (perhaps catalyzed by the unusual AT domain of module 2) as has been suggested for 6MSAS (Beck et al, 1990; Spencer and Jordan, 1992).

PItB PltC module 1 module 2 module 3

......

OH 0 H Pyoluteorin

Figure 3-6. Proposed route for formation of the pyoluteorin resorcinol moiety. The structure of the starter unit derived from proline is unknown and is designated as R. Condensation of each C2 extender unit and the modification of the 0-carbonyl are catalyzed by each of the three modules within the PKS. Release of the fully extended substrate is hypothesized to occur by condensation between the two carbon centers identified by the arrow. The aromatization of the resulting structure (conversion of I -4 II) is likely a non-enzymatic reaction. 67

3.2.3 Identification and sequence analysis of the pyoluteorin halogenases

Nucleotide sequence analysis of pltA, pltD, and orf-S. Three ORFs in the pyoluteorin biosynthetic gene cluster, designated pl t A, pltD , and orf-S exhibit similarity to loci encoding chlorinating enzymes required for tetracycline biosynthesis (cts4) within Streptomyces aureofaciens (Dairi et al, 1995) and for pyrrolnitrin biosynthesis (prnC) within Ps. fluorescens (Hill et al, 1994). pitA (1.3 kb) and pltD (1.6 kb) were sequenced completely; both genes are transcribed in the same orientation as those encoding the pyoluteorin PKS (nucleotide sequences for pitA, pltD, and orf-S are in Appendices 4, 5, and 6, respectively). pitA and pltD are very closely linked to the PKS genes, and putative Shine-Dalgarno sequences are present immediately upstream of the start codon for each ORF. In contrast, a coding region exhibiting similarity to both pltA and pltD is located at the extreme 5' end of the pyoluteorin region (orf-S) and is transcribed divergently to the other genes within the cluster. Because orf-S is at the terminal end of the pJEL1938 cosmid insert, the complete nucleotide sequence was not determined. A Shine-Dalgarno sequence is not present within the orf-S sequence; however, the first four bases within this gene overlap the putative regulatory gene pltR (see Chapter 4) suggesting that translation of the gene products may be translationally coupled. Transposon insertions within pitA and pltD inactivate pyoluteorin production (Figure 3-1) indicating a role for their translated products in pyoluteorin biosynthesis.

Comparison of PItA, PltD, and Orf-S to halogenases involved in tetracycline and pyrrolnitrin biosynthesis. Dairi et al (1995) have identified a 1.4 kb ORF (cts4) encoding the chlorinating enzyme required for tetracycline biosynthesis. An A+G rich region (AGAAGAG) between positions -20 to -14 upstream from the postulated start codon was identified as the putative Shine - Dalgarno sequence of cts4. Several discrepancies between the nucleotide and deduced peptide alignments for cts4 , pltA, and pltD prompted a closer examination of the cts4 nucleotide sequence. In light of the unconvincing placement and sequence of the Shine-Dalgarno ribosomal binding site proposed for cts4 and because the Pf-5 sequences aligned to a region extending 68

approximately 300 nts 5' of cts4, it was considered likely that the beginning of the cts4 ORF was incorrectly identified. Indeed, codon preference analysis of the tetracycline biosynthetic gene cluster revealed that the reported sequence contains a frame shift error (data not shown). Therefore, the Cts4 protein has had an additional 100 amino acids appended to the N-terminal region for all subsequent sequence alignments.

PrnC megkspane. hclarthFDVII 10e3maGtql4 gaiLkIgAft VliiEessbP RftICHtas/P WralmnriIa dRYgIrildh itsfystgry Cts4 mtdttadqtr hgdrpYDVVI ialsGlaGtmL geiLAIChgfr VasSt.Gittr ImILDgahhP RfaV0ItstIr cOlvvLrlIs dRYgc.Xian lasfgdylan Vssah.GgSs orf-S angyIPVII 1.0aGiaGalt gavLAAmiln PltD ...mndvqsg kapehYDIII VULDeaghP RteVOSaatP eselLrIls WdIptiay lagpdkiigh VgaBacGiXl agnaisvimL eacLArnkyr VglArnrgptP pdltnatIP yTsmiFelIa dRYOOSikn iartrdiggk Pick msdhdYDVVI iGg0paGstM asyLAXegvk VapS.sGvIck cavrEkelfe ROMAISaIVP aTtpviaeIg vmekI.Ekan fpkkfgaawt sadagpedItm Consensus nor-I -0-G--G------LAK- - -- V--L P R--VOS--/P -T---L--I- - RY -I -t V--S--G-I- PrnC nF.GFvFlikp ggehdpkeft QcvIpelpag pesSYYlopv DaYLAgakik yOcledsOktt Vteyh.adol giraVttagge r...ftgrym Cts4 nF.GFartird geepdpnets OfrIpsiv.g naalFFItgEt DsYWhaAvr IDcggprapL yOcdarQyyr Venie.fddg OftesgadOs t...vraryl VDaragfrspL orf-S gP.arattlige napsspdhl. ...Vatclkv pesILFIcIU DyFaLaiAlk heaesrgnik Irrpa.akl. PltD nL.GFLYHdr sravdIggal GInVpash.. genILFItrEd DaYLLaakig ygagIveidn spevl.veds glavatalOr w...vtadfm PitA gFgILdhdfr saeilfnerk QegVdrdf.. .tfltvdttgkf VOgaggggvL DriLLehkgs lOakvfOgve Iadveflapg nVienaklak rsveiWOanv VDasgrnvIL Consensus - F -GP -PH Q--V SLFS-D- D-YLL--A -0- --0 V-v- --0 VD PrnC atkfIcLreep crfktheral ythmlgvkpf ddifkvlopqr w. awhegt Cts4 argIgLreep arlkhhersi ftlattvgvdsi ddhvdmpael IhheFeggta wvipFretapr atnnIvIlegl gldprvIp.k tdisagdeFd rppypandst mhhiFergwm wiipFnobsg atnplcSVgi glderrYpar orf-S la:1143.0,Fr P1tD argagLvaga stqktrtlef athmlgvvpf decvggd... .fpggtehggt lhhvFdpgwv gvipPlalx$1 arnplv0V1v alredICp.. PitA grrIgLrokdtvfngfaihs wfdnfdrksa cgs .strtigdqvla pdk vdyiribflp mantWvwgip itetitaVgv vtqkgnYt.n Consensus sdatyeefFw P - - - - BY Y F- PrnC eflarfpeig aq?rdavpVR dWyktdrIgd asnacygetrY cLmlhangfi DpIFsrglen tavtihalaa rlikkIrdiD FsperFeyie rlgqicIldhn Cts4 shydrfpavg rgLkgarsVA eWvrtdtmgv ssartvgerW cLmshaagti DpIFIrglsn tceiinalme rbashlredD FaverFayve elegglIdea orf-S PltD glielypglg rhLagarrvIt eterlrqpprg vyrtalerrC IMfdegaaan DIIFsrklan aaelvlalah rlikAahsgD YrapaLndfv PitA eavkcrenlh daLkasegVI prldceadysy gmkevcgdeF vLigdaartv Itgdsiials DpiFsagvav alnsariasg diiekvknnD FslosaFtbye gmirngiknw Consensus VI -H L D--F A- -D F- -F

PrnC ddfveccyta FsdfrkWdaf hrlw...avg tiltmarlvg aharfrasrn egdldhLdnd ppylOylcsal meeyyglind skaevEaysa grkpadeaaa Cts4 dklvnnsfis FahypLWnsa fri4...asa eviggkriin altrtketgd dahcgaLd.d npypOlwcp. ldfykeafde ItelcEavda ghttaeeaar orf-S P1tD drialaayva FrdpeLWnaf arvwllgsia atitarkind afakdldpry PitA yefit...ly YrIniLFtaf vg fdeidgLaed ....0fwega yrgykdilnt tIgIcDdyks akvsaahaaa dpry rldilgLIgg dvysOkrlev lammaiiaa vesdpEhlwh kylgdmgypt Consensus F - - - -LW L E

PrnC rihaliderd fatertafgfgy citgdkpqln nskyslItam rbaywtqtra paevkkyfdy Cts4 nianfallkay ittriglalk llegrvread wmIpalgin. ..dsdthhin ptadkmi... riaewatghh rpeir ...... elaasa eevraamrvk p orf-S PltD sifaelanas fvtpifdfa. PitA akpaf nphar vycattIrkl kalwtvplmqv nsevgrlify refrkpaIrk es. Consensus

Figure 3-7. The deduced peptide sequence alignments for the putative halogenases PitA, PltD, and Orf-S with PrnC and Cts4. Residues in UPPER CASE letters signify conserved substitutions. Those in BOLD UPPER CASE letters are invariant among the available sequences (only the N-terminal 151 residues are available for Orf-S). The consensus sequence was calculated with a plurality of 4. Residues involved in forming necessary secondary structure for binding NAD co-factors are indicated by asterisks (*) within the the boxed region. 69

Multiple sequence alignment of the Cts4 and PrnC halogenases with the three putative halogenases of the pyoluteorin gene cluster show a considerable amount of similarity throughout the protein sequences (Figure 3-7). Of the 78 conserved residues among the deduced proteins, the majority are concentrated in the N-terminal region. The most intriguing aspect of these sequences is that, with the exception of P1tD, they each possess a core GxGxx(G/A) sequence believed to form the 13a--f3 secondary structure required for NAD co-factor binding within many oxidoreductases (Scrutton et al, 1990). The NAD co-factor binding site had not been recognized previously within the reported halogenase sequences. This result suggests that the halogenases are a subclass of the oxidoreductase enzymes.

Speculation on the parameters of biological halogenations. While there appear to be no obvious mechanisms that would account for biological halogenations, it is possible to consider some of the chemical parameters for the enzyme substrates that may guide future research. These points are merely speculative at this point; although the genes have been identified, subcloned, and theoretically can be easily expressed, it is virtually impossible to undertake any mechanistic studies without knowing the substrate requirements of the enzyme. The solvation of chloride (i.e. C1) within protic solvents generally makes it a fairly poor nucleophile. Therefore, chlorination is likely to occur via an electrophilic addition or substitution mechanism. Such a reaction would require the electrophilic equivalent of a chloronium species (i.e. Cl'). Although results from previous research involving the haloperoxidases suggest this species might be HOC1, this conclusion is probably not biologically relevant because in vitro halogenation of a susbtrate requires the use of high H202 concentrations. Within biological systems, reactive oxygen compounds like 1-1202 are efficiently eliminated; therefore, it is unlikely that H202 could reach sufficient concentrations in vivo to effect halogenation. Alternatively, a chloronium electrophile may arise directly from metal-catalyzed reaction with molecular oxygen. If the halogenation reaction does involve an electrophilic addition or substitution, the substrate necessarily must be a good nucleophile. As a ic 70 excessive heterocycle, pyrrole is quite facile at undergoing electrophilic attack. Furthermore, the aromaticity of the pyrrole carbon skeleton allows charge delocalization throughout the ring and could conceivably assist in stabilizing reactive intermediates and accounting for adding of chlorine atoms at the l and the 3-positions on the ring. Within the pyoluteorin pathway, therefore, halogenation is likely to occur after the pyrrolidine moiety has been oxidized to the pyrrole.

3.3 CONCLUSIONS

Several loci have been identified within the Pf-5 pyoluteorin biosynthetic gene cluster by sequence analysis, and the function for each of the encoded gene products serves either in a biosynthetic or a regulatory capacity. PltB and PltC encode a Type I PKS required for assembly of the pyoluteorin resorcinol moiety. Because many PKS sequences have been analyzed and the catalytic mechanisms of these enzymes are well understood, there is little doubt as to the assigned function of PltB and PltC. Nevertheless, the absence of a loading domain and the unusual AT domain make this PKS worthy of further study. It appears that three loci within the pyoluteorin gene cluster potentially encode halogenases. Furthermore, sequence alignment of PltA, PltD, and Orf-S with two known halogenases has identified a potential NAD co-factor binding region that may play a role in catalytic activity. The presence of three halogenases in the pyoluteorin region is quite surprising because only two chlorine atoms are present within pyoluteorin. Several possible explanations might account for this observation: i) Pf-5 may produce several (but as of yet unidentified) pyoluteorin homologues that differ only in the extent of halogenation. ii) One of the halogenases may be required for the biosynthesis of the presumed P1tR co-inducer. iii) One of the halogenases may be non-functional. iv) There may be a cryptic chlorination involved in pyoluteorin biosynthesis. v) Several alternative routes to pyoluteorin may exist requiring chlorination of various precursors. vi) orfS may not be involved in pyoluteorin biosynthesis, as no mutants in this region have been derived. In the midst of this speculation however, it should be stressed 71

that the catalytic functions for Orf-S, PltA, and PltD have been assigned tentatively since this class of enzymes is relatively unknown (nucleotide sequence data has been reported for only two other halogenases). In light of the novelty of these enzymes, characterization of the halogenation reaction required for pyoluteorin biosynthesis promises to be very exciting.

3.4 EXPERIMENTAL METHODS

Nucleic Acid Manipulations. Six DNA fragments from the cosmid pJEL1938, which contains the genomic pyoluteorin biosynthetic region of Ps. fluorescens Pf-5 (Kraus and Loper, 1995), were subcloned into pUC19 using the restriction sites identified in Figue 3-3 and transformed into Escherichia coli DH5a using standard methods (Sambrook et al, 1989; Ausubel et al, 1992). E. coli strains were routinely cultured in Luria-Bertani medium at 37 °C (Sambrook et al, 1989). Double stranded DNA was isolated for automated sequencing by either QlAprep -Spin plasmid mini-prep kits (Qiagen, Chatsworth, CA) or by Promega Wizard maxi-prep kits (Promega, Madison, WI) and quantified by UV spectrophotometry.

DNA Sequence Analysis. Sequence analysis of each subcloned region was initiated from the pUC primer sites and subsequently completed by "primer walking" in each direction across the entire length of the subclone. Automated DNA sequence analysis and primer synthesis were performed by the Center for Gene Research Central Services Laboratory at Oregon State University and by Macromolecular Resources Sequi-net Division at Colorado State University using dideoxynucleotide chain-termination (Sanger et al, 1977) on Applied Biosystems models 373A and 377 sequencers. Compilation, manual editing, and analysis of the sequence data were done using the University of Wisconsin Genetics Computer Group (Genetics Computer Group, 1991) programs (GCG). Open reading frames were identified within the DNA sequences by codon usage analysis utilizing the codon preference 72 frequencies compiled for Ps. aeruginosa (West and Iglewski, 1988). Protein functions were initially assigned on the basis of sequence similarity detected upon comparison to the Gen Bank or EMBL databases using BLAST (Altschul et al, 1990) and FASTA (Pearson, 1990) algorithms, respectively.

Deduced protein sequence analysis. The catalytic domains within P1tB and P1tC were identified by multiple sequence alignments (Pile Up) and by profile sequence alignments (Profile Gap) using standard parameters. The alignment analysis used various domain sequences from the following representative Type I polyketide and fatty acid synthases (Gen Bank accession numbers are in parenthesis): the erythromycin PKS of Saccharopolyspora erythraea (M63676 and M63677), the rapamycin PKS of Streptomyces hygroscopicus (X86780), the soraphen A PKS of Sorangium cellulosum (U24241), the mycocerosic acid synthase of Mycobacterium tuberculosis (M95808), the 6-methylsalicylate synthase of Penicillium patulum (X55776), and chicken fatty acid synthase of Gallus gallus (J04485). Interdomain sequences within the modules of P1tB and P1tC were manually aligned after comparison of several Gap generated alignments for these regions. Statistical analysis of the alignment between the domains in either P1tB or P1tC and the profile sequences above was performed using the Best Fit algorithm with the RANDOMIZATION command option. Scores were calculated for each alignment and compared to the average alignment score for ten random sequences that maintained base composition and length. Scores that were 5 standard deviations above the mean were considered indicative of significant similarity. Those scores that were 1-5 standard deviations above the mean were considered of limited similarity. Conserved residues were identified within PltA, P1tD, and Orf-S by multiple sequence alignment with Cts4 of St. aureofaciens (D38214) and PrnC of Ps. fluorescens (U74493). The putative NAD co-factor binding site was identified manually. 73

Chapter 4. Nucleotide Sequence Analysis of a Putative Transcriptional Activator Regulating Pyoluteorin Biosynthesis

4.1 INTRODUCTION

Secondary metabolite production is regulated by many factors including cell growth rate, extracellular signalling metabolites, and nutritional factors such as carbon, nitrogen, and phosphate sources (Demain et al, 1983; Grafe, 1989; Liras et al, 1990; Horinouchi and Beppu, 1992; Vining, 1995). These regulatory factors do not necessarily operate independently, nor does each factor elicit an identical response for each antibiotic pathway that may exist within a single organism. For example, S. coelicolor A3(2) cultures containing nitrate as the sole nitrogen source exhibit differential production of the pigments undecylprodigiosin and actinorhodin in late log phase and stationary phase, respectively (Hobbs et al, 1990). The complex regulatory networks affecting antibiotic production appear to operate on two distinct levels. Global regulatory loci have pleiotropic effects that can either directly or indirectly regulate antibiotic production, whereas regulatory loci directly linked to biosynthetic genes generally mediate gene transcription within that gene cluster (Martin, 1992; Hutchinson et al, 1993). Antibiotic production within Pf-5 is globally regulated, in part, by the sigma factor RpoS (as), which controls the transcription of many genes in response to starvation or the onset of stationary phase (Kolter et al, 1993; Lange and Hengge-Aronis, 1994; Hengge-Arronis, 1996). Inactivation of the rpoS locus within Pf-5 results in the over-production of both pyoluteorin and the antifungal metabolite 2,4-diacetylphloroglucinol (Sarniguet, et al, 1995). Concurrent research involving a Ps. fluorescens strain that is strikingly similar to Pf-5, has demonstrated that increasing the gene dosage of the essential "housekeeping" sigma factor RpoD (a') also resulted in the over production of both antibiotics (Schnider et al, 1995). Both results support the hypothesis that RpoS and RpoD act in concert to regulate secondary metabolism possibly via competition for RNA polymerase. Antibiotic production within Pf-5 is also globally regulated by a two- component regulatory system involving the sensor kinase ApdA (Corbell 74 and Loper, 1995) and the response regulator GacA (Laville et al, 1992). Within the two-component paradigm, an ATP-dependent auto-phosphorylation of ApdA (the sensor kinase) occurs in response to an environmental stimulus. Subsequently, ApdA phosphorylates GacA (the response regulator) which in its activated form, is capable of interacting with the target gene sequences to induce gene transcription (for general reviews of two-component regulatory systems, see Albright et al, 1989; Stock et al, 1990; Parkinson and Kofoid, 1992). Mutations within either apdA or gacA result in several pleiotropic effects including complete loss of antibiotic production. Although both ApdA and GacA are necessary for gene expression in Pf-5, neither the environmental stimulus nor the genes targeted by GacA have been identified. Antibiotic pathways within other Pseudomonas spp. are locally regulated by gene products that are translated concurrently with the required biosynthetic genes (see Chapter 1, section 1.2). Recently, the gene cluster responsible for 2,4-diacetylphloroglucinol production in Ps. fluorescens Q2-87 has been characterized (Bangera and Thomashow, 1996), and a 6.5 kb DNA fragment was shown to confer 2,4-diacetylphloroglucinol production within heterologous hosts. However, larger fragments containing flanking DNA sequence were unable to transfer antibiotic production. A linked gene encoding a 23 kDa repressor protein has since been identified via sequence analysis and accounts for the lack of expression in heterologous hosts (Thomashow et al, 1996). In the course of sequencing the pyoluteorin biosynthetic gene cluster, an ORF was identified that is divergently transcribed from and located upstream from the pltABCD loci. The similarity of this gene to the LysR family of transcriptional activators (Henikoff et al, 1988; Schell, 1993) and the presence of a putative operator region upstream of pltA provides compelling evidence at this time for the role of the translated protein in the regulation of the pyoluteorin biosynthetic genes. Unfortunately, there is no experimental evidence that inactivation of pltR negatively affects pyoluteorin production within Pf-5. Therefore, the conclusions discussed below are tentative pending further genetic analysis of pltR. 75

4.2 RESULTS AND DISCUSSION

4.2.1 Nucleotide sequence analysis of pltR

A 1.0 kb ORF (pltR), divergently transcribed from p1tABCD (Figure 4-1), was identified within the Pf-5 pyoluteorin gene cluster (nucleotide sequence of pltR is in Appendix 7). The translation initiation site as well as a possible Shine-Dalgarno sequence (positions -11 to -8 relative to the start codon) were identified 766 nt upstream of pltA. An alternative start codon and Shine- Dalgarno sequence exist approximately 60 nt upstream of the identified initiation site, but it is unlikely that this is the site of translation initiation due to the presence of an in-frame stop codon 12 nt downstream. No alternative start sites are present within 100 nt downstream of the identified initiation site. Codon preference analysis detected an atypical codon usage and low G+C bias patterns within the first 500 nt of pitR. Whereas other genes within the pyoluteorin cluster exhibit G+C bias in the range of 75-85%, G+C bias for the pltR 5' region was approximately 35-45%. The low G+C content within pltR may have a regulatory consequence for pyoluteorin production (via infrequently used tRNAs) or it may simply reflect a residual structural motif that is required for regulatory function.

Hinal 4236 4296 4128 4274 4175 4366 EcoRI 1 EcoRl EcoRI t tEcoRI HindlIl f I EcoRI I 1 I t I 1 sa mat-sunamimssimummounemulm,minammangsansm.. pltR pitA pltB pltC pltD

1.3 kb

Figure 4-1. Genomic region required for pyoluteorin production and the regulatory gene identified by nucleotide analysis. All transposon insertions are numerically labelled. Tn5 insertion locations are designated by black diamonds. The orientation of a Tn3nice reporter insertion (4366) is is indicated by the open flag. Shaded ORFs were described in Chapter 3. 76

4.2.2 Similarity of P1tR to LysR-type transcriptional activators

The deduced P1tR peptide sequence (344 aa) exhibits significant similarity to amino acid sequences of more than 20 members of the LysR family of transcriptional regulators (Henikoff et al, 1988; Schell, 1993). LysR-type regulators influence an extremely wide range of phenotypes and have several common features: i) They generally act as transcriptional activators and require a co-inducer for activity. ii) The gene encoding a LysR homolog is often closely linked to and transcribed divergently from the regulated target gene(s). iii) A highly conserved N-terminal domain containing a helix-turn-helix (HTH) motif, a somewhat variable co-inducer binding domain, and a conserved C-terminal domain are present within all LysR-type regulators. iv) LysR regulators generally recognize target promoters containing an inverted repeat within the -75 to -55 region (relative to the transcription initiation site). v) Expression of lysR- homologs is often negatively auto- regulated. Despite these similarities, there are no consistent biochemical relationships between either the required co-inducers or the proteins encoded by the regulated target genes. Based on the initial database search results, the highly conserved N- terminal domain that is characteristic of all LysR-type proteins was readily apparent within P1tR (Figure 4-2). Further analysis using a calibrated weight matrix (Dodd and Egan, 1990), predicted with absolute certainty that a HTH motif exists between residues 20-41 within the N-terminal region of P1tR. In addition to the N-terminal domain, a co-inducer binding and a C-terminal domain were assigned via comparison of the P1tR sequence with a 70% consensus sequence for each of the three domains common to LysR-type proteins (Schell, 1993). Furthermore, many amino acid residues along the entire length of a profile sequence compiled from ten LysR-type proteins representing several phylogenetic groupings (Schlaman et al, 1992) also were conserved within P1tR. 77

Figure 4-2. Gap alignment of P1tR and a LysR profile sequence. Identified domains are indicated by bars above the sequence alignment. Domains were identified from the 70% consensus sequences described by Schell (1993) as shown. The helix-turn-helix motif (residues 20-41) within the N-terminal region was calculated using the method of Dodd and Egan (1990). The consensus sequence was calculated from the PltR-LysR profile alignment and adheres to the following conventions: Amino acid residues in UPPER CASE are identical within both sequences. Conservative substitutions are designated as follows: g, hydrophobic residues (F, W, Y, V, I, L, M); f, hydrophilic residues (N, D, E). 78

helix-turn- Sfrnflr. P1tR Mk aLgvvndidl YisVtktgSF setgrlLgIp LysR profile mpevqtdhpe taeasmoqMr hLd.lrqlry FeaVaeenSL traaeeLhVs Consensus M- L S--V -SS L S-

helix

ofASstrocSff lef.1g..lf .r.fr.f... to PltR pssVmRrIns LEkELetcLF nRstkcliLT EtG11F1Eha knisrcighA LysR profile qpaVsRqVrr LEdDLgvpLF eRtgrgvkLT EaGeeLwEev qralqaldnA Consensus --V-R-S-- LE-fL---LF -R LT E-G- S-E A

Co-inducer binding

1.i g 1p p 1 P1tR rsEVkEhtas tlGvLkVsaP vaFgr..rhv ap.LLsriLn hhPglkieFs LysR profile mdEVeElgpg qsGrLwIayP tsLastiywf 1pqLLdafMq qyPdvelsLt Consensus --EV-E---- G- L -13 -P LL -S -P S-

domain lf..fSdSS P1tR lnD.kaLdps idnvdIcikL gilP..dsnL IptkLadmrr VicasPeyir LysR profile tgDseeLvea lragrIdlaL sgvpprciagL VaveLlqedm VVvvaPnlph Consensus -D -L I -L -P S -L V P

P1tR q4Gcpqt1E. .DLyqbacli hstcsnfSlt WqfkVDgL1K klmpssRlSv LysR profile plGpqkavEl aDLqeeefim 1prggrsS.. WrdaVErFwK qagiqpRiSm Consensus -G E- -DL W- Vf -S-K R-S

C -term. domain vP L P1tR nssEL.1vDg alq.GIGIih aPtwiVhEqI asgqLVsLld EYceadpqqg LysR profile eveDLeavVg lvaaGVGVav vPesvVeDeV epvrLVkLpf EWmldkraif Consensus --fL---V ----GSGS-- P---V-f-S ----LV-L-- ES

P1tR aLYalRar.. ...SsvVpAk trLFiNE1kR sigStpyWdl pfekeipqtl LysR profile 1LFvaRrhmk rrvSplVqAf weLLaBEavR 11sSldqFq Consensus -LS--R------S--V-A- -LS-fE- R -S S

P1tR atlhfdtamh salSrATTlq dksqts LysR profile SsATTa. Consensus S-ATT

Figure 4-2 79

4.2.3 A putative PltR binding site within the pyoluteorin gene cluster

Promoters targeted by LysR regulators contain an inverted repeat sequence centered near the -65 promoter region that exhibits a conserved TNTNA-N7-TNANA motif (Schell, 1993) and is essential for transcription mediated by LysR-type activator proteins (Ebright, 1986). An inverted repeat containing the TNTNA-N7-TNANA motif was identified within the pyoluteorin gene cluster and is centered 45 nts 5' to the translation start codon of pltR (Ebright box; Figure 4-3). It should also be noted that sequences flanking the identified Ebright box are particularly A+T rich, an attribute of some LysR-regulated promoters (Vale et al, 1991 and references therein). The proximity of the putative PltR binding region to the presumed pltR promoter region may serve to auto-regulate expression of PltR, as is characteristic of LysR regulons. Other potential PltR binding regions within the pyoluteorin gene cluster were identified by partial similarity to the TNTNA-N7-TNANA motif; however, these sequences did not possess the symmetrical inverted repeat characteristic (data not shown).

ATCAATTTTCCCAAAGCCAACTAATCATGGCCCCAGCACTGCTTGACAAACAAAAAGCCT

TAGTTAAAAGGGTTTCGGTTGATTAGTACCGGGGTCGTGACGINACTGTIrTGTTTTTCGGA pl to -10

AGCCGTACTAAGGAGGCTGTCATGATCTATTT GTAAT TT GC CTT TACK-1AAATAAGACAC

TCGGClaiTGATTtCTCCGACAGTACTAGATAAACATTAAACGGAAATGTTTTTATTCTGTG -35 RB S TAAAAATTCTAAAAGGATTTAGGAATGAAGGCGCTAGGCGTTGTCAATGACATAGACCTT

ATTTTTAAGATTTTCCTAAATCCTTACTTCCGCGATCCGCAACAGTTACTGTATCTGGAA PltR M K A L G V V N D I DL

Figure 4-3. The putative promoter region of pltR. The identified Ebright box containing the conserved motif TNTNA-N7-TNANA is in bold type. The ribosomal binding site (RBS) and the N-terminal sequence of PltR are also shown. Tentative -35 and -10 promoter regions for the pitA transcript are also indicated by the boxed nucleotides. 80

The presence of the Ebright inverted repeat sequence has tentatively allowed identification of the pltA promoter region (Figure 4-3). Although the -35 and -10 RNA polymerase recognition sites within the pltA promoter region do not conform strictly to known Pseudomonas promoter regions (Deretic et al, 1989), the identity and positioning for 7 out of 12 nts within the pitA promoter regions are identical within three a"-dependent promoters. Nevertheless, it must be stressed that in the absence of functional characterization and identification of the pitA promoter sequences, presumed similarities within the tentatively assigned -35 and -10 regions are speculative.

4.2.4 Models for PltR-dependent regulation of pyoluteorin biosynthesis

Identification of P1tR as a putative transcriptional activator is a significant contribution when viewed in the context of the global regulatory loci known to influence pyoluteorin production. In addition to the global regulators ApdA, GacA, RpoS and RpoD described above, a La protease has been identified recently as a negative regulator of pyoluteorin production (C. Whistler and I.E. Loper, unpublished). La protease induction is often associated with the heat shock response (Neidhardt and VanBogelen, 1987) and is involved in the proteolytic regulation of cellular development (Gottesman and Maurizi, 1992). Although secondary metabolism is not formally viewed as a developmental stage within bacterial metabolism, it is, nevertheless, closely coupled with transition into stationary phase. It is possible that the identified Pf-5 La protease may regulate pyoluteorin biosynthesis via P1tR degradation. Recent studies also suggest that ApdA and GacA influence the accumulation of RpoS (C. Whistler and J.E. Loper, unpublished); however, it is unlikely that this two-component system regulates pyoluteorin biosynthesis via RpoS because both ApdA- and GacA- mutants fail to produce pyoluteorin. Alternatively, ApdA and GacA may be required for production of the P1tR- binding co-inducer. Because pyoluteorin biosynthesis is also affected significantly by available carbon sources (Nowak-Thompson et al, 1994), ApdA and GacA may exert their control either by sensing carbon sources or by regulating co-inducer biosynthesis in response to an unknown 81

signal. In light of the number of regulatory loci required for pyoluteorin biosynthesis and the many possible interactions between their gene products, it is apparent that pyoluteorin biosynthesis is controlled by a complex regulatory network, an understanding of which is in its infancy.

4.3 CONCLUSIONS

The divergent orientation of PltR in relationship to the other pyoluteorin biosynthetic genes, the identified LysR-type domains (especially the N- terminal HTH motif), and the presence of the inverted repeat promoter binding region strongly suggest that PltR is a LysR-type regulator. Because pltR is also closely linked to p1tABCD, it is likely that pyoluteorin biosynthesis is regulated by PltR. Whereas several global regulatory loci controlling pyoluteorin production have been identified, this is the first report of a possible pathway- specific regulatory mechanism. In keeping with the LysR regulatory model, PltR most likely requires a co-inducer to effect transcription of the pyoluteorin gene cluster. The absence of this co-inducer within other Ps eudomonas species could account for low transcription levels of the gene cluster within heterologous hosts (B. N.-T., J. Kraus, and J.E. Loper, unpublished results). Because the biocontrol properties of Pf-5 are, at least in part, dependent upon pyoluteorin production and since transcription of the pyoluteorin gene cluster varies with respect to the plant host, the discovery of the pltR locus holds great promise for improvement of biocontrol activity.

4.4 EXPERIMENTAL METHODS

The methods used for nucleic acid manipulations and DNA sequence analysis of pltR were identical to those described in Chapter 3, section 3.4.

Deduced protein sequence analysis. The described domains within PltR (see section 2.3) were identified via Profile Gap using standard parameters. The LysR-profile used was constructed from the following representative 82

LysR-type regulators (Gen Bank accession numbers are in parenthesis): PtxR (U35068) and TrpI (X51868) of Ps. aeruginosa , TcbR of an unidentified Ps. species (M80212), RcbR of Chromatium vinosum (M64032), NahR (J04233) and CatR (U12557) of Ps. putida , LysR (J01614), LeuO (J03862), and IlvY (M14492) of E. coli, and G1tC of Bacillus subtilis (M28509). The presence of a helix-turn-helix motif was confirmed using the program Helix-turn-helix v. 1.0.5 (December, 1993) available via anonymous ftp from the ebi.ac.uk server. The putative P1tR promoter sequence was identified using the search routine in the GCG sequence editor (SeqEd). 83

Chapter 5. Concluding Summary

5.1 PYOLUTEORIN BIOSYNTHESIS

The results of the work described herein have demonstrated that proline is the primary precursor leading to formation of the dichloropyrrole moiety of pyoluteorin. Although other putative intermediates were tested, the lack of incorporation of these labelled compounds into pyoluteorin have precluded any assessment of the biosynthetic pathway. Nevertheless, identification and characterization of several genes residing within a genomic region that was previously found to be required for pyoluteorin production (Kraus and Loper, 1995) has provided some insight as to the assembly of pyoluteorin, as well as establishing this region as the biosynthetic gene cluster. The Type I polyketide synthase (PKS) encoded by pltB and pltC contains several functional domains that each catalyze a single reaction required for forming the resorcinol moiety of pyoluteorin. Some structural features of the pyoluteorin PKS were unexpected; namely the absence of both a loading and a thioesterase domain. With the exception of the rapamycin PKS (Aparicio, 1996), most Type I PKSs possess each of these domains, which are responsible for the initiation and termination, respectively, of polyketide assembly. Although these domains are not absolutely required for PKS activity, their absence makes it difficult to otherwise reconcile polyketide chain initiation and termination. Concurrent work in our lab has identified and partially characterized genes encoding a CoA-ligase and a thioesterase within the pyoluteorin gene cluster (N. Chaney and J. Loper, unpublished). The CoA ligase exhibits similarity to many amino acid activating synthetases that are required for antibiotic biosynthesis in a variety of organisms and may function in the initiation of polyketide biosynthesis by transferring the activated starter unit (perhaps prolyl-CoA) to the pyoluteorin PKS. Likewise, the thioesterase also shares similarity to a variety of thioesterases found within several biosynthetic gene clusters and may catalyze the release and cyclization of the terminal PKS substrate. Nevertheless, future research must be directed at verifying their presumed functions, because the proteins encoded by these 84

genes were identified based on sequence similarity alone. In contrast to the polyketide portion of pyoluteorin, the transformation of proline to the dichloropyrrole moiety is not easily discerned from gene sequences within the biosynthetic gene cluster. Three loci have been identified that encode halogenases that are likely responsible for the chlorination of one or more pyoluteorin pathway intermediates. It is not immediately obvious, however, why three halogenases exist within a pathway that resolves to a dichioro- functionality. Nevertheless, chlorination is required for the biological activity of pyoluteorin (B.N.-T., unpublished result), suggesting that these genes are required for the antifungal properties of pyoluteorin. Prior to this study, only three other halogenases had been identified, making this a poorly characterized class of enzymes. An NAD co-factor binding site was identified tentatively via protein sequence alignment of the halogenases described herein, suggesting that a formal oxidation or reduction may be required for halogenation. In light of the fact that chloride ion is a poor nucleophile, halogenation most likely occurs electrophilically via covalent bond formation between a chloronium equivalent (i.e. CI+) at an electron rich carbon center. Therefore, oxidation of the proline-derived pyrrolidine moiety to the pyrrole may be a prerequisite to chlorination due to the ability of pyrrole to undergo electrophilic reactions. One chemically feasible mechanism that accounts for the formation of the pyrrole and the subsequent halogenation is based on the identification of a gene encoding an acyl CoA-dehydrogenase within the pyoluteorin gene cluster (N. Chaney and J. Loper, unpublished). It is possible that this enzyme may catalyze the initial oxidation of the pyrrolidine ring to the A2-pyrroline derivative, 5.1 (Scheme 5-1). In a subsequent NAD+-dependent reaction catalyzed by either PltA or Orf-S, the A2-pyrroline would be further oxidized to 5.2, introducing the chlorine atom at C5 via proton abstraction from C4. Addition of the C4 chlorine atom would occur analogously via C5 proton abstraction mediated by PltD. Removal of the second proton from C4 would afford the 4,5-dichloropyrrole moiety of pyoluteorin, 5.3. 85

Scheme 5-1

PItA N 0 H NAD+

5.1

4 a a PltD fyFX AIT,E7X N N a O 0 H 5.3

5.2 REGULATION OF PYOLUTEORIN BIOSYNTHESIS

The translated gene production of the pltR locus, which has been identified within the pyoluteorin biosynthetic gene cluster, exhibits sequence similarity with other LysR transcriptional regulators. The divergent orientation of pltR from the pltABCD gene cluster, the identification of a helix-turn-helix motif within the N-terminal region of P1tR, and an inverted repeat sequence within the presumed promoter region of the pltA locus also suggest that P1tR is a LysR-type regulatory protein. The proximity of pltR to several pyoluteorin biosynthetic genes further implies that P1tR specifically regulates the transcription of the biosynthetic gene cluster. Because most LysR-type proteins regulate transcription only in response to a co-inducer molecule, it is likely that pyoluteorin production also requires an as of yet unknown, co-inducer. This is possibly the first report of a LysR-type regulatory gene controlling antibiotic biosynthesis (Doull and Vining, 1995). Future studies will be directed towards demonstrating the function of the gene product as in relationship to pyoluteorin production. 86

Bibliography

Adams, P. B. 1990. The potential of mycoparasites for biological control of plant diseases. Ann. Rev. Phytopath. 28:59-72.

Albright, L. M., E. Huala, and F. M. Ausubel. 1989. Prokaryotic signal transduction mediated by sensor and regulator protein pairs. Annu. Rev. Gen. 23:311-336.

Al-Douri, N. A., and P. M. Dewick. 1988. Biosynthesis of the 3-ethylchromone phytoalexin lathodoratin in Lathyrus odoratus. Phytochem. 27:775-783.

Aparicio, J. F., I. Molnar, T. Schwecke, A. Konig, S.F. Haydock, L.E. Khaw, J. Staunton, and P.F. Lead lay. 1996. Organization of the biosynthetic gene cluster for rapamycin in Streptomyces hygroscopicus: analysis of the enzymatic domains in the modular polyketide synthase. Gene. 169:9-16.

Aragozzini, F., R. Craveri, E. Maconi, G. S. Ricca, and C. Scolastico. 1988. Thermorubin biosynthesis: Evidence for the involvement of both salicylic acid and an undecaketide. J. Chem. Soc., Perkin Trans. 1. 1865-1867.

Armstrong, S. M., and T. R. Patel. 1993. Abiotic conversion of dihydrophloroglucinol to resorcinol. Can. J. Microbiol. 39:899-902.

Bangera, G., and L. Thomashow. 1996. Characterization of a genomic locus required for synthesis of the antibiotic 2,4-diacetylphloroglucinol by the biological control agent Pseudomonas fluorescens Q2-87. Mol. Plant- Microbe Interact. (MPMI). 9:83-90.

Beck, J., S. Ripka, A. Siegner, E. Schiltz, and E. Schweizer. 1990. The multifunctional 6-methylsalicylic acid synthase gene of Penicillium patulum. Its gene structure relative to that of other polyketide synthases. Eur. J. Biochem. 192:487-498. 87

Bedford, D. J., E. Schweizer, D. A. Hopwood, and C. Khosla. 1995. Expression of a functional fungal polyketide synthase in the bacterium Streptomyces coelicolor A3(2). J. Bacteriol. 177:4544-4548.

Bencini, D. A., C. R. Howell, and J. R. Wild. 1983. Production of phenolic metabolites by a soil pseudomonad. Soil Biol. Biochem. 15:491-492.

Bennett, G. J., and H. -H. Lee. 1988. The biosynthesis of mangostin: the origin of the xanthone skeleton. J. Chem. Soc., Chem. Commun. 619-620.

Birch, A. J. 1957. Biosynthetic relations of some natural phenolic and enolic compounds. Prog. Org. Chem. Nat. Prod. 14:186-213.

Birch, A. J. 1967. Biosynthesis of polyketides and related compounds. Science. 156:202-206.

Birch, A. J., P. Hodge, R. W. Rickards, R. Takeda, and T. R. Watson. 1964. The structure of pyoluteorin. J. Chem. Soc. 2641-2644.

Bockholt, H., G. Udvarnoki, and J. Rohr. 1994. Biosynthetic studies on the xanthone antibiotics lysolipins X and I. J. Org. Chem. 59:2064-2069.

Bosshard, H. H., R. Mory, M. Schmid, and H. Zollinger. 1959. Eine methode zur katalysierten herstellung von carbonsaure- and sulfosaure-chloriden mit thionylchlorid. Hely. Chim. Acta. 42:1653-1658.

Brobst, S. W., and C. A. Townsend. 1994. The potential role of fatty acid initiation in the biosynthesis of the fungal aromatic polyketide aflatoxin B1. Can. J. Chem. 72:200-207.

Brown, B. R. 1967. Biochemical aspects of oxidative coupling of phenols, p. 167-202. In Taylor, W. I., Battersby, A. R. (ed.), Oxidative Coupling of Phenols, Marcel Dekker, Inc., New York.

Brown, M. R. W., and R. M. E. Richards. 1965. Effect of ethylenediamine tetraacetate on the resistance of Pseudomonas aeruginosa to antibacterial agents. Nature (London). 207:1391-1393. 88

Bull, A., and J. Slater. 1982. Microbial interaction and community structure, In Bull, A., Slater, J. (ed.), Microbial Interactions and Communities, Academic Press, New York.

Burns, M. K., J. M. Coffin, I. Kurobane, and L. C. Vining. 1979. Biosynthesis of chlorflavonin in Aspergillus candidus: A novel fungal route to flavonoids. J. Chem. Soc., Chem. Commun. 426-427.

Cane, D. E., G. Luo, C. Khosla, C. M. Kao, and L. Katz. 1995. Highly efficient incorporation of polyketide chain elongation intermediates into 6 deoxyerythronolide B in an engineered Streptomyces host. J. Antibiot. 48:647-651.

Cane, D. E., and C.-C. Yang. 1987. Macrolide biosynthesis. 4. Intact incorporation of a chain-elongation intermediate into erythromycin. J. Am. Chem. Soc. 109:1255-1257.

Carbiddu, S., S. Melis, P. P. Piras, and F. Sotgiu. 1979. Metalation reactions. IV. Metalation of o -alkyl- and o,o '-dialkyl-substituted alkylthiobenzenes. j. Organomet. Chem. 178:291-300.

Carbiddu, S., A. Maccioni, P. P. Piras, and A. Plumitallo. 1981. Metalation reactions. VI. Steric limits in the preparation of 1,2,3-trisubstituted benzene derivatives. Gazz. Chim. Ital. 111:123-127.

Carmeli, S., R. E. Moore, G. M. L. Patterson, and W. Y. Yoshida. 1993. Biosynthesis of tolytoxin. Origin of the carbons and heteroatoms. Tetrahedron Lett. 34:5571-5574.

Carson, J. R., and N. M. Davis. 1981. Acid-mediated rearrangement of acylpyrroles. J. Org. Chem. 46:839-843.

Carter, G. T., J. A. Nietsche, J. J. Goodman et al. 1989. Direct biochemical nitration in the biosynthesis of dioxapyrrolomycin. A unique mechanism for the introduction of nitro groups in microbial products. J. Chem. Soc., Chem. Commun. 1271-1273.

Chen, B., G. Choi, and D. Nuss. 1994. Attenuation of fungal virulence by synthetic infectious hypovirus transcripts. Science. 264:1762-1764. 89

Chen, T. S., B. H. Arison, V. P. Gullo, and E. S. Inamine. 1989. Further studies on the biosynthesis of the avermectins. J. Ind. Microbiol. 4:231-238.

Collie, J. N. 1907. Derivative of the multiple keten group. J. Chem. Soc., Trans. 91:1806-1813.

Cook, R. J., Baker, K. F. 1983. The Nature and Practice of Biological Control of Plant Pathogens, APS Press, St. Paul, Minn.

Corbell, N., and J. Loper. 1995. A global regulator of secondary metabolite production in Pseudomonas fluorescens Pf-5. J. Bacteriol. 177:6230-6236.

Cortes, J., S. F. Haydock, G. A. Roberts, D. J. Bevitt, and P. F. Leadlay. 1990. An unusually large multifunctional polypeptide in the erythromycin producing polyketide synthase of Saccharopolyspora erythraea. Nature (London). 348:176-178.

Cue, B., J. Dirlam, L. Czuba et al. 1981. A practical synthesis of pyoluteorin. J. Heterocyclic Chem. 18:191-192.

Cuppels, D. A., C. R. Howell, R. D. Stipanovic et al. 1986. Biosynthesis of pyoluteorin: A mixed polyketide-tricarboxylic acid cycle origin demonstrated by [1,2-13C2] acetate incorporation. Z. Naturforsch., C: Biosci. 41c:532-536.

Dairi, T., T. Nakano, K. Aisaka, R. Katsumata, and M. Hasegawa . 1995. Cloning and nucleotide sequence of the gene responsible for chlorination of tetracycline. Biosci. Biotech. Biochem. 59:1099-1106.

Defago, G. 1993. 2,4-Diacetylphloroglucinol, a promising compound in biocontrol. Plant Pathology. 42:311-312.

Defago, G., and D. Haas. 1990. 1990. Pseudomonads as antagonists of soilborne plant pathogens: Modes of action and genetic analysis, In Bollag, J-M, Stotzky, G. (ed.), Soil Biochemistry, Marcel Dekker, Inc., New York. deLorenzo, V., and A. Aguilar. 1984. Antibiotics from Gram-negative bacteria: do they play a role in microbial ecology? TIBS. 266-269. 90

Demain, A., Y. Aharonowitz, and J. Martin. 1983. Metabolic control of secondary biosynthetic pathways, p. 49-72. In Vining, L. (ed.), Biochemistry and Genetic Regulation of Commercially Important Antibiotics, Addison- Wesley, London.

Deretic, V., W. M. Konyecsni, C. D. Mohr, D. W. Martin, and N. S. Hibler. 1989. Common denominators of promoter control in Pseudomonas and other bacteria. Bio/Tech. 7:1249-1254.

Difco. 1984. Difco Manual, Difco Labs, Detroit, Mich.

Dimroth, P., H. Walter, and F. Lynen. 1970. Biosynthese von 6 methylsalicylsaure. Eur. J. Biochem. 13:98-110.

Dodd, I. B., and J. B. Egan. 1990. Improved detection of helix-turn-helix DNA- binding motifs in protein sequences. Nucl. Acids Res. 18:5019-5026.

Donadio, S., and L. Katz. 1992. Organization of the enzymatic domains in the multifunctional polyketide synthase involved in erythromycin formation in Saccharopolyspora erythraea. Gene. 111:51-60.

Donadio, S., M. Stayer, J.B. McAlpine, S. Swanson, and L. Katz. 1991. Modular organization of genes required for complex polyketide biosynthesis. Science. 252:675-679.

Doull, J. L., and L. C. Vining. 1995. Global physiological controls, p. 9-64. In Vining, L. C., Stuttard, C. (ed.), Genetics and Biochemistry of Antibiotic Production, Butterworth-Heinemann, Boston.

Dutton, C. J., A. M. Hooper, P. F. Leadlay, and J. Staunton. 1994. Avermectin biosynthesis. Intact incorporation of a diketide chain-assembly intermediate into the polyketide macrocyclic ring. Tetrahedron Lett. 35:327-330.

Ebright, R. H. 1986. Proposed amino acid-base pair contacts for 13 sequence- specific DNA binding proteins, p. 207-219. In Oxender, D. L. (ed.), Protein Structure, Folding, and Design, Alan R. Liss, Inc., New York. 91

Farrand, S. K. 1990. Agrobaterium radiobacter strain K84: a model biocontrol system, p. 679-691.1n Baker, R. R., Dunn, P. E. (ed.), New Directions in Biological Control, Alan R. Liss, Inc., New York.

Fenical, W. F. 1993. Chemical studies of marine bacteria: developing a new resource. Chem. Rev. 93:1673-1683.

Fenton, A. M., P. M. Stephens, J. Crowley, M. O'Callaghan, and F. O'Gara. 1992. Exploitation of gene(s) involved in 2,4-diacetylphloroglucinol biosynthesis to confer a new biocontrol capability to a Pseudomonas strain. Appl. Environ. Microbiol. 58:3873-3878.

Fravel, D. R. 1988. Role of antibiosis in the biocontrol of plant diseases. Ann. Rev. Phytopathol. 26:75-91.

Fulton, R. W. 1986. Practices and precautions in the use of cross protection for plant virus disease control. Ann. Rev. Phytopath. 24:67-81.

Gatenbeck, S., and K. Mosbach. 1959. Acetate carboxyl oxygen (180) as donor for phenolic hydroxyl groups of orsellinic acid produced by fungi. Acta Chem. Scand. 13:1561-1564.

Gaucher, G. M., and M. G. Shepherd. 1968. Isolation of orsellinr acid synthase. Biochem. Biophys. Res. Commun. 32:664-671.

Geissman, T. A., Crout, D. H. G. 1969. Organic Chemistry of Secondary Plant Metabolism, Freeman, Cooper, and Company, San Francisco.

Ghisalba, 0., and J. Nuesch. 1981. A genetic approach to the biosynthesis of the rifamycin-chromophore in Nocardia mediterranei. IV. Identification of 3-amino-5-hydroxybenzoic acid as a direct precursor of the seven carbon amino starter unit. J. Antibiot. 34:64-70.

Gilman, H., and P. R. van Ess. 1933. The preparation of ketones by the carbonation of organolithium compounds. J. Am. Chem. Soc. 55:1258 1261. 92

Gilman, H., H. B. Willis, T. H. Cook, F. J. Webb, and R. N. Meals. 1940. Dibenzofuran. XVIII. Isomeric metalation products of some phenols and their methyl ethers. J. Am. Chem. Soc. 62:667-669.

Gotesman, S., and M. R. Maurizi. 1992. Regulation by proteolysis: energy- dependent proteases and their targets. Microbiol. Rev. 56:592-621.

Gottlieb, D. 1976. The production and role of antibiotics in soil. J. Antibiot. 29:987-1000.

Gould, W. D. 1990. Biological control of plant root diseases by bacteria, p. 287-317. In Nakas, J. P., Hagedorn, C. (ed.), Biotechnology of Plant-Microbe Interactions, McGraw-Hill, Inc., New York.

Grafe, U. 1989. Autoregulatory secondary metabolites from Actinomycetes, In Shapiro, S. (ed.), Regulation of Secondary Metabolism in Actinomycetes, CRC Press, Boca Raton, Florida.

Green, T. W., Wuts, P. G. M. 1991. Protective Groups In Organic Synthesis, 2nd Ed., John Wiley & Sons, Inc., New York.

Gribble, G. W. 1996. Naturaly occurring organohalogen compounds - A comprehensive survey. Prog. Chem. Org. Nat. Prod. 68:1-498.

Gutterson, N. 1990. Microbial fungicides: Recent approaches to elucidating mechanisms. Crit. Rev. Biotechnol. 10:69-91.

Gutterson, N., T. Layton, J. Ziegle, and G. Warren. 1986. Molecular cloning of genetic determinants for inhibition of fungal growth by a fluorescent Pseudomonad. J. Bacteriol. 165:696-703.

Gutterson, N., J. Ziegle, G. Warren, and T. Layton. 1988. Genetic determinants for catabolite induction of antibiotic biosynthesis in Pseudomonas fluorescens HV37a. J. Bacteriol. 170:380-385. 93

Haas, D., C. T. Bull, C. Keel, J. Laville, M. Maurhofer, A. Natsch, P. Sacherer, C. Voisard, C. von Schroetter, and G. Defago. 1993. Regulation of secondary metabolites relevant to biocontrol in Pseudomonas fluorescens CHAO, p. 36. Fourth International Symposium on Pseudomonas: Biotechnology and Molecular Biology, Abstract, Vancouver, BC, Canada.

Hailes, H. C., S. Handra, P. F. Leadlay, I. C. Lennon, S. V. Ley, and J. Staunton. 1994a. Biosynthesis of tetronasin: Part 2. Identification of the tetraketide intermediate attached to the polyketide synthase. Tetrahedron Lett. 35:311 314.

Hailes, H. C., C. M. Jackson, P. F. Leadlay, S. V. Ley, and J. Staunton. 1994b. Biosynthesis of tetronasin: Part 1. Introduction and investigation of the diketide and triketide intermediates bound to the polyketide synthase. Tetrahedron Lett. 35:307-310.

Harrison, L. A., L. Letender, P. Kovacevich, E. Pierson, and D. Weller. 1993. Purification of an antibiotic effective against Gaeumannomyces graminis var. tritici produced by a biocontrol agent, Pseudomonas aureofaciens. Soil. Biol. Biochem. 25:215-221.

Hengge-Arronis, R. 1996. Back to log phase: as as a global regulator in the osmotic control of gene expression in Escherichia coli. Mol. Microbiol. 21:887-893.

Henikoff, S., G. W. Haughn, J. M. Calvo, and J. C. Wallace. 1988. A large family of bacterial activator proteins. Proc. Natl. Acad. Sci. U. S. A. 85:6602 6606.

Hewson, W. D., and L. P. Hager. 1980. Bromoperoxidases and halogenated lipids in marine algae. J. Phycol. 16:340-345.

Hill, D. S., J. I. Stein, N. R. Torkewitz, A. M. Morse, C. R. Howell, J. P. Pachlatko, J. 0. Becker, and J. M. Ligon. 1994. Cloning of genes involved in the synthesis of pyrrolnitrin from Pseudomonas fluorescens and role of pyrrolnitrin synthesis in biological control of plant disease. Appl. Environ. Microbiol. 60:78-85. 94

Hill, S., S. T. Lam, P. E. Hammer, and J. Ligon. 1995. Cloning, characterization, and heterologous expression of genes from Pseudomonas fluorescens involved in the synthesis of pyrrolnitrin. Phytopathology. 85:1187 (abstract).

Hobbs, G., C. M. Frazer, D. C. J. Gardner, F. Flett, and S. G. Oliver. 1990. Pigmented antibiotic production by Streptomyces coelicolor A3(2): kinetics and the influence of nutrients. J. Gen. Microbiol. 136:2291-2296.

Hodge, P., and R. W. Rickards. 1963. Improved synthetic routes to pyrrole-2 carboxylic acid and its derivatives. J. Chem. Soc. 2543-2545.

Hopwood, D. A., and D. H. Sherman. 1990. Molecular genetics of polyketides and its comparison to fatty acid biosynthesis. Ann. Rev. Genet. 24:37-66.

Horinouchi, S., and T. Beppu. 1992. Autoregulatory factors and communication in Actinomycetes. Ann. Rev. Microbiol. 46:377-398.

Howell, C. R. 1982. Effect of Gliocladium virens on Pythium ultimum, Rhizoctonia solani, and damping-of of cotton seedlings. Phytopathology. 72:496-498.

Howell, C. R., and R. D. Stipanovic. 1979. Control of Rhizoctonia solani on cotton seedlings with Pseudomonas fluorescens and with an antibiotic produced by the bacterium. Phytopathology. 69:480-482.

Howell, C. R., and R. D. Stipanovic. 1980. Suppression of Pythium ultimum induced damping-off of cotton seedlings by Pseudomonas fluorescens and its antibiotic, pyoluteorin. Phytopathology 70:712-715.

Howell, C. R., and R. D. Stipanovic. 1983. Gliovirdin, a new antibiotic from Gliocladium virens, and its role in the biological control of Pythium ultimum. Can. J. Microbiol. 29:321-324.

Howie, W., D. Matsubara, N. Gutterson, and T. Suslow. 1989. Directed enhancement of biocontrol in Pseudomonas by constitutive antibiotic biosynthesis. Phytopathology. 79:1160 (Abstract). 95

Howie, W. J., and T. V. Suslow. 1991. Role of antibiotic biosynthesis in the inhibition of Pythium ultimum in the cotton spermosphere and rhizosphere by Pseudomonas fluorescens. Mol. Plant-Microb. Interact. 4:393-399.

Hrazdina, G., F. Kreuzaler, K. Hahlbrock, and H. Griseback. 1976. Substrate specificty of flavanone synthase from cell-suspension culture of parsley and structure of release products in vitro. Arch. Biochem. Biophys. 175:392 399.

Hutchinson, C. R., and I. Fujii. 1995. Polyketide synthase gene manipulation: a structure-function approach in engineering novel antibiotics. Ann. Rev. Microbiol. 49:201-238.

Hutchinson , C. R., H. Decker, K. Madduri, S. L. Otten, and L. Tang. 1993. Genetic control of polyketide biosynthesis in the genus Streptomyces. Antonie van Leeuwenhoek. 64:165-176.

James, D. W., and N. I. Gutterson. 1986. Multiple antibiotics produced by Pseudomonas fluorescens HV37a and their differential regulation by glucose. Appl. Environ. Microbiol. 52:1183-1189.

Jeffs, P. W., and D. McWilliams. 1981. Carbon-13 nuclear magnetic resonance study of the biosynthesis of cycloheximide. Stereospecific incorporation of malonate into the glutarimide ring. J. Am. Chem. Soc. 103:6185-6192.

Jones, C. A., P. J. Sidebottom, R. J. P. Cannell, D. Noble, and B. A. M. Rudd. 1992. The squalestatins, novel inhibitors of squalene synthase produced by a species of Phoma. J. Antibiot. 45:1492-1498.

Jordan, P. M., and J. B. Spencer. 1990. Stereospecific manipulation of hydrogen atoms with opposite orientations during the biosynthesis of the polyketide 6-methylsalicylic acid from chiral malonates in Penicillium patulum. J. Chem. Soc., Chem. Commun. 238-242.

Kado, C. I., and M. G. Heskett. 1970. Selective media for isolation of Agrobacterium, Corynebacterium, Erwinia, Pseudomonas, and Xanthomonas. Phytopathology. 60:969-976. 96

Kaiser, H.-P., and J. M. Muchowski. 1984. Catalytic hydrogenation of pyrrole at atmospheric pressure. J. Org. Chem. 49:4203-4214.

Katz, L., and S. Donadio. 1993. Polyketide synthesis: prospects for hybrid antibiotics. Ann. Rev. Microbiol. 47:875-912.

Kawamura, N., R. Sawa, T. Takahashi, T. Sawa, and H. Takeuchi , T. Naganawa. 1996. Pyralomycins, novel antibiotic from Microtetraspora spiralis: Biosynthesis of Pyralomycin la. J. Antibiot. 49:657-660.

Keel, C., U. Schnider, M. Maurhofer, C. Voisard, J. Laville, U. Burger, P. Wirthner, D. Haas, and G. Defago. 1992. Suppression of root diseases by Pseudomonas fluorescens CHAO: Importance of the bacterial secondary metabolite 2,4-diacetylphloroglucinol. Mol. Plant-Microbe Interact. (MPMI). 5:4-13.

Kerr, A. 1980. Biological control of crown gall through production of agrocin 84. Plant Disease. 64:25-30.

Khosla, C., and R. J. X. Zawada. 1996. Generation of polyketide libraries via combinatorial biosynthesis. Tech. Biotechnol. 14:335-341.

Kibby, J. J., I. A. McDonald, and R. W. Rickards. 1980. 3-Amino-5 hydroxybenzoic acid as a key intermediate in ansamycin and maytansinoid biosynthesis. J. Chem. Soc., Chem. Commun. 768-769.

Kindl, H. 1985. Biosynthesis of stilbenes, p. 349-377. In Higuchi, T. (ed.), Biosynthesis and Biodegradation of Wood Components, Academic Press, Inc., New York.

King, E. 0., M. K. Ward, and D. E. Raney. 1954. Two simple media for the demonstration of pyocyanin and fluorescin. J. Lab. Clin. Med. 44:301-307.

Kleinkauf, H., and H. von Dohren. 1995. Linking peptide and polyketide biosynthesis. J. Antibiot. 48:563-567.

Kleinkauf, H., and H. von Dohren. 1996. Review: A nonribosomal system of peptide biosynthesis. Eur. J. Biochem. 236:335-351. 97

Kobayashi, H., S. Iwasaki, E. Yamadam, and S. Okuda. 1986. Biosynthesis of the antimitotic antitumour antibiotic rhizoxin by Rhizopus chinensis; Origins of the carbon atoms. J. Chem. Soc., Chem. Commun. 1702-1703.

Kolter, R., D. Siegele, and A. Tormo. 1993. The stationary phase of the bacterial life cycle. Ann. Rev. Microbiol. 47:855-874.

Kraus, J., and J. Loper. 1992. Lack of evidence for a role of antifungal metabolite production by Pseudomonas fluorescens Pf-5 in biological control of Pythium damping-off of cucumber. Phytopathology. 82:264-271.

Kraus, J., and J. Loper. 1995. Characterization of a genomic region required for production of the antibiotic pyoluteorin by the biological control agent Pseudomonas fluorescens Pf-5. Appl. Environ. Microbiol. 61:849 854.

Kreuzaler, F., and K. Hahlbrock. 1975. Enzymic synthesis of an aromatic ring from acetate units. Partial purification and some properties of flavanone synthase from cell-suspension cultures of Petroselinum hortense. Eur. J. Biochem. 56:205-213.

Kreuzaler, F., H. Ragg, W. Heller, R. Tesch, I. Witt, D. Hammer, and K. Hahlbrock. 1979. Flavanone synthase from Petroselinum hortense: Molecular weight, subunit composition, size of messenger RNA, and absence of pantetheinyl residue. Eur. J. Biochem. 99:89-96.

Lam, S. T., T. D. Gaffney, R. A. Frazelle, K. Gates, J. DiMaio, N. Torkewitz, J. Ligon, S. Hill, S. Goodwin, and H. -. Kempf. 1993. Two genes which regulate the coordinated expression of antifungal activities in Pseudomonas fluorescens, p. 209. Fourth International Symposium on Pseudomonas: Biotechnology and Molecular Biology, Abstract, Vancouver, BC, Canada.

Lange, R., and R. Hengge-Aronis. 1994. The cellular concentration of the as subunit of RNA polymerase in Escherichia coli is controlled at the levels of transcription, translation, and protein stability. Genes. Develop. 8:1600 1612. 98

Lanz, T., S. Tropf, F. -J. Marner, J. Schroder, and G. Schroder. 1991. The role of cysteines in polyketides synthases: Site directed mutagenesis of resveratrol and chalcone synthases, two key enzymes in different plant- specific pathways. J. Biol. Chem. 266:9971-9976.

Laville, J., C. Voisard, C. Keel, M. Maurhofer, G. DeFago, and D. Haas. 1992. Global control in Pseudomonas fluorescens mediating antibiotic synthesis and suppression of black root rot of tobacco. Proc. Natl. Acad. Sci. U. S. A. 89:1562-1566.

Lee, J. J., P. W. Dewick, C. P. Gorst-Allman, F. Spreafico, C. Kowal, C. -J. Chang, A. G. McInnes, J. A. Walter, P. J. Keller, and H. G. Floss. 1987. Further studies on the biosynthesis of the boron-containing antibiotic aplasmomycin. J. Am. Chem. Soc. 109:5426-5432.

Lee, J. P., S. -W. Tsao, C. -J. Chang, X. -G. He, and H. G. Floss. 1994. Biosynthesis of naphthamycin A in Streptomyces collinus. Can. J. Chem. 72:182-188.

Leisinger, T., and R. Margraff. 1979. Secondary metabolites of the fluorescent pseudomonads. Microbiol. Rev. 43:422-442.

Leive, L. 1968. Studies on the permeability change produced in coliform bacteria by ethylenediaminetetraacetate. J. Biol. Chem. 243:2373-2380.

Letellier, G., and L. P. Bouthillier. 1957. Synthesis of 2-pyrrolecarboxylic (14CO2H) acid and study of its metabolism in the intact rat. Can. J. Biochem. Physiol. 35:811-817.

Levy, E., F. J. Gough, K. D. Berlin, P. W. Guiana, and J. T. Smith. 1992. Inhibition of Septoria tritici and other phytopathagenic fungi and bacteria by Pseudomonas fluorescens and its antibiotics. Plant Pathology. 41:335 341.

Ligon, J. M., S. T. Lam, T. D. Gaffney, D. S. Hill, P. E. Hammer, and N. Torkewitz. 1996. Biocontrol: genetic modification for enhanced antifungal activity, p. 457-462. In Stacey, G., Mullin, B., Gresshoff, P. M. (ed.), Biology of Plant-Microbe Interactions, ISMPMI, St. Paul, Minn. 99

Lipman, D. J., W. J. Wilbur, T. F. Smith, and M. S. Waterman. 1984. On the statistical significance of nucleic acid similarities. Nucl. Acid. Res. 12:215 226.

Liras, P., J. Asturias, and J. Martin. 1990. Phosphate control sequences involved in transcriptional regulation of antibiotic biosynthesis. TIBTech. 8:184-189.

Loper, J. E., and J. S. Buyer. 1991. Siderophores in microbial interactions on plant surfaces. Mol. Plant-Microbe Interact. (MPMI). 4:5-13.

Loper, J., N. Corbell, J. Kraus, B. Nowak-Thompson, M. Henkels, and S. Carnegie. 1994. Contribution of molecular biology towards understanding mechanisms by which rhizosphere Pseudomonads effect biological control, p. 89-96. In Ryder, MH, Stephens, PM, Bowen, GD (ed.), Improving Plant Productivity with Rhizosphere Bacteria, CSIRO Division of Soils, Glen Osmund, South Australia.

Lowden, P. A. S., G. A. Bohm, J. Staunton, and P. F. Leadlay. 1996. The nature of the starter unit for the rapamycin polyketide synthase. Angew. Chem. Int. Ed. Engl. 35:2249-2251.

Lugtenberg, B. J. J., L. A. de Weger, and J. W. Bennett. 1991. Microbial stimulation of plant growth and protection from disease. Curr. Opin. Biotechnol. 2:457-464.

MacDonald, W. L., and D. W. Fulbright. 1991. Biological control of chestnut blight. Use and limitations of transmissible hypovirulence. Plant Disease. 75:656-661.

Mann, J. 1987. Secondary Metabolism, Clarendon Press, Oxford.

Maplestone, R., M. Stone, and D. Williams. 1992. The evolutionary role of secondary metabolites - a review. Gene. 115:151-157.

Martin, J. 1992. Clusters of genes for the biosynthesis of antibiotics: regulatory genes and overproduction of pharmaceuticals. J. Indust. Microbiol. 9:73-90. 100

Maurhofer, M., C. Keel, D. Voisard, D. Haas, and G. Defago. 1992. Influence of enhanced antibiotic production in Pseudomonas fluorescens strain CHAO on its disease suppressive capacity. Phytopathology. 82:190-195.

Maurhofer, M., C. Keel, D. Haas, and G. Defago. 1994. Pyoluteorin production by Pseudomonas fluorescens CHAO is involved in the supression of Pythium damping-off of cress but not of cucumber. Eur. J. Plant Pathol. 100:221-232.

Maurhofer, M., C. Keel, D. Haas, and G. Defago. 1995. Influence of plant species on disease suppression by Pseudomonas fluorescens strain CHAO with enhanced antibiotic production. Plant Pathol. 44:40-50.

Mazzola, M., R. J. Cook, L. S. Thomashow, D. M. Weller, and L. S. Pierson . 1993. Contribution of phenazine antibiotic biosynthesis to the ecological competence of fluorescent Pseudomonads in soil habitats. Appl. Environ. Microbiol. 58:2616-2624.

Mazzola, M., D. K. Fujimoto, L. S. Thomasow, and R. J. Cook. 1995. Variation in sensitivity of Gaeumannomyces graminis to antibiotics produced by fluorescent Pseudomonas sp. and effect on biological control of take-all of wheat. Appl. Environ. Microbiol. 61:2554-2559.

McCarthy, J. E. G., and C. Gualerzi. 1990. Translational control of prokaryotic gene expression. Trends Genet. 6:78-85.

McDaniel, R., S. Ebert-Khosla, H. Fu, D. A. Hopwood, and C. Khosla. 1994. Engineered biosynthesis of novel polyketides: Influence of a downstream enzyme on the catalytic specificity of a minimal aromatic polyketide synthase. Proc. Natl. Acad. Sci. U. S. A. 91:11542-11546.

McDaniel, R., S. Ebert-Khosla, D. A. Hopwood, and C. Khosla. 1995. Rational design of aromatic polyketide natural products by recombinant assembly of enzymatic subunits. Nature (London). 375:549-554.

McGeady, P., and R. Croteau. 1993. A mild and convenient method for tritium labelling of activated aromatic compounds using BF3-Et20 and tritiated water. J. Chem. Soc., Chem. Commun. 774-776. 101

Meyers, A. I., and W. B. Avila. 1980. Solvent and base studies on the site of aryl metalation of 2-( 3, 5- dimethoxyphenyl )- 4,4- dimethyl- 2- oxazoline. Tetrahedron Lett. 21:3335-3338.

Moore, B. S., H. Cho, R. Casati, E. Kennedy, K. A. Reynolds, U. Mocek, J. M. Beale, and H. G. Floss. 1993. Biosynthetic studies on ansatrienin A. Formation of the cyclohexanecarboxylic acid moiety. J. Am. Chem. Soc. 115:5254-5266.

Moore, L. W., and G. Warren. 1979. Agrobacterium radiobacter strain 84 and biological control of crown gall. Ann. Rev. Phytopath. 17:163-179.

Neidhardt, F. C., and R. A. VanBogelen. 1987. Heat shock response, p. 1334-1345. In Neidhardt, F. C. (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, ASM, Washington, D.C.

Neidleman, S., and J. Geigert. 1987. Biological halogenation: Roles in nature, potential in industry. Endeavour. 11:5-15.

Nelson, E. B., W. L. Chao, J. M. Norton, G. T. Nash, and G. E. Harman. 1986. Attachment of Enterobacter cloacae to hyphae of Pythium ultimum: Possible role in the biological control of Pythium preemergence damping- off. Phytopathology. 76:327-335.

Nelson, E. B., and A. P. Maloney. 1992. Molecular approaches for understanding biological control mechanisms in bacteria: Studies of the interaction of Enterobacter cloacae with Pythium ultimum. Can. J. Plant Pathol. 14:106-114.

Nowak-Thompson, B., S.J. Gould, J. Kraus, and J.E. Loper. 1994. Production of 2,4-diacetylphloroglucinol by the biocontrol agent Pseudomonas fluorescens Pf-5. Can. J. Microbiol. 40:1064-1066.

O'Hagan, D. 1995. Biosynthesis of fatty acid and polyketide metabolites. J. Nat. Prod. 12:1-32.

Omura, S., A. Nakagawa, K. Shibata, and H. Sano. 1982. The structure of hitachimycin, a novel macrocyclic lactam involving 13-phenylalanine. Tetrahedron Lett. 23:4713-4716. 102

O'Sullivan, D. J., and F. O'Gara. 1992. Traits of fluorescent Pseudomonas spp. involved in suppression of plant root pathogens. Microb. Reviews. 56:662-676.

Parkinson, J., and E. Kofoid. 1992. Communication modules in bacterial signaling proteins. Ann. Rev. Genet. 26:71-112.

Paulitz, T. C. 1990. Biochemical and ecological aspects of competition in biological control, p. 713-724. In Baker, R. R., Dunn, P. E. (ed.), N ew Directions in Biological Control, Alan R. Liss, Inc., New York.

Paulitz, T. C., C. S. Park, and R. Baker. 1987. Biological control of Fusarium wilt of cucumber with nonpathogenic isolates of Pusarium oxysporum. Can. J. Microbiol. 33349-353.

Pelletier, I., and J. Altenbuchner. 1995. A bacterial esterase is homologous with non-haem haloperoxidases and displays brominating activity. Microbiology. 141:459-468.

Penfold, C. N., C. L. Bender, and J. G. Turner. 1996. Characterization of genes involved in biosynthesis of coronafacic acid, the polyketide component of the phytotoxin coronatine. Gene. 183:167-173.

Pfender, W. F., J. Kraus, and J. Loper. 1993. A genomic region from Pseudomonas fluorescens Pf-5 required for pyrrolnitirn production and inhibition of Pyrenophora tritici-repentis in wheat straw. Phytopathology 83:1223-1228.

Pierson, L. S., T. Gaffney, S. Lam, and F. Gong. 1995. Molecular analysis of genes encoding phenazine biosynthesis in the biological control bacterium Pseudomonas aureofaciens 30-84. FEMS Microbiol. Lett. 134:299-307.

Pierson, L., V. Keppenne, and D. Wood . 1994. Phenazine antibiotic biosynthesis in Pseudomonas aureofaciens 30-84 is regulated by PhzR in response to cell density. J. Bacteriol. 176:3966-3974.

Pierson, L. S., and E. A. Pierson. 1996. Phenazine antibiotic production in Pseudomonas aureofaciens: role in rhizosphere ecology and pathogen supression. FEMS Microbiol. Lett. 136:101-108. 103

Pierson, L. S. III, and L. S. Thomashow. 1992. Cloning and heterologous expression of the phenazine biosynthetic locus from Pseudomonas aureofaciens 30-84. Mol. Plant-Microbe Interact. 5:330-339.

Plater, R., and W. R. Strohl. 1994. Polyketide biosynthesis: antibiotics in Streptomyces, p. 61-91. In Ellis, B. E. et al. (ed.), Recent Advances in Phytochemistry: Genetic Engineering of Plant Secondary Metabolites, Plenum Press, New York.

Rao, K. V., and G. C. Reddy. 1990. Synthesis and herbicidal activity of the halo analogues of pyoluteorin. J. Agric. Food Chem. 38:1260-1263.

Reville, W. P., M. J. Bibb, and D. A. Hopwood. 1995. Purification of a malonyltransferase from Streptomyces coelicolor A3(2) and analysis of its genetic determinant. J. Bacteriol. 177:3949-3952.

Robinson, J. A. 1991. Chemical and biochemical aspects of polyether-ionophore antibiotic biosynthesis. Prog. Chem. Org. Nat. Prod. 58:1-81.

Rohr, J. 1992. Comparison of multicyclic polyketides by folding analysis: A novel approach to recognize biosynthetic and/or evolutionary interrelations of the natural products or intermediates and its exemplification on hepta-, octa-, and decaketides. J. Org. Chem. 57:5217 5223.

Rohr, J. 1995. Combinatorial biosynthesis - an approach in the near future? Angew. Chem. Int. Ed. Engl. 34:881-885.

Sarniguet, A., J. Kraus, M.D. Henkels, A.M. Muehlchen, and J.E. Loper. 1995. The sigma factor as affects antibiotic production and biological control activity of Pseudomonas fluorescens Pf-5. Proc. Natl. Acad. Sci. U.S.A. 92:12255-12259.

Schell, M. 1993. Molecular biology of the LysR family of transcriptional regulators. Ann. Rev. Microbiol. 47:597-626.

Schlaman, H. R. M., R. J. H. Okker, and B. J. J. Lugtenberg. 1992. Regulation of nodulation gene expression by NodD in Rhizobia. J. Bacteriol. 174:5177 5182. 104

Schnider, U., C. Keel, C. Blumer, J. Troxler, G. Defago, and D. Haas. 1995. Amplification of the housekeeping sigma factor in Pseudomonas fluorescens CHAO enhances antibiotic production and improves biocontrol abilities. J. Bacteriol. 177:5387-5392.

Schroder, G., J. W. S. Brown, and J. Schroder. 1988. Molecular analysis of resveratrol synthase; cDNA, genomic clones, and relationship with chalcone synthase. Eur. J. Biochem. 172:161-169.

Schulman, M. D., D. Valentino, and 0. Hensens. 1986. Biosynthesis of the avermectins by Streptomyces avermitilis. Incorporation of labelled precursors. J. Antibiot. 39:541-549.

Scott, A. I., L. C. Bead ling, N. H. Georgopapadakou, and C. R. Subbarayan. 1974. Biosynthesis of polyketides. Purification and inhibition studies of 6-methylsalicylic acid synthase. Bioorg. Chem. 3:238-248.

Scrutton, N. S., A. Berry, and R. N. Perham. 1990. Redesign of the coenzyme specificity of a dehydrogenase by protein engineering. Nature (London). 343:38-43.

Serre, L., E. C. Verbree, Z. Dauter, A. R. Stuitje, and Z. S. Derewenda. 1995. The Escherichia colt malonyl-CoA:acyl carrier protein transacylase at 1.5-A resolution. J. Biol. Chem. 270:12961-12964.

Seto, H., H. Sato, S. Urano, J. Uzawa, and H. Yonehara. 1976. Utilization of 13C-13C coupling in structural and biosynthetic studies. VII The structure and biosynthesis of vulgamycin. Tetrahedron Lett. 48:

Shanahan, P., J. D. Glennon, J. J. Crowley, D. F. Donnelly, and F. O'Gara. 1993. Liquid chromatographic assay of microbially derived phloroglucinol antibiotics for establishing the biosynthetic route to production, and the factors affecting their regulation. Anal. Chim. Acta. 272:271-277.

Shanahan, P., D. J. O'Sullivan, P. Simpson, J. D. Glennon, and F. O'Gara. 1992. Isolation of 2,4-diacetylphloroglucinol from a fluorescent pseudomonad and investigation of physiological parameters influencing its production. Appl. Environ. Microbiol. 58:353-358. 105

Shim, J.-S., S. K. Farrand, and A. Kerr. 1987. Biological control of crown gall: construction and testing of new biocontrol agents. Phytopathology. 77:463 466.

Simpson, T. J. 1995. Polyketide biosynthesis. Chem. Ind. 407-411.

Slocum, D. W., and C. A. Jennings. 1976. Directed metalation reaction. 6. Competition of substituents for ortho direction of metalation in substituted anisoles. J. Org. Chem. 41:3653-3664.

Sonnett, P. E. 1971. Synthesis of the trail marker of the Texas Leaf-Cutting Ant, Atta texana (Buckley). J. Med. Chem. 15:97-98.

Spencer, J. B., and P. M. Jordan. 1990. Use of chiral malonates to determine the absolute configuration of the hydrogen atoms eliminated during the formation of 6-methylsalicylic acid by 6-methylsalicylic acid synthase from Penicillium patulum. J. Chem. Soc., Chem. Commun. 1704-1706.

Spencer, J. B., and P. M. Jordan. 1992a. Purification and properties of 6 methylsalicylic acid synthase from Penicillium patulum. Biochem. J. 288:839-846.

Spencer, J. B., and P. M. Jordan. 1992b. Stereospecific elimination of hydrogen atoms with opposite absolute orientations during the biosynthesis of orsellinic acid from chiral malonates in Penicillium cyclopium. J. Chem. Soc., Chem. Commun. 646648.

Staley, A. L., and K. L. Rinehart. 1991. Biosynthesis of the streptovaricins: 3-Amino-5-hydroxybenzoic acid as a precursor to the me t a-C7N unit. J. Antibiot. 44:218-222.

Stanisich, V. A., and B. W. Holloway. 1972. A mutant sex factor of Pseudomonas aeruginosa. Genet. Res. 19:91-108.

Still, W. C., M. Kahn, and A. Mitra. 1978. Rapid chromatographic technique for preparative separations with moderate resolution. J. Org. Chem. 43:2923-2925. 106

Stipanovic, R. D., and C. R. Howell. 1982. The structure of gliovirin, a new antibiotic from Gliocladium virens. J. Antibiot. 35:1326-1330.

Stock, J. B., A. M. Stock, and J. M. Mottonen. 1990. Signal transduction in bacteria. Nature (London). 344:395-400.

Stone, M. J., and D. H. Williams. 1992. On the evolution of functional secondary metabolites (natural products). Mol. Microbiol. 6:29-34.

Strunz, G. M., R. E. Wall, D. J. Kelly, and M. A. Holder-Franklin. 1978. Phloroglucinol derivatives from Aeromonas hydrophila. J. Antibiot. 31:1201-1202.

Summers, R. G., A. Ali, B. Shen, W. A. Wessel, and C. R. Hutchinson. 1995. Malonyl-coenzyme A:acyl carrier protein acyltransferase of Streptomyces glaucescens: A possible link between fatty acid and polyketide biosynthesis. Biochemistry. 34:9389-9402.

Swan, D.G., A.M. Rodriguez, C. Vilches, C. Mendez, and J.A. Salas. 1994. Characterization of a Streptomyces antibioticus gene encoding a type I polyketide synthase which has an unusual coding sequence. Mol. Gen. Genet. 242:358-362.

Taguchi, H., U. Sankawa, and S. Shibata. 1969. Biosynthesis of natural products. VI. Biosynthesis of usnic acid in lichens. A general scheme of biosynthesis of usnic acid. Chem. Pharm. Bull. 17:2054-2060.

Takeda, R. 1958. Structure of a new antibiotic, pyoluteorin. J. Am. Chem. Soc. 80:4749-4750.

Thiericke, R., and J. Rohr. 1993. Biological variation of microbial metabolites by precursor-directed biosynthesis. J. Nat. Prod. 10:265-89.

Thiericke, R., A. Zeeck, A. Nakagawa, S. Omura, R. E. Herrold, S. T. S. Wu, J. M. Beale, and H. G. Floss. 1990. Biosynthesis of the manumycin group antibiotics. J. Am. Chem. Soc. 112:3979-3987.

Thomas, R., and D. J. Williams. 1983. Oxytetracycline biosynthesis: Origin of the carboxamide substituent. J. Chem. Soc., Chem. Commun. 677-679. 107

Thomashow, L. S., M. G. Bangera, R. F. Bonsall, D. Kim, J. Raaijmakers, and D. M. Weller. 1996. 2,4-Diacetylphloroglucinol, a key antibiotic in soilborne pathogen suppression by fluorescent Pseudomonas spp., In Stacey, G., Mullin, B., Gresshoff, P. M. (ed.), Biology of Plant-Microbe Interactions, ISMPMI, St. Paul, Minn.

Thomashow, L. S., D. M. Weller, R. F. Bonsall, and L. S. Pierson. 1990. Production of the antibiotic phenazine-l-carboxylic acid by fluorescent Pseudomonas species in the rhizosphere of wheat. Appl. Environ. Microbiol. 56:908-912.

Thomashow, L., and D. Weller. 1988. Role of phenazine antibiotic from Pseudomonas fluorescens in biological control of Gaeumannomyces graminis var. tritici. J. Bacteriol. 170:3499-3508.

Thomashow, L., and D. Weller. 1995. Current concepts in the use of introduced bacteria for biological control: Mechanisms and antifungal metabolites, In Stacey, G., Keen, N. T. (ed.), Plant-Microbe Interactions, Vol. 1, Chapman and Hall, New York.

Townsend, C. A., S. B. Christensen, and K. Trautwein. 1984. Hexanoate as a starter unit in polyketide biosynthesis. J. Am. Chem. Soc. 106:3868-3869.

Tsou, H. -R., Z. H. Ahmed, R. R. Fiala, M. W. Bullock, G. T. Carter, J. J. Goodman, and D. B. Borders. 1989. Biosynthetic origin of the carbon skeleton and oxygen atoms of the LL-F2849, a potent antiparasitic macrolide. J. Antibiot. 42:398-406.

Tu, J. C. 1980. Gliocladium virens, a destructive mycoparasite of Sclerotinia sclerotiorum. Phytopathology. 70:670-674.

Tu, J., and Vaartaja. 1981. The effect of the hyperparasite (Gliocladium virens) on Rhizoctonia solani and on Rhizoctonia root rot of white beans. Can. J. Bot. 59:22-27.

Vaara, M. 1992. Agents that increase the permeability of the outer membrane. Microbiol. Rev. 56:395-411. 108

Vaara, M. 1990. The effect of oligolysines Lys-3, Lys-4, and Lys-5 on the outer membrane permeability of Pseudomonas aeruginosa. FEMS Microbiol. Lett. 67:15-20. van Pee, K.-H. 1996. Biosynthesis of halogenated metabolites by bacteria. Ann. Rev. Microbiol. 50:375-399.

Via le, A. M., H. Kobayashi, T. Akazawa, and S. Henikoff. 1991. rcbR, a gene coding for a member of the LysR family of transcriptional regulators, is located upstream of the expressed set of the ribulose 1,5-bisphosphate carboxylase/oxygenase gene in the photosynthetic bacterium Chromatium vinosurn. J. Bacteriol. 173:5224-5229.

Viljanen, P., H. Kayhty, M. Vaara, and T. Vaara. 1986. Suseptibility of Gram- negative bacteria to the synergistic bactericidal action of serum and polymyxin B nonapeptide. Can. J. Microbiol. 32:66-69.

Vincent, M. N., L. A. Harrison, J. M. Brackin, P. A. Kovacevich, P. Mukerji, D. M. Weller, and E. A. Pierson. 1991. Genetic analysis of the antifungal activity of a soilborne Pseudomonas aureofaciens strain. Appl. Environ. Microbiol. 57:2928-2934.

Vining, L.C. 1992. Secondary metabolism, inventive evolution and biochemical diversity - a review. Gene. 115:135-140.

Vining, L.C. 1995. Global regulation of secondary metabolism. Proc. Ninth Symp. Actinomycetes. 183-190.

Voisard, C., C. Keel, D. Haas, and G. Defago. 1989. Cyanide production by Pseudomonas fluorescens helps suppress black root rot of tobacco under gnotobiotic conditions. EMBO J. 8:351-358.

Walters, D. S., R. Craig, and R. 0. Mumma. 1990. Fatty acid incorporation in the biosynthesis of anacardic acids of geraniums. Phytochem. 29:1815-1822.

Wat, C. -K., A. G. McInnes, D. G. Smith, and L. C. Vinings. 1972. Use of 13C in biosynthetic studies. Location of isotope from labelled acetate and formate in shanorellin by '3C nuclear magnetic resonance spectroscopy. Can. J. Biochem. 50:620-624. 109

Weller, D. M., and L. S. Thomashow. 1993. Microbial metabolites with biological activity against plant pathogens, p. 173-180. In Lumsden, R. D., Vaughn, J. L. (ed.), Pest Management: Biologically Based Technologies Proceedings of Beltsville Symposium XVIII, Beltsville, MD, May 2-6, 1993, American Chemical Society, Washington D. C.

Weller, DM. 1988. Biological control of soilborne plant pathogens in the rhizosphere with bacteria. Ann. Rev. Phytopathol. 26:379-407.

West, S.E.H., and B.H. Iglewski. 1988. Codon usage in Pseudomonas aeruginosa. Nucleic Acid Research. 16:9323-9335.

Williams, S. T., and J. C. Vickers. 1986. The ecology of antibiotic production. Microb. Ecol. 12:43-52.

Xu, G. W., and D. C. Gross. 1986. Selection of fluorescent pseudomonads antagonistic to Erwinia carotovora and supressive of potato seed piece decay. Phytopathology. 76:414-422.

Yu, J. -H., and T. J. Leonard. 1995. Sterigmatocystin biosynthesis in Aspergillus nidulans requires a novel Type I polyketide synthase. J. Bacteriol. 177:4792 4800.

Yue, S., J. S. Duncan, Y. Yamamoto, and C. R. Hutchinson. 1987. Macrolide biosynthesis. Tylactone formation involves the processive addition of three carbon units. J. Am. Chem. Soc. 109:1253-1255.

Zamir, L. 0. 1980. The biosynthesis of patulin and penicillic acid, p. 224-268. In Steyn, P. S. (ed.), The Biosynthesis of Mycotoxins: A Study in Secondary Metabolism, Academic Press, New York. 110

APPENDICES 111

Appendix 1. Production of 2,4-diacetylphlorogiucinol by the biocontrol agent Pseudomonas fluorescens Pf-5

NOTE: The material presented within this appendix was written under joint authorship with Drs. J. Kraus, J.E. Loper, and S.J. Gould and was published in Can. J. Microbiol. (1994) 40:1064-1066. B.N.-T. was the principal author of this work and responsible for the isolation and structural confirmation of 2,4-diacetylphloroglucinol and assisted in the quantitative analysis of antibiotic production. Dr. J. Kraus identified bioactivity attributed to 2,4 diacetylphloroglucinol production and was responsible for the quantitative analysis of antibiotic production within this strain.

Antibiosis has been established as an important mechanism of biological control against plant pathogens through the application of modern genetic techniques (Lugtenberg et al. 1991; O'Sullivan and O'Gara 1992; Weller and Thomashow 1993). Current research has focused on metabolites produced by fluorescent pseudomonads and has implicated oomycin A (Howie and Suslow 1991), phenazine-l-carboxylate (Pierson and Thomashow 1992; Thomashow and Weller 1991), 2,4-diacetylphloroglucinol (Keel et al. 1992; Vincent et al. 1991), pyoluteorin (Maurhofer et al. 1992), and hydrogen cyanide (Voisard et al. 1989) as factors contributing to the suppression of plant disease. Pseudomonas fluorescens strain Pf-5, originally isolated from a cotton rhizosphere, is an effective biocontrol agent against Pythium damping-off of cotton (Howell and Stipanovic 1980) and cucumber (Kraus and Loper 1992) and Rhizoctonia damping-off of cotton (Howell and Stipanovic 1979). Pf-5 also inhibits ascocarp formation by Pyrenophora tritici-repentis on wheat straw residue (Pfender et al. 1993). The antibiotics pyrrolnitrin and pyoluteorin have been isolated from cultures of Pf-5 and, when these metabolites were applied to seed surfaces, seedling damping-off caused by Rhizoctonia solani and Pythium ultimurn was suppressed (Howell and Stipanovic 1979, 1980). In addition to these metabolites, Pf-5 produces hydrogen cyanide, ammonia, a pyoverdine siderophore, and a previously uncharacterized antibiotic (A3), 112

which inhibits growth of R. solani and P. ultimurn (Kraus and Loper 1992). We report herein the isolation and identification of A3, the effect of various carbon sources on A3 production, and its antibiotic activity against several phytopathogenic organisms. Antagonism of P. ultimum by Pf-5 in culture was attributed earlier to the production of pyoluteorin (Howell and Stipanovic 1980). Pf-5 mutants deficient in pyoluteorin production, however, inhibit growth of P. ultimum on glucose-supplemented nutrient agar (NAG1c) (Kraus and Loper 1992), on which Pf-5 produces A3. Cultures of Pf-5 grown on NAG1c were extracted with aqueous acetone (80% v/v acetone), and A3 was partially purified by thin layer chromatography (TLC) (silica gel developed in chloroform/acetone, 9/1, v/v). Because 2,4-diacetylphloroglucinol (Phl) is produced by many strains of P.fluorescens that exhibit biocontrol activity (Defago 1993; Harrison et al. 1993; Levy et al. 1992; Shanahan et al. 1992) and inhibits mycelial growth of P. ultimum (Keel et al. 1992), we considered it a likely candidate for the identity of A3. A synthetic sample of Phl co-migrated with A3 on TLC and both compounds formed a brilliant yellow derivative when TLC plates were visualized with diazosulfanilic acid. Furthermore, analysis by C18 reverse phase HPLC (0.8 x 10 cm Waters Nova-pak radial compression cartridge eluted with H20/MeCN/Me0H, 45:30:25 at 1.5 mL/min) revealed that both compounds exhibited identical retention times (6.4 min) and possessed identical UV spectra (UV., 270 nm). To identify the structure of A3 unequivocally, we isolated a sufficient quantity of the compound for spectroscopic analysis from a Tn5 mutant of Pf-5 (JL3985) that over-produced A3 regardless of the carbon source present in culture media (Pfender et al. 1993; Sarniguet et al. 1994). The bacterium was cultured in 1.5 L of glycerol-Phytone medium (5.0 g/L glycerol, 20 g/L Phytone Peptone (Becton and Dickinson, Cockeysville, MD), 1.5 g/L K2HPO4, 1.5 g/L MgSO4 7 H2O, pH 7.0) at 20 °C with agitation (200 rpm). After 48 h, the culture supernatant was acidified to pH 2 with 3M HC1, extracted with ethyl acetate (3 x 150 mL), and the crude extract (0.5 g) was fractionated on a flash-silica gel column (5 x 15 cm) eluted with toluene/acetone (3/1, v/v). The fractions containing A3 (250 mg) were combined and re-chromatographed on a flash-silica gel column (2.5 x 15 cm) eluted with chloroform/methanol (9/1, v/v). Fractions containing pure A3 (180 mg) were combined and analyzed 113

by standard 1H and 13C NMR spectroscopy. The chemical shifts in both the 1H and the 13C NMR spectra of A3 were identical with those observed for authentic Phl. (A3 was recrystallized from acetone/water: mp 173-174°C (lit. Strunz et al. 1978, 168-172°C); 1H NMR (400 MHz, acetone-d6) 8 16.30 (br s, 1H), 5.85 (s, 1H), and 2.60 (s, 6H); 13C NMR (100 MHz, acetone-d6) 8 204.56, 172.69, 170.06, 104.67, 95.62, and 32.76.) The spectrum of antibiotics produced by Pf-5 was affected by the carbon source composition of the culture medium (Table 1).When Pf-5 cultures were grown in nutrient broth (Difco Laboratories, Detroit, MI) supplemented with glucose (20 g/L), Phl was produced almost exclusively. This was consistent with the earlier observation that antibiotic A3 is produced on NAG1c. This result contrasts with the report by Shanahan et al. (1992) that Phl production by P. fluorescens F113 was reduced in response to glucose in culture media, and may reflect either differences in medium formulation or a fundamental difference in the regulation of Phl production between the two Pfluorescens strains. In contrast to Phl, pyoluteorin and pyrrolnitrin production by Pf-5 was diminished if nutrient broth was supplemented with glucose but enhanced if the medium was supplemented with glycerol (Table 1). The differential regulation of antibiotic production within a given strain of Pseudomonas sp. is not without precedent and has been described for P. fluorescens strain HV37a (James and Gutterson, 1986).

Table 1. Concentrations of antibiotics produced by P. fluorescens Pf-5 in culture.

Antibiotic concentration (ggimL culture) 2,4-diacetyl- Culture medium phloroglucinol pyoluteorin pyrrolnitrin

Nutrient broth <0.02 0.50 <0.09 Nutrient broth + 2% glycerol <0.02 5.75 2.07 Nutrient broth + 2% glucose 16.60 0.19 <0.09

NOTE: Five mL cultures were incubated at 20°C for 48 h with shaking, acidified with HC1 to pH=2 and extracted twice with 5 mL ethyl acetate. The ethyl acetate fractions were concentrated and analyzed by photo-diode array HPLC in comparison with authentic standards. The values presented are an average of three replicate cultures. 114

The profound effect of carbon source on Phl production by Pf-5 in culture media suggests that the chemical composition of habitats that the bacterium occupies in the spermosphere or rhizosphere will also effect Phl production. Phl has been detected in the rhizosphere of gnotobiotic plants inoculated with P. fluorescens CHAO (Keel et al. 1992), indicating that the rhizosphere is conducive to the production of Phl, at least under certain conditions. However, not all occupied sites or environmental conditions may be conducive to the in situ production of this metabolite at the time or in the concentrations needed to suppress a fungal pathogen. Although the role of Phl in disease suppression by Pf-5 has not yet been investigated, isolates of plant pathogens that are suppressed by Pf-5 are sensitive to Phl in culture. Inhibition of P. ultimum N1 (Kraus and Loper 1992), R. solani J1 (Howell and Stipanovic 1979), and P. tritici-repentis 6R180 (Pfender et al. 1993) were observed at Phl concentrations of 128 ug/mL in potato dextrose agar (Difco), and Erwinia carotovora subsp. carotovora W3C105 and E. carotovora subsp. atroceptica W3C37 (Xu and Gross 1986) were inhibited at 250 ug/mL of Phl in nutrient yeast broth (Stanisich and Holloway 1972). These data are in agreement with previously reported inhibitory activities of Phl (Keel et al. 1992; Levy et al. 1992). Indirect evidence for the in situ production of antibiotics by Pseudomonas spp. and evaluation of these metabolites in disease suppression is usually obtained by comparing the biological control activities of antibiotic-deficient mutants to that of a parental strain. Identification of mutants defective in antibiotic production has commonly been attempted by screening mutants for loss of inhibition in culture. The array of inhibitory metabolites produced by Pf-5, however, necessitates the use of chemical analysis to screen mutants because the presence of an unidentified but biologically-active compound, such as Phl, could mask the existence of a biosynthetic mutant blocked in a single antibiotic pathway. The case-in-point is a study conducted in our laboratory (Kraus and Loper 1992) in which individual culture extracts from several thousand Tn5 insertion mutant cultures were assayed by TLC for pyoluteorin and pyrrolnitrin production. Pyoluteorin-deficient mutants did not inhibit P. ultimum on certain media but were inhibitory against the fungus in other culture media, including NAGlc, which we now know to be conducive to Phl production by Pf-5. The ability of Pf-5 to produce Phl may 115 explain our initial failure to obtain pyoluteorin or pyrrolnitrin non-producing mutants by screening for loss of inhibition of P.ultimum or R. solani. Indeed, the common approach of screening mutants by inhibition assays can not successfully detect biosynthetic mutants if the phenotypes of such mutants are masked by the production of an unsuspected antibiotic(s). 116

Appendix 2. Nucleotide Sequence and Translated Product of pltB

CCCAGTCGAAGGAGATGACTGAATGGATGCTCGTGCGCCCATGGATTTTGAACCCATTGC 1 MDARAPMDFEPIA+ 60

AATCATCGGGAGTGGATGCCGCTTTGCCAAAGGTGCATCGACCCCCGAGGCATTCTGGGA 61 IIGSGCRFAKGASTPEAFWE+ + + + + + 120

GCTCCTTCGTGCTGGGACCGAiiiiGTCGGACCGGTTCCCGCAGAGCGCTGGGACACAGC 121 L LRAGTDFVGPVPAERWDTA+ + + + + + 180

GGCGATCTACGATGAGAGCGCTGCTGAAACCGGTACGACCTACAGCAAGGTGGGGGCGTT 181 AIYDESAAETGTTYSKVGAF+ + + + + + 240

CCTCGAGCACATCGATCGCTTCGATGCGCATTACTTTGGCATCTCGGCCTCTGAAGCCAA 241 L EHIDRFDAHYFGISASEAK+ + + + + + 300

GGAAATGGACCCCCAGCAGCGCCTGCTGCTGGAGGTCGCCTGTGAAAGCGTGGCCCGTGC 301 E MDPQQRLLLEVACESVARA+ + + + + + 360

CGGCCTGACCCGCGAGCAGCTCAAAGGCAGCAGGACCGCGGTCTATGTCGGCATGCTGGG 361 G LTREQLKGSRTAVYVGMLG+ + + + + + 420

CATGGATTACCTGGCACTGCACTCCCGCGAGGCCGGCATCGAACAGATCAACCCCTACTA 421 MDYLALHSREAGIEQINPYY+ + + + + + 480

TGCCGCCGGCAAGGAATTCAGCTTCGCCGCCGGGCGTATTGCCTACCACCTGGGTGTTCA 481 AAGKEFSFAAGRIAYHLGVH+ + + + + + 540

TGGGCCGGCAATGACCGTGACCACTGCATGCTCTTCTTCCCTGGTGGCCATGCACCTGGC 541 G PAMTVTTACSSSLVAMHLA+ + + + + + 600

CTGTCGCGCCTTGCAGGCAGGGGAGGCGGACATGGCCCTGGCCGGAGGGGTGAACCTGAT 601 CRALQAGEADMALAGGVNLM+ + + + + + 660

GCTGGCGCCGGACCTGACGATCTACATGAGCCAGATCAGGGCGATCTCGCCCAGTGGTCG 661 L APDLTIYMSQIRAISPSGR+ + + + + + 720 117

CTGCCGGGTATTCGATGCGGCGGCCGACGGGATCGTCCGGGGCGAAGGCTGCGGGGTCAC 721 + + + + + + 780 C R V F D A A A D G I V R G E G C G V T

GGTGCTCAAGCGCCTGGCAGATGCCCTACGCGATGGCGATCCGATCCAGGCGGTGATCAG 781 + + + + + + 840 L K R L A D A L R D G DP I Q A V I R

GGGCTCGGCGATCAACCAGGATGGCGCCAGTGCCGGCCAGACCGTGCCCAACGCCAATGC 841 + + + + + + 900 G S A INQ DG A S A GQ T V PN AN A

CCAGGCTGCAGTGATCAGCCAGGCCCTGAAAGTGGCCGGCTTGAGCGTGGACGACATCGA 901 + + + + + + 960 Q AAV I S Q A L K V A G LSVDD I D

CTATGTCGAGGCCCACGGCACCGGTACGCCCCTGGGCGATCCGATCGAACTCTCTTCGCT 961 + + + + + + 1020 Y V E A H GT G T P L G D P I IEL S SL

GGACAGTGCCTTCCAGGGGCGTGAGCGGCCGCTGTGGGTGGGCTCGGTCAAGGCCAACAT 1021 + + + + + + 1080 D S A F Q G R E R P L W V G S V K A N M

GGGCCATCTGGATGCGGCCGCCGGAATGGCCAGCGTGATCAAGACCATGATGGTTCTCAA 1081 + + + + + + 1140 G H L D A A A G M A S V I K TMMV L K

GCACGCCGAGGTGCCTGCGCAACTGCATCTTGCACAGTTGAACCCACTGGTGGACTGGAA 1141 + + + + + + 1200 H A E V P A Q L H L A Q L N P LVDWK

GCGTTCCCGGCTGGCGGTGCCCACGGCGATCGAGTCGTTGCCCGACCGGCCGCGCCTGGC 1201 + + + + + + 1260 R S R L A V P T A IESL PDR PR L A

TGGAATCAGTGGTTTCGGGCTCAGTGGCACCAACGTCCACATGATTCTCGAAGATGCCAG 1261 + + + + + + 1320 G I S G F G L S G T N V H M I L E D A S

CGTCTATCGGCAGGCACAGCCGCAGCAGGAACGATCGGCACAGGGCCGGCCCTGGGTCCT 1321 + + + + + + 1380 Y R Q A Q P Q Q E R S A Q G R P W V L

GCCGGTATCGGCCAGGTCCGCGCAGGCCGTGGTGGAGCAGGCCAGGGCCTATGCCGTTCA 1381 + + + + + + 1440 P V S AR S A Q A V V EQ AR A Y A V H

TCTGCCGCAGCAGGACGATGGGCAATTGCAAGCCTTTGTCGCCAGCGCCATTCATCGTCG 1441 L PQQDDGQLQAF+ + + VA+ S A + IHRR + 1500 118

TGATCACTTTCCCTACCGGTCGGCCGTGGTGGGTGCGAACGCCGGGCAATTGAAGAGCCA 1501 + + + + + + 1560 D H F P Y R S A V V G A N A G Q L K S Q

GCTGGAGCAGCTGCCGGCCCCGACCTTGGCCTGCACGACGGACGAAGAGGATCGGCGCGG 1561 L EQL+ PAP+ T L AC+ T TDEEDRR+ + G + 1620

GCCCGTGCTGGTGTTTACCGGGCAGGGCGCCCAGTGGGTGGGAATGGGCCGCGACCTGCT 1621 + + + + + + 1680 P V L V F T G Q G A Q W V G M G R D L L

TGAACGGGAGCCGGCCTTCCTGGCGATGATCCGGCGTTGCGACCAGGCGCTGGCCCAGTG 1681 + + + + + + 1740 E R E P A F L A M I R R C D Q A LAQW

GGCAAGCTGGTCGGTGGAGGCCGAGCTGCGCAGCGACGCCAGTGGGTCCCGCCTGCACCT 1741 + + + + + + 1800 A S W S V E A E L R S D A S G S R L H L

GACCGAATTTGCCCAGCCCTGCATCTTTGCCATCCAGGTGGCGATCAGTGAATGCCTGCG 1801 + + + + + + 1860 T E F A Q P C I F A I Q V A I SECLR

CCAGTGGGGTGTGATCCCGGCGGCGGTGGTCGGCCACTCCATGGGGGAAGTGGCCGCGGC 1861 + + + + +- + 1920 Q W G V I P A A V V G H S M G E V A A A

CTATTGCGCCGGAGCACTGGACCTGGAGTCGGCGGTGCGGGTCATCCACCATCGGGCCCA 1921 + + + + + + 1980 Y C A G A L D L E S A V R V I H H R A Q

GGCCATGAAGGACACCCTGGGGCAGGGGCGGATGCTGGTGGTGGGCTTGCCCGCGCCGAC 1981 + + + + + + 2040 A M K D T L G Q G R M L V V G L P A P T

GCTCCAGTCGCGTCTGGCGAACAACCCGCAACTGGAACTGTCGGTGGTCAACAGTCGAAA 2041 + + + + + + 2100 L Q S R L A N N P Q L E L S V V N S RN

CAGCTGTGTGGTGTCCGGTAGCCCGCAGGCTGTACAGGCTCTGGATCAGCAACTGCGTGA 2101 + + + + + + 2160 S CVVSGS PQAVQA LDQQLR D

CGAAGGCATCTTCACTTACCTGATGCCGGCGGAATATGCCTTCCACTCCTGCCAGATGGA 2161 E G IF+ T YLMPAEY+ + AFHSCQMD+ + + 2220

TGAGTGCCTGACGCAGATCCGCGCGGGGCTGGAGGATTTGCCGGTGGTTGCCGCGCACAC 2221 + + + + + + 2280 E C L R A G L E D L P V V A A H T 119

GCCCTGGATTTCCACCAGCGCGATGCCCGAGGAGCCGATCCTGGCAGATGCCGACTACTG 2281 + + + + + + 2340 P W I S T S A M P E E P I L A D A D Y W

GGCGAGGAATGCCCGTGGCATCGTGCGCTTCGATCGCGCCATCGAACAGTTGATCGAGCA 2341 + 2400 ARNARG IVRF DR A IEQL I E Q

GGGGCACCGGCTGTTCGTCGAGATCGGCCCGCACACCGTGTTGGCGGCGTCGATCAACCA 2401 + + + + + + 2460 G H R L F V E I G P H T V L A A S INQ

GGCCCTGGCCGACAAGGG'AACCCAGGGACTGGTGTGCGGCGCCTTGCACAAGCAGGGCGA 2461 + + + + + + 2520 A L A D K G T Q G L V C G A L H K Q G D

CGCCGCCCTGGAGCTGGCCAGTATTGTTGCCCGCCTGTACGAGTGGGGCGCAGGTCCCGA 2521 + + + + + + 2580 A A L E L A S I V A R L Y E W G A G P D

CTGGCAAGCCTTCCAACCCAGGGAGGCGGCGCTGGAGCTGCCGGCCTATCCCTGGCAGCA 2581 + + + + + + 2640 W Q A F Q P R E A A L E L P A Y P WQQ

GGAGCGTTTCTGGTTTGCCCCGGCGCCGCGGCCGCAGCCGGCAGGGCTCGTGTCCCAGCT 2641 + + + + + + 2700 E R F W F A P A P R P Q P A G L V S Q L

GCGGGCGCAGGTACTGGTGTATGACGCCCAGGGCAATCTCTGCGCCCAGGCCAACGATGT 2701 + + + + + + 2760 R A Q V L V Y D A Q G N L C A Q A N D V

GGCGCTGAGCGTGCCGCAGCTGGCACAAGTCGCTGTGCCGGCACCGGCCAAGGTGTCCGC 2761 + + + + + + 2820 A L S V P Q L A Q V A V P A P A K V S A

AGCGGCGCAGCCGGTGGGGGATGTGCGGGCGCAGATTGGTGCACTGCTGACGCAGATCAT 2821 + + + + + + 2880 A A Q P V G D V R A Q I G A L L T Q I I

Tn5 insertion mutant 4296 CGGCGTGGCCTGTGCC GACCCGGACCCGGATCGAGGCTTCTTCGAACTGGGCCTGAGCTC 2881 +2940 G V A C A D P D P D R G FF E L GL S S

GATTTCCCTGGTTGAATTCAAGCGCATGCTGGAGCGCCAGTTCGCCCTCAAGCTGTCGGC 2941 + + + + + + 3000 S L V E F K R M L E R Q F A L K L S A A 120

GACCGTGGGCTTCGACTATCCCACCATCAACCGGCTGGGCCAGTACCTGGAAGGGCTGCT 3001 + + + + + + 3060 T V G F D Y P T INRLGQYLEGL L

GTCCCGGGAGCCGGCCAGCACCCCGGTAACGGTCGATGCCGGGGCGACCGACGCCGCGGG 3061 + + + + + + 3120 S R E P A S T P V T V D A G A T D A A G

GAGCGTCGCCGTGGTGGCCATGGCGTGCCGGTTTCCCCAGGCCGACAGCCCCGAGGCACT 3121 + + + + + + 3180 S V A V V A M A C R F P Q A D S P E A L

CTGGAAGCTGATGCTGGAACAGACGGACACCGTGGGGCCGGTACCGCCGTCCCGGCTCGC 3181 + + + + + + 3240 W K L M L E Q T D T V G P V P P S R L A

CGGCGCTAAGCCGGAGGAAACCTTCCCGCGGTTTGCCAGCCTTATCCAGCGCCCCGAAGG 3241 + + + + + + 3300 G A K P E E T F P R F A S L I Q R P E G

GTTCGACGAAGCGTTCTTCCGCATTTCCCCCAAGGAAGCCCGGAGCATGGAC.CCCCAGCA 3301 + + + + + + 3360 F D E A F F R I S P K E A R S M DPQQ

ACGGCTGTTGCTGATGGTGGCCTGGGAGGCTCTGGAACGGGCCGGTATCCCTCAGGAGAA 3361 + + + + + + 3420 R L L L M V A W E A L E R A G I PQEK

GCTGCTGGAACAGAGGGTCGGGGTGTTTGTCGGGGCCAACTCCCACGACTACGAAACCCG 3421 + + + + + + 3480 L L E Q R V G V F V G A N S H D Y E T R

GGTCCTGGGTAGCGCGCAAGGCGTGGATGCCCACTACGGCACTGGCAGTTCGTTTTCGGC 3481 + + + + + + 3540 L G S A Q G V D A H Y G T G S SF S A

GATCTGCGGGCGCCTCTCGCA1 rri CTCGGCGTGCGTGGGCCGAGCCTGACGGTGGACAC 3541 + + + + + + 3600 1 C G R L S H F L G V R G P S L T V D T

CGCATGCTCGTCATCGCTCACGGCCATCCATCTGGCGTGCAACAGCTTGCGTGCCGCGGA 3601 + + + + + + 3660 A C S S S L T A I H L A C N S L R A A E

GTGCGACATCGCGATTGTGGGTGGGGTGAATGTCATCGCCTCGGCGTCGATCTTTCAATC 3661 + + + + + + 3720 C D I A I V G G V N V I A S A S IF QS

CATGGGGCAGGCCGGCGCATTGGCCCCGGACGGCATCAGCAAGGCCTTCGACGACAGTGC 3721 + + + + + + 3780 M G Q A G A L A P D G I S K A F D D S A 121

CGACGGTTACGGTCGAGGGGAGGGCTGCGGCGTGGTCATTCTCAAGCGCCAGGCCCAGGC 3781 + + + + + + 3840 D G Y E G C G G R G V V I L K R Q A Q A A

CGAGCGGGAGCGTGACCCGATTGTCGCGACGATTCTCGGTTCGGCGGTCAATCACGACGG 3841 + + + + + + 3900 E R E R D P IVA T IL G S AVNHDG

TGCCTGTGCCGGGCTGACAGTGCCCAATGGTCCGGCGCAAGAGGCGCTGATCAGCGAAGC 3901 + + + + + + 3960 A C A G L T V P N G P A Q E A L I S EA

ACTGGCCAACGCCGGCGTGCATCCGGGGCAAGTCAGCTACGTGGAGGCCCATGGCACCGG 3961 + + + + + + 4020 L A N A G V H P G Q V S Y V E A H G T G

CACGGTGCTGGGTGACCCCATCGAGCTCAACGCCCTGCACAACGCCTATCGCCAGGCCAG 4021 + + + + + + 4080 T V L G D P I E L N A L H N A Y R Q A A S

CCCCGACAGTCCGCCGCTGACGGTGGCCTCGGTCAAGGCCAATATCGGGCACCTCGAGGC 4081 + + + + + + 4140 P D S P P LTVA A S V K A N IGH L E A

GGCAGCGGGCATCGCTTCGCTGATCAAGGCCTGCCTGGTGGTGGAGCACGGCCGGATTGC 4141 + + + + + + 4200 A A G I A S L I K R I IA

CCCGCAAGCCCATCTGCAGCGGGCCAACACCCGTGTCGACTGGGCGGCCATGAACCTCAA 4201 + + + + + + 4260 P Q A H L Q R A N T R V D W A A M N L K

GCTGGCGCATCAGGCCATGGACTGGCCGGGCCGGCCGGAGTCGCGGGTGGCCGGGGTCAG 4261 + + + + + + 4320 L A H Q A M D W P G R P E S R V A G V S

TGCCTTTGGTTTCACCGGGACCAACGTGCATGTGCTGCTCAAGGGCTATACGGCGCCCGC 4321 + + + + + + 4380 A F G F T G T N V H V L L K G Y T A P A

GACAGCGCCTTTGCCACCCGCCACAGCACCAGTGGCCTTGTGCCTGTCCGCGGCCACCCC 4381 + + + + + + 4440 T A P L P P A T A P V A L C L S A A T P

GGCGGCATTGGCGGAGCTGGCCCAGCGCTATGTGTCCTTCCTGGGCGCTACCGAGCACTG 4441 + + + + + + 4500 A A L A E L A Q R Y V S F L G A T E H C

CCCACAGACCATCTGCTACAACGCGCTGATGCGGCGCACGGCG1 i CAAGGAACGCCTGGT 4501 + + + + + + 4560 P Q T I C Y N A L M R R T A F K E R L V 122

CGTCCACGGCCAGGATTGCCGCGAGCTGGCTCAGGCACTACAGGCCTGGCTGGCCGGTAG 4561 + + + + + + 4620 H G Q D C R E L A Q A L Q A W L A G S

CCCCATCGCCAATGACCGCAAACCCGCGGCCGGCGAACCCTGGGCAACCCTGGCCGAGGC 4621 + + + + + + 4680 P I A N D R K PAAGE P W A T L A E A

CTTTGGCCGCGGCGCCCAGTCGCCGGGCCCCGAGCGCTTGCCGGACGGCTGTCAGGCCAT 4681 + + + + + + 4740 F G R G A Q S P G P E R L P D G C Q A I

CGGGTTGCCCACCTATCCCTGGCAGCTCAACGACTACTGGATCGATGCTGGCCAGCCGGC 4741 + + + + + + 4800 G L P T Y PWQLNDYW IDA G Q P A

TACGGCGGTGCAGCCCGCCCGGGCAGCCTCGGGCCATCCTTGCCTGCAGGGGCTGGTACG 4801 + + + + + + 4860 T A V Q P A R A A S G H P C L Q G L V R

GCCGGCCGGGCAGCTCTGGTACTGGTCGGGGGCTCTCGCTCCCCAGGCCGGGCATTACGA 4861 + + + + + + 4920 P A G Q L J W Y W S G A L A P Q A G H Y D

CCCGCTGGGCGAGCAGGGGTACAGGGTCAAGACTCACCTGTTGCTGGATGCCGTGCTCCA 4921 + + + + + + 4980 P G L E Q G Y R V K T H L LL D A V L Q

GGCGGTGCGGGAAACACCACGGGGCGTGCAGCAGATCCGCGACTTGCAGATTGCCCAGCT 4981 + + + + + + 5040 A V R E T PRGVQQIRDLQ I A Q L

GCGCCTGCGCGGCGAACAGCACCTCACTTCGCACCTGAGCATCCACCTCACCCAGGCGCC 5041 + + + + + + 5100 R L R G E Q H L T S H L S I H L T Q A P

TGACGCCTGCTTCGAACTGGCCTTGCAAGGGGCCGGCGACGAGCGCCGCCAGGTCTGCAT 5101 + + + + + + 5160 D A C F L A L Q G A G G D E R R Q V C M

GAGCGGAACCCTGGTTGACTGTGCAGCGCAGTTGCAGGAGGAAACCCTGTGCGGGGTGTC 5161 + + + + + + 5220 S G T L V D C A A Q L Q E E T L C G V S

GATGCTCGATGAGCCTTCACCGCCGGCGCCGGACGTTGGCCTGTGCCCATGGTCCGGGTG 5221 + + + + + + 5280 M L D E P S P P A P D V G L C PWSGC

TGCGGCCACGGGCGGGCAGCGGGCCTTGTACCGTTTTGCCCACTCCCTGACGGCGGCCGA 5281 + + + + + + 5340 A A T G G Q R A L Y R F A H S L T A A E 123

GCGCCAGACCCAGTTGCTGGCCAGTGTGGTCGAGCTGTTTGAAGGCGCCCACGCCGCCGC 5341 + + + + + + 5400 R Q T Q L L A S V V E L F E G A H A A A

GCTGGTCGGGTTCTCCGGGCTCCAGGTATGGGCAGACCTGCCGGCGCAGGTATGGATTGT 5401 + + + + + + 5460 L VGF SGLQVWADL PAQVW I V

GCTGGCCGGCCATGACGCCGACAAACCCGACAGCCTGCAGGTAGTGGATGCCCGGGGTTG 5461 + + + + + + 5520 L D A D K P D LQVVDARGC GC

CCAGCTGGCGCTGTTCGAAGGTCCGCAGTTTGGCCATCCTGGTTCGTGGTACTTGCCCGA 5521 + + + + + + 5580 Q L A L F E G P Q F G H P G S W Y L PD

CCTCCAGACCGCACCCCTCGACCTGCCGATGATCGCTCGCCAATGGCAGGACTATCCGAT 5581 + + + + + + 5640 L QT A PLDL PMIARQWQDY PM

GCCGGGCGAGGGCGCTCGGCAACGCGAGGGCTACTGGATGGTGCTGGCCTGGAGCACTGC 5641 + + + + + + 5700 P GEGARQREGYWMVLAWS T A

CGAGGTCCAGCCACTGGCTGCGGCGTTTGCGGCTGAACAGCGTCCGGTTGAAGTCATCGA 5701 + + + + + + 5760 P E V Q L A A A F A A E Q R P V E V I I E

GCTGCACGCCGGGCAACAACCTCTGGCCCGCAAACTGTCCTCGGCGTTGCGCGGGGCAGT 5761 + + + + + + 5820 L H A G Q Q P L A R K L S S A L R G A V

GGCCGATCCGTCCTGCCTGGGGGTGATCGTTGCCGGCGTTGAGGCTCAGGAGGCCGACGG 5821 + + + + + + 5880 A D P S C L G V I V A G V E A Q E A DG

CCTCGGTATATCCCTGGTGGCCTCTGCGGCGCTGGTACAGGCCTTCGCCGGTGCGATTGC 5881 + + + + + + 5940 L G S L V A S A A L V Q A F A G A A I A

CAGCGTCGGAACACCGGCAAAGCCGGTCTGGTTCGCCCTTCATGCCAGTGATGCTGCGTC 5941 + + + + + + 6000 S V G T P A K P V W F A L H A S DA A S

GCCGGCCATGGCCGCTGTCCAGGCCACCTGGCAGGGCGCCGCGCATATCTTCGCCCTGGA 6001 + + + + + + 6060 P A MA AVQ A T WQG A A H IF A L E

GCATCCTGCCTGGTGGGGCGGTCTGGTGACTTTGCAGGGCAGCGACAGAAGAAGCTACGC 6061 + + + + + + 6120 H P A W W G G L V T L Q G S D R R S Y A 124

CAGCCTCTGCCGGCTGTTGCACGGCCAACCTGGCCATGATCACTTTGCGATCAGCGGCGC 6121 + + + + + + 6180 S LCRLLHGQPGHDHF A I SG A

CCGGGTCGAGGTGCAATACCTGGTGGAGGATCAAGCCGACCCGCTACAGCGTCTGGAGCC 6181 RVEVQYLVEDQADPLQR+ + + + LE+ P+ 6240

GCCGGCGCTGAACGGAACCGTTGTGCTGCATGCCGTCCCGGGATCTGACCTGGAGACTGT 6241 + + + + + + 6300 P A L N G T V V L H A V P G S D L T V

GCTGACGGCGCTCGGGCAGCGGGGTGTGCAGCGTGTGCTGCTGCTCTGCGAGGCCCCCGG 6301 + + + + + + 6360 L T A L G Q R G V Q R V L L L C E A P G

GCAACTGCACATGCCTGAGCGGATGCCCGAAGCGATGGCGATCAGCAGCCTCTCGGACCT 6361 Q LHMPERMPEAMA+ + + IS+ SL+ SDL + 6420

GAGCCGGGAAAACCTGGCTGACACCTTCGCGACGCTGCGAGCGCAAGACCGCATCGCCGG 6421 + + + + + + 6480 L A D T F A T L R AAQDR Q D R I A G

CTTTATCCACCTCGATCTTGACTGGCGCACGGTTGCGCTCAAGGAGCCAGAGTTTGTCGT 6481 + + + + + + 6540 F H L D L D W R T V A L K E P E F V V V

Tn5 insertion mutant 4128 GCGAATGCAGGAAGGCGTTCGGCC GCTTGAGGTCTTGCAGCAGGTTCACCAGCTGATCGA 6541 + + + + + +6600 R M Q E G V R P L E V L Q Q V H Q L ID

TGACCCTGAAGCGTTCTTCCTGATCCTGGGCAGCGTGTCTTCCCTGCTTGGCGGGGCCGG 6601 + 4- + + + + 6660 D PEAFF L ILGSVSSLL G G A G

CTTCGCCCGCTCGGCGATTGCCGATGCATATGCCTTGTGGGTGCATGCACAGCGCCGGCG 6661 + + + + + + 6720 F A R S A I A D A Y A L W V H A Q R R R

CCAGGGGTTGAACTGCCAGCTGCTGCACCTGACGCAGAGCGAACAGGAGCTTGAGCAGGA 6721 Q GLNCQLLHL+ + + TQSEQELEQD+ + + 6780

CGCCGCAGCCCGCACGGCCATGCAAGGCAGCGGCCTGCAGCCTTTGCAGCGCTCGCAGAT 6781 + + + + + + 6840 A A A R T A M Q G S G L Q P L Q R S Q I 125

AGTGCAGGCCATAGCCCGCGTGCTGGGTGGCCAGGGCCAGTGCGGGCTGCTCAATGTCGA 6841 + + + + + + 6900 Q A I A R V L G G Q G Q C G L L N V D

CTGGCAACAACTCAAAGGGCTGTACCTGAGCGTGCTGCCCTGGCCCTTGCTCGAACACCT 6901 W QQLKGL+ + YLSVLPWPLLEHL+ + + + 6960

GGGTGCAGCGGATAGCGCAGCGGATCAGCGTCTGGCCGAGCTGATTGGCCTGCCACCGCT 6961 + + + + + + 7020 G A A D S A A D Q R L A E L I G L PPL

GCAGCAGCGCCGGGCCATGCAGGCGCTGGTCTGCGAGGTAGTGGGGCAGGTGTTCGGCGT 7021 + + + + + + 7080 Q Q R R A M Q A L V C E V V G Q V F G V

TGCCGATGGCCTGGAGCTCGATGTCAGGAAGGGCTTCTTCGACATGGGCATGTCCTCGGT 7081 + + + + + + 7140 A D G L E L D V R K G FF D M G M S S V

CATGTCGCTGGACCTGCGCTCACGGCTGGGCCGGGCGTTGAGCATCGACCTGCCTTCGAC 7141 + + + + + + 7200 MSLDLRSRLGR AL SIDL P S T

CTTCGGCTTCGAATACACCTCCATCGAACAGGTCACGGACTACCTCATGGGCCAGCTCCT 7201 + + + + + + 7260 F G F E Y TSIEQVT TDYLMGQL L

GGCGCCTGAAACCCGGGAGCCGGTGGCGGCGCCCGAACCTGTGTCGCCAGCGTCGCGGCA 7261 + + + + + + 7320 A P E T R E P V A A P E P V S PASRH

TCAGGACCTGCATGAACTGTCACGGGCCGAGCTGATTGGTGCTCTCGAAGACGAGCTGCG 7321 + + + + + + 7380 Q D L H E L R A E EL IGALEDELR

CGATATCGCCAATTACTAGCCCTGGGGCTGGATGACAACTGCCAGAGAATCACTGATCGA 7381 + + + + + + 7440 D I A N Y * 126

Appendix 3. Nucleotide Sequence and Translated Product of pltC

CCrGGGGCTGGATGACAACTGCCAGAGAATCACTGATCGAGGCCTTTATGGATAACGATG 1 + 60 W GWMT T ARES L IEAF MDN D V

TCCGCGATGTAAGCAAGGAACAGCTGCAAGAGAGTCTTGCTCAGGCAATCACTACCATCC 61 + + + + + + 120 R D V S K E Q L Q E S L A Q A I T T I R

GTGCGCTCAAGGAAAAGGTGGCCGGCAAGAGCTCGGCGCCTGTCGAACCGATCGCCGTAG 121 + + + + + + 180 A L K E K V A G K S SA PVEP I A V V

TGGGCCTGGGGTGCCGGCTACCGGGCAGTGCCGACACACCGAAGCGGCTGTGGAGCCTGC 181 + + + + + + 240 G L G C R L P G S A D T P K R L W S L L

TGAAACATGCCACCGATGCGGTGGGCGACATGCCCAGCGACCGTCTGTACGGCACCGACT 241 + + + + + + 300 K H A T D A V G D M P S D R L Y G T D Y

Tn5 insertion mutant 4274 ATTACCATCCTGATCCCCAGGCACCCGGCAAGG CCTACGTCATGCGCGGTGGCTTCATCG 301 + + 360 Y H P D P Q A P G K A Y V M R G G F I E

AGGGGGTGGATCAGTTCGACCCGGGCTTCTTCGGCATTTCGCCCAAGGAAGCCGAAGGCA 361 + + + + + + 420 G V D Q F D P G F F G I S P K E A E G M

TGGACCCCCAGCAGCGCCTGGCCCTGGAGGTTGCCTGGGAGGCCCTGGAGAACGCCGCGA 421 + + + + + + 480 D P Q Q R L A L E V A W E A L E N A A I

TCGCCCCCGACAGCCTGCATGGCAAGAAGCTCGGCGTGTTCATGGGGGTCAGTACCAATG 481 + + + + + + 540 A P D S L HG K K L G V F M G V S TN D

ATTACGTGCGCCTGCGCCAGCAGTTGGGCGCGGTCGAGGACGTCAACGCCTACCAGTTCT 541 + + Y V R L R Q Q L G A+ V E + + + 600 D V N A Y Q F Y Y

ATGGCGAAACCAGCTTCGTGGCCGGGCGCATTGCCTACACCCTGGGCTCCAGGGGCCCGG 601 + + + + + + 660 G E T S F V A G R I A Y T L G S R G P A

CGGTGGTGCTCGACACCTCCTGCTCCTCATCCCTGGTGGCCCTGCACCAGGCCTGCAACA 661 + + + + + + 720 V L D T S C S S S L V A L H Q A C N S 127

GCCTGCGCAGCCGCGAGAGCGAGCTGGCGCTGGCCGGTGGGGTCAACCTGATCCTGTCGC 721 + + + + + + 780 L R S R E S E L A L A G G V N L I L S P

CCTACGGTTTCATCCTGGTCAGCAAGCTGCGGGCCGTGGCCCCCGATGGCCGCTGCAAGA 781 + + + + + + 840 Y G F I L V S K L R A V A PDGRC K T

CCTTCGACGCGGCGGCCGATGGCTACGGGCGCGCCGAAGGCTGCGTGATCCTTGCGCTCA 841 + + + + + + 900 F D A A A D G Y G R A E G C V I L A L K

AGCGGCTGAGCGATGCGGTACGCGACCAGGACCCGGTGCTGGCCGTGATCGAGGGTAGTG 901 + + + + + + 960 R L S D A V R D Q D P V L A V I E G S A

CGGTCAACAACGACGGCGCCAGCAGCGGCATCACCGTGCCCAACATCCACGCCCAGGAAG 961 + + + + + + 1020 N N D G A S S G I T V P N I H A Q E E

AGGTGATCAGGCTGGCGCTCGGCCAGGCCGGGCTCCAGGGCAGCGAGGTCGACTATGTCG 1021 + + + + + + 1080 I R L A L G Q A G L Q G S E V D Y V E

AGGCCCATGGCACCGGCACCGCGCTGGGCGACCCGATCGAACTGCACGCCCTGCATGCGG 1081 + + + + + + 1140 A H G T G T A L G DP I E L H A L H A V

TACTGGGCAAGCAGCGGCCCGTCGATGCACCGCTGCTGGTGGGCTCGATCAAGGCCAACA 1141 + + + + + + 1200 L G K Q R P V D A P L L V G S I K ANM

TGGGGCATCTCGAACCGGTCGCCGGGGTGACCGGGCTGGCCAAGGTCCTGCTGTGCCTGC 1201 + + + + + + 1260 G HL E PV A G V T G L AK V L L C L Q

AACAAGAGGCCCTGGTGCCCCAGGTGCACTTCAACACGCCCAACCCGCGGATCGAATGGG 1261 + + + + + + 1320 Q E A L V P Q V H F N T PNPR I E W D

ATCGCCTGGCCCTGAAGGTGGTCACCGAATCCACGCCCTGGCCACGCCAGGGCAAGGCGC 1321 + + + + + + 1380 R L A L K V V T E S T P W P R Q G K A R

GGCACGCCGGTGTCAGTTCGTTTGGTGTCACCGGCACCAACGCCCATGTGCTAGTGGGCG 1381 + + + + + + 1440 H A G V S S F G V T G T N A H V L V G D

ACGCGCCGCTGCGCGAACGTGCCCAGGGGCGCGACAACCCCTGGCAGCTGATCACTCTGT 1441 + + + + + + 1500 A P L R E R A Q G R D N P W Q L I T L S

CGGCCAAGGGCGAGACGCCCCGGCGCCAGATTGCCGGACGCTATGAACGTTTTATCGCCG 1501 + + + + + + 1560 A K G E T P R R Q I A G R Y E R F I A D 128

ACAACAGCCAGCTCGAACTCAAGGACCTGTGCTACACGGCGAACGTCGGGCGGGCGCACT 1561 + + + + + + 1620 N S Q L E L K D L C Y T A N V G R A H F

TTGGCCATCGTTTCGCGGCCGTGGCCGATAGCCGTGAGGGGCTGCGCGAGCAACTGGCAG 1621 + + + + + + 1680 G H R F A A V A D S R E G L R E Q L A A

CCTATGCGTCACGCAAGGTCGTGGGGCATGTATTCGAAGGGCGCTGCCAGGGAGCGGCGG 1681 + + + + + + 1740 Y A SRK VVGHVF EGRCQG A A A

CGCCGCTGGTGATGCTCTTTCCGGGGCAGGGCTGCCAGTACCGGGCAATGGCCCAGGCAC 1741 + + + + + + 1800 P L V M L F P G Q G C Q Y R A M A Q A L

TGTATGACAGCGAACCATTCTTCAAGGCGCAGATCGATGAATGCCGCGCCCTGTTGCAGC 1801 + + + + + + 1860 Y D S E P F F K IDECRALLQ LQ P

CGCTGATGGACGTGGACCTGCTGACCCTGGTGCTGGACGCGGGTGCGGCCAGTGACAGCT 1861 + + + + + + 1920 L M D V D L L T L V L D A G A A S D S Y

ACCTGCAACAGACCCGCTATGCCCAGCCGGCGATATTCGCGGTCGAATATGCCCTGGCGC 1921 + + + + + + 1980 L Q Q T R Y A Q P A I F A V E Y A L A R

GGTTGTGGATGCATTGGGGGGTCGCTGCCGATGCGCTGTTCGGACACAGTTTCGGCGAGA 1981 + + + + + + 2040 L W M H W G V A A D A L F G H S F G E I

TCAGTGCAATCTGTGTAGCCGGGGCGGTATCCCTGGCTGACGCGCTGCGTATGGTGGAGG 2041 + + + + + + 2100 S A C V A G A V S S L A D A L R M V E A

CGCGTGGGCGCCTGGCCCAGCAACTGATGACGGCCGGCGGCGCGATGTACGCACTGGGCA 2101 + + + + + + 2160 R L A Q Q L M T A G G A M Y A L GMG M

TGAGCGAGGCGCAACTGCTGGAGCTGCTCAAGGACCGGCCCGGCAGCGCGATCGAACTGG 2161 + + + + + + 2220 S E A Q L L E L L K D R P G S A I E L A

CGGCGGTCAACAGCCCGCAGGACGTGGTGGTGGCCGGGCCGCAAGCCGAGGTCCAGGCGC 2221 + + + + + + 2280 A V N S PQDVVV A G PQA E V Q A L

TGGCCGAAGCGGCGCTCGCCAGCGGTTGCAAGGTCAAGAAGCTTGCGGTTTCCCATGCCT 2281 + + + + + + 2340 A E A A L A SGCK VKK L A V S H AF

TCCATACCGCAGCCACCGAGCCGATGCTTGAAGCGTTCCGCCAGACGGTGGCGCAGATCA 2341 + + + + + + 2400 H E P M L T A A T E A F R Q T V A Q Q I T 129

CCTTCAGCGAGCCGCGCTTGCCGGTCATCAGCAGTGTCACTGGCCGGGTGCATACGCTCT 2401 + + + + + + 2460 F SEPRLPVISSVTGRVHT L S

CCAGCCTGAGCTCGCCCGATTACTGGTGTACGCACACCCGCCAGGCCGTGAGGTTCTGCG 2461 + + + + + + 2520 S L SS PDYWC THTRQAVRF CE

AAGGGGTCAACACCCTGATCGCAGAGCTGGGGGTGAAAACCTTCCTTGAAGTGGCCTCCG 2521 + + + + + + 2580 G VNTLIAELGVK T F LEV A S D

ATGCTGTGTTGACGCCGCTGATCGGCCGTCACCCCCTGGCGGACGGCAGCCTGGTCCTGG 2581 + + + + + + 2640 A V L T P L L A ADG SLVL A

CCAGCCTGCGCCGGGCCGGCGATCCGTCCAGGGACCTGCGCCTGGCCGCCGCGCAGCTCT 2641 + + + + + + 2700 S L R R A G D P S RD L R L A A A Q L Y

ACGTGGGCGGCCATAACCTCGACTGGGCCCGGCTGCACGAGCATGACGGGGCCTTGCGCC 2701 + + + + + G G H N L D W A R L H EHDG A LRQ + 2760

AGGCCTTGCCCGGGTATGCGTTTCAGCGCCAGCGCTACTGGTTCGACAACGCCAGCGGCT 2761 + + + + + + 2820 A L P G Y A F Q R Q R Y W F DNA SG S

CGCCCCTGCAACAGGTCGCTGCGGGAGCGGTCGGCCGGCTGCTGGGGCATTC.C4GTCAATG 2821 + + + + + + 2880 P L Q Q V A A G A V G R L L G H S V N A

CACCCACTCCGGCATTCGAGTCGACGCTCGATAGCGCGCTCCTGCAGGTTGTGGGCGGCG 2881 + + + + + + 2940 P T P A F E S T L D S A L L Q V V G G E

AGATCCGCGATGACCTGGCGCTGCTGCGTCCTGACCGGTTGCTGGCAGCCCTGAGCGATG 2941 RDDLALLRP+ + P+ DRL + LA AL+ SDE + 3000

AGCTGGCCGGCCACTTGCAGCTGGACACCTATGGGGTGAGCCTCGCCAGCATCACCCCGG 3001 + + + + + + 3060 L A G H L Q L D T Y G V S L A S I T PA

CCCTGGCGTTCCATGTCGATGACCAGTTGCATCTGTTCACCGAACTCAAGCCGCTGTCCG 3061 + + + + + + 3120 L A F H V D D Q L H L F T E L K P L S G

GGGCGGCCTGGGAGGTCAACTGTTCGGCCCTGAGCGCAGCGGCCAAGGTTGCCGGTGCCG 3121 + + + + + + 3180 A A W E V N C S A L S A A A K V AG AD

ATTGGCAGCCGGTCCTGTCGCTGACCCTGGAAGGCCTGCCGGCGGCTGTCCGGGGCGGCC 3181 + + + + + + 3240 W Q P V L S L T L E G L P A A V R G G L 130

TGCCGGACCCGGGCCATGCCGCCACCCAGGACGCAGGCTTTGTCTACCACATGCAGTTGC 3241 + + + + + + 3300 P DPGHA A TQDAGF VYHMQL P

CGGCAGAACCCGATCCGCACGGCAATCACCTTGGGCAGGTCCTCGAGTACCTGGGGCAGG 3301 + + + + + + 3360 A E P D P H G N H L G Q V L E Y L G Q A

CGGTGCCGGGCGATGCGCCGGGCTCGATAAGCGGCATCCGCCGCTGGGTCGCCAGCCGGT 3361 PGDAPGSISGIRRWV+ + + + A + SR S + 3420

CCGCTTCGCCCCGGGCACAAAGCCTGGTGATTGCCACGAGCCAGGCACATCCGCAGCAGT 3421 + + + + + + 3480 A S P R A Q S L V I A T S Q A H PQQF TCGACTGCGCCTTGTACGATGCGCACGGCCAGTGTGTAGGCAGCCTGGAAGGGCTGACCC 3481 + + + + + + 3540 D C A L Y D A H G Q C V G SLEGL T L

TGGCTGCCGCCCCGAGTGAGGAGGCGCTGCGTGGCATGCTTTACCAGCCCGACGTGCTCT 3541 + + + + + + 3600 A A A P S E E A L R G M L Y Q P D V L Y

ACAGCCTGGATTGGCTCGAACGCCCTCGCCAGCGCGCCGAAGTGCCGCAGCAGGACCGCT 3601 S LDWLER+ + PRQRAEVPQQDRL + + + + 3660

Tn5 insertion mutant 4175 TGTTCACCCTGGTCAGCCGCTCGCCGGAGACCGCCGCTCCCCT GGTCGAACACCTGCAGC 3661 +3720 F T L V S R S P E T A A PL VEHLQR

GTGGCGGGCATCGTGCCCGGGTCCTGTGCCCGCAGCAGTTGCTCGATAGCGCGCGGAGGG 3721 + + + + + + 3780 G G H R A R V L C P Q Q L L D S A R R V

TGCTGGCCCAGGACCCGAGCCAGGCATTGACCGATGAAGGCGAGGCCTTGGCCGCCGATA 3781 + + + + L AQDPSQAL T DEGEAL +A A DI + 3840

TCATCGTGCTCGATGGCAAAGACATCGAGGATGCCGGCTCGACCACGCTGCAGAGCCTGT 3841 IVLDGKDIEDAGS+ + + +T T LQSLS + + 3900

CGAGCCTGCAGGCGACACTGTTCCACCCGTTGCTGGAGATGGTCCAGGCGCTCATCGAAC 3901 S LQA+ TLFHPLLEMVQAL+ + + + IEL + 3960 TGGGCCCGCGCGGTGGCCGGTTGTGGCTGGTGACCCAAGGGGCAAACCCTGTTGGCCTGG 3961 + + + + + + 4020 G P R G G R L W L V T Q G A N P VG L D 131

ATCGCGACCAGCCCCTGCAAGTGGCGACCGGGCCGTTGTGGGGGCTGGGCAAGACGCTCG 4021 + + + + + + 4080 RDQPLQVA TGPLWGLGK T L A CCCTGGAGCACCCCGAACACTGGCGCGGGCTGATCGACCTGGCCCCCGACGATCCTCACT 4081 + + + + + + 4140 L IDLI D AP DDPHW GGGCAAGGGCGCTGGCCGAAGAAGTCAGCGACTCCGATGGCGACGACAAGATCTGCCTGC 4141 + + + + + + 4200 A R A L A E E V S DSDGDDK IC L R

GCCCGGGCAAGCGTTATGTCCAGCGCCTGAACCATTTCAGCGCTGCGCAGTTGCCAGCAC 4201 + + + + + + 4260 P GKRYVQRLNHFSAAQL P A Q AGGCCTATGCACCTTGCCCGCAGGGCAGTTACCTGATTACCGGCGGCATGGGTGGAATTG 4261 A Y +A PC +PQGSYL + I T+ GGMGGIG + + 4320 GCCTGGCCATGGCCCAGTGGCTTCTGGATAAAGGCGCGGGCGCTGTGCTGATCACCGGCC 4321 + + + + + + 4380 L A M A Q W L L D K G A G A V L ITGR

GGCGGCCACTGGAGGATGTCGCCACGGGCCTGGAGCGCTTCGGCGCTGCGGCATCGCGGG 4381 + + + + + + 4440 R P L E D V A T G L E R F G A A A S R V

TGAGTTACGTCCAGGCCGACATCACCTCGCCCCAGGACATGCAGCGGCTGTTCTGCGGTC 4441 + + + + + + 4500 S YVQADI T SPQDMQRLF C G L TGGAGGCGTCGGGCGCGAGCCTGAAGGGCATCTTTCATGCCGCCGGGATTTCAATTCCGC 4501 + + + + + + 4560 E A S G A L K G GIF HAAG IS I PQ

AGGATCTCAAGGACGTCGACCGTGACAGTTTCGACCAGGTGATGCGGCCCAAGGTCGAGG 4561 + + + + + + 4620 D L K D V D R D S F D Q V M R P K V E G

GCACCTGGTTGTTGCACGAACTGTCCCTGGGGCTGGACCTGGACTTCTTTGTCCTGTGTT 4621 TWLLHELSLGLDLDFF+ + + + V + LC S + 4680 CGAGCATCGCCAGTGTCTGGGGTTCGCAGCATGTCGCCAGCTACGCAGCCGCCAATCAGT 4681 + + + + + + 4740 S IA S V W G S Q H V A S Y A A A N Q F

TCCTCGACAGCCTGGCGTGGCAGCGTCGGGCCATGGGCCTGAGTGCCCTGGTGATCGACT 4741 + + L DS L AWQRRAMGL+ SA+ L +V I DW + 4800 GGGGGCTGTGGGCCGGCGGCAGTCACCTGTTCGACGAGCAGGTGCTGAATTTCCTCACCA 4801 + + + + + + 4860 G L W A G G S HL F D E Q V L N F L T S 132

GCGTTGGCTTGAAACAGATCGCCCCGGTACAGAACGTCGGCCTGCTGTCGCGGATTCTGG 4861 + + + + + + 4920 G L K Q I A P V Q N V G L L S R IL A

CCAGCGAGCTGCCCCAGATGGTGGTGTCAGGGGTGGACTGGAATCGCTTCAAGCCACTGC 4921 + 4980 S E L P Q M V V S G V D W N R F K PL L

TCGAATCCCGCGGCCCGCAACCGCTGCTGCAGTACATCCGCAGCCAGGCTCCGACCGCCA 4981 + 5040 E S R G PQPLLQY IR S QA P T A R

GGGCCGGCGACAGCAGCAACGTGGAGATCCTGCAGCAACTGGCGGGCGCCGATGAAGCCG 5041 + 5100 A G DS SNVEILQQL A G A DEA A

CTGCCCTGGCATTGCTGGATGACTACGTGTGGGAGCAGTACGCGCAGTTGCTCGGGGTCA 5101 + + + + + + 5160 A L A L L D D Y V W E Q Y A Q L L G V K

AGACCGAACAGGTGCGGGCCAAGCTCGAGGATGGCGGCAGCCTGATGGACTACGGCCTGG 5161 + 5220 T E Q V R A K L E D G G SLMDY GL D

ACTCGCTGCTGGTGATGGACATGGTCGCCCGCTGCCGGCGCGATCTGAAGCTGGAGATCA 5221 + + + + + + 5280 S L L V M D M V A R C RRDLK L E I K

AGGCCCGCGAGTTTCTTGAGTGTCCGGGCCTGATGTGGCCGGACTTCCTGGCCCGTTCGA 5281 + + + + + + 5340 A REF LECPGLMWPDF L AR S I

TAAAGGAACAGGGCTGCGTGGCCGAGGCCTGACGGGCCGCGGGCAGTGCAA 5341 + + + + +- 5391 K E Q G C V A E A * 133

Appendix 4. Nucleotide Sequence and Translated Product of pltA

AAGACCTGAATGCAATATCCAGGACTGTCGAGCAACTAAAGGCCGAGTGCGCCTAACAGG 1 + 60

GAGTGGGCAATGAGCGATCATGATTATGATGTAGTGATTATCGGTGGCGGGCCGGCGGGT 61 + + + + + + 120 M S D H D Y D V V I I GGG PA G

TCGACCATGGCCTCCTACCTGGCAAAAGCCGGTGTCAAATGCGCGGTGTTCGAAAAAGAA 121 + + + + + + 180 S T M A S Y L A K A G V K C A V F E K E

CTGTTCGAGCGCGAGCATGTTGGCGAGTCGCTGGTACCGGCCACCACTCCGGTGCTGCTG 181 + + + + + + 240 L F E R E H V G E S L V P A T T PVL L

GAAATCGGGGTGATGGAAAAGATCGAGAAAGCCAACTTCCCGAAGAAGTTCGGCGCTGCC 241 + + + + + + 300 E I G V M E K I E K A N F P K K F G A A

TGGACCTCGGCAGATTCCGGCCCCGAAGACAAGATGGGCTTCCAGGGGCTGGACCACGAT 301 + + + + + + 360 W T S A D S G P E D K M G F Q G L D H D

TTCCGTTCGGCGGAAATCCTCTTCAACGAGCGCAAGCAGGAAGGGGTCGATCGCGACTTC 361 + + + + + + 420 F R S A E L F N E R K Q E G V D R D F F

ACGTTCCACGTCGACCGCGGCAAGTTCGACCGCATTCTTCTGGAGCACGCAGGTTCGCTG 421 + + + + + + 480 T F H V D R G K F D R I L L E H A G S L

GGGGCCAAGGTCTTCCAGGGCGTGGAGATCGCTGACGTCGAGTTTCTCAGCCCGGGCAAT 481 + + + + + + 540 G A K V F Q G V E I A D V E F L S P GN

GTCATTGTCAATGCCAAGCTGGGCAAGCGCAGCGTGGAGATCAAGGCCAAGATGGTGGTG 541 + + + + + + 600 I V N A K L G K R S V E I K A K M V V

GATGCCAGCGGTCGCAACGTGCTGCTGGGCCGCCGGCTGGGCTTGCGAGAAAAGGACCCG 601 + + + + + + 660 D A S G R N V L L G R R L L R E K D DP

GTCTTCAACCAGTTCGCGATTCACTCCTGGTTCGACAACTTCGACCGCAAGTCGGCGACG 661 + + + + + + 720 F N Q F A I H S W F D N F D R K S A T 134

CAAAGCCCGGACAAGGTCGACTACATCTTCATTCACTTCCTGCCGATGACCAATACCTGG 721 + + + + + + 780 Q S P D K V DDY IF I H F L PM TNT W

GTCTGGCAGATCCCGATCACCGAAACCATTACCAGCGTGGGCGTGGTTACGCAGAAGCAG 781 + + + + + + 840 W Q I P I T E T I T S V G V V T Q K Q

AACTACACCAACTCCGACCTCACCTATGAAGAGTTCTTCTGGGAAGCGGTGAAGACCCGG 841 + + + + + + 900 N Y T N S D L T Y E E F F W E A V K T R

GAAAACCTGCATGACGCGCTGAAGGCATCGGAGCAGGTCCGCCCGTTCAAGAAAGAGGCG 901 + + + + + + 960 E N L H D A L K A S E Q V R PF K K E A

GACTACAGCTACGGCATGAAAGAAGTCTGTGGCGACAGCTTCGTGCTGATCGGCGATGCC 961 + + + + + + 1020 D Y S Y G M K E V C GDSF V L I G D A

Tn5 insertion mutant 4236 GCACGGTTCGTCGACCCGATCTTCTCCAGCGGCGTCAGCGTTGCACTCAAC AGTGCGCGC 1021 +- + 1080 A R F V D P I F S S G V S V A L N S A R

ATCGCCAGCGGCGACATCATCGAGGCGGTGAAGAACAACGACTTTAGCAAGTCCAGTTTC 1081 + + + + + + 1140 I A S G D I I E A V K N N D F S K S S F

ACTCACTACGAAGGCATGATCAGGAATGGCATCAAGAACTGGTATGAGTTCATCACGCTC 1141 + + + + + + 1200 T HYEGMIRNG IKNWYEF I T L

TATTACCGCCTGAACATCCTCTTCACCGCGTTCGTTCAAGACCCACGCTACCGCCTGGAC 1201 + + + + + + 1260 Y Y R L N L F T A F V Q QDPR Y R L D

ATCCTGCAATTGCTGCAAGGGGACGTCTACAGCGGCAAGCGCCTGGAAGTGCTGGACAAG 1261 + + + + + + 1320 I L Q L L Q G D V Y S G K R L E V L D K

ATGCGCGAAATCATCGCTGCGGTTGAAAGCGACCCGGAACACCTCTGGCACAAGTACCTG 1321 + + + + + + 1380 M R E I IAAVE S DPEHLWHK Y L

GGCGACATGCAGGTTCCTACCGCCAAACCCGCGTTCTAAACACTAAACACCCAGTCGAAG 1381 + + + + DMQVPTAK K P A F* + + 1440 135

Appendix 5. Nucleotide Sequence and Translated Product of pltD

CGCGGGCAGTGCAACCCAGACCCATGATCTGTGATTGAGGTGGTTATGAACGATGTGCAG 1 + + + + + + MN DV Q 60

TCTGGCAAGGCGCCAGAGCATTACGACATTCTCTTGGCGGGCAACAGCATCAGCGTGATC 61 + + + + + + 120 S GK A PEHY DI L L AGNS I S V I

ATGCTCGCCGCCTGCCTGGCCCGGAACAAGGTCCGGGTCGGTTTGTTGCGCAACCGGCAG 121 + + + + + + 180 M L A A C L A R N K V R V G L L RNR Q

ATGCCCCCCGACCTTACCGGTGAGGCGACGATTCCCTATACCTCGATGATTTTCGAGCTG 181 + + + + + + 240 M P P D L T G E A T I P Y T S M IF EL

ATTGCCGACCGCTATGGCGTGCCGGAAATAAAGAATATCGCCCGCACCCGGGATATCCAG 241 + + + + + + 300 I A D R Y G V P E I K N I A R T R D I Q

Tn3-nice insertion mutant 4366 CAGAAGGTGATGCCGTCTTCCGGGGTCAAGAAGAACCTCGGGTTC ATCTATCACCAGCGC 301 + + + + + - -- + 360 Q K V M P S S G V K K N L G F IYHQR

AGCCGGGCGGTGGACCTGGGCCAGGCGCTGCAATTCAACGTGCCCTCCGAGCATGGCGAG 361 + + + + + + 420 S R A V D L G Q A L Q F N V P SEHGE

AACCATCTGTTCAGGCCCGATATCGATGCCTATCTGCTGGCGGCGGCCATCGGTTATGGC 421 + + + + + + 480 N HL F R PD I DA Y L L A A A I G Y G

GCGCAGCTGGTGGAGATCGATAACAGCCCAGAGGTGCTGGTCGAGGACAGCGGGGTCAAG 481 + + + + + + 540 A Q L V E I D N S P E V L V E D S CVK

GTAGCTACGGCACTGGGGCGCTGGGTCACTGCCGATTTCATGGTTGATGGCAGCCAGGGC 541 + + + + + + 600 A T A L G R W V T A D F M V D G SQG

GGCCAGGTGCTGGCGCGGCAGGCTGGCCTGGTCAGCCAGGCTTCGACGCAGAAGACCCGG 601 + + + + + + 660 G Q V L A R Q A G L V S Q A S T Q K T R 136

ACCCTGGAATTCTCCACTCATATGCTCGGGGTGGTGCCGTTCGATGAGTGCGTGCAGGGC 661 + + + + + + 720 T L E F S T H M L G V V P F DEC VQG

GATTTTCCCGGCCAGTGGCATGGCGGCACTCTGCATCACGTGTTCGATGGGGGCTGGGTG 721 + + + + + + 780 D F P GQWHGG T LHHVF DG GW V

GGGGTCATCCCGTTCAACAACCATCAGCACTCGCGCAACCCTTTGGTCAGCGTGCTGGTT 781 + + + + + + 840 G V I P F N N H Q H S R N P L V S V L V

TCACTGCGTGAGGACCTCTGCCCGAGCATGGACGGCGACCAGGTCCTGGCCGGCCTGATC 841 + + + + + + 900 S L R E D L C PSMDGDQVLAGL I

GAGCTGTACCCCGGCCTGGGGCGGCACCTGTCCGGCGCCCGGCGGGTGCGCGAGTGGGTG 901 + + + + + + 960 E L Y P G L G R H L SGARRVREWV

CTGCGCCAGCCGCCCCGGCAGGTCTATCGCACGGCGCTCGAACGCCGCTGCCTGATGTTC 961 + + + + + + 1020 L R Q P P R Q V Y R T A L E R R C LMF

GACGAGGGCGCCGCGAGCAACGATCTGTTGTTCTCGCGCAAGCTGTCCAATGCTGCGGAA 1021 + + + + + + 1080 D E G A A S N D L L F S R K L S N A A E

CTGGTTCTGGCCCTGGCGCACCGGCTGATCAAGGCGGCGCACAGCGGTGACTACCGCAGC 1081 + + + + + + 1140 L V L A L AHRL I K A A H S G D Y R S

CCGGCCCTGAATGATTTTGTCCTGACCCAGGACAGCATCATCAGCTTGAGTGACCGGATC 1141 + + + + + + 1200 P A L N D F V L T Q D S I I S L S DR I

GCCTTAGCGGCTTATGTGTCGTTTCGCGACCCCGAGTTGTGGAATGCCTTCGCCCGTGTC 1201 + + + + + + 1260 A L A A Y V S F R D P E L W N A F A R V

TGGCTGCTGCAGTCGATTGCCGCCACCATCACCGCGCGCAAGATCAACGATGCCTTTGCC 1261 + + + + + + 1320 W L LQS IAA T I T ARK INDAF A

AAGGACCTGGACCCGCGAGTGTTCGATGAAATCGACCAGCTCGCAGAGGACGGTTTCTGG 1321 + + + + K DLDPRVF DE IDQL +A E DGF+ W 1380

ATGCCTCTGTATCGGGGGTACAAGGATATTCTCAACACTACGCTGGGCCTTTGTGATGAC 1381 + + + + + + 1440 M P L Y R G Y K D I L N T T L G L C DD 137

GTCAAAAGCGCCAAGGTCTCTGCTGCGCACGCGGCGAGCAGCATCTTTGCGGAGCTTGCC 1441 + + + + + + 1500 K S A K V S A A H A A S S F A E L A A

AACGCCAGTTTTGTTCCGCCTATTTTTGATTTTGCTAATCCTCACGCTCGTGTCTATCAA 1501 + + + + + + 1560 N A S F V P P I F D F A N P H A R V Y Q

CTGACCACCTTGAGAAAGCTCAAGGCGCTCTGGTGGGGCCTGATGCAAGTGCCCTCAGAG 1561 + + + + + + 1620 L T T L R K L K A L W W G L M Q V P S E

GTCGGACGGCTGATTTTCTATCGATCCTTCAGAAAACCTTCCCTGCGCAAGGAGAGTTGA 1621 + + + + + + 1680 G R L I F Y R S F R K P S L R K E S *

AATGGACTTCAACTACGACGATACCCAGAAAAAACATGCGGCCATGATCGCCCAGGTGTG 1681 + + + + + + 1741 138

Appendix 6. Nucleotide Sequence and Translated Product of orf-S

GCGACCACCTTGCAAGACAAGAGCCAGACATCATGAATCAGTACGACGTCATTATCATCG 1 + 60 MNQYDV I I I G

GTAGTGGTATCGCCGGCGCGCTGACCGGCGCCGTCCTCGCGAAGTCCGGGCTGAACGTTC 61 + + + + + + 120 S G I A G A L T G A V L A K S G L N V L

TGATCCTCGACTCGGCCCAGCACCCACGATTCTCCGTCGGCGAAGCGGCGACACCGGAAA 121 + + + + + + 180 I L D S A Q H P R F S V G E A A T PE S

GCGGTTTTCTGCTGCGTTTGCTCTCAAAGCGCTTCGACATCCCTGAAATCGCCTACCTCT 181 + + + + + + 240 G F L L R L L S K R F D I P E IAYL S

CGCACCCCGACAAGATCATCCAGCACGTCGGTTCGAGCGCCTGCGGGATCAAGCTGGGCT 241 + + + + + + 300 H P D K I I Q H V G S S A C G IK LGF

TCAGTTTTGCCTGGCATCAAGAGAACGCGCCGTCGTCCCCCGACCACCTTGTGGCCCCGC 301 + + + + + + 360 S F A WHQENA PSS PDHL 4 A P P

CGCTGAAGGTGCCGGAAGCCCATC ii ri CCGGCAGGACATCGACTATTTCGCCCTGATGA 361 + + + + + + 420 L K V P E A H L F R Q D I D Y F A L M I

TTGCCCTGAAACACGGCGCCGAATCCAGACAGAACATCAAGATCCGTCGACCTGCAGCCA 421 + + + + + + 480 A L K H G A E S R Q N I K I R R PAAK

AGCTT 481 485 L 139

Appendix 7. Nucleotide Sequence and Translated Product of pltR

AGCCGTACTAAGGAGGCTGTCATGATCTATTTGTAATTTGCCTTTACAAAAATAAGACAC 1 + 60

TAAAAATTCTAAAAGGATTTAGGAATGAAGGCGCTAGGCGTTGTCAATGACATAGACCTT 61 + + + + + + 120 M K A LGVVN D I D L

TATATATCCGTAACTAAAACGGGTAGTTTCTCCGAAACAGGAAGACTACTTGGCATCCCA 121 + + + + + -- + 180 Y I S V T K T G S F S E T G R L L G I P

CCTTCTTCGGTGATGCGCAGAATCAATAGCCTCGAAAAAGAACTTGAAACCTGTCTATTC 181 + + + + + + 240 P S SVMRR IN S LEK E L E T C LF

AATCGATCGACAAAATGCTTGATACTTACCGAAACAGGCCTGCTCTTTCTCGAACATGCA 241 + + + + + + 300 N R S TK C L L T E T G GLLF L EH A

AAGAACATCTCGAGATGCATTGGGCATGCTAGATCAGAAGTAAAAGAACACACTGCATCG 301 + + + + + + 360 K N I S R C I G H A R S E V K E H T A S

ACCCTTGGCGTTCTCAAGGTTTCAGCACCCGTGGCCTTTGGACGGCGCCATGTAGCGCCT 361 + + + + + + 420 T L G V L K V S A P V A F G R R H V A P

TTATTGAGCAGAATACTGAACCATCATCCCGGTTTGAAAATTGAATTCTCGCTTAACGAC 421 + + + + + + 480 L L S R I LNHHPGLK IEF S LND

AAAGCTCTCGACCCCAGCATCGATAACGTCGATATCTGTATCAAGCTGGGCATATTGCCG 481 + + + + + + 540 K A L D P S I D N V D I C I K L G IL P

GACAGTAATCTGATTCCAACGAAACTTGCGGATATGCGGCGCGTGCTCTGCGCAAGCCCG 541 + + + + + + 600 D S N L I P T K L A D M R R V L C A S P

GAATACATCCGGCAACACGGCTGTCCGCAAACTCTTGAAGACCTGTACCAACATGCCTGC 601 + + + + + + 660 E Y I R Q H G C P Q T L E D L Y Q H A C

CTGATCCACAGCACCTGCAGCAACTTCTCACTGACCTGGCAATTCAAGGTCGATGGCCTG 661 + + + + + + 720 L I H S T C S N F S L T W Q F KVDG L 140

CTCAAGAAGCTCATGCCCAGCAGCCGGCTATCGGTCAACAGTTCCGAGTTGCTGGTGGAC 721 + + + + + + 780 L K K L M P S S R L S V N S S E L L V D

GGCGCCCTTCAGGGCATAGGAATCATCCACGCGCCCACCTGGCTGGTGCATGAACAGATC 781 + + + + + + 840 G ALQG I G I I H A P T W L V H EQ I

GCCAGCGGCCAACTGGTGTCGCTGCTCGACGAATATTGCGAGGCCGATCCGCAACAGGGA 841 + + + + + + 900 A S G Q L V S L L D E Y C E A D P QQG

GCGATCTACGCACTGAGGGCACGCAGCAGCGTTGTCCCGGCCAAGACCCGCCTGTTCATC 901 + + + + + + 960 A I Y A L R A R S S V V P A K T R L F I

AATGAGTTGAAACGCTCGATCGGCAGCACGCCTTACTGGGACTTGCCGTTTGAAAAGGAA 961 + + + + + + 1020 N E L K R S I G S T P Y W D L P F E K E

ATACCGCAGACCCTGGCCACCCTGCATTTCGATACCGCAATGCATTCAGCGC'i 11CCAGA 1021 + + + + + + 1080 I P Q T L A T L HF D T A M H S A I J S R

GCGACCACCTTGCAAGACAAGAGCCAGACATCATGAATCAGTACGACGTCATTATCATCG 1081 + + + + + + 1140 A T T L Q D K S QT S *

GTAGTGGTATCGCCGGCGCGCTGACCGGCGCCGTCCTCGCGAAGTCCGGGCTGAACGTTC 1141 + 1200