Information to Users

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note wül indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

UMI University Microfilms international A Bell & Howell Information Company 300 Nortfi Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9130589

Molecular and biochemical studies on thiopeptide antibiotic biosynthesis

Woodman, Robert Harvey, Ph.D.

The Ohio State University, 1991

U-M 300 N. Zeeb RA Ann Arbor, MI 48106

MOLECULAR AND BIOCHEMICAL STUDIES ON THIOPEPTIDE ANTIBIOTIC BIOSYNTHESIS

A DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By Robert Harvey Woodman, B.S.

********

The Ohio State University

1991

Dissertation Committee: Approved by Heinz G. Floss Edward J. Behrman rXirvo ^ Adviser George M. Milo The Ohio State Biochemistry Program Richard T . Sayre William R. Strohl Copyright by Robert Harvey Woodman 1 9 9 1 To Dr. Stella Elakovich, Dr. William R. Eure, Mr. Blanchard Hinton, Mr. Dean Hubbard, Mr. and Mrs. Thomas H. Smith III, and my parents, Robert C. and Carolyn B. Woodman for all they have done through the years.

11 ACKNOWLEDGEMENTS

I thank Drs. Heinz G. Floss and William R. Strohl — my official and unofficial advisors, respectively — for allowing me to work on, and learn from, this project. I especially thank Dr. Strohl for his generous allowances of time, lab space, guidance, and advice in the course of my research.

I express thanks to all the members of the Strohl lab, past and present, for their advice, help, humor, and tolerance. I especially thank Tracey Miller and Li Yun for their advice and assistance and Don Dosch for his friendship and several good pieces of advice. I also thank Mark and

Cheryl Thompson for their friendship and gracious hospitality in the last days of this work; their help cannot be too highly praised. Thanks go also to all of the members of the Reeve lab, especially Diane Stroup and Vanessa Steigerwald, for friendship and splendid advice during my research.

Finally, I thank all of the people to whom this work is dedicated. Each of these, in some significant way, has influenced my life and my career choices. (I am especially appreciative of my mother's frequent reminder "This, too, shall pass.") To them, I offer my heartfelt thanks for their friendship, help, and encouragement in my life.

Ill VITA

.March 8, 1962...... Born - Smithville, Missouri

August 8, 1985...... B. S., cum laude, University of Southern Mississippi, Hattiesburg, Mississippi

Sept. 1985 - Dec. 1986...... Graduate Teaching Associate The Ohio State Biochemistry Program The Ohio State University Columbus, Ohio

Jan. 1987 - present...... Graduate Research Associate Department of Microbiology The Ohio State University Columbus, Ohio

PUBLICATION

Woodman, R.H. 1985. Study of a precipitate from Brasenia schreberi. Bachelor of Science Thesis. Hattiesburg, Mississippi: Honors College, University of Southern Mississippi.

PUBLISHED ABSTRACTS

1. Li, Y., R.H. Woodman, D.C. Dosch, W.R. Strohl, and H.G. Floss. 1989. Nucleotide sequence of a nosiheptide resistance gene from Streptomyces actuosus. Abstr. Ann. Meet. Am. Soc. Microbiol. 014, p. 306.

2. Li, Y., D. Dosch, R. Woodman, and W.R. Strohl. 1990. Regulation of nosiheptide production and resistance in Streptomyces actuosus. Ann. Meet. Soc. Industr. Microbiol. Abstract SlOO, p. 82.

IV FIELDS OF STUDY

Major field of study: The Ohio State Biochemistry Program Studies in the biochemistry of microorganisms. TABLE OF CONTENTS

ACKNOWLEDGEMENTS...... i ü

VITA...... iv

LIST OF TABLES...... ix

LIST OF FIGURES...... xi

LIST OF ABBREVIATIONS...... xiv

CHAPTER I: INTRODUCTION...... 1

CHAPTER II: REVIEW OF THE LITERATURE...... 7

Streptomyces and Secondary Metabolites...... 7 Antibiotic Synthesis Regulation: Introduction.... 8 Antibiotic Synthesis Regulation: Classes of Regulatory Genes...... 9 Antibiotic Synthesis Regulation: Pleiotropic Regulators...... 12 Antibiotic Synthesis Regulation: Pathway- Specific Regulators...... 16 Antibiotic Synthesis Regulation; Antibiotic Resistance...... 18 Peptide Antibiotics: General...... 21 Thiopeptide Antibiotics: Thiostrepton and Nosiheptide...... 26 Molecular Biology of Thiostrepton and Nosiheptide Biosynthesis...... 33

CHAPTER III: SPECIFIC AIMS OF THE RESEARCH...... 36

CHAPTER IV: METHODOLOGY...... ,...... 39

Chemicals and Biochemicals...... 39 Media Used in This Work...... 41 Bacterial Strains and Plasmids...... 46 Antibiotics Usage...... 46 Culturing of Organisms...... 55 Preparation of Streptomyces Spores...... 58 N-methyl-N'-nitro-N-nitrosoguanidine Mutagenesis of Streptomyces Spores...... 58

vi Nitrous Acid Mutagenesis of Streptomyces Spores... 59 Isolation of Mutant Strains of Streptomyces...... 60 Protoplast Fusion of Mutant Organisms...... 62 Preparation of Chromosomal DNA...... 62 Plasmid DNA Isolation Methods...... 65 Cesium Chloride Purification of Plasmids...... 70 Preparation of E. coli Competent Cells...... 71 Preparation of Protoplasts of Streptomyces Strains for Cloning...... 71 Transformation of E. coli with Plasmids...... 73 Transformation of Streptomyces Protoplasts...... 74 Restriction Endonuclease Digestions...... 74 Construction of Standards for Autoradiography 75 DNA Probes for Hybridization...... 75 Isolation of Specific DNA Fragments...... 78 DNA Ligations with Plasmids...... 78 Radioactive Labelling of DNA...... 79 Southern Hybridization...... 81 Hybridization of DNA Probes to Colonies of Recombinant E . c o h ...... 82 Autoradiography and Scintillation Counting...... 84 Fermentation of S. azur eus...... 84 Isolation of High-Purity Thiostrepton...... 85 Isolation of High-Purity Nosiheptide...... 86 Quick Extractions of Thiostrepton and Nosiheptide. 88 Reduction of the Specific Activity of ^®S-L-cysteine...... 89 Protocol for the First Synthesis Study of Thiopeptide Antibiotics...... 89 Protocol for the Second Synthesis Study of Thiopeptide Antibiotics...... 91

CHAPTER V: RESULTS AND DISCUSSION...... 96

Mutants of Thiopeptide Producers...... 96 Southern Hybridization Analysis of Thiopeptide Producers and Thiostrepton Non-producing Mutants...... 99 Transformation of S. azureus Mutants...... 133 Protoplast Fusion of Two Thiopeptide Antibiotic Non-producing Mutants and the Resulting Restoration of Production...... 134 Discovery of Plasmid-Borne Thiostrepton Resistance in Streptomyces azureus...... 136 Cloning of the nsM-Equivalent Gene from S. azureus...... 181 Analysis of the Function of NshA...... 184 Time-course Fermentation: Analysis of S. azureus DNA...... 194 Study of the Mode of Biosynthesis of Thiostrepton and Nosiheptide by Chloramphenicol Inhibition of Antibiotic Production...... 197

vii CHAPTER VI: SUMMARY AND CONCLUSIONS...... 227

LITERATURE CITED...... 231

vixi LIST OF TABLES

Table 1: Predicted Number of Genes for the Biosynthesis of Thiostrepton and Nosiheptide...... 32

Table 2: Bacterial Strains Used in This Research..... 47

Table 3: Mutant Strains of Thiopeptide Producers Used m This S t u d y 51

Table 4: Plasmids Used in This Project...... 53

Table 5: Autoradiography Standards for DNA Hybrid ization to Thiostrepton Resistance...... 76

Table 6: Autoradiography Standards for nshR, nshA, and the 8.5 kbp NshR DNA Cloned from S. actuosus...... 77

Table 7: Analysis of S. azureus Mutational Frequency Using MNNG as the Mutagen...... 98

Table 8: Sequences of the Oligonucleotide Probes Used in this Study...... 104

Table 9: Streptomyces Strains Hybridizing to Probe 1 (nshA)...... 109

Table 10: Size of tsr- and nsJiR-hybridizing DNA Fragments in Thiopeptide Producers...... 113

Table 11: Size of BamHI-digested DNA Fragments from Thiopeptide Antibiotic Producers Hybrid izing to Oligonucleotide Probe 4PAS1...... 125

Table 12: DNA Fragments Hybridizing to Probe 4 from Selected Saz Mutants, Using Various Restriction Endonuclease Digestions...... 128

Table 13: DNA Fragments Hybridizing to Probe 1 from Selected Saz Mutants, Using Various Restriction Endonuclease Digestions...... 129

XX Table 14; Summary of Experimental Approaches Used to Clone Thiopeptide Antibiotic Resistance Genes, Thiostrepton Biosynthesis Genes, or Associated (nshfl-like) DNA...... 137

Table 15: Dry Weights of Cells and Net Weights of Antibiotic Recovered...... 204

Table 16: Radioactivity Incorporated into TCA- insoluble Material...... 205

Table 17: Amount of Radioactivity in Antibiotics..... 206

X LIST OF FIGURES

Figure 1: The life cycle of Streptomyces spp...... 10

Figure 2: The structure of the peptide antibiotic gramicidin S ...... 23

Figure 3: The structure of the thiopeptide anti biotics thiostrepton (A) and nosiheptide (B)...... 27

Figure 4: Restriction map of the 8.5 kbp BamHI- digested DNA fragment which confers resistance to nosiheptide, and a closer view of the sequenced region (from the left-hand BamEI to the PvuII site im mediately to the right of orfC)...... 101

Figure 5: Autoradiogram of chromosomal DNA from various strains of Streptomyces which hybridize to probe 1 (nshA)...... 107

Figure 6: Autoradiograms generated by Southern hy- ization of S. azureus DNA to four DNA probes which hybridize to DNA putatively associated with thiopeptide antibiotic biosynthesis...... 115

Figure 7: Restriction maps of S. azureus DNA surrounding the gene encoding the nshA- like protein (A) and the resistance genes of thiostrepton (B)...... 119

Figure 8; Autoradiogram of Streptomyces chromosomal DNA hybridizing to the oligonucleotide probe 4PAS1, a probe to the consensus se quence of the 4'-phosphopantetheine binding site in Streptomyces [ 154 ]...... 123

Figure 9: Hybridization of BamHI-digested DNA from S. azureus and Saz mutants to the 8.5 kbp DNA probe (probe 4) cloned from S. actuosus, encoding thiopeptide antibiotic resistance

XI DNA, and putative thiopeptide antibiotic regulatory and biosynthesis DNA...... 131

Figure 10: The plJlOl-specific probes used in this study...... 141

Figure il; Typical results from plasmid preparations of S. lividans protoplasts into which DNA from S. azureus cloned into various plasmids has been introduced by transformation...... 146

Figure 12: Hybridization of restriction endonuclease- digested DNA fragments from pIJ702 and two other plasmids to a 2,180 bp fragment of DNA internal to the essential DNA of plasmid PIJ702...... 152

Figure 13: Photograph of the DNA of a plasmid prepara tion of S. azureus DNA which had been cloned into plasmid pKC505...... 157

Figure 14: The 6.1 kbp tsrP-containing DNA cluster cloned from S. azureus into pKC505...... 159

Figure 15: Autoradiogram showing hybridization of the nshR DNA probe (derived from pANT400) to DNA isolated from several strains of Strepto- myc^s spp•••••••••••••••••••••••••••••••••••• 164

Figure 16: Autoradiogram showing hybridization of the Sstll "minimal replicon" DNA probe to DNA isolated from several strains of Strepto- siyc&s spp•••••••••••••••••••••••••••••••••••• 166

Figure 17: Autoradiogram showing hybridization of the pANT425-derived probe (containing the Pstl- BamHI DNA fragment of the pIJlOl replication region) to DNA isolated from various strains of Streptomyces spp...... 168

Figure 18: Restriction endonuclease map from Saz-l/ATCC of the chromosomal region hybridizing to tsrR and nshR...... 172

Figure 19: Restriction endonuclease map of pANT423, the plasmid that appears to correspond to the 6.1 kbp BamHI fragment of DNA in the S. azureus chromosome to which tsrR and nshR hybridize...... 174

Figure 20: Restriction endonuclease map of pANT424, the plasmid that appears to correspond to the

Xll 4.5 kbp BamHI fragment of DNA in the S. azureus chromosome to which tsrR and nshR hybridize...... 176

Figure 21: Restriction map of pANT408, the initial clone containing DNA which hybridized to probe 1 (nshA)...... 182

Figure 22: Restriction map of pANT409, the subcloned construct of pANT408 containing only the 2.8 kbp Pstl-Bglll DNA fragment from S. azureus which hybridizes to probe 1 (nshA)...... 185

Figure 23: Restriction map of pANT407 used for the site-specific mutagenesis, via plasmid replacement, of nshA in S. actuosus...... 187

Figure 24: Autoradiogram of Southern hybridization analysis of 10 strains of S. actuosus into which pANT407 had been introduced by trans formation and subsequently lost...... 191

Figure 25: The dry weights of fermentation of S. azureus (log scale)...... 195

Figure 26: Hybridization of S. azureus DNA digested with BamHI and SstI to probe 2 (nshR) at different time points in fermentation of the organism...... 198

Figure 27: Total protein analysis as an approximation of the growth curve of S. laurentii, S. actuosus and S. azureus...... 209

Figure 28: Radioactivity, as cpm, in the TCA-precipitable material collected on glass fiber filters...... 212

Figure 29: Total cpm of the ==S-L-cysteine incorporated into nosiheptide and thiostrepton extracted from 20 mL aliquots of cultures in medium DPM3 at specific time points after the addition of radiolabelled cysteine...... 214

Figure 30: Total Incorporated into thiostrepton and nosiheptide in the presence and absence of chloramphenicol...... 217

Figure 31: Thin-layer chromatography of S. actuosus antibiotic extracts over time...... 220

X l l l Figure 32: Thin-layer chromatography of S. azureus antibiotic extracts over time...... 222

Figure 33: Thin-layer chromatography of 3. laurentii antibiotic extracts over time...... 224

XIV LIST OF ABBREVIATIONS

General Abbreviations

ATCC American Type Culture Collection, Rockville, MD ATP adenosine triphosphate bp base pairs of DNA cpm counts per minute DNA 2'-deoxyribonucleic acid dCTP 2'-deoxyribose cytosine triphosphate dpm destructions per minute EMS ethylmethane sulfonate hr hour(s) kbp Kilobase pairs of DNA kd kilodaltons M molar liCi microcurie Hg microgram /XL microliter mCi millicurie mg milligram ml milliliter mM millimolar min minute(s) nCi nanocuries NRRL Northern Regional Research Laboratories, Peoria, IL MNNG N-methyl-N'-nitro-N-nitrosoguanidine RNA ribonucleic acid sec second(s) UV ultraviolet

Buffers and Related Items

BME B-Mercaptoethanol DTT Dithiothreitol (Cleland's Reagent) EDTA ethylenediamine tetraacetate HEPES N-2-Hydroxyethylpiperazine-N-2'-ethane- sulfonic acid IPTG isopropylthio-B-galactoside MOPS 3-(N-Morpholino)propanesulfonic acid SDS Sodium Dodecyl Sulfate SSC IX = 0.015 M Na^citrate, 0.15 M NaCl, pH 7.0

XV SSPE IX = 10 mM Na^HPOa, 150 mM NaCl, 2 mM EDTA, pH 7.4 STE IX = 10 mM Tris, 1 mM EDTA, 100 mM NaCl, pH 8.0 TE Tris-EDTA IX = 10 mM Tris, 1 mM EDTA, pH 8.0 TBE Tris-Borate EDTA IX = 0.089 M Tris 0.089 M Borate, 2 mM EDTA TES N-tris[hydroxymethyl]-2~aminoethanesulfonic acid TPE Tris-Phosphate EDTA IX = 0.08 M Tris-Phosphate, 2 mM EDTA Tris Tris(hydroxymethyl)aminomethane X-gal 5-bromo-4-chloro-3-indolyl-6-D-galactoside

Other, less frequently used abbreviations, will be defined where and when they appear in the text.

XVI CHAPTER I

INTRODUCTION

The original intent of this research was to clone genes associated with the biosynthesis, regulation, or resistance mechanisms involved with production of the modified thiopeptide antibiotics, thiostrepton and nosiheptide. The

initial strategy was to isolate mutants of Streptomyces azureus and Streptomyces actuosus (producers of thiostrepton and nosiheptide, respectively), blocked in the formation of their respective antibiotics, which could be used as recipients for the cloned DNA. By "mixing and matching" genes encoding reactions of the thiostrepton (Tsr) and nosiheptide

(Nsh) pathways, the long-term goal of the overall project in our laboratory is to produce novel, "hybrid," thiopeptide structures. Keys to this original strategy were the facts that: (i) a cluster of genes encoding the entire actinorhodin biosynthetic pathway had been cloned from Streptomyces coelicolor and expressed in a heterologous host [116]; (ii)

"hybrid" antibiotics of the benzoisochromanequinone family had been generated by interspecies cloning of the actinorhodin pathway genes into strains of Streptomyces producing different benzoisochromanequinones [reviewed in 24,27,51]; and (iii) 2 biosynthetic genes for antibiotics produced by Streptomyces have, in all 24 examples known to date, occurred in single DNA clusters with their respective antibiotic resistance genes on the chromosome [24,27], Thus, the initial approach used in attempts to clone Tsr and Nsh genes was to isolate the gene or genes conferring resistance to the thiopeptide antibiotics along with contiguous DNA. This DNA, which should theoretically contain antibiotic biosynthesis genes, would then be used in the analysis of the pathways and to form hybrid thiopeptide metabolites. This tactic was considered reasonable since, as stated, every previous example has shown that genes encoding proteins used in the biosynthesis of antibiotics are clustered around genes encoding proteins which provide resistance to the antibiotic produced.

When this project began, however, very little was known about the biochemistry or genetics of the thiopeptide producers, S. azureus and S. actuosus. Therefore, I began these studies with initial experiments to determine the factors and characteristics of these strains which were relevant to the study of genes encoding proteins involved in thiopeptide antibiotic biosynthesis. The factors and characteristics included; (i) determination of the transformation frequency of these strains; (ii) generation of blocked mutants of these strains; (iii) analysis of these mutants; and (iv) cloning of the Tsr and Nsh resistance genes, along with contiguous DNA. 3

During the middle stages of this project, Dr. Don Dosch, a post-doctoral associate in our laboratory, cloned an 8.5 kbp

DNA fragment which contained the Nsh resistance (nshR) gene, plus other genes possibly related to nosiheptide biosynthesis,

from S. actuosus [45,109]. The resistance gene was subsequently found to be the second of a two-gene operon in S. actuosus [109].

At this point, I was able to use this 8.5 kbp DNA

fragment, and subclones derived from it, to probe the S. azureus mutants blocked in the biosynthesis of thiostrepton to determine if any noticeable changes had occurred in the DNA

fragments of the mutants which hybridized to the cloned S. actuosus DNA, as compared to the DNA fragments which hybridized in the wild-type S. azureus to the cloned DNA from

S. actuosus. These DNA probes were also used to generate restriction endonuclease maps of DNA in S. azureus which (i) hybridized to the genes encoding resistance to thiopeptide antibiotics and (ii) hybridized to the gene found upstream in the resistance operon in S. actuosus. The purpose of these restriction maps was to facilitate cloning of the DNA

surrounding both the resistance gene of S. azureus and the gene in S. azureus which was homologous to the upstream gene

in the Nsh resistance gene operon of S. actuosus.

The project was altered from its original course, however, due to the emergence of difficulties in cloning DNA

from S. azureus, and of the inherent interest generated by the 4 discovery of the nshA-nshR operon in S. actuosus. The upstream gene in the nosiheptide resistance gene operon, and the protein product which it putatively encoded, have been partially analyzed and found to contain features suggestive of a regulatory gene (see Chapter II, pages 33, 34). A fragment of this upstream gene, nshA, was used as a probe against

Southern blots of chromosomal DNA isolated from various species of Streptomyces. The gene was found to be both widespread and highly conserved among Streptomyces spp. The gene encoding the protein for nosiheptide resistance, however, appeared only among thiopeptide producers when similarly analyzed. These results made it appear important to accomplish the following two goals: (i) interrupt the S. actuosus nshA gene and attempt to determine what effect this mutation has on the production of nosiheptide; and (ii) clone the S. azureus homolog of nshA.

In numerous attempts to clone large DNA fragments containing the thiostrepton resistance gene from S. azureus, a series of highly unusual, but very reproducible results was obtained which interfered with the cloning attempts. In about

50 separate experiments using 13 different experimental approaches (see Table 14, Chapter V), including experiments using large vectors (e.g. > 18 kbp) for cloning, the resultant plasmids were nearly always either ca. 5 kbp or ca. 6.7 kbp in size and contained DNA encoding Tsr resistance. Therefore, a major goal of this project became the elucidation of the 5 origin of these unusual plasmids which should assist in the development of future cloning strategies using these strains.

The results of these studies indicate a major new finding that plJlOl-like replicon DNA is integrated into the S. azureus chromosome associated with the resident thiostrepton resistance genes.

Two final goals of this project were: (i) to study the general stability of the DNA putatively associated with the biosynthesis of thiostrepton in S. azureus; and (ii) to determine the general mechanism by which the precursor molecules of thiostrepton and nosiheptide are formed.

Over the past ten years, several researchers have found that DNA in streptomycetes can be highly unstable. Unstable amplifications and deletions occur in many species of

Streptomyces, particularly in and around regions of DNA which contain genes encoding enzymes involved in the biosynthesis of antibiotics [reviewed in 26,27,64]. The approach adopted for the stability study was to conduct a time-course fermentation of S. azureus. Time-dependent DNA samples were probed with a fragment of the gene encoding nosiheptide resistance to discern any rearrangements of the DNA surrounding the fragment of DNA to which the probe hybridized.

To determine the mechanism by which the precursor molecules of thiostrepton and nosiheptide were formed, four different approaches were taken. First, the DNA from organisms producing thiostrepton and nosiheptide was analyzed 6 for the presence of DNA sequences which could hybridize to oligonucleotide probes containing the sequence of DNA which could encode the putative precursor peptides of the thiopeptide antibiotics. Second, the DNA from organisms producing thiopeptide antibiotics was analyzed for the presence of DNA sequences which could hybridize to an oligonucleotide probe containing the sequence of DNA encoding the conserved amino acid sequence of the binding site for

4 '-phosphopantetheine. This binding site is not only used in fatty acid synthetase, but if a peptide synthetase were used in the synthesis of the thiopeptide antibiotic precursors, this binding site would likely be part of the peptide synthetase (see Chapter II, pages 21 to 26). Third, the synthesis of thiostrepton and nosiheptide was measured in the presence and absence of chloramphenicol using an "endpoint analysis" experiment. Finally, the synthesis of thiostrepton and nosiheptide was measured in the presence and absence of chloramphenicol using a time-course procedure in which samples of cells were removed from the cultures of S. azureus, S. actuosus, and S. laurentii at defined intervals and analyzed. CHAPTER II: REVIEW OF THE LITERATURE

Streptomyces and Secondary Metabolites

Streptomyces is a genus containing mycelial bacteria whose members produce a wide variety of compounds classified as "secondary metabolites." These secondary metabolites, however, are not secondary in their importance: among these compounds are almost sixty percent (60%) of the 5,500 different antibiotics now known [51,119]. Streptomyces coelicolor A3(2), the most widely studied species of

Streptomyces, is known to produce at least four antibiotics

[27]. The purpose of this vast number of secondary metabolites is not completely clear; however, circumstantial evidence suggests that, while not essential to the differentiation process of the Streptomyces life cycle

(discussed later), antibiotics have value in providing protective and regulatory functions during the differentiation phase of the life cycle [27].

Secondary metabolites, including antibiotics, may be defined as follows [43,55,115,158,172]:

(a) Secondary metabolites are required neither for growth of

the producing organism nor for its survival; they are 8

dispensable with respect to the producing organism, at

least in the laboratory.

(b) Secondary metabolites are normally produced only during

the stationary phase of the growth cycle, and production

is usually suppressed by growth-inducing substrates.

complex and multibranched, often including both

converging and diverging reaction paths.

(d) The compounds classified as secondary metabolites

typically have complex, unusual structures not found in

compounds required for growth.

(e) Enzymes catalyzing secondary metabolic reactions

frequently display a noticeably reduced chemical

specificity for substrates compared to enzymes of the

primary metabolic pathways.

Antibiotic Synthesis Regulation; Introduction

The intricacy of antibiotic synthesis regulation in the streptomycetes is largely a reflection of the general complexity of this family of bacteria. The genus Streptomyces

(and indeed, virtually all the Actinomycetales) is composed of a large number of species which share a complex and highly regulated life cycle. These Gram-positive bacteria begin as single spores. The spores germinate and begin producing mycelia which push downward into the substrate upon which they reside. Later in the life cycle, these "substrate mycelia" 9 undergo lysis and concomitant with this lysis, mycelia which grow vertically from the substrate surface begin to form; these are called "aerial mycelia." Both substrate and aerial mycelia are septated, multinuclear, multibranched structures.

The last phase of this intricate life cycle involves the

formation, from the aerial mycelia, of unicellular spores which are dispersed to new locations, thereby renewing the life cycle [23,28; Figure 1]. Thus, cells in streptomycete colonies exhibit both temporal and spatial regulation in their life cycles [23,28,151]. This fact has been cleverly exploited by some researchers to further the understanding of developmental regulation in Streptomyces [151].

Illustrations of the complexity of this life cycle have been observed both in the patterns of the RNA and ribosomal proteins of Streptomyces granaticolor during the process of spore germination [122] and in the number of genes involved in the formation of spores of S. coelicolor A3(2). In the latter case, there are no fewer than eight genes involved in the production of spores from aerial mycelia, the primary regulator of which is the gene whiG [21,121].

Antibiotic Synthesis Regulation; Classes of Regulatory Genes

Regulation of antibiotic formation may be placed into two classes. The first class of regulatory genes includes those which have "pleiotropic" effects. These genes display wide- ranging regulatory influence across multiple Figure 1: The Life Cycle of Streptomyces spp.

10 Free Spore

Aei;ial Coiling Sporulation M y c e liu m Substrate Mycelium

Figure 1

t-* 12 developmental and biosynthetic pathways. The second class of regulatory genes includes "pathway-specific" genes, i.e., genes which are clustered with the biosynthetic genes of a particular antibiotic and exert their regulatory effects only on that biosynthetic pathway.

Antibiotic Synthesis Regulation: Pleiotropic Regulators

The archetypical pleiotropic regulatory system for antibiotic biosynthesis in Streptomyces is found in the organism Streptomyces griseus, the producer of streptomycin.

This organism produces a compound called A-factor

(2-S-isocapryloyl-3-S-hydroxy-methyl-Y-butyrolactone) which putatively acts like a pheromone. In A-factor-deficient mutants, neither sporulation nor antibiotic production occur normally; however, the addition of small amounts of A-factor to deficient cells will restore both normal sporulation and normal antibiotic production [22].

Horinouchi and Beppu [71,72] have shown that when

A-factor is added to the culture medium of an A-factor- deficient mutant, that mutant exhibits a protein profile in which ten protein bands, always observed in the wild type strain but never in the deficient mutant, appear in the deficient mutant with A-factor added. Removal of the A-factor again causes the disappearance of the ten protein bands. It has also been shown that A-factor in S. griseus stimulates the transcription of the streptomycin resistance gene [60]. 13

Many other species of Streptomyces also produce substances like A-factor, but the role of the compounds in the regulation of physical and/or chemical differentiation is still unclear. For example, in S. coelicolor A3(2), the role of the A-factor-like compound is apparently not as critical for sporulation and antibiotic production as it is in S. griseus [22].

A gene, called afsB, has been cloned from S. coelicolor and sequenced. This gene has been found to control both the biosynthesis of A-factor and the production of the pigmented antibiotics actinorhodin and undecylprodigiosin. The protein product of the gene is 243 amino acid residues long, and the gene itself is constitutively expressed. Furthermore, the gene for A-factor regulation appears to be widely distributed among species of Streptomyces, as determined from Southern blot hybridization [77]. In addition to afsB, afsA, the structural gene which is primary for the synthesis of

A-factor, has been cloned from Streptomyces bikiniensis and S. griseus and sequenced [75,76,78]. Furthermore, a gene which enhances the effects of afsB, called afsC, has been found in

S. coelicolor A3(2) [71].

In addition to pleiotropic mutants of S. griseus which require exogenous A-factor, an A-factor-independent pleiotropic mutation has been found in that organism affecting both sporulation and streptomycin production. This mutation was found to be in the protein which bound A-factor, and 14 studies of this A-factor-binding protein and its mutant suggest that, in the absence of A-factor, the protein represses genes involved in both the synthesis of streptomycin and the onset of sporulation. It was further determined that the intracellular concentration of A-factor is the basis for the temporal regulation of A-factor-regulated genes, by the derepression of genes which are bound by A-factor-binding protein [74].

Another regulatory gene found in S. coelicolor, named afsR, also has been found. This gene, which is different from afsB, stimulates, at high copy number, the production of the pigmented antibiotics, particularly actinorhodin, in non- pigmented mutants of both S. coelicolor A3(2) and S. lividans

[74,160].

Another form of pleiotropic regulation in Streptomyces has been described in a review by Keith Chater [22]. Seven classes of pleiotropic mutants in S. coelicolor fail to produce aerial mycelia and, except for one particular mutant, show partial or even complete impairment of antibiotic biosynthesis. Because of the failure to produce the aerial hyphae, these classes of mutants are called "bald mutants"

(bid). Thus far, the most interesting of bid mutants belong to the class, bldA. The bldA gene product is a tRNA essential for the translation of the UUA codon [104], a codon which is exceedingly rare in Streptomyces. In an analysis of 19,460 codons from Streptomyces genes (personal communication from 15

M.J. Bibb cited in [22]), only four codons were of the TTA

(leucine) type (Figure 2 in [22]). Moreover, genes virtually

unexpressed in bldA mutants, due to regulation by the rare tRNA, could be altered by site-directed mutagenesis of the TTA

codon to a different leucine codon to effect abundant

expression in the bldA mutants [105].

Streptomycetes have multiple RNA polymerase a factors whose use and expression in Streptomyces is developmentally

regulated [18,19,121]. It is clear that a factors can play a major developmental role in the life cycle of Streptomyces

[25]. As an example, the a factor used in the transcription

of sporulation genes of Streptomyces coelicolor A3(2), ,

is encoded by the whiG gene and is critical to the process of

sporulation. It is not unreasonable to suggest that the potential for high transcriptional flexibility in these

organisms makes possible pleiotropic regulation of antibiotic biosynthesis by the use of differential ct factors [19,22].

Recently, Champness and her coworkers discovered a new pleiotropic regulator of antibiotic biosynthesis [ 1 ]. A chromosomal locus was identified in which mutations resulted

in the loss of biosynthesis of all antibiotics known to be

produced by S. coelicolor A3(2), including methylenomycin, which is encoded by the S. coelicolor A3(2) plasmid SCPl,

[27,92,93]. There was no apparent effect on sporulation.

Though no genes from this locus (named absA) have yet been cloned, the chromosomal location has been established, and it 16 maps to a location distinct from any of the antibiotic biosynthesis loci and from the known regulatory loci of S. coelicolor.

A gene recently cloned in this laboratory from

Streptomyces insignis and sequenced, asaA (for activation of sporulation and antibiotic), appears also to be a novel global regulatory gene (Y. Li, P.L. Bartel, R.H. Woodman, and W.R.

Strohl, in preparation).

Antibiotic Synthesis Regulation: Pathway-Specific Regulators

A pathway-specific regulatory gene governs the production of an antibiotic by controlling specific genes in the biosynthesis of that antibiotic, without affecting other pathways such as those of primary metabolism, sporulation, or other antibiotics. Chater has pointed out [24] that, to date, only positive regulatory elements (both pleiotropic and pathway-specific) have been cloned and analyzed. It is likely, however, that pathway-specific negative regulators exist, but are as of yet undiscovered. In the case of methylenomycin, for example, antibiotic biosynthesis increases if a 5 kbp region of DNA adjacent to the cluster of genes essential for antibiotic biosynthesis is deleted.

Furthermore, cloning of the entire methylenomycin biosynthetic pathway, sans this negative regulatory region, onto a low copy number plasmid resulted in higher production of the antibiotic than when this regulatory region was included in the plasmid 17 clone [L. Woodburn, cited as a personal communication, in

(24)].

The positive regulatory elements found so far operate by unknown mechanisms to increase the expression of genes in the biosynthetic pathway cluster. Examples thus far, include actll in the actinorhodin pathway [24,70], strR in the streptomycin pathway [137], and redD in the pathway synthesizing undecylprodigiosin [49]. When additional copies of these genes are introduced into the producing organism, even at low copy number, dramatic overproduction of the antibiotic occurs.

Pathway-specific regulators can be isolated by mutant analysis. Examples may be drawn from the identification of the regulatory mutants in actinorhodin [116] and undecylprodigiosin [49]. In both of these cases, antibiotic nonproducing mutants were created by chemical mutagenesis methods. Cosynthesis experiments, in which different mutants were grown together to see if the two in combination could cause the biosynthesis of the antibiotic end product, were carried out. This procedure allowed the various mutants to be sorted into classes. With the two examples, the regulatory mutants were identified by their inability to cosynthesize parental antibiotics with any other class of mutant.

Molecular cloning of the appropriate DNA into these regulatory mutants restored production of the antibiotic [116,128]. 18

Antibiotic Synthesis Regulation: Antibiotic Resistance

Secondary metabolite production in streptomycetes is rigorously controlled. For its own survival, an antibiotic- producing organism must have an effective defense against the lethal compound which it is producing. Antibiotic resistance mechanisms in Streptomyces, therefore, have been extensively studied, and comprehensive reviews on the subject are available in the literature [27,28,34,36,66]

The arrangement on the chromosome of the genes involved in antibiotic synthesis is probably important in both the regulation of the antibiotic resistance genes and in the use of the antibiotic resistance gene(s) as a regulatory tool for other antibiotic (biosynthesis) genes. In every case found so far, the genes for antibiotic biosynthesis are located in large clusters of DNA on the chromosome or plasmid.

Furthermore, the antibiotic resistance gene (or genes) is embedded in this cluster of DNA [27,51,66]. It should be noted that antibiotic resistance markers in streptomycetes have facilitated cloning of large DNA fragments coding for the enzymes catalyzing antibiotic biosynthesis. Thus, transformation of a sensitive host with plasmid libraries containing DNA from the desired DNA donor, followed by selection for resistance to the appropriate antibiotic marker can theoretically yield large DNA fragments containing multiple antibiotic biosynthesis pathway genes

[24,51,66,140,147] . 19

Davis and Maas [40] listed seven biochemical mechanisms by which an organism could display resistance to antibiotics.

These mechanisms are as follows:

1. alteration of the antibiotic to inactivate it;

2. alteration of the target site of the antibiotic to make

it impervious to the action of the antibiotic;

3. block the transportation of the antibiotic;

4. bypass of the antibiotic-sensitive step by replacing

the target with a different, but functionally

equivalent, entity;

5. titration of the antibiotic by increasing the level of

an inhibited enzyme;

6. production of an endogenous or exogenous product which

spares the cell from the antibiotic-sensitive step;

7. production of a metabolite which antagonizes the action

of the inhibiting antibiotic.

Davies [39] has shown that among all antibiotic-resistant organisms isolated to date, the first six of these proposed mechanisms has been found. It has been also shown that

interstrain (and perhaps interspecies) transmission of

resistance genes has occurred among procaryotes, leading to

the evolutionary ascension of numerous organisms with

antibiotic resistance [46,147].

Cundliffe [36] has pointed out that in antibiotic- producing streptomycetes, three general resistance mechanisms

(out of the seven listed) are employed, singularly or in 20 combination, in self-defense against the produced antibiotic.

These mechanisms are: (a) modification of the antibiotic target site in the cell; (b) intracellular inactivation of the antibiotic (with the possibility of a reactivation step as the antibiotic leaves the cell); and (c) active transport of the antibiotic out of the cell, thereby giving the lethal compound little or no chance at damaging the cell.

D. A. Hopwood and K. Chater [24,27,66] have suggested that antibiotic resistance genes (some paths have more than one, [156]) may, in part, regulate the biosynthesis of antibiotics. If the resistance mechanism is not operative, the organism cannot safely produce the antibiotic. Various forms of regulation between resistance and biosynthetic genes have been suggested based on studies of gene transcription in the biosynthetic cluster. For example, in the methylenomycin cluster the resistance gene (mmr) is read as a monocistronic transcript from only one point of initiation [131, and reviewed in 27]. The promoter for mmr is, however, placed back-to-back with a promoter for a divergent transcript (not confirmed as such, but believed to be, a biosynthetic gene transcript) which initiates 81 bp away from the start site for the resistance gene. Also, the downstream end of the resistance gene transcript overlaps with a transcript from the other direction, known to encode a biosynthetic gene, and

forms a stable, stem-loop RNA structure which could act as a transcription terminator in both directions in the region of 21 the overlap [27,160]. It appears that such interrelation of gene transcripts could provide a mechanism for the coordinate regulation of antibiotic biosynthesis, by providing the means of allowing expression of the resistance gene at the same time, or even before, expression of the physically proximal biosynthesis genes [27].

By no means is mmr the only example of a resistance gene with divergent promoters. Other examples include the erythromycin resistance gene (ermE) from Saccharopolyspora erythraea (formerly Streptomyces erythreus), the neomycin resistance gene (aph) from Streptomyces fradiae [both reviewed in refs. 11,27], the streptomycin resistance gene (sph) from

S. glaucescens [27], and the redD locus of S. coelicolor A3(2)

[128].

Peptide Antibiotics ; General

Peptide antibiotics are known to be widely produced in prokaryotes and lower eukaryotes. Two general classes of peptide antibiotics exist. The first class consists of those which are synthesized by peptide synthetase enzymes (or enzyme complexes). The antibiotics gramicidin S from Bacillus brevis

[118,152], cyclosporin A and its variants from the fungus

Beauveria nivea (formerly Tolypocladium inflatum) [103,174], and valinomycin from Streptomyces levoris [3,144,149] are the most prominent examples. The second class is composed of those antibiotics which are synthesized on ribosomal templates 22 and posttranslationally modified. Nisin from Streptococcus lactis [84,86] and epidermin from Staphylococcus epidermidis

[87] are the most well-known examples in this group. Several excellent, comprehensive reviews on the structure and synthesis of peptide antibiotics are available

[95,96,97,98,112]. In general, it is believed that the number of peptide antibiotics synthesized on an enzyme template is far greater than the number of peptide antibiotics which are ribosomally synthesized and posttranslationally modified

[98,144].

The most intensively studied peptide antibiotic is gramicidin S from Bacillus brevis ATCC 9999 [96]. This antibiotic is a cyclic decapeptide with the structure shown in

Figure 2.

Formation of gramicidin S, as well as many other peptide antibiotics in which synthesis is enzymatically directed, involves a multienzyme thiotemplate mechanism catalyzing the following reactions: (i) amino acid activation; (ii) enzyme aminoacylation; (iii) amino acid racemization; and (iv) peptide bond formation. Elongation appears to involve the enzyme cofactor 4'-phosphopantetheine in most, if not all, enzymatic synthesis systems. The enzyme (or complex) responsible for the initial peptide synthesis is called a

"peptide synthetase" [96,97,98]. Gramicidin S synthetase consists of two multifunctional enzymes, GSl and GS2, which are 120 kd and 280 kd, respectively. The two enzymes in the Figure 2: The structure of the peptide antibiotic gramicidin S. Adapted from Figure 5.3 of [96]

23 24

D-Phe- -►Pro------►Val- -►Orn- -►Leu ▲

Leu<- -Orn-*- -Pro ■* D-Phe

Figure 2 25

synthetase act to catalyze the formation of the peptide D-Phe-

Pro-Val-Orn-Leu, followed by the cyclization. GSl activates

and epimerizes phenylalanine while GS2 activates the amino

acids in the sequence Pro-Val-Orn-Leu. A mutant of GS2 which

contains no 4'-phosphopantetheine is unable to accept

D-phenylalanine from GSl [lOl]. Similar types of synthetases

and reactions appear in other enzyme-directed peptide

antibiotic syntheses, although there obviously must be some

differences which lead to different end-products. The

bacitracin peptide synthetase, for example, is composed of

three large multifunctional proteins [97,98].

The biosynthetic genes for gramicidin, grsA and grsB, which encode the enzymes GSl and GS2, and grsT, which encodes

a 29 kd protein of unknown function, have recently been cloned

and partially sequenced [100,101]. The genes are apparently transcribed on an operon (see Figure 1 in [100]). The pro

teins putatively encoded by these genes are quite similar to tyrocidine synthetase 1 (grsA being 56% identical to tycA) and vertebrate fatty acid thioesterase II enzymes (grsT being 30% homologous to the fatty acid thioesterases). Although a definite function was not known for the protein encoded by grsT, the peptide motif GHSXG was found in the putative protein product of the gene. This motif is found at the active site of thioesterases and lipases and in fatty acid synthetases associated with acyl group transfer (using the

4 '-phosphopantetheine prosthetic group). The highest homology 26 of the GrsT protein was to rat mammary gland thioesterase II and to duck uropygial gland thioesterase II, both of which use their active site serine to release fatty acid moieties from the thioester link to 4'-phosphopantetheine. Such clues may suggest a function for GrsT in the biosynthesis of gramicidin

S [100].

The site of attachment for 4'-phosphopantetheine in acylcarrier proteins involved in fatty acid biosynthesis is also conserved. Hale and Leadlay [58] exploited this fact to find discrete bands of DNA hybridizing to oligonucleotide probes for this motif (GADSLDTVE). These researchers were studying erythronolide synthase, involved in the synthesis of erythromycin, and were working on the hypothesis that, since the enzyme has some notable similarities to fatty acid synthase, it might be possible to find the erythronolide synthase gene by searching for DNA which hybridized to a probe coding for the 4'-phosphopantetheine attachment site. Their initial attempts, however, only netted them a fatty acid acylcarrier protein [57]. This attachment site, as it is found and conserved in species of Streptomyces, has been further refined by Sherman et ai [154], and that refinement was used in this study (see Chapter V).

Thiopeptide Antibiotics; Thiostrepton and Nosiheptide

Thiostrepton and nosiheptide (Fig. 3) are two members of a group of sulfur-rich, highly-modified, peptide antibiotics Figure 3: The structure of the thiopeptide antibiotics thiostrepton (A) and nosiheptide (B).

27 Id w Io (2)1 |p«olo(3)|

H N NHj = 0

D t o l a l l ) |T h z (5 )|

H j I Ala (1)1

CH2CH3 H HH3C N . 0 K 0 ^ “ HI H N " ^ 0 H N

, HO = CH3 |Th:(2)| H

Figure 3

CO 00 29 containing multiple thiazole rings (the "thiostrepton group") which inhibit Gram-positive bacteria [102]. Members of this group of antibiotics include thiostrepton (originally called bryamycin) [31,102,141], nosiheptide (also called multhiomycin) [47,146], sporangiomycin [164], siomycin [138], micrococcin P [171], althiomycin [94], the thiocillins [155], and thiopeptin [63]. Nosiheptide is produced by Streptomyces actuosus ATCC 25421, Streptomyces species ATCC 31463, and

Streptomyces glaucogriseus NRRL 13832 LL-BP 189. It is used as an animal feed additive in Europe and Japan to promote weight gain [7] because of its coccidiostatic properties.

Thiostrepton is produced by Streptomyces azureus ATCC 14921,

Streptomyces hawaiiensis ATCC 12236, and Streptomyces laurentii ATCC 31255; it is used commercially as a topical veterinary antibiotic and is also the most widely used selectable marker in the molecular biology of Streptomyces

[53,169].

The mechanism of action of the thiostrepton group of antibiotics is thought to be similar among most of its members

[167]. In organisms not expressing a gene conferring resistance to this group of antibiotics, thiostrepton inhibits protein synthesis by binding to 23S rRNA and to the ribosomal protein L-11, thus interfering with the action of ribosomal elongation factors EF-Tu and EF-G [32,33,35]. Thiostrepton interferes also with the synthesis of the stringent response factor, ppGpp [132]. The mechanism of resistance of the 30 thiostrepton producers to their toxic product is the modification of the target site to prevent antibiotic binding.

Specifically, the thiostrepton resistance gene product methylates adenosine residue 1,067 of 238 rRNA to 2'-0-methyl- adenosine which inhibits binding of the antibiotic to the rRNA

[32,33,167], The resistance gene product also prevents inhibition of stringent response factor production [132].

The structures of thiostrepton [2,126] and nosiheptide

[127,142] have been elucidated using chemical degradation.

X-ray crystallography, and 2-D NMR analysis, and many aspects of the biosynthesis of both antibiotics are known

[52,54,81,82,173]; some of these aspects will be discussed below.

Experiments utilizing both in vivo feeding of stable isotopes and NMR product analysis have been carried out in S. actuosus ATCC 25421 for nosiheptide and S, laurentii ATCC

31255 for thiostrepton. The overall biosynthetic process appears to proceed as follows: (i) assembly of all the precursor amino acids into a linear peptide (perhaps using peptide synthetases); (ii) post-assembly modifications of the peptide, including the formation of the macrocycle (see discussion below); (iii) modification of tryptophan to either indolic acid (nosiheptide) or quinaldic acid (thiostrepton)

(see discussion below); and (iv) attachment of this modified tryptophan to the backbone of the antibiotic. Recently, the number of gene products required for biosynthesis of 31 nosiheptide and thiostrepton has been predicted [50]. These predictions are shown in Table 1.

It has been shown that all components of these highly modified peptides originate from L-amino acids. The primary

(or linear) precursors of nosiheptide and thiostrepton have been identified (except for possible additional C-terminal residues). The precursors are as follows: HgN-ser-cys-thr- thr-cys-glu-cys-cys-cys-ser-cys-ser-ser(?)-COOH for nosiheptide; andHaN-ile-ala-ser-ala-ser-cys-thr-thr-cys-ile- cys-thr-cys-ser-cys-ser-ser-ser(?)-COOH for thiostrepton.

The thiazole rings in the antibiotics (Fig. 3) are formed from a molecule of cysteine and the carboxyl group of the adjacent amino acid. In both antibiotics, dehydroalanine is formed from the dehydration of serine and dehydrobutyrine from the dehydration of threonine. Hydroxyglutamic acid is derived from glutamine, while thiostreptine is a derivative of isoleucine. The formation of the residue which produces the macrocyclic structures of nosiheptide and thiostrepton

(pyridine in nosiheptide and piperideine in thiostrepton) are generated by a unique process in which the side chains of two serine residues, separated by ten amino acid residues in the peptide chain, are joined.

One of the interesting aspects of the biosynthesis of these antibiotics is the formation of the loop structures.

The indolic acid group forming the loop structure of nosiheptide and the quinaldic acid group, which forms the 32

Teüsle 1: Predicted Number of Genes for the Biosynthesis of Thiostrepton and Nosiheptide

Number of genes predicted to be required to encode functions for nosiheptide and thiostrepton biosynthesis:

I Predicted enzyme activity required Thiostrepton I Nosiheptide for biosynthesis of antibiotic

Minimum Host Likely Maximum Minimum Host Likely Maximum

I Peptide synthetase® 1 2 3 1 1 2 1 Eacemase 1 1 1 1 1 1 Peptide hydrolase 1 1 1 1 1 1 Amide synthesis 1 1 1 1 1 I Serine/threonine dehydratase 1 2 2 1 2 1 Thiazole formation 2 2 10 2 2 9

j Thiostreptine formation 2 3 3 _b -- Piperideine formation 2 3 5 - -- 1 Quinaldic acid formation 4 4 6 - -- Quinaldic acid attachment 3 3 3 - --

Hydroxyglutamate formation -» - 1 1 1 ■ 1 Pyridine formation -- 1 3 4 6 1 Indolic acid formation -- ! 4 4 5 1 Indolic acid attachment -- - 1 1 2 2

1 Total Genes Predicted 22 16 20 32 I “ 1

“Based on other prokaryotic peptide synthetases.

*Tfot applicable to this antibiotic. 33 larger loop structure of thiostrepton, are both derived from tryptophan, though by different enzymatic mechanisms. In nosiheptide, the indolic acid moiety is formed sequentially by intramolecular rearrangement, méthylation, and hydroxylation, probably using four to five enzymatic steps. The quinaldic acid moiety of thiostrepton is formed by a different intramolecular rearrangement and a different méthylation, probably using between four and six enzymatic steps. The tryptophan modifications occur separately from the modification of the peptide precursor chain of these two antibiotics and are added to the backbone of the antibiotic at a later stage in the biosynthesis.

Molecular Biology of Thiostrepton and Nosiheptide Biosynthesis

Both the thiostrepton resistance gene (tsrJ?) [8,165,166] and the nosiheptide resistance gene (nshR) [45,109] have been cloned. The nosiheptide resistance gene was initially cloned by other workers in this author's laboratory in 1988, then subsequently, in 1989, by Dary et al [38] and Cho et al [29].

Both tsrR and nshR have been sequenced [8,109]. The gene products of the genes encoding resistance to thiostrepton and nosiheptide are 74% similar (increased to 79% similarity when conservative substitutions are taken into consideration). The nucleotide sequences of these two genes are 72% identical

[109]. Expression of either nshR or tsrR confers resistance 34 to both antibiotics equally [45], which is not surprising considering their similarities.

Transcriptional analyses of both genes have been carried out [11,109], and there are startling differences between the

DNA structures surrounding these two resistance genes. First, the promoters of the two sequenced resistance genes are quite different [111]. Second, tsrR is transcribed from two tandem promoters [11], whereas nshR has a weak promotor and is also transcribed by partial readthrough of an upstream transcript from a gene designated nshA (also called orf699). The nshA gene contains two promoters, one strong and one relatively weak, and the more potent of the two (promotor P2) has a -10 region which is different from any other streptomycete promotor yet analyzed [109]. The upstream gene in S. actuosus

(nshA, also called orf699) is separated from the downstream gene (nshR) by a strong stem-loop structure (possessing a AG° of formation of -24.0 kcal/mol; see Fig. 6 in [109])

Transcription, however, apparently reads through this terminator about 35% of the time in S. actuosus. No readthrough of this terminator was observed, however, in SI

Nuclease mapping of transcription in S. lividans TK24.

Despite the fact that no readthrough occurs in S. lividans, the NshR phenotype is still expressed in that organism, indicating that the promoter for nshR is utilized [45,109].

Transcription and translation of nshR appear to be unique with respect to genes found in Streptomyces spp.; both appear to 35

initiate from the same nucleotide [ 109 ]. This unusual feature is shared by only six other genes among the streptomycetes : promotor PI of the erythromycin resistance methylase (ermE-pl) from Saccharopolyspora erythraea (formerly Streptomyces erythraeus) [10]; promotor PI of the aminoglycoside phosphotransferase gene (aph-pl) from Streptomyces fradiae

[11]; the afsA promotor from S. griseus [78]; the streptothricin acetyltransferase promotor (sta-p) from S. lavendulae [73]; the promotor for the ribostamycin phosphotransferase gene (rph-p) from S. ribosidificus [80]; and the promotor for the aminocyclitol acetyltransferase gene

(aacC7-p) from S. rimosus forma paromomycinus [113]. CHAPTER III; SPECIFIC AIMS OF THE RESEARCH

The overall goal of this research is to gain a better understanding of the genetics of the organisms producing thiopeptide antibiotics using S. azureus and S. actuosus, which produce thiostrepton and nosiheptide, respectively, as model systems. Elements of this project include: (i) the generation of blocked mutants of S. azureus and S. actuosus unable to produce thiopeptide antibiotics; (ii) cloning of genes known or suspected of being involved in the biosynthesis of the thiopeptide antibiotics; (iii) Southern blot hybridization analysis of the wild-type and mutant forms of the producing organisms to determine the restriction map of the resistance gene area in the wild-type S. azureus and to determine if deletions or other obvious aberrations occurred in this region of the chromosome; and (iv) the study of the mode of biosynthesis of both thiostrepton and nosiheptide.

The generation of mutants which are stably blocked in the biosynthesis of thiostrepton and nosiheptide was attempted using the mutagens ethyl methane sulfonate (EMS), ultraviolet light (UV), nitrous acid (HNO^), and N-methyl-N'-nitro-N- nitrosoguanidine (MNNG). Attempts were also made to find mutants which are auxotrophic for amino acids incorporated

36 37 into the thiopeptide antibiotics. The most efficacious mutagen was MNNG; other mutagens which were tried were much less successful in the generation of stable mutants.

The S. azureus homolog of the putative regulatory gene, nshA, first isolated from S. actuosus, was cloned in the course of this research. Furthermore, the effect of nshA on the production of nosiheptide in S. actuosus was studied using site-specific mutation of the nshA gene followed by "gene replacement" using homologous recombination. Cloning of genes encoding resistance to thiopeptide antibiotics from producers of thiostrepton was attempted using a variety of methods, described in Table 14 of Chapter V. The great difficulties encountered in these cloning attempts led to a startling new discovery regarding resistance-gene-linked, chromosomally- integrated plasmids in the organisms producing thiopeptide antibiotics.

Southern hybridization analysis of the DNA of the wild- type and mutant forms of 5. azureus was carried out to: (i) understand the chromosomal arrangement of the putative biosynthetic clusters in the wild-type organism; and (ii) analyze the nature of mutations which occurred in the course of generating stable mutants. In preliminary experiments, possible deletions were observed in three S. azureus mutants, showing the requirement for this approach to map the DNA isolated from these mutants. 38

The study of the mode of synthesis of the antibiotics was considered crucial to future strategies in the cloning and analysis of the thiostrepton and nosiheptide biosynthesis pathways. The question of whether the precursor molecules of these complex antibiotics are synthesized on the ribosome or on an enzyme template had been addressed before [81,128], but the results were inconclusive. Cloning strategies could conceivably be altered based on the answer to this question.

Therefore, two studies of the means of antibiotic biosynthesis were conducted. A preliminary study used an endpoint analysis of biosynthetic products in the presence and absence of chloramphenicol. A later, more detailed investigation involved the analysis of antibiotic synthesis over time in the presence and absence of chloramphenicol. The data generated from these studies will assist future attempts to clone DNA encoding genes for the synthesis of thiostrepton and nosiheptide, because the results delineate the nature of the problems which will be encountered in future cloning work as well as the types of experiments which will be most likely to yield, in the future, the most productive results with respect to cloning the genes encoding proteins involved in the biosynthesis of thiostrepton. CHAPTER IV: METHODOLOGY

Chemicals and Biochemicals

Amberex (an industrial grade yeast extract) was obtained from Universal Foods (Milwaukee, WI). Soluble yeast extract,

Bacto-peptone, Bacto-tryptone, casaminoacids, and malt extract were purchased from Difco Laboratories (Detroit MI).

Trypticase soy broth was purchased from BBL Microbiology

Systems (Cockeysville, MD). Cotton seed meal for the production medium was obtained from Traders Protein (Memphis,

TN). Agar for plate media was purchased primarily from United

States Biochemicals (Cleveland, OH) and Sigma Chemical Company

(St. Louis, MO). Nosiheptide was obtained from other members of the H.G. Floss laboratory, having been biosynthesized in the labs and compared against a standard from Rhône-Poulênc

(France) [81]. Thiostrepton was initially received as a gift of E.R. Squibb & Sons (Princeton, NJ), but was later purchased from Sigma Chemical Company. Apramycin sulfate was a gift of

Eli Lilly & Co. (Indianapolis, IN). Bovine serum albumin

(Fraction V) used as a spectrophotometric standard was obtained as a prepared solution at a concentration of 2.00 mg/mL from Pierce Chemical Company (Rockford, IL).

Dithiothreitol, trichloroacetic acid, Scintiverse™ E (a

39 40 multipurpose scintillation cocktail), glycine, glycerol, and most reagent-grade solvents (for thin-layer chromatography) were purchased from Fisher Scientific Co. (Fairlawn, NJ). The radioactive materials, ^®S-L-cysteine and a-^^P-dCTP, and the fluorographic spray, EN^HANCE, were purchased from Dupont-New

England Nuclear (Boston, MA). Crude y-^^-ATP for kinase reactions was purchased from ICN Biochemicals (Irvine, CA).

Random, hexameric, oligonucleotide, DNA primers for the random-primed oligonucleotide labelling method of radioactively labelling DNA were purchased from Promega

Biotech (Madison, WI), as were also the complete kits for replacement vector cloning and in vitro packaging using

AEMBL4. All restriction endonucleases, T4 polynucleotide kinase, T4 DNA ligase, Klenow Fragment of DNA Polymerase I, SI

Nuclease, IPTG, X-gal, and agarose were obtained from

GIBCO/BRL (Gaithersburg, MD). Nitrocellulose used in Southern hybridization experiments and in colony hybridizations experiments was purchased from Schleicher & Schuell Co.

(Keene, NH). Silica Gel G thin-layer chromatography plates

(0.2 mm thickness, aluminum-backed, without fluorescent indicator) were purchased from E. Merck AG, Darmstadt,

Germany. Oligonucleotides used for hybridization analyses were synthesized by the OSU Biochemical Instrumentation

Center. Cesium chloride (CsCl) used for the ultracentrifuge purification of plasmid DNA was purchased in the 99% pure, optical grade category from Research Organics, Inc. 41

(Cleveland, OH). Kodak GBX developer, GBX fixer, and X-OMAT film were purchased from Hale Medical Systems (Columbus, OH).

Other chemicals, if not specifically purchased from one of the above-named companies, were of the highest available purity.

Media Used in this Work

All media were sterilized by autoclaving for 20 to 25 minutes at > 121°C, with steam pressure of ca. 20 psi, unless otherwise stated. All media were adjusted to a final pH of

7.2 ± 0.1, except for T Media and LB Media which were adjusted to pH 7.5 ± 0.1.

A. Yeast Malt Broth fYMB) (modified from [5])

Yeast Extract 5.0 gm / L (soluble or Amberex) Malt Extract 10.0 gm / L (must be soluble) Peptone 5.0 gm / L Glucose 10.0 gm / L lOX Trace Elements Solution 0.200 mL / L casaminoacids 0.250 gm / L dHaO to 1.0 L

After autoclaving, add the following sterile solutions;

MgSO^ (1.0 M) 5.0 mL / L Glycine (12.5% w/v) 40.0 mL / L

B. Yeast Malt Broth with Sucrose fYEMEl (modified from [5])

Yeast Extract 5.0 gm / L (soluble or Amberex) Malt Extract 10.0 gm / L (must be soluble) Peptone 5.0 gm / L Glucose 20.0 gm / L Sucrose 340 gm / L 42 lOX Trace Elements Solution 0.200 mL / L dHaO to 1.0 L

After autoclaving, add the following sterile solutions:

MgSO^ (1.0 M) 5.0 mL / L Glycine (12.5% w/v) 40.0 mL / L

C. Luria-Bertani fLBl Medium [117]

Tryptone 10 gm / L Yeast Extract 5 gm / L NaCl 10 gm / L MgSO* 2.4 gm / L dH^O to 1.0 L

LB Aaar. used in cloning experiments, is made by the addition of 15 gm of agar per liter of medium. For bacteriophage A transfections in A-EMBL4 cloning experiments, 20% filter- sterilized maltose was added after autoclaving to a final concentration of 0.4%. This medium was called T. T Top

Agarose was made by the addition of 7.0 gm of agarose per liter of medium to T Media.

D. Minimal Medium (adapted and modified from [20])

K2HPO4 0.476 gm / L KHsPOa 0.200 gm / L Na^HPO* 0.170 gm / L (NHa)=SO* 1.321 gm / L NaNOa 0.850 gm / L dH^O to 800 mL Agar 22 gm / L

After autoclaving, add the following sterile solutions: lOX trace elements sol'n 200 fiL / L MgS0 *.7H20 (1.0 M sol'n) 1000 /xL / L CaCl2 .2H20 (0.25 M sol'n) 1000 pL / L D-glucose (25% w/v sol'n) 100 mL / L HEPES or MOPS (0.25 M, pH 7.2) 100 mL / L 43

E. R2YE (modified from [67])

SOLUTION I

K=SO* 0.2 gm / L MgCla*6 H^O 10.12 gm / L Glucose 10.0 gm / L Casaminoacids 0.10 gm / L Agar 22-25 gm / L dH=0 to 600 mL

SOLUTION II

103 gm of sucrose added to 200 mL of dH^O and swirled until dissolved.

DIRECTIONS

Solution II was added to Solution I while both were still fairly hot, then the following sterile solutions were added;

K=HPOa (0.5% w/v) 10 mL / L CaCl2*2H20 (3.68%) 80 mL / L L-Proline 15 mL / L (20% w/v; optional) TES, HEPES, or MOPS Buffer 100 mL / L (0.25 M, pH 7.2) lOX Trace Elements Solution 0.200 mL / L NaOH (1 N) 5 mL / L (doesn't need sterilization) Yeast Extract (10% w/v) 50 mL / L (soluble or Amberex)

Approximately 25 mL of medium per 82 cm (diameter) Petri plate was dispensed. Before use of these plates for any work involving protoplasts, the media was partially dried by the incubation of the plates under a biological safety hood (with the UV sterilizing light on and the lids off) for 1.5 - 2 hr or by incubation upright at room temperature for five days

[67]. 44

R5 Medium [67]

Sucrose 103 gm / L KaSOa 0.2 gm / L MgCla.6 HaO 10.12 gm / L Glucose 10.0 gm / L Casaminoacids 0.10 gm / L lOX Trace Elements Solution 0.200 mL / L Yeast Extract 5.0 gm / L MOPS Buffer 5.25 gm / L or HEPES Buffer 5.95 gm / L dH=0 to 1 ,000 mL Agar 25 gm / L

After autoclaving, add the following sterile solutions:

KH2PO4 (0.5% w/v) 10 mL / L CaCla (5.0 M) 4 mL / L NaOH (IN) 7 mL / L

R5 Medium was not used with protoplasts. It was dried at room temperature only until there was no condensation on the plate surface.

G. Production Medium [5]

Glucose 30 gm / L Cotton Seed Meal 10 gm / L NaCl 3 gm / L lOX Trace Elements Solution 0.100 mL / L CaCOa 3 gm / L The medium was adjusted to pH 7.0 prior to the addition of the

CaCOa. Aliquots of 50 mL of medium per 250 mL Erlenmeyer flask were used for growth of organisms.

H. Defined Production Medium I fPPMl] modified from [20,123]

K=HPOa 10.5 gm / L KH2PO4 4.5 gm / L NUa*citrate*2 H^O 0.5 gm / L NH^NOa 2.0 gm / L ddHgO to 600 mL 45

After autoclaving, add the following sterile solutions:

MgSO^ (IM) 5.0 mL/L CaCl=.2 HzO (0.25M) 0.5 mL/L lOX Trace Elements Solution 0.25 mL/L MOPS Buffer (0.25M, pH 7.2) 100 mL/L 30% Glucose Solution 250 mL/L Yeast Extract Solution (10% w/v) 50 mL/L (optional for poor-growing strains)

I. Revised Defined Production Medium fPPMS) (modified from DPMI plus [81])

K=HPOa 0.50 gm / L KHaPOa 0.20 gm / L Naa*citrate*2 E^O 0.50 gm / L lOX Trace Elements Solution 0.20 mL / L MOPS Buffer 5.25 gm / L L-glutamate 5.00 gm / L L-arginine 1.00 gm / L L-aspartate 1.00 gm / L Soluble Yeast Extract 1.00 gm / L (can be considered optional) ddHaO to 750 mL

After autoclaving, add the following sterile solutions:

MgSO^ (l.OM) 5.0 mL / L CaCl2*2 HgO (0.25M) 0.5 mL / L 20% Glucose Solution 250 mL / L

J. Trace Element Solution fNormal Strength Formulation1 [67]

ZnCla 40 mg / L FeCla-SHaO 200 mg / L CuCl2 «2H 20 10 mg / L MnCl2 *4H20 10 mg / L NagBaO^.lOHaO 10 mg / L (NH*)6M0 7 0 2 a'4HaO 10 mg / L ddH=0 to 1.0 L

For convenience, this solution was routinely made at ten times the normal formula strength shown above. 46

Bacterial Strains and Plasmids

The sources of strains of bacteria used in this research are listed in Table 2. All mutant strains used in this study are listed in Table 3. Most mutant strains were derived from

Streptomyces azureus ATCC 14921 (Saz-1), a producer of thiostrepton obtained from the ATCC. Two mutant strains were derived from S. actuosus ATCC 25421. One mutant strain, S. species ATCC 31463, was obtained from the ATCC. Plasmids used in this study are listed in Table 4. Plasmids pIJ6l, pIJ702, pIJ941, pIJ915, pIJ2315 were the gift of D.A. Hopwood of the

John Innes Institute. Plasmid pKC505 was a gift of R. Baltz of Eli Lilly & Co. Plasmids pSKN02 and pSKN04 were obtained from G. Volckaert [130]. Plasmid pWHM333 was the gift of C.R.

Hutchinson of the University of Wisconsin, Madison, WI.

Antibiotics Usage

The antibiotics needed for selection of plasmids are defined in Table 4. The amounts needed for selection of the plasmid in the various organisms in which they were used are as follows: ampicillin and carbenicillin (for E. coli), 100 jug/mL; apramycin (for E. coll), 20 pg/mL; apramycin (for S. lividans), 40 - 50 /xg/mL; chloramphenicol (for E. coli), 25 ng/xoL; chloramphenicol (for Streptomyces), 125 jug/mL; cycloheximide, 30 /xg/mL; and nosiheptide and thiostrepton, 50 fxg/mL. Stock solutions of these antibiotics were made as follows: ampicillin and carbenicillin were dissolved in ddH^O 47

Table 2: Bacterial Strains Used In This Research

Name of Strain Relevant Characteristics or Purpose of Use Source of Strain

Staphylococcus Sensitive to thiostrepton and nosiheptide. B. Swoager" aureus ATCC 6538P Used for the assay of these antibiotics.

Micrococcus Sensitive to thiostrepton and nosiheptide. B. Swoager lysodeikticus Used for the assay of these antibiotics.

Pseudomonas Used as a negative control in hybridization D. Galloway aeruainosa POl studies with nshA.

E. coli HBlOl Used as a negative control in hybridization B. Swoager studies with nshA.

E. coli DH5a (1) high transformation efficiency; V. Steigerwald (2) blue-white screening ability with pUC19; (3) RecA" to avoid plasmid re arrangement; (4) Clean mini-prep DNA due to lack of Endonuclease 1.

E. coli TBl Same as DH5a, except for the presence of V. Steigerwald Endonuclease 1 in cells.

E. coli HH538 Derivative of LE392. Permissive host for spi Promega Biotech selection of A (Lej., no P2 lysogen in cells). (Madison, WI)

E. coli NH539 LE392 derivative. Restrictive host for Promega Biotech AEHBL4 spi selection (Le^ the cells con (Madison, WI) tain a lysogen of P2).

E. coli P2392 LE392 derivative. Restrictive host for K. Kendrick AEHBL4 spi selection. AEHBL4 recombinant plaques are larger than with HH539.

Planomonospora Used in hybridization studies of pshA ATCC*” parontaspora and thiostrepton/nosiheptide resistance subsp. antibiotic,Î genes. Sporangiomycin (thiostrepton analog) ATCC 23864 producer. Actinomycete, but not a streptomyciite.

Streptomyces sp. Mutant nonproducer of nosiheptide. Used in ATCC ATCC 31463 protoplast fusion with S. azureus mutant and for study of resistance methylases and nshA 48

Table 2; (continued)

Name of Strain Relevant Characteristics or Purpose of Use Source of Strain

S. actuosus Nosiheptide producer. Source of WIA for clonin 1 ATCC ATCC 25321 in several experiments. Source of nshA

S. alboniaer Used for hybridization studies with nshA B. Swoager

S. albus G Used for hybridization studies with nshA. D.P. Taylor

S. aureofaciens Used for hybridization studies with nshA B. Swoager ATCC 107621

S. azureus Thiostrepton producer. Source of DNA for ATCC ATCC 14921 cloning in most experiments. Used to study nshA

S. banberaiensis Used for hybridization studies with nshA B. Swoager ATCC 13879

StreptonYces C5 Used for hybridization studies with nshA. P. Bartel®

S. californicus Used for hybridization studies with nshA B. Swoager ATCC 3312

S. coelicolor K3C ;) Used for hybridization studies with jisM D. A. Hopwood

S. coeruleorubiduj Used for hybridization studies with nshA B. Swoager HBBL 13-2569

S. aalileus Used for hybridization studies with nshA ATCC ATCC 31133

S. aalileus Used for hybridization studies with nsM ATCC ATCC 31671

S. oriseus Used for hybridization studies with nshA B. Swoager m o 3499

S. oriseus Used for hybridization studies with nshA K. Kendrick N m B-2682

S. oriseoruber Used for hybridization studies with nshA ATCC ATCC 17919 49

Table 2: (continued)

Name of Strain Relevant Characteristics or Purpose of Use Source of Strain

S. hawaiiensis Thiostrepton producer. Used for hybridization ATCC ATCC 12236 studies with resistance genes and nshA

S. insionis Used for hybridization studies with nshA ATCC ATCC 31913

S. laurentii Thiostrepton overproducer. Source of DNA for ATCC ATCC 31255 cloning in a few experiments. Hybridization studies with nshA.

S. lincolnensis Used in hybridization studies with nsM ATCC ATCC 25466

S. lividans Devoid of plasmids SLP2" and SLP3"; thiostreptoi D. A. Hopwood TK24 nosiheptide sensitive. Used in cloning and for nshA studies.

S. lonaisporus Used in hybridization studies with nshA B. Swoager ATCC 23931

S. lucensis Used in hybridization studies with nshA B. Swoager ATCC 17804

S. matensis Used in hybridization studies with gsbA B. Swoager ATCC 23935

S. noaalater Used in hybridization studies with nghA ATCC ATCC 27451

S. pactum Used in hybridization studies with nshA B. Swoager ATCC 27456

S. parvulus Used in hybridization studies with nsfeA ATCC ATCC 12434

S. peucetius Used in hybridization studies with nshA ATCC ATCC 29050

S. purpurescens Used in hybridization studies with nshA ATCC ATCC 25489 50

Table 2: (continued)

Name of Strain Relevant Characteristics or Purpose of Use Source of Strain

S. rimosus Used in hybridization studies with nshA . ATCC ATCC 10970

S. siovaensis Siomycin (thiostrepton analog) producer. ATCC ATCC 13989 Used in studies of resistance genes and QshA

S. steffisburaens s Used in hybridization studies of nshA ATCC ATCC 27466

S. violaceoruber Used in hybridization studies of pshA ATCC ATCC 14980

S. viridochro- Used in hybridization studies of nshA ATCC moaenes ATCC 21343

“B. Swoager is the curator of the culture collection for the Department of Microbiology, The Ohio State University, Columbus, OH.

“’ATCC; American Type Culture Collection, Rockville, HD.

®rhis strain was obtained by P. Bartel [5] from the Frederick Cancer Research Center, Frederick, MD. 51

Table 3: Mutant Strains of Thiopeptide Producers Used in this Study

Name of Mutagen Used or Special Notes or Mutant Mutant Source Characteristics

Saz-2 NTG from J. J. Lee thiostrepton sensitive

Saz— 6 NTG from J. J. Lee poor sporulation

Saz-7 NTG, this study poor sporulation cannot protoplast cannot prepare chromosomal DNA

Saz-8 NTG, this study

Saz-9 NTG, this study

Saz-10 NTG, this study

Saz-11 NTG, this study

Saz-12 NTG, this study

Saz-13 NTG, this study spontaneous deriva tive of Saz-12

Saz-14 this study spontaneous deriva tive of Saz-12

Saz-15 NTG, this study extremely slow growth rate. poor sporulation.

Saz-16 this study spontaneous deriva tive of Saz-2. poor sporulation. thio strepton sensitive.

Saz-17 this study spontaneous deriva tive of Saz-6

Saz-18 NTG, this study 52

Table 3: (continued)

Name of Mutagen Used or Special Notes or Mutant Mutant Source Characteristics

Saz-19 NTG, from J. J. Lee

Sac-11 site-specific mu carries stable mu tation of nshh and tation in nshA gene replacement in wt strain.

Sac-53 same as Sac-11 carries stable mu tation in nshA

Ssp31463 ATCC purchased from ATCC as a producer of nosiheptide, but carries a mutation of unknown type which prevents the synthesis of nosi heptide . 53

Table 4: Plasmids Used in this Project

Plasmid Relevant Characteristics® Source

PÜC19 hiah coDV number E. coli vector for clonina. amp'. lacZ GIBCO/BRL

pIJ61 low CODV number Streptomyces vector, thio'. neo' D. A. Hopwood

PIJ702 hiah CODV number Streptomyces vector, thio'. mer D. A. Hopwood

pIJ915 sinole copv Streptomvces vector, vio' D. A. Hopwood

pIJ941 sinale codv Streotomyces vector, thio', hy

PIJ2315 sinale copy Streptomyces vector, thio'. contains part D. A. Hopwood of act biosynthetic pathway.

PKC505 low copy Streptomyces/high copy E. coli shuttle cosmid. H. Richardson apr'

PWHH333 apif derivative of pKC505. contains large part of C. R. Hutchinson dnr pathway.

PAHT29 neo'. derivative of pIJ61. BamHI cloning site. D. C. Dosch

PÀHT34 E. coli vector derived from pBE322. contains entire D. 0. Dosch 8.5 kbp piece of DNA conferring resistance to nosiheptide. amp'

PANT38 Streptomyces vector, contains 2.6 kbp piece of DNA with D. C. Dosch nshA. nshR. and 3' end of orfB. contains vio' from pIJ915. Markers are nslf and vio'.

pANT39 E. coli vector. Contains the same DNA as in dANT38. D. C. Dosch but constructed in pDCl9. Used in the autoradiographic size standard construction for DNA fragments hybridizin 1 to labelled pANT34 DNA (see Table 6).

pm400 E. coli vector containina fragment of nshR with no Y. Li nshA DNA present, amp'

PAHT401 E. coli vector containina internal fraament Ï. Li of nshA. amp'

pMT402 apr' derivative of pKC505. contains entire this study 8.5 kbp insert from pANT34.

pANT 407 vio', nsh' pIJ702 derivative, contains nshA D.C. Dosch & gene with site-specific deletion. this study 54

Table 4: (continued)

Plasmid Relevant Characteristics® Source

pAHT408 PÜC19 containing 2.8 kbp nshA-equivalent DHA from this study azureus plus extra, non-contiguous DNA. amp", lacZ"

pAHT409 subclone of pANT408 containing 2.8 kbp Pstl-BolII this study S. azureus DNA with nshA-equivalent DNA.

PAHT410 nsh", vio". Contains nshA and ns)g. plJ702 derivative. D. C. Dosch Balll cloning site. Partial tsrR gene deletion.

pANT419 vio". M I and B g i H cloning sites. Derivative of this study pANT410 in which nshA and psM genes completely deleted.

PANT420 vio", nsh". pAHT419 derivative. Contains 8.5 kbp insert this study from pANT34 cloned into B g l K site. Insert rearranged spontaneously.

PANT421 same as pANT421 above, but the insert is not rearranged. this study

PANT422 apr", thio". 6.1 kbp MH I thiostrepton resistant this study fraament from S. azureus cloned into pKC505. Spontan eous rearrangement of one end of insert occurred.

pSKN02 hiah-copy-number E. coli/Streotomyces shuttle vector, 6. Volkaert [130] cam" selection in E. coli.

pSKN04 hiah-copy-number E. coli/Streptomvces shuttle vector, G. Volkaert [130] cam" selection in E. coli, neo" selection in E. coli and Streotomyces spp.

“Abbreviations: amp*", ampicillin resistance; thio*, thiostrepton resistance, encoded by ts^; nsh'', nosiheptide resistance, encoded by nshR; neo", neomycin resistance; vio", vionycin resistance; hyg", hygromycin resistance; apr", apramycin resistance; cam", chloramphenicol resistance; dpE, daunorubicin biosynthesis genes; lacZ/ «-fragment of the lacj gene; mel, melanin pigment production; 55 at a concentration of 100 mg/mL; apramycin was dissolved in ddHaO at a concentration of 20 mg/mL; chloramphenicol stock solutions were dissolved in 95% ethanol at concentrations of

25 mg/mL (for use in E. coli) and 125 mg/mL (for use in

Streptomyces); cycloheximide was dissolved in ddH^O at concentrations of 10 mg/mL; nosiheptide and thiostrepton were dissolved in the purest available grade of DMSO at 50 mg/mL concentrations. Stock solutions dissolved in ddH^O were filter-sterilized with 0.2 /xm pore-size Acrodisc* syringe filters (Gelman Sciences, Ann Arbor, MI). For E. coli, antibiotics were added to the plates as needed just prior to use and were allowed to soak into the media thoroughly by incubating the plates at 30°C for one to three hours. For

Streptomyces transformations, antibiotics were added to the plates about 12 - 20 hours after transformation (see below).

Culturing of Organisms

Cycloheximide was added to media only as needed to combat fungal contamination. Cultures in liquid media were incubated at 250 rpm on a circular rotary incubator. The incubation temperatures were 30 °C for Streptomyces and either 30°C or

37“c for E. coli strains. Staphylococcus aureus, and

Micrococcus lysodeikticus. E. coli was routinely cultured on either LB Agar or Trypticase Soy Agar (TSA; Trypticase Soy

Broth plus 15 gm/L agar) and incubated at 30°C or 37°C. E. coli used in the blue/white selection method of cloning with 56 pUC19 was grown only on LB Agar with 100 ng/mL of ampicillin or carbenicillin, 40 /iL of X-gal (stock solution in dimethylformamide was 2% w/v, giving a final concentration on the plate of 0.0032%), and 10 fiL of IPTG (stock solution in ddHgO was 100 mM, giving a final concentration on the plate of

40 fiU). All species of Streptomyces were routinely cultured on R5 medium, as it was observed that most strains used grew and sporulated well on this medium. The actinomycete

Planomonospora parontospora var. antibiotica would grow only on LB and TSA. Protoplasts of Streptomyces were regenerated on dried R2YE as described by Hopwood et al [67] ; R5 was tried as a regeneration medium and found to be quite unsatisfactory in terms of the efficiency and rate of protoplast regeneration. S. aureus and M. lysodeikticus were routinely cultured on TSA and incubated at either 30°C or 37°C.

For plasmid preparations, E. coli was grown in LB medium with the appropriate antibiotic added in the appropriate concentration. For small scale preparations, the bacteria were grown in 5.0-10.0 mL of LB in 16 mm X 150 mm glass culture tubes. E. coli HBlOl, used as a negative control in the Southern blot experiments with orf699, was grown in 400 mL

TSB.

Streptomyces grown for chromosomal DNA isolation, large- scale plasmid isolation, or protoplast formation were grown in

YMB as a starter culture (50 mL in a 250 mL flask, 30°C, 250 rpm) using an inoculum of either 1.0 mL of a dense spore 57 preparation (> 1*10^ spores per mL) or a single, isolated colony of the organism from plate media. The organism was allowed to grow for 36 to 72 hours, until the culture was dense. An aliquot was transferred to a second, larger flask of YEME (if the organism formed large, mycelial balls in YMB) or YMB (if the organism grew with its mycelia dispersed in

YMB). The size of the aliquot transferred was ca. one-tenth of the total volume of the second flask of media. Typically,

S. laurentii was grown in YMB, though the scale-up culture medium contained no casaminoacids. S. lividans and S. actuosus were variable and unpredictable with respect to dispersed growth in YMB. All other organisms grew as mycelial balls in YMB; therefore, they were grown in YEME for scale-up work. A coiled, stainless steel spring was added to flasks of

YMB and YEME to promote dispersed growth [41].

For general analysis of products of fermentation of various species of Streptomyces, organisms were grown in production medium (Medium G). Addition of a stainless steel spring to the culture flask was not needed to promote dispersed growth.

The first and second experiments designed to ascertain whether nosiheptide and thiostrepton are ribosomally or enzymatically synthesized used the media DPMI and DPM3, respectively. The flasks contained stainless steel springs and were incubated at 30“C and 250 rpm. 58

Preparation of Streptomyces Spores

Spores of species of Streptomyces were prepared from agar plates of the fully sporulated, densely populated organism.

The procedure used was essentially as described by Bartel [5].

The medium on which the Streptomyces strains were grown was either Minimal Medium (Medium D) or R5 (Medium F). Spores to be used within one week of their preparation were kept at 4°C in ddHaO. For longer term storage, spores were frozen at

-20°C in aliquots of 1.0 mL in 10% sterile glycerol in ddH^O.

Once thawed, spores in glycerol had to be used immediately or discarded.

N-methyl-N'-nitro-N-nitrosoauanidine

Mutagenesis of Streptomyces spores

Mutagenesis of spores was carried out as described by

Dellc et al [42], as modified by Hopwood et al [67], and P. L.

Bartel [5]. Approximately 1.5 mg of MNNG was used at a final concentration of 3.0 mg/mL in Tris-Maleate buffer (TM; 0.05 M

Tris, 0.05 M tris-maleate, pH 9.0). Spores were treated for

90 min. at 30°C in this buffer. The titer of surviving spores was determined by serial dilution in sterile water, followed by the assay of the number of surviving spores at each dilution. After determination of the titer of survivors, the spores were pelleted by centrifugation and resuspended in 1.0 mL sterile 10% glycerol and stored in 100 /iL aliquots at

-20°C. 59

All materials which might have become contaminated by the

MNNG and which could not be discarded were washed carefully several times with a 1 .0% solution of bleach (sodium hypochlorite), followed by distilled water rinses. Materials which could be discarded were soaked overnight in a solution of 1.0 N NaOH which was in a 4.0 L plastic beaker placed inside the fume hood and were discarded the following day.

Nitrous Acid Mutagenesis of Streptomyces Spores

The procedure employed to mutagenize spores of

Streptomyces by nitrous acid was modified from J.H. Miller

[123]. Nitrous acid (0.10 M) was prepared by dissolving 69.0 gm of NaNOg in 10.0 mL of acetate buffer (0.101 M Na*acetate adjusted to pH 4.6 with glacial acetic acid). Between 5.0*10® and 5.0*10^ spores were collected by centrifugation in a 1.5 mL microcentrifuge tube. The spores were washed twice with acetate buffer, and then were resuspended in 300 ph of nitrous acid. The spores were incubated 15 min. at 30°C with gentle vortexing every 5 min. to keep them suspended. The nitrous acid was neutralized by dilution in 700 /iL of ”5X Minimal A

Salts" (230 mM K^HPO^, 165 mM KHaPO*, 38 mM (NHU)2S0 a, 8.5 mM

Na^'Citrate, and 25 mM MgSO* at pH 7.0 from [123]). The spores were collected by centrifugation, and were then washed once with 1.0 mL of "5X Minimal A Salts" and once with 1.0 mL of sterile ddH^O. Spores were titered and stored in glycerol as described in the MNNG mutagenesis procedure. 60

Isolation of Mutant Strains of Streptomyces

The attempts to isolate auxotrophic mutants involved making duplicate plates of mutagenized Streptomyces spores, using one plate of nutrient-rich R5 medium (Medium F), and one plate of Minimal Medium (MM; Medium D). Colonies from spores which could grow on R5, but not on MM, were isolated by the use of sterile toothpicks and rescreened to confirm the putative auxotrophic phenotype. All thiostrepton nonproducing mutants isolated were also screened in this manner.

The procedure for isolating stable, non-thiopeptide- producing mutants was developed by P. Bartel [5] for the isolation of anthracycline antibiotic mutants from

Streptomyces C5, with slight modifications being made for this study. Sterile, 96-well, plastic, tissue culture plates

(Corning Laboratory Sciences Co., Corning, NY) were filled with ca. 1 mL of R5 Medium per well. Spores were plated out on Petri dishes of R5 medium at ca. 100 spores/plate and allowed to regenerate by incubation at 30“c. After regeneration and sporulation (about 10 days), the colonies resulting from each spore were picked over to both a single well of the tissue culture plate and to a master plate containing R5 medium and incubated 7 days at 30°C. Colonies which were obviously bald (i.e., no aerial mycelia) were discarded. After incubation, the cultures in the tissue culture plate and on the master plate were assayed to see which were bald and which failed to grow. All non-bald. 61 viable colonies from the tissue culture plates were then picked over to a separate plate seeded with either S. aureus or M. lysodeikticus, made from fresh, overnight cultures in

TSB. Molten TSA was cooled to about 45°C, and 2.5 mL of the fresh, overnight culture was added to 250 mL of the agar medium, which was quickly swirled to mix and then plated out on sterile Petri plates to give lawns of sensitive test organisms. After solidification, the agar plugs from the tissue culture plates were picked in their entirety to the seeded plates (using 20 plugs per seeded plate) and were incubated overnight at 30”c. After incubation, the plugs were removed, and the spots where they had been were visually screened for zones of inhibition of the sensitive organism.

Zones lacking inhibition were subcultured from the master plate onto R5 Medium.

A candidate mutant was entered into the mutant strain collection only if it passed three subsequent attempts to assay it, using overlay assays, for inhibition of the sensitive organisms. The candidate mutants were transferred to R5 medium using a toothpick; no more than twelve colonies were placed on any one plate. After seven days of incubation, the plates were overlaid with 100 /xL of a fresh, overnight culture of a sensitive organism in 3.5 mL of molten "top R5"

(R5 with only 7.0 gm/L of agar). The candidate mutants were grown 7 days with shaking in Production Medium (Medium G) and then extracted for the antibiotic as described in a later 62 section. Even after storage, mutants were assayed for a lack of bioactivity ca. every six months, to insure that no reversion to a thiostrepton-producing phenotype had occurred.

Protoplast Fusion of Mutant Organisms

Protoplast fusion was carried out exactly as described by

Hopwood et al [67]. Approximately equal numbers of protoplasts of the mutant strains Saz-17 and Ssp31463 (see

Table 3) were mixed in a solution of 50% (w/v) PEG 1000 in P

Buffer (10.3% sucrose, 1.5 mM K^SO*, 15 mM MgSO^, 0.20 mL/L of lOX trace element solution, 0.05% KH^PO^, 0.025 M CaCl^, and

0.025 M MOPS Buffer, pH 7.2, from Hopwood [67]) and incubated

2 min. at room temperature. The protoplasts were spread carefully onto R2YE plates and incubated for 7 days at 30“c.

After incubation, speculating colonies were transferred by sterile toothpicks to R5 plates. For this experiment Saz-17,

Ssp31463, and Sac-1 (wild type S. actuosus) were grown on R5 medium as control cultures. These plates were incubated 7 days at 30°C and then overlaid with 100 nL of a fresh, TSB, overnight culture of S. aureus in 3.5 mL of top R5. After overnight incubation zones of inhibition were scored the next day.

Preparation of Chromosomal DNA

Chromosomal DNA was prepared from all organisms listed in

Table 2, except E. coli strains TBl, DH5a, NM538, NM539, and 63

P2392, the two organisms used as thiopeptide-sensitive test organisms. The Pseudomonas aeruginosa PAOl DNA was a gift of

Darrell Galloway.

Chromosomal DNA was isolated essentially as described by

Hunter [83]. E. coli HBlOl was grown in 400 mL of TSB. The strains of Streptomyces listed in Table 2 were grown in 2 L flasks with 400 mL of YEME, except for S. laurentii which was grown in 400 mL of YMB without casaminoacids added.

Planomonospora parontospora var. antibiotica was grown in 400 mL of TSB. The cells were harvested by centrifugation at

15.000 X g, for 10 minutes at 4°C. The cell pellet was resuspended in 20 mL of TE after which 20 to 25 gm of T4 lysozyme were added. After 20 min. of incubation at 30°C, the cells were lysed by the addition of 2.0 mL of 20% sodium dodecyl sulfate (SDS), followed by the immediate addition of

3.0 mL of 5.0 M NaCl in TE and 25 mL of neutral phenol- chloroform. (Note: neutral phenol-chloroform is a 1:1 w/v mixture of phenol and chloroform into which has been dissolved

5.0 mg of 8-hydroxyquinoline per mL. This mixture is equilibrated with an equal volume of 1.0 M Tris, pH 8 .8 , followed by an equal volume of 0.1 M Tris, pH 8.0. The final solution is stored at 4°C in amber bottles beneath a layer of

0.1 M Tris, pH 8.0). The mixture containing lysed cells, SDS,

NaCl solution, and phenol-chloroform was mixed quickly, divided equally into two 30 mL Oak Ridge tubes, and mixed by inversion for 20 min. The contents of the tubes were then 64 separated into phases by centrifugation for 30 minutes, at

15,000 X g (4°C). The viscous upper layer was pipetted into a fresh Oak Ridge tube, an equal volume of chloroform was added, and the tube mixed by inversion for 10 min. The phases were separated as described above and the upper layer was extracted again with chloroform. After the second chloroform extraction, the DNA was pipetted into a clean glass beaker, an equal volume of isopropanol was added and the DNA was collected by spooling onto Pasteur pipets with heat-sealed tips. The spooled DNA was dried under vacuum in a desiccator at RT and then resuspended in 5.0 mL of TE. RNase A, which had been DNase-inactivated by heat treatment at 50°C for 30 min., was added to the DNA to a final volume of 40 /xg/mL, and the mixture was incubated 90 min. at 50“C. After incubation with the RNase A, Proteinase K was added to the mixture to a final concentration of 200 /xg/mL; NaCl and SDS were also added to final concentrations of 100 mM and 0.4%, respectively. The solution was incubated at 37°C for at least 1 hr. An equal volume of neutral phenol-chloroform was then added, and the purification procedure was repeated from the extraction steps to the spooling step. The DNA was dried and then resuspended in a minimum volume of TE. The concentration was approximated by absorbance at a wavelength of 260 nm, and the concentration of the DNA was adjusted with TE to 2.0 mg/mL to 3.0 mg/mL.

Typical yields of DNA from Streptomyces ranged from 1.0 to 20 mg of DNA per 400 mL of culture. 65

Plasmid DNA Isolation Methods

Plasmid DNA was isolated by three different methods.

High-copy-number plasmids from Streptomyces were initially isolated by the alkaline lysis method described by Hopwood et al [67], but later were isolated by a better method developed by M. Babcock [4], which worked well for both E. coli and

Streptomyces plasmids. Large, low-copy-number plasmids from

Streptomyces were isolated by a method described by Hunter

[83], or by a variation on the alkaline lysis method described by Maniatis et al [117].

A. Alkaline Lvsis Method 1 (from Hopwood et al [67])

The alkaline lysis method was used to isolate plasmids from streptomycete cultures grown in 50 mL of YMB or YEME.

Cells were collected by centrifugation at 12,000 x g, for 10 minutes (4“C), and then were resuspended in 5.0 mL of lysozyme solution ( 2 - 3 mg of T4 lysozyme per mL of a solution containing 0.3 M sucrose, 25 mM EDTA, and 25 mM Tris, pH 8.0) and incubated for 30 min. (37°C). After incubation, cells were lysed by the addition of 2.5 mL of 0.3 M alkaline-SDS

(0.3 N NaOH and 2% SDS, made freshly from 1.0 N and 20% stock solutions, respectively). Immediately upon addition of the alkaline-SDS, the cells were mixed by repeated inversion. The lysed cells were incubated for 35 min. (55°C) and then allowed to cool 5 min., after which the caps of the tubes were tightened and the tubes then were shaken vigorously. Then, to 66 each tube 1.0 mL of acid phenol-chloroform (1:1 phenol:chloroform, w/v, which had not been equilibrated with

Tris) was added, and the tubes were mixed by vortexing for about 1 min. The phases were separated by centrifugation at

10,000 X g for 10 minutes (4°C). The aqueous upper phase was removed to a sterile Oak Ridge or Corex* tube and precipitated by the addition of 0.1 vol. of 3 M unbuffered sodium acetate and an equal volume of isopropanol. The tube was mixed by inversion and allowed to stand 5 min. at RT. The DNA was then pelleted by centrifugation at 10,000 x g for 10 minutes (4°C).

All liquid was removed from the pellet, and the pellet was washed once with 70% ethanol (-20°C) and once with 95% ethanol

(-20°C). The DNA was then vacuum-dried at RT.

If the DNA was to be purified by CsCl centrifugation, it was resuspended in 5.0 mL of TE buffer and passed through a

0.45 /im nitrocellulose filter, after which 250 /xL of 100 mM spermine-HCl was added. The solution was mixed and allowed to stand at room temperature for 5 min., and then was collected by centrifugation as before. The supernatant was decanted and the DNA pellet was resuspended in 3.0 mL of 0.3 M sodium acetate plus 10 mM MgCl^. Seven mL of 95% ethanol was added, the mixture was shaken, and then allowed to stand for 1 hr. at room temperature. The DNA was pelleted as before by centrifugation and redissolved in TE buffer. It was then further purified by centrifugation in a CsCl gradient. 67

B. Isolation of Large. Low Copy StreptomvcBS Plasmids.

1. PEG Precipitation (from Hunter [83])

Mycelia, harvested from 200 mL of YEME broth by centrifugation at 12,000 x g for 10 minutes (4°C), were resuspended in 25 mL of TE buffer, after which 5.0 mL of 0.25

M EDTA and 75 mg of T4 lysozyme were added. The cells were incubated for 20 min. (30”C), and then lysed by the rapid, sequential addition of 25 mL of TE buffer, 15 mL of 0.25 M

EDTA, and 4.2 mL of 20% SDS. This mixture was rapidly stirred, after which 16.8 mL of 5.0 M NaCl in TE buffer was added. The solution was allowed to stand at 4"c for 4 hr. and the phases were then separated by centrifugation at 25,000 x g, for 30 min. (4°C). The volume of the supernatant was measured in a sterile, graduated cylinder and 0.5 vol. of sterile, 30% PEG 8000 was added. The solution was mixed by gentle inversion; the liquid was poured into a 250 mL centrifuge bottle, and the mixture was incubated 8 hr. at 4°C.

The DNA was pelleted by centrifugation at 5,000 x g for 10 min. (4°C) and the supernatant decanted. The pellet was resuspended in TE buffer for purification through CsCl. 68

2. Modified Alkaline Lysis

(modified from Maniatis et al [117])

LB was used to grow E. coli, and YEME was used to grow

the streptomycetes for plasmid extraction. Cells were

collected from 500 mL of medium by centrifugation at 7,000 x

g (E. coli) or 12,000 x g (Streptomyces), for 10 min. (4°C).

The supernatant was decanted, and any remaining liquid was

removed by a Pasteur pipet. The cells were resuspended in 10 mL of Solution I (20% sucrose, 10 mM EDTA, 25 mM Tris, pH

8.0), 50 mg of T4 lysozyme were added, and this mixture was

incubated in Solution I for 30 minutes at 37°C. After

incubation, 10 mL of Solution II (0.2 N NaOH, 1% SDS) were

added, the bottles were mixed by gentle, repeated inversion,

and then were placed on ice for 10 minutes. To this solution,

7.5 mL of ice-cold 5 M potassium acetate (3 M potassium

acetate adjusted with glacial acetic acid to pH 4.5 - 4.8, so

that the final solution is 3 M with respect to potassium and

about 5 M with respect to acetate) was added; the bottles were

covered with Parafilm®, mixed completely by inversion, and

allowed to stand on ice for 10 minutes. Cell debris was then

pelleted by centrifugation at 15,000 x g for 20 minutes (4°C).

The supernatants were recovered into fresh, sterile 250 mL

centrifuge bottles, the volumes were measured, and 0.6 vol. of

isopropanol was added. This was mixed by inversion and

allowed to stand at RT for 30 minutes. The DNA was pelleted

by centrifugation at 15,000 x g for 20 minutes (4°C), and the 69 supernatants were decanted. The pellets were solubilized in

4.5 mL of TE buffer for further purification in a CsCl gradient.

C. Rapid Plasmid Preparation for E. coli and Streptomyces

This method, obtained from Babcock [4], was the most successful small-scale preparation procedure for high-copy- number plasmids in E. coli and Streptomyces. It could also be scaled up to a 50 mL culture without any real problems. The method as described here is for the preparation of plasmid from a 2.0 mL culture.

An overnight culture of E. coli or a 2 to 4 day culture of Streptomyces spp. was collected by centrifugation in a microcentrifuge at 10,000 rpm for 5 min. The cell pellet was resuspended in 100 to 400 nL of MP buffer (10 mM EDTA, 25 mM

Tris, pH 8.0). For Streptomyces spp., lysozyme was added to a final concentration of 2.5 mg/mL to 5.0 mg/mL, and the cells were incubated for 30 min. at 37°C. No lysozyme was needed to lyse E. coli. To the cells 0.4 vol. of a freshly made solution containing 0.3 M NaOH and 2% SDS was added. The solution was mixed vigorously by vortexing to lyse the cells and then incubated for 30 min. (55°C). After incubation, the mixture was allowed to cool slowly to RT, after which, 0.4 vol. of chloroform was added, and the mixture was mixed by vortexing. After that, 0.4 vol. of 5 M potassium acetate

(made as described above) was added, and the solution was 70 again mixed by vortexing. The mixture was separated by centrifugation 5 min., the aqueous phase was removed to a clean tube, and 0.6 vol. of isopropanol was added. Complete mixing followed and the plasmid was pelleted by centrifugation for 5 min. The pellet was washed once with 70% ethanol

(-20°C) and once with 95% ethanol (-20”C). The DNA was dried in a vacuum desiccator at RT, then resuspended in 50 /iL TE buffer.

Cesium Chloride Purification of Plasmids

Plasmid DNA used for cloning and hybridization studies was always purified by ultracentrifugation through a cesium chloride (CsCl) gradient. For E. coli-based vectors, the DNA was resuspended in 4.5 mL of TE buffer, to which 4.5 gm of ultrapure CsCl and 250 /iL of ethidium bromide solution (10 mg/mL) was added. For E. coli-Streptomyces shuttle vectors, the DNA was resuspended in 4.5 mL of TE and 4.95 gm CsCl and

265 /iL of ethidium bromide solution (10 mg/mL) was added. For

Streptomyces-based. vectors, the method of Hopwood et al [67] was followed in which 1.05 gm/mL of CsCl was added to a measured volume of DNA in TE buffer. When the CsCl was dissolved, the volume was again measured, and 50 /iL/mL of ethidium bromide (10 mg/mL) was added. The refractive index of the solution was then adjusted to exactly 1.3925 with TE buffer. 71

All vectors were loaded into Beckman Quick-Seal* ultracentrifuge tubes the plasmids were separated from the chromosomal DNA, protein, and RNA by centrifugation for 20 hours at 55,000 rpm (20°C) using a Beckman VTi 65 or VTi 65.2 rotor. The plasmids were removed from the tubes using an 18- gauge needle. Ethidium bromide was removed from the DNA by repeated extraction with NaCl-saturated isopropanol. The aqueous phase was then loaded into previously boiled dialysis tubing. The DNA was dialyzed for 24 hr. at 4°C with constant stirring against 1,000 to 2,000 vol. of TE buffer with changes every 6 hr. The DNA was then ethanol-acetate precipitated, pelleted by centrifugation, and resuspended in a minimum volume of TE buffer. The concentration was determined by measuring the absorbance at 260 nm using a spectrophotometer.

Preparation of Competent E. coli Cells

The preparation of E, coli cells competent for cloning was essentially as described by Hanahan [59]. Long-term storage of competent cells was in an ice-cold, sterile solution of 30% glycerol (w/v), 50 mM CaCl^, and 10 mM Tris, pH 8.0. The cells were then "flash frozen" in an ethanol-dry ice bath and stored at -70°C until needed.

Preparation of Protoplasts of Streptomyces Strains for Cloning

Preparation of protoplasts followed the procedure described by Hopwood et al [67], with a few modifications. 72

Organisms protoplasted in this study were S. actuosus, S. azureus, S. laurentii, S. lividans TK24, S. species ATCC

31463, and the following S. azureus mutants: Saz-2, Saz-6 ,

Saz-7, Saz-14, Saz-16, Saz-17, and Saz-18.

For protoplast preparations, strains of Streptomyces were always grown from YMB starter cultures. Except for S. laurentii, after the growth of the culture was dense (36 to 72 hours), 5.0 mL of culture broth was transferred to 50 mL of

YEME. For S. laurentii, the transfer was to 50 mL of YMB with no casaminoacids.

The second cultures were grown for 36 to 40 hours with rotary shaking at 250 rpm (30'C), and then cells were harvested by centrifugation. Harvested cells were resuspended in 8.0 mL of lysozyme solution, (a filter-sterilized solution of P Buffer containing 2 to 5 mg/mL of T4 lysozyme). The cells were incubated for 30 to 60 min. at 30°C. The solutions were triturated three times, and the cells were incubated for a further 15 min. The lysozyme solution was diluted with an equal volume of P Buffer and incubated further for 15 min.

The cells were then passed through a sterile 20 mL syringe containing sterile, tightly packed, non-absorbent cotton within the barrel. Nonprotoplasted mycelia remained with the cotton, but protoplasts passed through and were collected in a sterile centrifuge tube. The protoplasts were gently sedimented by centrifugation at 3,000 x g, for 10 min. (RT), and then were gently resuspended in 5.0 to 20 mL fresh P 73

Buffer depending on protoplasting efficiency. Protoplasts were stored and used as described by Hopwood et al [67].

Transformation of E. coli with Plasmids

Competent E. coli cells were usually used immediately after being made competent. However, if frozen competent cells were to be used, they were thawed quickly by incubating them for 2 to 5 min. in a 37°C water bath. Once thawed, competent cells could not be refrozen for later use.

Cells were transformed essentially as described by

Hanahan [59]. After being made competent, the DNA (up to a volume of 50 iXL in TE or diluted ligation mix) was added to

200 |xL of ice-cold competent cells. The transformation procedure was followed as described [59], and the cells were spread on plates of LB media containing appropriate amounts of antibiotic. If pUC19 cloning was being conducted, X-gal and

IPTG in the appropriate amounts were also added to the plates to allow for blue-white screening of transformants. The cells were incubated at 37°C overnight.

For shotgun cloning, it was desirable to add one to three micrograms of DNA from the ligation reaction (depending on the amount used for ligation). For recloning of specific, already-purified plasmids, or for specific subcloning, it was necessary only to add a few tenths of a ng of DNA, and usually no more than 500 ng of DNA. 74

Transformation of Streptomyces Protoplasts

The transformation procedure for Streptomyces spp. was essentially that of Hopwood et al [67]. Frozen protoplasts were quickly thawed by incubation in a 37°C water bath for 2 to 5 min. before beginning the transformation procedure.

Approximately 0.1 mL aliquots were spread onto dried plates of

R2YE. The cells were allowed to soak into the media for about

10 to 20 min., then the plates were incubated at 30°C. After

12 to 20 hr. incubation, the cells were overlaid with 3.5 mL of top R5 media containing the proper amount of the appropriate antibiotic. The cells then continued incubation for another 5 to 7 days to allow time for transformed cells to recover, grow, and sporulate.

Restriction Endonuclease Digestions

Use of restriction enzymes was according to instructions supplied by the manufacturer (GIBCO/BRL). The DNA from digestion reactions which were to be used directly in cloning were precipitated in either ethanol-acetate or isopropanol- acetate, as described previously. Digestion reactions which were to be separated by electrophoresis were first stopped by the addition of loading buffer/tracking dye ("blue juice”; 5 mM EDTA, 0.3 M sucrose, and 0.7 mg/mL broraophenol blue, from

Maniatis et al [117]). Reaction mixtures could be stored at

4°C for up to one year in loading buffer/tracking dye before needing to be discarded. 75

Construction of Standards for Autoradiography

For hybridizations of labelled DNA probes to chromosomal

DNA, it was necessary to know the sizes of chromosome fragments hybridizing to the probes. In some cases, the size standards were A DNA digested with Hindlll. The gel was photographed with a fluorescent ruler (Diversified Biotech,

Newton Centre, MA). When the filter was hybridized to a radioactive probe and exposed to film for autoradiography, a phosphorescent ruler (also from Diversified Biotech) was used to correlate the photo of the gel and size standards with the hybridizing bands found in the autoradiogram.

Autoradiographic standards for the probes to the genes for thiostrepton and nosiheptide resistance, nstiA, and the 8.5 kbp nosiheptide-resistance-encoding DNA fragment were constructed by digestion of plasmids containing the relevant hybridizing DNA (pIJ941 [67], pANT34 and pANT39 [45,109]).

Tables 5 and 6 show the restriction endonucleases used for digestion and the sizes of hybridizing DNA fragment with the probes.

DNA Probes for Hybridizations

The design of the oligonucleotide probes used in this study is discussed in the following chapter. The larger probes were genes known or thought to be involved in the regulation and biosynthesis of thiostrepton or nosiheptide.

These probes included: orf699 for the nshA gene, a putative 76 regulatory upstream orf in the nosiheptide resistance transcript; n s h R for the nosiheptide resistance gene; C1+C2 for the two orfs downstream of nshR; and 8.5K for the probe made from the entire 8.5 kbp BamHI piece of DNA containing the above-mentioned ORPs plus other, as yet unsequenced DNA; and tsrR for the thiostrepton resistance gene. Hereafter, these probes will be referred to as probe 1, probe 2, probe 3, probe

4, and probe 5, respectively.

Table 5; Autoradiography Standards for DNA Hybridization to Thiostrepton Resistance

Plasmid Restriction Enzyme Size of DNA Digested® Digestion Hybridizing

PIJ941 Xhol, PstI 10.9 kbp

PIJ941 PstI, Bglll 9.2 kbp

pIJ941 EcoRV, Bglll 6.1 kbp, 18.8 kbp^

pIJ941 Sstl 3.9 kbp

pIJ941 SstI, EcoRV 1.5 kbp, 2.5 kbp

“This set of autoradiography standards was developed by D.C. Dosch [45,109].

^This band showed up only at medium (e.g. 0.4X SBC, 0.1% SDS, 42°C) or low stringencies of washing. At high stringency (e.g. 0.2X SBC, 0.1% SDS, 55°C), the band did not appear. 77

The probes were prepared as follows: probe 5 was a probe constructed from the whole 1,790 kbp cloned thiostrepton resistance gene fragment [165,166]; probes 1, 2, and 3 were generated by nested exonuclease III deletions during the course of sequencing these genes [111]. Probe 4 was simply the entire 8.5 kbp BamHI-digested DNA fragment originally cloned from S. actuosus which conferred resistance to nosiheptide. Plasmids containing the DNA used for probes were cloned into E. coli (DH5a or TBl), grown in large-scale quantities, prepared using standard plasmid preparation

Table 6 : Autoradiography Standards for nshR, nshA, and the 8.5 kbp NshR DNA Cloned from S. actuosus

Size of DNA Hybridizing (kbp)“ Plasmid Restriction Enzyme Digested Digestion to nshR to nshA

pANT34 EcoRV 12.8 12.8

PANT34 BamHI 8.5 8.5

PANT34 Mlul 8.9, 3.9 8.9, 3.9

PANT34 BamHI, PstI 2.6 2.6

PANT34 SphI, PvuII 1.1 1.1

PANT39 Ndel 5.0 5.0

“These bands appeared also when the entire 8.5 kbp NshR fragment [45] was used as an autoradiographic probe. 78 methods, purified on CsCl gradients, and digested with restriction enzymes to cut out the desired DNA fragment.

Isolation of Specific DNA Fragments

The isolation of specific DNA fragments was done according to the method of Danner [37]. DNA was separated by electrophoresis through TBE agarose (0.7% to 1.0% w/v).

Lambda DNA digested with HindiII was loaded into wells flanking the DNA of interest to serve as a size standard. DNA was stained in the gel with a solution of aqueous ethidium bromide (1.0 /xg/mL), briefly destained in distilled water, and then observed under UV light. The DNA of interest was excised with a razor blade as quickly as possible to avoid prolonged exposure to the UV light and thus avoid excessive DNA damage.

Electroelution and purification of the isolated DNA fragment from the agarose gel then proceeded using the method of Danner

[37], with no modifications.

DNA Ligations with Plasmids

Chromosomal DNA and plasmid DNA were digested with the appropriate restriction endonuclease (or endonucleases), ethanol precipitated, dried, and mixed in water in a ratio of ca. 3:1 of chromosomal DNA:plasmid DNA. These DNA fragments were not treated with alkaline phosphatase. Ligations were carried out normally in a total volume of 50 /xL; the DNA in water comprised 34 /xL of the volume. To the DNA, 5 nL of lOX 79 ligation buffer was mixed. This was then treated at 55°C for

15 minutes followed immediately by cooling on ice for 15 minutes. After cooling, 5 /nL of 100 mM DTT and 5 fxL ATP were added and mixed. Ligation was initiated by the addition and immediate mixing of l.O /iL of T4 DNA Ligase when "sticky ends"

(single-stranded, overlapping, complementary DNA at the DNA's digested ends) existed and 4.0 /xL of T4 DNA Ligase when the ends were "blunt". All of the solution was collected at the bottom of the tube by a brief spin in the microcentrifuge, after which the ligation reaction mixture was incubuted at

16°C or RT for 4 to 16 hours. For subsequent transformation of Streptomyces, the ligation mixture was ethanol precipitated, dried, and resuspended in TE. For transformation of E. coli the ligation mixture was diluted by the addition of 4 volumes of TE buffer, and transformation performed with 50 /iL of this (per 200 /xL of cells).

Radioactive Labelling of DNA

A. Oligonucleotide Purification and Labelling

Oligonucleotides were synthesized by the Biochemical

Instrument Center of the Ohio State University.

Oligonucleotides were radioactively labelled with crude, y-3=p-ATp (icN Biochemicals, Irvine, CA; 7,000 Ci/mmol). The labelling procedure employed was that of Maniatis et al [117].

Normally, 0.2 to 0.3 /xg of oligonucleotide DNA was used in the reaction. After the T4 kinase reaction had been completed. 80 the reaction mixture was diluted with an equal volume of STE

(100 mM NaCl, 10 mM Tris, pH 8.0, 1 mM EDTA) and passed through a spun column twice by centrifugation.

B. Radioactive Labelling of Large DNA Fragments

Large fragments of DNA (defined by this author as stretches of DNA longer than 200 bp) in this work were labelled by using the method of random oligonucleotide labelling as described by Feinberg and Vogelstein [48].

Though kits for execution of this procedure are available, all buffers and other materials were purchased separately and made in the lab. The DNA used was isolated specifically from restriction endonuclease digested plasmids which carried the

DNA of interest. The DNA was digested in such a way that only the fragment of interest would be excised for isolation from an agarose gel.

C . Spun Columns

In order to separate unincorporated nucleotides from labelled DNA probes, the labelling mixture was passed through a Sephadex G-50 column by centrifugation. These spun columns were made exactly as described by Maniatis et al [117]. The columns were stored at 4°C, standing upright, with the tops covered with Parafilm® and the bottom tips capped by a plastic cover which came with the syringes. 81

Southern Hybridization

Southern Hybridization procedures were those described by

Southern [157], as modified by Maniatis et al [117]. A few modifications of significance were employed here. For oligonucleotide hybridizations, the most important modification was that the agarose concentration of the gel used in electrophoretic separation of the DNA was at least

1.0%. Also, the gels were made in TPE buffer and separated by electrophoresis through TPE buffer. (For DNA hybridizations using large fragments of DNA, the agarose concentrations of the gels varied between 0.5% and 1.0% agarose. The buffer for the gels and for electrophoresis was THE.) For both oligonucleotides and larger DNA fragments, gels were heavily stained in ethidium bromide, lightly destained in water, and exposed to UV light for at least two minutes. Probe-specific

DNA autoradiography standards or bacteriophage X DNA which had been digested with HindiII were used always for the electrophoresis of the DNA (see previous discussion).

For hybridization, different buffers were used according to the type of probe used. For large DNA fragments, the hybridization buffer was as follows: 50% formaraide, 5X SSC,

0.5% SDS, 5X Denhardt's Solution [117], and 100 /itg/mL sheared, single-stranded DNA from salmon sperm. When oligonucleotides were used as probes, SSPE was substituted for SSC at all steps as described by Vinci [170]. The buffer was also modified to the following composition: 25% formamide, 6X SSPE, 0.5% SDS, 82

5X Denhardt's Solution, 200 nq/mL sheared, single-stranded DNA from salmon sperm.

Hybridization of bound DNA on nitrocellulose filters was carried out as described by Maniatis et al [117] using the

Omni-Blot* system of American Bionetics (purchased via Fisher

Scientific). Hybridization buffer with the salmon sperm DNA, but sans probe, was added to the filter to 150 to 175 /iL buffer per cm^ of filter. The filters were prehybridized at

42°C for 2 to 8 hr. for large DNA probes and 8 to 12 hr. for oligonucleotide probes. The probes were then added, and hybridization under the same conditions continued for another

8 to 12 hr.

After hybridization, the filter-bound DNA was washed twice, at RT, using a buffer composed of 2X SSC and 0.1% SDS, for 30 min. each wash, then washed twice at a concentration of salt and level of temperature calculated to create a specific level of stringency as described [117,153,170]. When oligonucleotides were used as probes, all washes were at 2X

SSPE, 0.1% SDS, and the temperature was varied to increase stringency [170].

Hybridizations of DNA Probes to

Colonies of Recombinant E. coli

DNA cloned into E. coli, using the plasmid pUC19, was screened using a method developed by V. Steigerwald [159].

After transformation of E. coli with pUC19 plus the DNA of 83 interest for cloning, the transformed E. coli (growing on LB plates having the appropriate amounts of ampicillin, IPTG, and

X-gal added) were visually inspected for blue and white colonies. White colonies were presumed to contain inserts.

All white colonies were picked over to plates of LB, with 100 pg/mL of ampicillin added, using sterile toothpicks. The colonies were picked to a gridded plate of LB with 50 to 120 colonies per plate. Each plate was triplicated and all were incubated at 37°C overnight. One plate of each triplicate was designated the master plate and stored at 4°C. The other two plates were chilled one hour at 4°C. Nitrocellulose circles

(82 cm diameter) were aligned with some orienting mark on the plates, laid over the colonies, and left for five minutes.

The circles were then placed colony side up on Whatman 3MM paper soaked in 1% SDS, 5 mM EDTA and left there five minutes.

The circles were then transferred to Whatman 3MM paper soaked in denaturing solution (0.5 M NaOH, 1.5 M NaCl) and left five minutes. Afterwards, the filters were transferred to Whatman

3MM paper soaked in neutralizing solution (1.0 M Tris, pH 7.9,

1.5 M NaCl), left for five minutes, and then transferred to

3MM paper soaked in 2X SSPE. The filters were air-dried on

3MM paper for one hour, then baked in vacuo for two hours at

80”c. Filters were prehybridized in hybridization buffer at

42°C, overnight, then hybridized to the appropriate probe overnight at 42°C. 84

Autoradiography and Scintillation Counting

Autoradiography was performed by placing filters in X-ray cassettes and exposing them to Kodak X-OMAT film for an empirically determined length of time. Film was developed using Kodak GBX* Developer and Fixer according to the instructions in literature supplied by the manufacturer.

All scintillation counting was done, using various protocols, described later, with the scintillation cocktail

Scintiverse™ E. Efficiency of counting ^^P-labelled samples was assumed to be 0.90 based on previous analyses by

Steigerwald [159]. For ^^S-labelled material, the efficiency of counting was measured by using a known standard of since and have comparable energies of their B- particles and are, therefore, counted in the same window of the scintillation counter. The efficiency was calculated by dividing the counts recorded by the known counts in the standard.

Fermentation of S. azureus

A 50 mL culture of S. azureus in DPM3, started from 1«10^ spores, was grown for 48 hr as described previously. The entire contents were transferred to 500 mL of the same medium, and this secondary seed culture was grown exactly 24 hr under the same conditions. The entire contents were then transferred to 11 L of DPM3 medium (containing 5.0 mL of the antifoam agent MAZU) in a 14 L fermentor. Sterile air was 85 pumped through the culture at the rate of 1.0 L of air/L of media/min. The contents of the fermentor were stirred at 300 rpm, and the incubation temperature was 30°C. At the initial time point (To), two 50 mL samples were taken. One sample was for dry weight analysis, and one sample was for DNA analysis.

At defined time points from 0 hour through 120 hours, samples were taken of the culture. One sample of 30 mL volume was for

DNA analysis. A second, 20 mL sample was for dry weight analysis. The DNA isolation and analysis was conducted as described previously. Dry weight analysis of cells was conducted as described by Dekleva et al [41], with the exception that ddH^O was substituted for saline in the washing of cells.

Isolation of Hiah-Purity Thiostrepton

In the preliminary ^®S-L-cysteine-labelling experiment designed to study the method of synthesis of thiopeptide antibiotics, a method for the isolation of high-purity thiostrepton, obtained from Mocek and Floss [128], was used to isolate thiostrepton.

S. azureus was grown in DPMI for seven days as described previously, after which an equal volume of CHCI3 was added to the culture flask. The flask was returned to the rotary incubator and shaken for 30 minutes; the contents of the flask were poured into a centrifuge bottle, and the phases were separated by centrifugation at 10,000 x g, for 20 min. (0°C). 86

The organic layer was removed from the aqueous layer, and the organic phase was taken to dryness in a rotary evaporator

(35°C to 40°C). The dried material was dissolved in a minimum volume of tetrahydrofuran (THF). The THF solution was removed from the rotary evaporator flask to a pre-weighed, glass

Corex* centrifuge tube with a screw-top lid. An equal volume of hexane was added, and the mixture was kept at 4°C for 12 hr. to precipitate thiostrepton. The antibiotic was collected by centrifugation as before, the supernatant was decanted, and the pellet was dried by a slow stream of air (under the hood).

The tube containing the precipitate was weighed to determine the yield of thiostrepton. The pellet was then dissolved in a minimal volume of CHCI3 or THF. The solution contained quite pure thiostrepton.

Isolation of Hiah-Purity Nosiheptide

In the preliminary ==S-L-cysteine-labelling experiment designed to study the method of synthesis of thiopeptide antibiotics, a method for the isolation of high-purity nosiheptide, described by Houck [81] and modified by Mocek and

Floss [128], was used to isolate nosiheptide.

The cells were harvested by centrifugation at 10,000 x g, for 20 min. (0°C). The supernatant was completely removed, and the cells were resuspended in 0.025 vol. (relative to the total volume of the medium into which the cells were inoculated) of phosphate-citrate buffer. (Phosphate-citrate 87 buffer is a sterile buffer consisting of 1.0 M citrate and 0.5

M phosphate at pH 6.0.) Also, 0.85 vol. THF was added, and the mixture was shaken vigorously for 2 min. To this, 0.15 vol. of mixed hexanes was added, and the bottle was shaken again.

The phases were separated as before, after which the separated phases were carefully poured into a separatory funnel, the organic phase was collected in a round-bottom, 500 mL flask. The aqueous phase was extracted again using a mixture of .4 vol. THF and 0.4 vol. mixed hexanes, the phases were separated by centrifugation as before, and the organic phases combined. The combined organic phase was then dried overnight (4°C) by the addition of 0.25 by weight (wt.; relative to the original fermentation volume) of anhydrous

MgSO^, after which the MgSO* was filtered from the THF extract. The extract was then taken to dryness on the rotary evaporator at <35°C. The dried residue was dissolved in .10 vol. of THF, and placed in a centrifuge tube. An equal volume of hexane was added, and the mixture was kept at

4°C for 12 hr. to precipitate the nosiheptide, which was then collected by centrifugation as before. The supernatant was decanted into a round-bottom flask and taken to dryness by rotary evaporation as before, and the remaining nosiheptide was precipitated as before using a mixture of .025 vol. THF and 0.025 vol. mixed hexanes. The precipitates were combined and dried under a steady stream of air under the fume hood. 88

The crude nosiheptide was dissolved in a minimal amount of CH^Cl2 :ethanol (4:1) and filtered to remove solid impurities. The filtrate was placed into a pre-weighed centrifuge tube and 2 vol. of peroxide-free, diethyl ether

(relative to the volume of the CHzClgZethanol) were added.

The mixture was kept at 4°C for 4 hr. to induce precipitation of nosiheptide, and the antibiotic was collected by centrifugation as before. Nosiheptide was dried as before, and the yield was determined.

Quick Extractions of Thiostrepton and Nosiheptide

For the second ^®S-L-cysteine-labelling experiment, and for all routine extractions of thiopeptide antibiotics, this procedure was employed. In the ==S-L-cysteine labelling experiment, the medium DPM3 was used, while in routine extractions, the crude production medium (Medium G) was used.

Thiostrepton and nosiheptide were extracted directly from the culture media by the addition to the culture flask of an equal volume of a solution of 95:5 CHCl3 :methanol. The entire contents of the flasks were then transferred to a centrifuge bottle and shaken until emulsified. Alternatively, the solution was added in an equal amount directly to the flask, which was then returned to the incubator and shaken at 250 rpm for 30 min. (30°C). The phases were then separated by centrifugation at 10,000 x g, for 20 min. (0°C). The organic phase was removed and dried by rotary evaporation at 35°C to 89

40 ”c. The residue was then redissolved in a minimal volume of

CHCI3. Although several UV fluorescing materials were extracted from the cells in the crude production medium, almost the only visible or UV fluorescing compounds which extracted from the defined media (DPMI and DPM3) were the antibiotics themselves. Two poorly fluorescing, orange-colored compounds were always extracted from S. actuosus with nosiheptide.

Reduction of the Specific Activitv of =^S-L-cysteine

The specific activity of ==S-L-cysteine was reduced by the addition of 1.9 mL of a nonradioactive solution of 0.121 mg/mL L-cysteine and 10 mM DTT to 0.101 mL of the labelled cysteine. For the first labelling experiment performed, this reduced the specific activity of the labelled cysteine to ca.

500 mCi/mmol. For the second labelling experiment, the specific activity was reduced to ca. 600 mCi/mmol.

Protocol for the First Synthesis

Study of Thiopeptide Antibiotics

Two organisms, S. actuosus and S. azureus, were tested.

Three flasks of 100 mL DPMI were inoculated with 10 mL of cells from a starter culture (from 1*10^ spores for S. azureus, and from an isolated, fully sporulated colony on R5 for s. actuosus) of DPMI. A total of six flasks were inoculated. The 100 mL cultures were grown for exactly 73 90 hours as described above. To one flask in each set of three

(the control flask) 100 /xL of 95% ethanol was added. To a second flask 100 /xL of 25 mg/mL chloramphenicol freshly dissolved in 95% ethanol was added. To the third flask 100 /xL of 125 mg/mL chloramphenicol freshly dissolved in 95% ethanol was added. The flasks were shaken for exactly 1 hour from this point, and then 25 /xCi of 500 mCi/mmol ^®S-L-cysteine was added to each flask. Exactly 23 hours from the addition of the cysteine, all the cultures were extracted. From each flask, 10 mL of cells were taken for dry weights; three 1.0 mL samples were taken from each flask, and each sample was separately added to a tube of ice-cold 10 % (w/v) trichloroacetic acid (TCA); the remaining 87 mL was worked up for antibiotic using the procedure for obtaining high-purity material. THF was added to each sample of antibiotic recovered so that the final concentration of antibiotic in each sample was 100 /xg/mL. Using thin-layer chromatography,

2.0 mg of each sample was separated with the solvent system

CHCI3 :n-heptane:isopropanol (5:5:3). The remainder was redissolved in Scintiverse™ E and counted by scintillation.

A. Samples extracted with TCA

The samples in TCA were mixed by vortexing and then adsorbed onto glass fiber filters on a vacuum manifold. Before adsorbing a sample onto a filter, the filter was washed with

5.0 mL of ice-cold TCA. After the contents of each tube of 91 cell material were placed onto the filters, the sample tubes were each rinsed with 5.0 mL of ice-cold TCA, and each rinse was passed over the appropriate filter. The filters were placed into scintillation vials and baked dry at 80°C for 12 hr. Each vial was cooled to RT, and 10 mL of Scintiverse™ E was added. The filters were counted in a scintillation counter to determine incorporated amounts of

B. Thin-Laver Chromatoaraphv of Samples

The types of plates used for thin-layer chromatography have been described earlier in this chapter. Suitable solvents for thin-layer chromatography are CHCI3 :n-heptane: methanol (6:3:1), CHCl3 :n-heptane:methanol (5:5:1), and

CHCl3 :n-heptane:isopropanol (5:5:3). The latter was the one used most often by this author. For radiochromatography, the solvent system CHCI3 :n-heptane:isopropanol (5:5:3) was used exclusively. After chromatography of the samples, the plates were air-dried for at least 4 hours, then sprayed with

ENTRANCE from New England Nuclear (to increase the signal from the ^^S) and exposed to film at -70°C for an empirically determined length of time.

Protocol for the Second Synthesis

Study of Thiopeptide Antibiotics

For the second study of incorporation of ^^S-L-cysteine into thiopeptide antibiotics, the thiostrepton-producing 92 organism, S. laurentii, was analyzed in addition to S. actuosus (nosiheptide), and S. azureus (thiostrepton). For each organism, a 50 mL starter culture was grown in DPM3 for

48 hours, as described above. The S. azureus culture was started from 1»10^ spores, while the S. laurentii and S. actuosus cultures were started from isolated, fully sporulated colonies. At 48 hours incubation, two 20 mL aliquots of each starter culture were inoculated separately into 300 mL of DPM3 medium in 1.0 L flasks. At specified time points, beginning with 0 hours, and continuing ca. every 6 hr. through the 84th hour, two 1.0 mL aliquots of each large flask were removed for total protein analysis (a substitute for dry weight analysis).

At exactly 48 hr of incubation, 100 nL of 125 mg/mL chloramphenicol (freshly dissolved in 95% ethanol) was added to one of the flasks and 100 /iL of 95% ethanol was added to the other flask. At exactly 49 hours of incubation, 150 /iCi of ^^S-L-cysteine (~600 mCi/mmol) was added to every flask.

Beginning at 55 hours after inoculation, and in addition to the aliquots taken for total protein analysis, a 20 mL aliquot of each culture was taken, ca. every 6 hours, for antibiotic extraction. Also, two 1.0 mL aliquots were taken and mixed separately with 1.0 mL of ice-cold 10% TCA to measure the total amount of incorporated into cells. The experiment was stopped at 83 hours.

The antibiotics were extracted using the quick extraction method, and subjected to scintillation counting in optically 93 clear 1 dram vials with 4 mL of Scintiverse™ E added.

Initially, ca. 80% of the total sample extracted was dried in these vials and then counted, but the counts were too high for accurate correction of potential sample quenching of signal

(quench correction). The remaining 20% of sample (in CHCI3) was split in half, and half was counted (with the exception of

S. laurentii of which only 0.1 to 0.2 of the remaining sample was counted) and corrected for quench. The remaining half was committed to thin-layer chromatography analysis as described above. The TCA-mixed samples were processed and counted as described above. The counts here were also too high to conduct quench correction.

A. Total Protein Analvsis

Coomassie stain was made as follows: 100 mg of Coomassie

G-250” was dissolved in 50 mL of 95% ethanol, and the resulting deep blue solution was mixed with 150 mL of 85% phosphoric acid (H3PO4 ) (causing the solution to turn brown); this was then diluted with 850 mL of ddH^O and stored in an amber bottle at 4°C until needed.

Aliquots of cells taken for total protein analysis were placed in microcentrifuge tubes and pelleted by centrifugation for 5 minutes at 10,000 x g (RT). The supernatant was removed; the cells were resuspended in 500 juL of 0.1 N NaOH and boiled for 10 minutes in a dry heating block set at 110°C. 94

After boiling, the cells were allowed to cool slightly, then

500 /zL of 0.1 N HCl was added to neutralize the solution.

For analysis, a Beckman Spectronic 2000 spectrophotometer was set to read absorbance at 595 nm (Asgs). To 980 /zL of

Coomassie stain in 12 mm X 75 mm glass tubes was added 20 /zL of the lysed cells, BSA of a known concentration (from Pierce

Chemical Co.), or a 1:10 dilution of either cells or standards. This was briefly mixed by vortexing and allowed to react. The color of the solution became blue with increasing amounts of protein, and the change in color was measured spectrophotometrically.

The spectrophotometer was zeroed against Coomassie stain, then the standards were analyzed. After this, the samples from the cells were analyzed. The reference cuvette was never removed from the holder; rather the stain was drawn out with a Pasteur pipette. In this way, the baseline would not have to be reset. When the protein concentration in the cells got too high to read, the cells were diluted 1:10 in ddH^O. The concentration of protein in mg/mL was calculated from the absorbance, based on curves drawn from the standard protein concentrations analyzed.

B. Quench Correction of Counts in the Antibiotics

Since the extracts contain both the desired antibiotics and some additional, colored material as described previously, and also since the antibiotics were resuspended in CHCla, 95 which can quench counts in scintillation cocktail [30], it was necessary to establish to what degree the sample counts were being quenched. A known amount of the antibiotic extract in

CHCl3 was placed into an optically clear 1 dram vial, 4.0 mL of Scintiverse™ E was added, the samples were counted. After counting, an exact amount of a known standard of ^"^C toluene was added to every sample, and the samples were counted again.

Counter efficiency was also determined as described previously. The difference between recorded counts and known counts added to the sample was taken as the amount of quench in that sample. The total radioactivity in the sample was then calculated by correcting both for quench and for counter efficiency, converting to disintegrations per minute (dpm), and then converting to microcuries (pCi; 10“® Curies; 2.2*10® dpm) or nanocuries (nCi; 10“® Curies; 2.2*10® dpm). CHAPTER V: RESULTS AND DISCUSSION

Mutants of Thiopeptide Producers

To understand the genetics of thiopeptide antibiotic synthesis, and to elucidate various features of the biosynthetic pathway, it is essential to have mutants which are blocked in the biosynthesis of those antibiotics.

Mutations other than those in the structural genes of the biosynthetic pathway can affect production of antibiotics

(with either beneficial or detrimental effects), as was described in the literature review chapter. Thus, non- sporulating mutants were discarded to avoid isolating pleiotropic mutants in which the DNA damage had been in and around the area of genes involved in global regulation.

In this work, mutagenesis of the thiopeptide producers using the mutagenic agents EMS, nitrous acid, and MNNG was attempted. In preliminary experiments, EMS had a very poor kill rate under the conditions used (data not shown), so it was no longer used as a mutagen. Nitrous acid, on the other hand, had average kill rates in S, azureus and S. actuosus of

99.987% and 99.998%, respectively; in both cases, 5* 10"^ spores were used. Several mutant strains were generated by this procedure, including one mutant (designated Sac 1.1.3) which

96 97 appeared to overproduce nosiheptide dramatically, both in overlay assays and in thin-layer chromatography studies of extracts. However, none of these mutants was stable. The overproducing mutant. Sac 1.1.3, eventually reverted to normal production levels, and the nonproducing mutants eventually reverted to production of nosiheptide. Thus, no stable mutants of either S. azureus or 5. actuosus were ever recovered using nitrous acid mutagenesis.

With MNNG as a mutagenic agent, the same strategy was employed which was utilized earlier, (i.e., only sporulating auxotrophs and sporulating non-producers of thiostrepton were sought. The decision was also made at this time to focus solely on S. azureus.

Approximately 1.8*10® freshly prepared S. azureus spores obtained from minimal medium (Medium D) were mutagenized with

MNNG. After mutagenesis, the surviving spores were titered and a kill percentage obtained using the following formula:

Equation 1

100 • I 1 - I numberimber of suisurviving spores | j number of total spores

Roughly 75,000 of the mutagenized spores survived the MNNG treatment, giving a kill percentage of 99.96%.

Over the course of several months of intense effort, a total of 5,376 mutagenized spores were analyzed for 98 thiostrepton formation, bald mutants, auxotrophs, etc. (see

Table 7).

Two of the mutants, Saz-13 and Saz-14 are spontaneous mutants derived from Saz-12; they are phenotypically dissimilar. Interestingly, Saz-14, which no longer produced thiostrepton, produces the peptide antibiotic, lavendomycin

[99], an antibiotic which S. azureus is not known to produce normally [8 8 ]. Saz-2, Saz-6 , Saz-16, and Saz-19 through

Saz-23 are mutants derived from the preliminary work of Dr.

Jung Lee. Saz-16 is a spontaneous mutant derived from Saz-2.

Mutants Saz-3, Saz-4, and Saz-5 reverted to thiostrepton production and were removed from the mutant collection.

Table 7: Analysis of S. azureus Mutational Frequency Using NTG as the Mutagen.

Total Number of Spores Screened: 5,376

Number of Candidate Auxotrophs Found: 23 Percentage of Total Spores Screened: 0.43% Number of Stable Auxotrophs Found: 0

Number of Bald Mutants in Spores: 1,008 Percentage of Total Spores Screened: 18.75%

Number of Candidate Blocked Mutants Found : 71 Percentage of Total Spores Screened: 1.32% Number of Stable Blocked Mutants Found: 10 Percentage of Total Spores Screened: 0.19% 99

Of all the mutants isolated, only two, Saz-2, and its derivative Saz-16, showed sensitivity to thiostrepton; all other blocked mutants were thiostrepton-resistant. As required by the screening protocol, all the blocked mutants isolated also sporulated; Saz-7, Saz-12, and Saz-13, however, sporulated poorly. Saz-2 sporulated well, but its spontaneous derivative, Saz-16, sporulated poorly. Saz-6 , obtained from

J. J. Lee, did not sporulate; however, Saz-17, a spontaneous derivative of Saz-6 , sporulated well.

The difficulty of isolating stable, thiopeptide antibiotic-nonproducing mutants and auxotrophic mutants in S, actuosus was emphasized in recent work by a Chinese group collaborating with this laboratory [143]. Twelve S. actuosus mutants which were supposed to be either auxotrophic or nosiheptide-nonproducing strains produced using either EMS or acridine orange and carefully screened, were sent to this lab for further analysis. Unfortunately, every one of those mutants reverted to apparent wild-type phenotype with respect to growth on minimal media and production of nosiheptide.

Southern Hybridization Analysis of Thiopeptide Producers and

Thiostrepton Non-Producina Mutants

One approach used to analyze the mutants was to determine the sizes of the chromosomal restriction fragments which hybridized to various available probes. The nomenclature of these probes and their construction are described in the 100 methods section. Figure 4 shows a restriction map of probes

1, 2, 3, and 4. Probe 5, the thiostrepton resistance gene from S. azureus, is a 1,790 bp BamHI-digested DNA fragment taken from the plasmid pIJ30 [165,166], given to this laboratory by D.A. Hopwood of the John Innes Institute.

Additional probes were used in attempts to dissect the biosynthetic pathways of thiopeptide antibiotic biosynthesis.

Three probes were designed to hybridize to DNA coding for the consensus binding site of 4'-phosphopantetheine [57,58,154].

The rationale for the construction of these probes was that if a peptide synthetase was involved in thiopeptide antibiotic biosynthesis, then one or more DNA fragments other than fatty acid synthetase would hybridize to at least one of these probes. Two probes were designed to hybridize to DNA which could code for the putative peptide precursors of these antibiotics. The last probe was an oligonucleotide probe designed to the consensus active site of Streptomyces serine proteases [170]. This probe was used as a control on the quality of the oligonucleotide labelling and hybridization.

All of these oligonucleotide probes were designed to minimize dilution of the radioactive signal from the probes by using codons which were unique in their coding, if at all possible. Heterogeneity of the probes was further reduced by the use of the known codon bias of Streptomyces genes [9,162].

The sequence of each probe and the purpose for which it was used are shown in Table 8 . Figure 4: Restriction map of the 8.5 kbp BamHI-digested DNA fragment which confers resistance to nosiheptide, and a closer view of the sequenced region (from the left-hand BamHI to the PvuII site immediately to the right of orfC). Note also the strong stem- loop region between nshA and nshR. The 2,326 bp sequenced region from the BamHI to the PstI has been published elsewhere [109]. Probe 4 is the entire 8.5 kbp BamHI-digested DNA fragment. Restriction endonuclease abbreviations are as follows: Ba, BamHI; Kp, Kpnl; Ml, Mlul; Pv , PvuII; Ps, PstI; Sc, S ad; Sa, Sail; Sm, Smal; Sp, SphI (From [110]).

101 nshA Transcription termination

nshR Transcription y initiation Box A

...(iAC(iAlXA(îUCCUüA(’A(XXÎCAVC(ÎC(Xi(: (IKiCacaCAUÜACXJCjAGCXX* ...I) IJ Q A * nshR SOObp

n s h A n s h R

Ml ScSa Pv Sa Sm SmSm 1kb

Pv Ps SpKp Pv Pv Ml Sa X = = b < i i = n s h A n s h R orfB orfC

H PROBE 1 PROBE 2 PROBE 3 O Figure 4 to 103

1. Hybridization to Large Gene Probes

DNA was isolated from the thiostrepton producers, S. azureus and S. laurentii, digested with BamHI, separated by agarose gel electrophoresis, transferred to nitrocellulose by

Southern blotting, and hybridized to probe 1, isolated from pANT401. DNA isolated from S. lividans TK24 and S. coelicolor

A3(2) were used concurrently as "negative controls" (i.e., strains which were not expected to have DNA that hybridized to nshA, since the initial hypothesis regarding nshA was that it was a gene specific to the biosynthesis, or regulation, of the thiopeptide antibiotics). Surprisingly, the DNA from these two "control organisms" also hybridized at moderate stringency to the probe (data not shown). Therefore, DNA isolated from several additional streptomycetes and digested with BamHI were tested, along with DNA from the plasmids pWHM333 and pIJ2315.

Though the plasmid DNA did not hybridize to the probe, all of the chromosomal DNA preparations did hybridize (data not shown). At this point, the DNA from numerous strains of

Streptomyces was isolated as described in the previous chapter. As controls, the DNA from E. coli HBlOl and

Pseudomonas aeruginosa PAOl were used. All preparations of

DNA were digested with BamHI, electrophoresed in a ERL Model

H4 gel apparatus, and transferred to nitrocellulose by the method of Southern [157]. Probe 1 was then hybridized to the nitrocellulose-bound DNA at a temperature of 35°C. The washes 104

Table 8 ; Sequences of the Oligonucleotide Probes Used in this Study.

Oligo Name Sequence Purpose

Gly-Asp-Ser-Gly-Gly-Pro-Leu Consensus serine Protease® 5 ' -GGC-GAC-TCC-GGC-GGC-CCG-CT-3 ' protease active G C site. Used for as control probe in hybridization.

Cys-Thr-Thr-Cys-Glu-Cys-Cys Putative nosihep NOSIl 5'-TGC-ACC-ACC-TGC-GAG-TGC“TGC-3' tide precursor G G sequence.

Cys-Thr-Thr-Cys-Ile-Cys-Thr-Cys Putative thio TSRl 5'-T6C-ACC-ACC-TGC-ATC-TGC-ACC-TGC-3' strepton precur G G G sor sequence.

Asp-Asp-Leu-Gly-Het-Asp-Ser-Leu-Asp Initial peptide ACPI" 5'-GAC-GAC-CTG-GGG-ATG-GAC-TCG-CTG-GAC-3' sequence for 4'- C C C C phosphopantetheine binding site.

-Het- Refined peptide Glu-Asp-Leu-Gly-M-Asp“Ser-Leu-Ala-Leu-Val-Glu sequence for 4'- Gln-Asp-Leu-Gly-Het-Asp-Ser-Ile-Asp-Leu-Leu-Glu phosphopantetheine 4PASr 5'-GAG-GAC-CTG-GGG-ATG-GAC-TCC-CTC-GCC-CTG-ATG-GAG-3' binding site. C G C C A A CG First of two con C sensus probes.

-Het- Refined peptide Glu-Asp-Leu-Gly-Tyr-Asp-Ser-Leu-Ala-Leu-Val-Glu sequence for 4'- Gln-Asp-Leu-Gly-Tyr-Asp-Ser-Ile-Asp-Leu-Leu-Glu phosphopantetheine 4PAS2° 5'-GAG-GAC-CTG-GGG-TAC-GAC-TCC-CTG-GCC-CTG-ATG-GAG-3' binding site. CGCC AACG Second of two con C sensus probes.

“Probe from V. Vinci [170].

‘’Consensus sequence from [58].

^Consensus sequence as modified by Sherman et al [154] 105 following the two washes of 2X SSC, 0.1% SDS varied in stringency. All washes were in 0.5X SSC, 0.2% SDS; stringency was varied by increasing temperature. The wash temperatures were at 35°C, 58“c, 68°C, and 80°C. After each wash, the filter was placed on a plastic backing and wrapped in Saran

Wrap® to keep wet and autoradiographed as described. The percentage of mismatch was calculated as described elsewhere

[153] using Equation 2.

Equation 2

T„ = 81 + 16.6(logioCi) + 0.4[%(G+C)] - 0.6(%formamide) - 600/n - 1.5 (%mismatch)

where 0 ^. is the concentration of salt in solution and n is

the length of the hybrid in base pairs. For Streptomyces

strains, the G+C content is assumed to be 73% [113].

Although this equation is normally used to calculate the temperature of hybridization needed to allow for a specified amount of mismatch to the probe in the hybridization buffer, it can also be used to calculate the mismatch allowed by specific wash conditions. This is accomplished through a rearrangement of Equation 2, as shown in Equation 3. The hybridization at 35°C allows a percent mismatch of ca. 28%

(the probe would hybridize to DNA which is as little as 72% homologous). 106

Using Equation 3 to calculate the stringency allowed by the washes (the percentage of formamide was 0 , since the wash buffers consisted only of SSC and SDS), the percent mismatches allowed were as follows: 39% at the wash temperature of 35°C,

23% at the wash temperature of 58°C, 17% at the wash temperature of 6 8 ”C, and 8 .8% at the wash temperature of 80°C

(Fig. 5).

Equation 3

%mismatch = I 81 + 16.6(logioCi) + 0.4[%G+C)] 1 I I - 0.6(%formamide) - 600/n - T„ II 1.5

where is the concentration of salt in solution and n is

the length of the hybrid in base pairs. For Streptomyces

strains, the G+c content is assumed to be 73% [113].

As can be seen in Figure 5, DNA from E. coli and P. aeruginosa do not hybridize to the probe. The DNA isolated from S. alboniger was apparently too degraded to specifically hybridize to the probe. Table 9 gives a complete list of organisms which hybridize to the probe at high stringency, along with the approximate sizes of the hybridizing DNA fragment or fragments as cut with BamHI. It should be noted that the data for Planomonospora parontaspora subsp. antibiotica, S. hawaiiensis, S. glaucogriseus, S. hawaiiensls,

S. sioyaensis, and S. species ATCC 31463 are not shown in

Figure 5. Figure 5; Autoradiogram of chromosomal DNA from various strains of Streptomyces which hybridize to probe l inshA). All DNA shown here has been digested with BamHI, separated by agarose gel electrophoresis, and blotted to nitrocellulose, as described in the chapter on methodology. From left-to-right, top row, the lanes are as follows: lane 1 , size standard (see Table 6 ); lane 2, Escherichia coli HBlOl; lane 3, Pseudomonas aeruginosa PAOl; lane 4, S. actuosus; leine 5, S. alboniger (badly degraded); lane 6 , S. albus; lane 7, S. antibioticus} lane 8 , S. aureofaciens; lane 9, S. azureusî lane 10, S. bambergiensis; lane 11, S. C5; lane 12, S. calif ornicus', lane 13, S. coelicolor; Icine 14, S. coerulorubidusî lane 15, 5. galileus ATCC 31133; lane 16, S. galileus ATCC 31671; lane 17, S. griseoruber; lane 18, S. griseus IMRU 3499; lane 19, S. griseus NRKL B-2682; lane 20, size standard (Table 6 ). From left-to-right, bottom row, the lanes are as follows: lane 1 , blank; lane 2 , size standard (Table 6 ); lane 3, S. insignis; leine 4, S. laurentii', lane 5, S. lividans TK24; lane 6 , S. longisporus; lane 7, S. lucensis', lane 8 , S. matensis; lane 9, S. nogalater; lane 10, S, pactum; lane 11, S. parvulus; lane 12, S. peucetius; lane 13, S. puniceus; lane 14, S. purpurescens; lane 15, S. rimosus; lane 16, S. steffisburgensis; lane 17, S. viridochromogenes; lane 18, size standard (Table 6 ); and lanes 19 & 20, blank. The filter was washed twice at 6 8 °c in a buffer of 0.5X SSC, 0.2% SDS, exposed to Kodak X-OMAT film for 12 hours at -70°C, then developed using Kodak GBX“ Developer and Fixer as described in literature supplied by the manufacturer.

107 108

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

2.8-

r i

1.1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

1.1

Figure 5 109

Table 9: Streptomyces Strains Hybridizing to Probe 1 (xishA)

Hybridizing Organism Hybridizing DNA Fragment Size*

E coli HBlOl None^

P. aeruginosa PAOl None*

S . actuosus 8.5 kbp

S. antibioticus 11.9 kbp

S. aureofaciens 12.5 kbp, 5.7 kbp

S. azureus 10.0 kbp

S. bambergiensis 1.8 kbp

S. C5 11.0 kbp

S. californicus 7.7 kbp°

S. coelicolor A3(2) 12.0 kbp

S. coerulorubidus None^

S. galileus ATCC 31133 10.5 kbp

S. galileus ATCC 31671 10.5 kbp

S. griseoruber 5.9 kbp

S. griseus IMRU 3499 8.1 kbp

S. griseus NRRL B-2686 8.1 kbp

S. insignis None^

S. laurentii 2.6 kbp^

S . lincolnensis 5.3 kbp, 4.3 kbp

S. lividans TK24 10.0 kbp 110

Table 9 (continued)

Hybridizing Organism Hybridizing DNA Fragment Size“

S. longisporus 10.0 kbp

S. lucensis 2.9 kbp, 1.9 kbp

S . matensis 7.2 kbp=

S. nogalater None‘=

S. pactum None^

S. parvulus 11.5 kbp, 5.1 kbp

S. peucetius 5.1 kbp, 3,8 kbp

S. puniceus 6.6 kbp

S. purpurescens 7.2 kbp

S. rimosus NOne^

S. steffisburgensis 3.2 kbp

S. viridochromogenes 3.8 kbp

Planomonospora paronta None^ spora antibiotica

S. glaucogriseus 8.9 kbp

S. hawaiiensis 6.3 kbp

S . sioyaensis 6.8 kbp°

S. species ATCC 31463 7.5 kbp, 7.0 kbp

“DNA hybridizing after washes which allow for no more than 17% mismatch. ^None means no DNA fragment was found to hybridize even at the lowest stringency of washing used (0.5X SSC, 0.2% SDS, RT). “A very faint band seen after 12 hours of autoradiographic exposure to film. *^A faint band of 4.5 kbp was seen at the lowest level of stringency after exposure of the film. I l l

The strong conservation (29 hybridizing strains of

Streptomyces out of 34 strains of Streptomyces tested) among

Streptomyces spp. of DNA hybridizing to probe 1 suggests that this DNA is perhaps a generally important gene for

Streptomyces. It also suggests that the DNA sequence is not well-conserved outside of the genus Streptomyces, since neither the Planomonospora strain nor the E. coli and

Pseudomonas aeruginosa controls hybridized to the probe.

DNA hybridization experiments were conducted also using probes 2, 3, 4, and 5. These DNA probes did not hybridize to any non-thiopeptide producer. Probe 3 hybridized only to the

DNA from the strains S. actuosus, S. glaucogriseus (quite weakly), and S. species ATCC 31463 (data not shown). These data suggest that the DNA cloned from S. actuosus is related to the biosynthesis of nosiheptide, and more generally to the biosynthesis of the antibiotics of the thiostrepton family.

Probes 5 (tsrR) and 2 (nshR), at both medium and high levels of stringency, (0.4X SSC, 0.1% SDS, 42°C; and 0.2X SSC,

0.1% SDS, 55°C, respectively) hybridized only to DNA from streptomycete thiopeptide producers. Weak hybridization was seen at medium stringency between probe 2 and Planomonospora parontaspora subsp. antibiotica. It was observed that of the thiopeptide producers screened, only S. actuosus, S. azureus, and S. glaucogriseus appeared to have multiple fragments of

DNA which hybridized to both probes. S. actuosus has been reported [45] to have a fragment of about 3 kbp which 112 hybridizes to probe 5, but in this work, such hybridization, though seen occasionally, was inconsistent, even using the reported conditions. However, a 3.0 kbp fragment of DNA did consistently hybridize using medium stringency wash conditions when probe 4 was used. A fragment of ca. 1.8 kbp also hybridized to probe 4 when S. actuosus was used; however, this hybridization was both weak and inconsistent. The data in

Table 10 give the size of fragments of DNA hybridizing to either probe 2 or probe 5.

Careful study was made of the two fragments of DNA in S. azureus which hybridized to both tsrjR (probe 5) and nshR

(probe 2). Though the data is not shown, it was found consistently that the larger of the two fragments, the 6.1 kbp

BamHI fragment, hybridized strongly to the tsr probe at both medium and high stringencies. The 4.5 kbp fragment hybridized almost not at all at high stringency (two washes with 0.2X

SSC, 0.1% SDS, 55°C). Autoradiographic exposures of more than

48 hours were required to see the fragment even faintly. The hybridizing DNA fragment could be seen well, however, after washing at medium stringency (two washes at 0.4X SSC, 0.1%

SDS, 4 2 ° C ). Interestingly, when probe 2 and medium stringency washing were used, the two fragments hybridized to the probe at about equal intensity following overnight autoradiographic exposure. Following a high stringency wash using probe 2, both DNA fragments were found to hybridize much less strongly to probe 2, but the 4.5 kbp fragment clearly hybridized better 113

Table 10: Size of tsr- and nshR-hybridizing DNA Fragments in Thiopeptide Producers

Hybridizing Strain Size of BamHI Fragments(kbp)

P. parontaspora subsp. antibiotica 7.5“

S. species ATCC 31463 6.0*='“

S . actuosus 8.5, 2.8“ , 1.8“

S . azureus 6.1, 4.5, 1.5

S . glaucogriseus 8.5“ , 6.0, 4.0“ , 2.0“

S . hawaiiensis 6.1

S. laurentii 2.6“

S. sioyaensis 7.5“

“An Actinomycete, not a Streptomycete. Hybridization only attempted to tsr under medium stringency washing (42“c, 0.4X SSC, 0.1% SDS) conditions.

^Non-producing mutant of nosiheptide. Hybridization only attempted to tsr under medium stringency washing (42°c, 0.4X SSC, 0.1% SDS) conditions.

“Hybridization of this band of DNA was weak using medium stringency washing conditions but was consistent with the probes used.

“Hybridization of this band of DNA was weak and inconsistent using medium stringency washing conditions. 114 than the 6.1 kbp fragment (data not shown). In addition to the finding of separate fragments hybridizing to probes 2 and

5 when the S. azureus chromosomal DNA was digested with BamHI, two separate fragments hybridizing to these probes were seen using several other restriction endonucleases. These data were generated in the course of generating a restriction endonuclease map of the DNA surrounding the resistance genes and the nshA-like gene in S. azureus (discussed below). It appears that two DNA sequences exist in S. azureus which may be rRNA methylase genes encoding resistance to thiopeptide antibiotics. With respect to conservation of sequence, one fragment of DNA appears to be more "thiostrepton-like" and the other more "nosiheptide-like." It is also interesting that the 6.0 kbp BamHI-digested DNA fragment from the nosiheptide producer, S. glaucogriseus, hybridizes more intensely to probe

5 after an overnight autoradiographic exposure following a medium stringency wash than do DNA fragments from the thiostrepton producers, S. hawaiiensis and S. laurentii, under the same conditions.

Using probes 1, 2, 4, and 5 and single and multiple restriction endonuclease digests of S. azureus DNA, restriction enzyme maps were established of the DNA surrounding the nshA-like gene and of the DNA surrounding sequences hybridizing to genes potentially encoding thiopeptide resistance methylases. Examples of the kinds of mapping data generated with S. azureus are shown in Figure 6 . Figure 6 : Autoradiograms generated by southern hybridization of S. azureus DNA to four DNA probes which hybridize to DNA putatively associated with thiopeptide antibiotic biosynthesis. 6A; S. azureus DNA digested with different restriction enzymes and Southern Blotted to probe 5 (tsrR). (Notez in this blot, the standards lanes do not appear, but they do appear upon longer exposure.) 6B; same as 6A but using probe 2 (nshR) cloned from S. actuosus. 6C: same as 6A but using probe 1. 6D: same as 6A but using probe 4 (the 8.5 kbp BamHI fragment). The outside lanes of each gel are size standard lanes. For the tsrR probe, the size standard used is from Table 5. For the other probes, the size standard used is from Table 6 . For the autoradiogram in 6A, the filter was washed twice at 55°C in a buffer of 0.2X SSC, 0.1% SDS. For the other three autoradiograms, the filters were washed twice in a buffer of 0.4X SSC, 0.1% SDS at 42°C. All four filters were exposed to Kodak X-OMAT film at -70“C for four hours and then developed using Kodak GBX® Developer and Fixer following the instructions supplied by the manufacturer. The lanes of restriction endonuclease-digested S. azureus DNA on each filter are, from left-to-right, as follows: BamHI, Mlul, SphI, PstI, SstI, PvuII, Bglll, Pstl+Bglll, SphZ+Pvull, BamHI+MIuI, Clal. The lane immediately to the left of the right-hand size-standard is a 6.1 kbp fragment of DNA randomly cloned from S. azureus and digested with BamHI. Note that this lane acts as a "negative control" since it does not hybridize to any of the probes.

115 0^ BamHI ^ Mlul

f 0 1 sphi

9 5stl PvuII B g r i l l

Pstl+Bqlll 0 Sphl+Pvull 0 BamHI+MluI 0 Clal BamHI

5 I BamHI (D • • 0 Mlul

0 t gpAl PstI SstI PvuII Bgrlil bd Pstl+Bglll Sphl+PvuII BamHI+Mlul Clal

BamHI

911 N> w r-] 00 w 4 o\ vo VI M H" 1 i * t f B I BamKI 0\ « Mlul 0 • t t SphI 1 # « e * : H- P'sti I Oi § : ; # $ ssti PvuII

n Bglll M l i . psti+Bcjiii 0 Sphl+PvuII g ^ 0 BamHI+MluI

Clal BamHI

Bamül Mlul

SphI PstI SstI

PvuII Bglll Pstl+BglII Sphl+PvuII # ^ BamRI+MluI

Clal

BamKI

LIT » fit tf 118

The restriction enzyme fragment maps generated by these data are shown in Figure 7. In stark contrast to the nshA- nshR structure found in S. actuosus, there is no close, physical linkage of the nshA-like gene of S. azureus to the tsrR and nsJiR-like genes. These genes may be linked on larger

DNA fragments, but they are at least 20 kbp apart according to the restriction endonuclease mapping data currently available.

When probe 4 was used to study the S. azureus chromosomal

DNA, it was found that, in addition to DNA fragments which hybridized to probes 1, 2, 3, and 5, many new hybridizing fragments of DNA appeared. Presumably, these new hybridizing

DNA fragments come from the ca. 5 kbp of DNA which lies to the right of the sequenced region of the cloned 8.5 kbp BamHI DNA fragment from S. actuosus. Since the genes encoding the enzymes for biosynthesis in all known antibiotic biosynthetic pathways found in Streptomyces to date occur in large, single clusters on the chromosome [24,27], this region of DNA from S. actuosus, and the S. azureus DNA to which it hybridizes, may contain genes encoding thiopeptide biosynthesis. This, however, is still speculative until further analyses of the

DNA and its functions are conducted. The data do suggest strongly that the arrangement of the putative biosynthesis genes encoded by the DNA comprising probe 4 is sharply different between S. actuosus and S. azureus. Figure 7; Restriction maps of S. azureus DNA surrounding the gene encoding the nshA-like protein (A) and the resistance genes of thiostrepton (B). The DNA in these two regions may contain genes which encode proteins involved in thiopeptide antibiotic biosynthesis and regulation. The data were generated from Southern hybridization data such as that seen in Figure 6 . This data was generated from 24 different Southern hybridization analyses, using 112 different exposure times of the radioactive nitrocellulose filters to X-ray film. Restriction endonuclease abbreviations are as follows; Ba, BamHI; Eg, Bglll; Cl, Clal; Ml, Mlul; Pv, PvuII; Ps, PstI; Sc, SacI; Sp, Sphl.

119 Sc Kjl 1^1 Sc Ml Ps Bg & Ba rn ID n s h A - l i k e

A. Region of DNA Surrounding the nshA-like Gene

I 1 1 . 0 k b p

p> Pv Sc Sc Ml Pv Cl P* Sc Sp Pv Ba Ba Sp Pv Ml Pv Bapv r I'y I u u u u u n s h R - l i k e tsrR

B. Region of DNA Surrounding the Two Thiopeptide Antibiotic Resistance Genes

Figure 7

H to o 121

2. Oligonucleotide Probe Analysis

The protease oligonucleotide was used to determine the accuracy of hybridization and washing conditions. It was considered essential as a control for conditions, but the data generated merely show that sequences of DNA which bind to the consensus serine protease active site of Streptomyces exist in the thiopeptide antibiotic producing organisms and in the

Streptomyces spp. used as controls.

Since peptide synthetases have been shown to contain 4'- phosphopantetheine prosthetic groups similar to those of fatty acid synthetases [57,58,154] oligonucleotide probes were constructed based on the amino acid sequence of the highly conserved 4'-phosphopantetheine binding site, using the known codon bias of streptomycetes [9,154,162]. Of the three probes used to probe for these sequences (see Table 8 ), only probe

4PAS1 hybridized to DNA isolated from the thiopeptide producers under any conditions. The probe consistently hybridized to DNA fragments isolated from the thiopeptide producing strains up to a temperature of 55“c (the highest temperature tested).

Using Equation 2, described above, the T„ of 4PAS1 in the washing buffer (2X SSPE, 0.1% SDS) is calculated to be ca.

83°C. At 55°C, six mismatches are allowed; thus, it is unclear upon simple inspection why the oligonucleotide probe

4PAS2, which is quite similar, does not hybridize also. Table

11 shows the sizes of the BamHI-digested DNA fragments from 122 the thiopeptide antibiotic producers that hybridize to oligonucleotide 4PAS1. Figure 8 shows an autoradiogram of

DNA, including DNA from strains of Streptomyces which do not produce thiopeptide antibiotics, in which the oligonucleotide hybridizes to the nitrocellulose-bound DNA.

It is assumed that at least one of the hybridizing fragments is that of the 4 ' -phosphopantetheine binding site of the acyl carrier protein of fatty acid synthetase [57,58,154],

Other cellular processes also use the 4'-phosphopantetheine binding site, likely including the synthesis of the thiopeptide antibiotics, which explains the presence of other hybridizing DNA fragments in the lanes.

One of the fragments of DNA hybridizing in S. actuosus to

4PAS1 was ca. 9 kbp. Thus, it was hypothesized that DNA encoding a 4 '-phosphopantetheine binding site to which 4PAS1 hybridized existed on the cloned 8.5 kbp DNA from S. actuosus.

Comparative hybridization of 4PAS1 to both S. actuosus chromosomal DNA and pANT34 DNA (containing the 8.5 kbp BamHI

DNA fragment in pBR322) indicated that the oligonucleotide did not hybridize to the cloned 8.5 kbp BamHI DNA fragment.

Because the autoradiographic bands from the hybridizing

DNA fragments in all organisms tested were both light in intensity and numerous, it was not possible to confidently state which of the bands seen in the autoradiogram corresponded to a DNA fragment potentially containing the thiopeptide synthetase. Therefore, no attempt was made to Figure 8 : Autoradiogram of Streptomyces chromosomal DNA hybridizing to the oligonucleotide probe 4PAS1, a probe to the consensus sequence of the 4'- phosphopantetheine binding Site in Streptomyces [154]. From left-to-right, the DNA samples in the lanes are as follows: bacteriophage A DNA cut with Hindlll; E. coli HBlOl cut with BamHI; S. actuosus cut with BamHI; S. azureus cut with BamHI; S. coelicolor cut with BamHI; S. glaucogriseus cut with BamHI; S. griseus IMRU 3499 cut with BamHI; S. hawaiiensis cut with BamHI; S. laurentii with BamHI ; S. lividans TK24 cut with BamHI; S. species ATCC 31463 cut with BamHI; and bacteriophage A DNA cut with Hindlll. The filters were treated, hybridized to the oligonucleotide probe, and washed at 55°C in 2X SSPE, 0.1% SDS as described in Chapter 4. The filter was exposed to film for 16 days at -70 °C then developed as described previously.

123 ■ W E. coli HBlOl

. m S. actuosus

- lite S. azureus S. coelicolor

2 S. glaucogriseus I S. griseus 00 S. hawaiiensis S. laurentii

S. lividans s g a ATCC 31463

X m:. «

t-> ro 125 clone a DNA fragment using DNA isolated from one of the thiopeptide antibiotic-producing organisms based on this hybridization data. This data may be useful in future

"reverse genetics" experiments, however, once the peptide synthetase for the antibiotic, if it in fact exists, is isolated.

3. Analvsis of S. azureus Mutants.

The structures of the DNA around the tsrR, nshR-like, and nshA-like genes of the S. azureus mutants used in this work were extensively analyzed by restriction enzyme mapping analysis as described in the previous section. DNA from

Table 11; Size of BamHI-digested DNA Fragments from Thiopeptide Antibiotic Producers Hybridizing to Oligonucleotide Probe 4PAS1

Hybridizing Strain Approximate Size of DNA Fragments™

S. actuosus 9.4, 6.3, 5.0, 3.0

S . azureus 9.4, 6.0, 5.5, 2.5

S. glaucogriseus 7.0, 4.4

S. hawaiiensis 9.4, 4.0, 1,8

S. laurentii 6.6

S. species 12.0, 8.5 ATCC 31463

“Sizes of DNA fragments which hybridized to either probe 2 or probe 5 are given in kbp. 126 mutant Saz-16 did not hybridize to probes 1 through 5, which contain DNA thought to be involved in thiopeptide antibiotic regulation, resistance or synthesis (data not shown). The progenitor strain of Saz-16, Saz-2, has a BamHI-digested DNA fragment which hybridizes to the tsrR probe (probe 5); however, hybridization of a second, BamHI-digested, nshR-like

DNA fragment was not observed using either probe 2 or probe 5.

Furthermore, when probe 4 is used to study DNA from Saz-2, an extra fragment of DNA which hybridizes to this probe is seen.

Moreover, all the sizes of the fragments of DNA which hybridize to these probes are shifted with respect to the DNA from the wild-type strain. Since Saz-2 contains significant alterations in the DNA which hybridizes to these probes, and since Saz-16 is a spontaneous derivative of Saz-2 which contains no DNA fragments hybridizing to these probes, this may indicate that there is gross linkage in S. azureus between the thiostrepton resistance methylase DNA genes and the nshA- equivalent gene.

In the analysis of the other mutants for which DNA suitable for Southern hybridization could be prepared, interesting observations were made. Most of the strains exhibited a change in the restriction fragment length of

BamHI-digested DNA which hybridized to either probe 1 (the nshA probe) or probe 4 (the cloned, 8.5 kbp BamHI DNA fragment). Several of the other strains also exhibited changes in restriction fragment length size of other DNA 127 fragments from the unsequenced portion of the cloned 8.5 kbp

BamHI DNA fragment from S. actuosus. In a couple of cases, the hybridizing fragments were completely missing. Figure 9 shows an example of the hybridization of S. azureus wild-type and mutant DNA to probe 4. Table 12 contains a summary of the various restriction endonuclease-digested DNA fragments of S. azureus which hybridize to probe 4. Table 13 describes some of the various DNA fragments from Saz mutants which hybridize to probe 1 .

There are two explanations for shifts in the mobility of these DNA fragments. The first explanation is that individual, specific restriction endonuclease sites are mutated by the MNNG and, hence, lost. The second explanation is that deletions are occurring due either to the action of

MNNG or to some instability in the DNA on which these DNA fragments reside. Since so many S. azureus mutants studied exhibit these apparent mobility shifts, it seems unlikely that the specific loss of restriction endonuclease sites is the correct explanation for the changes in the DNA fragments. The more likely explanation would seem to be that deletions of DNA linked to DNA fragments which hybridize to probe 4 is occurring, leading to the mobility shifts.

That deletions should occur when MNNG is the mutagenic agent is not surprising, since MNNG is one of the most potent alkylating agents known. Its method of mutagenesis is to act at the replication fork of DNA to produce 7-methylguanine, and 128

Table 12: DNA Fragments Hybridizing to Probe 4 from Selected Saz Mutants, Using Various Restriction Endonuclease Digestions*

Sizes of Hybridizing DNA Fragments (kbp) Saz Strain BamHI Mlul # 1 MI SstI PvuII Sail Saz-l** 10.0, 6.1, 10.7, 7.8, 9.5, 7.3, 13.0, 2.3 10.4, 9.2, 7.0, 4.7, 5.0, 1.8, 2.3, 0.8 4.8, 2.0 5.5, 2.6 6.1, 5.0, 1.7 0.8, 0.5 0.5

Saz-2 îlO.5, 8.0, None® ND® ND 7.9 7.5, 1.4 1.8 1 6.5, 2.6

Saz-6 jlO.l, 6.1, 7.8, 4.8, ND ND 9.4, 5.0 4.7, 1.7 5.0, 1.8, 0.8 2.0 0.8, 0.5

Saz-8 11.0, 6.1, 11.1, 4.8, 8.1, 6.7, 9.7, 8.9 10.2, 9.2, 7.0, 4.7, 1.8, 0.8, 2.3, 0.8 2.0 5.6, 2.2 4.6, 0.5 1.7 0.5

Saz-9 6.1, 2.3, 4.8, 2.0 ND ND 9.2, 4.6, 7.0, 4.7, 1.8, 0.8, 0.8 0.5 1.7 0.5

Saz-10 11.0, 6.1, 10.9, 8.1, ND ND 10.2, 9.2, 7.0, 4.7, 5.0, 1.8, 2.3, 0.8 4.8, 2.0 4.6, 0.5 1.7 0.8, 0.5

Saz-12 8.1, 6.1, 7.8, 4.8, 9.4, 6.9, ND 9.2, 4.6 4.7, 1.7 5.0, 1.8, 0.8 2.0 3.6, 1.9 0.8, 0.5

Saz-17 10.3, 6.1, 10.9, 4.3, ND ND 10.2, 9.2, 7.0, 4.7, 5.0, 1.8, 2.3, 0.8 2.3 4.6, 0.5 1.7 0.8, 0.5

Saz-18 11.2, 6.1, 11.0, 7.8, 10.6, 9.2 ND 10.2, 9.2, 7.0, 4.7, 5.0, 1.8, 2.3, 0.8 4.3, 2.0 7.9, 6.5, 4.4, 0.5 1.7 0.8, 0.5 1.7

Saz-19 10.5, 5.9, 11.0, 4.3, 9.8, 7.9, 9.9, 9.3 9.2, 4.4 7.0, 4.7, 5.0, 1.8 2.1, 0.8 2.0 6.3, 1.7 1.7

“Fragments hybridizing at medium stringency. The buffer for the two stringent washes was 0.4X SSC, 0.1% SDS, 42°C, 30 min. each wash.

^jild-type strain

°In some exposures, DNA fragments of 8.1 kbp and 5.3 kbp appeared to hybridize, but in most exposures, no Hlul-hybridizing DNA fragments appeared. 129

Table 13: DNA Fragments Hybridizing to Probe 1 from Selected Saz Mutants, Using Various Restriction Endonuclease Digests*

Sizes of Hybridizing DNA Fragments (kbp) Saz strain BamHI Hlul # 1 MI SstI Saz-1" 10.0 4.6, 0.8 9.7, 4.8 9.7 3.8, 3.0

Saz-8 11.0 4.6, 0.8 11.0, 4.6 9.7, 8.9 3.8, 3.0

Saz-12 8.0, 6.3 4.6, 0.8 11.0, 4.6 ND= HD

Saz-17 10.0 4.6, 0.8 ND ND ND

Saz-18 11.2 4.6 10.5, 4.6 ND ND

Saz-19 10.5 4.4 8.5, 4.6 9.3 3.5

“Fragments hybridizing at medium stringency. The buffer for the two stringent washes was 0.4X SSC, 0.1% SDS, 42°C, 30 min. each wash.

"wild-type strain

°HD = Not Determined 130 it frequently causes multiple mutations. Furthermore, MNNG has been known to induce deletions in some DNA [14]. It is also not surprising that deletions should occur, since it is well known that many strains of Streptomyces carry highly unstable regions of DNA [26,27,61,64,65,79,152].

Many, though by no means all, of these unstable regions also happen to be regions of DNA where antibiotic biosynthesis genes are clustered [26,27,51,61,64,65,66,79,152]. However, in a G:C rich organism, such as S. azureus, it might be expected that the use of MNNG to induce multiple mutations by the production of 7-methylguanine at the replication fork [14] would cause higher-than-normal losses of DNA sites for restriction endonucleases. The mutants isolated in this study were apparently normal except for antibiotic production. They could grow on minimal medium. All mutants sporulated to some degree, and most mutants sporulated normally. Almost all mutants also demonstrated normal color, colony size and shape, and other characteristics reminiscent of the wild-type S. azureus. It appears, therefore, that other characteristics of the mutants are quite similar to wild-type; however, this can be stated with certainty only when probes specific for genes responsible for these other characteristics become available.

At present, these other probes are not available. Through past experience, and with the presently available evidence, deletions appear to be favored as the explanation for the Figure 9: Hybridization of BainHI-digested DNA from S. azureus and Saz mutants to the 8.5 kbp DNA probe (probe 4) cloned from S. actuosus, encoding thiopeptide antibiotic resistance DNA, and putative thiopeptide antibiotic regulatory and biosynthesis DNA. From left-to-right, the lanes are as follows; size standard (Table 6 ), Saz-1 (wild- type), Saz-2, Saz-6 , Saz-8 , Saz-9, Saz-10, Saz-12, Saz-13, Saz-15, Saz-17, Saz-18, Saz-19, size standard (Table 6 ). Isolation, restriction endonuclease digestion, and Southern hybridization of these samples of DNA were conducted as described in the chapter on methodology. The stringency of the hybridization and washing was medium (0.4X SBC, 0.1% SDS, 42°C, for the washes). Exposure of the filter to film was for 8 hours at -70°C. The lanes containing the autoradiographic standards were slightly overloaded.

131 size std.

Saz-J

Saz-2

Saz —6

Saz-8 Saz-9 *4 H- Saz-10

I(D Saz-12 VO Saz-14 Saz-15 Saz-J .7

Saz - J B

Saz-1 9

size std.

w to 133 mobility shift aberrations seen in these mutants; nevertheless, no probable explanation can be yet either completely ruled out or confidently stated.

Some of the mutants were analyzed for hybridization to the 4 '-pantetheine binding site consensus oligonucleotide. In this experiment, all the mutants analyzed, including Saz-17, were missing the largest of the DNA fragments to which 4PAS1 hybridized using wild-type S. azureus DNA (data not shown).

This by no means proves that this missing DNA encodes the putative thiopeptide synthetase for thiostrepton; it simply shows that some deletion may have occurred in this region.

Transformation of S. azureus Mutants

Protoplasts of the following S. azureus mutants were made as described in the methods section: Saz-2, Saz-6 , Saz-12,

Saz-14, Saz-17, Saz-18, and Saz-19. The plasmid pANT402 was introduced by transformation into these protoplasts. Plasmid pANT402 (Table 4; methods section) contains the entire 8.5 kbp

DNA fragment cloned from S. actuosus and encoding the genes nshA, nshR, and possibly other thiopeptide antibiotic biosynthesis genes.

Transformants containing pANT402 were obtained for all protoplasts except Saz-14 and Saz-17, which appear to be naturally resistant to apramycin even though the wild-type S. azureus strain from which Saz-14 and Saz-17 are derived is sensitive to the antibiotic. The transformants obtained were 134 cultured on R5 medium for 5 days, then overlaid with the thiostrepton-sensitive organism, Staphylococcus aureus, to assay for transformants which had recovered the ability to produce an inhibitory substance. No transformants of Saz-2,

Saz-6 , Saz-12, Saz-16, Saz-18, and Saz-19 were found which could produce any inhibitory substance. Furthermore, thin- layer chromatography analysis of extracts from representative transformants grown in production medium (Medium G) showed no difference between transformed and non-transformed strains.

This may indicate that the aberrations of DNA seen in these mutants result in the loss of a greater number of functional genes related to biosynthesis of thiopeptide antibiotics than can be compensated for by the functional genes encoding proteins involved in reactions in thiopeptide antibiotic biosynthesis which are contained in the 8.5 kbp DNA fragment cloned from S. actuosus.

Protoplast Fusion of Two Thiopeptide Antibiotic Nonproducinq

Mutants and the Resulting Restoration of Production

Both Saz-17 and Ssp31463 are mutants incapable producing thiopeptide antibiotics. The latter strain was submitted to the ATCC as a strain of Streptomyces that produced nosiheptide; however, stringent examination of this strain indicated that it did not produce nosiheptide (data not shown). Saz-17 is a thiostrepton non-producing mutant which shows no apparent deletions in DNA thought highly likely to 135 contain genes encoding enzymes for thiostrepton biosynthesis.

There is a possible deletion in Saz-17 of a fragment of DNA hybridizing to oligonucleotide probe 4PAS1, but nothing is known about the missing fragment. Therefore, Saz-17 may be a point mutation.

Seiya Ogata and coworkers [135] used repeated rounds of protoplast fusion (with its attendant genetic recombination) among various mutant strains of S. azureus to generate a strain of S. azureus which produced large quantities of thiostrepton. It was hypothesized that protoplast fusion between Saz-17 and S. species ATCC 31463 might yield an organism producing thiostrepton, nosiheptide, or a novel thiopeptide antibiotic. These two strains were protoplasted and fused as described in the methods section. The number of regenerated protoplasts after this procedure was too numerous to count. Of these, 315 isolated colonies were picked over to plates of R5 medium (9 candidate colonies per plate). The organisms Saz-17, Ssp31463, and wild-type S. actuosus were also picked over to these plates for a total of 12 colonies per plate. All regenerated protoplasts were also picked over to a master plate of R2YE media. After five days of incubation, the test plates were overlaid with the thiopeptide-antibiotic-sensitive strain Staphylococcus aureus and assayed for inhibition of the test organism. As expected,

Saz-17 did not inhibit the S. aureus. Four of 35 colonies of

Ssp31463 showed inhibition zones of less than 1.0 mm in 136 diameter, and the remaining 31 showed no zones of inhibition.

All colonies of S. actuosus showed zones of inhibition, with the average zone being 4.3 mm in diameter.

Of the fusion candidates tested, 32 (ca. 10%) showed the production of zones of inhibition. The zones of inhibition ranged from 1.2 mm to 3.5 mm in diameter; the average diameter of the zones was 2.4 mm. The organisms producing an inhibitory substance were recorded and the master plates stored for future analysis.

Discovery of Plasmid-Borne Thiostrepton Resistance in

Streptomyces azureus

A. Introduction

Numerous attempts were made to clone 3 to 10 kbp DNA fragments (see Figure 7) contiguous to, and containing, the tsrR and nsJiR-like genes of S. azureus. Attempts also were made to clone DNA from S. laurentii, S. hawaiiensis, and S. glaucogrxseus to which probes 2 (nshR) and 5 (tsrR) hybridized

(presumably, these hybridizing DNA fragments encoded thiopeptide antibiotic resistance methylases). Additionally, attempts were made to clone the homolog of the S. actuosus nshA gene from S. azureus. A summary of the major attempts to clone S. azureus DNA encoding resistance genes, associated biosynthesis genes, or regulatory genes associated with thiopeptide antibiotics is found in Table 14. 137

Table 14: Summary of Experimental Approaches Used to Clone Thiopeptide Antibiotic Resistance Genes, Thiostrepton Biosynthesis Genes, or Associated (nshA-like) DNA.

Experiient Experinental Donor DNA and Vector Selection/Screening^ Result M e r Goal Cloning Kethod DE Host Strain Method 1= clone large D E S, azureus, S. lau ÀEHBL4 E. coli screening of recom No clones frapents around rentii Sau3AI par NH53B, NH539 binant 1 D E using found which tsr/nsbS genes tial digests and and P2392 probes 2 and 5. hybridized isolation of DNA to probes. frapents in 9-23 kbp size range 2*^ clone frapents of S. azureus, S. pANT29 S. lividans selection of thi(f, Recloned 8.5 D E containing, & actuosus, or S. TK24, Saz-2 neo^ transformants. kbp nstf D E contipous with, laurentii digested from S. act resistance methyl- with SamHI and uosus. Clonei ases. "shotgun" cloned. plJlOl-like 1 plasmids from all strains.

]C clone 3.0 kbp D E S. actuosus D E PANT29 S. lividans selection of thi(f, plJlOl-like encoding resistanc 1 digested with TK24, Saz-2 neif transformants plasmids iso methylase from S a M Q and selected lated in two S. actuosus for 2-4 kbp frag experiments; ments of DNA. no transform ants in four experiments.

4^ clone 3.0 kbp DNA S, actuosus DNA pKC505 S. lividans selection of thio'^, No clones encoding resistanc digested with TK24 apif transformants recovered in methylase from SamHI and selected four attempts S. actuosus for 2-4 kbp frag ments of D E .

5« clone DNA frapentî S. azureus DNA pK5G5 S. lividans selection of thio"^, plJlOl-like encoding resistanc partially digeste TK24 apr'- transformants plasmids wit methylases in with Sau3AI and thief selec S, azureus "shotgun" cloned. tion. No transformants with double selection.

6® clone DNA frapents S. azureus D E PKC505 S. lividans selection of thio'^, Transformants encoding resistanc; digested with TK24 apr^ transformants found; unable methylases in ScoRI, then to isolate S, azureus "shotgun" cloned. plasmids. 138

Table 14 (continued)

Experiient Experiiental Donor D M and Vector Selection/Screening^ Result U e r Goal Cloning Method D M Host Strain Method 7® clone D M fragient S. a m e n s D M pKC505 S. lividans selection of thicF, clones found encoding S. a m e n , digested with TK24 a p f transformants with correct thiostrepton resis Bglll and phenotype, tance methylases. "shotgun" cloned. but plasmids could not be recovered.

s'’ clone S, a m e n s S. aznrens d i - . p8KND2 S. lividans thio^ (PSNKD2); transformants thiostrepton resis gested with SamHI m i thi(f, neo® (pSKN04). of correct tance methylase and both "shot pSKN04 phenotype DM gun" cloned and found, but size-selected and plasmids not cloned. able to be isolated.

gb clone S, a m e n s S. a m e n s D M di pSK8D2 s. lividans selection of thi(f transformants thiostrepton resis gested with SphI m i (pSKH02) or thi(f, of correct tance methylase and size-selected PSKN04 ne(f (pSKH04) trans phenotype DM for D M fragments formants. found, but which can hybrid plasmids not ize to probes 2 able to be and 5. isolated.

10 clone S. a m e n s S. azureus D M di pIJ915 s. lividans selection of thio^, pIJlDl-like D M around resis gested with SamHI m i vi(f transformants. plasmids tance methylase and "shotgun" isolated, as genes. cloned. also found in experiments 3 and 5.

11 clone S. a m e n s S. a m e n s D M di PMIT419 s. lividans selection of thio’-, transformants D M around resis gested with SamHI m i vio® transformants. with the 1 tance methylase and size-selected correct phen genes. for fragments in otype found, the size ranges to but plasmids Éi c h probes 2 and could not be 5 hybridized. isolated.

12^ clone S. a m e n s S. a m e n s D M di pKC5D5 S, lividans selection of thi(f, plJlOl-like D M around resis gested with SasHI m i a p f transformants. plasmids and tance methylase and "shotgun" cloni d ^ 4 2 8 (see genes. or size-selected text) found. for 4-9 kbp fragmei ts. 139

Table 14 (continued)

Experiment Experimental Donor DNA and Vector Selection/Screening^ Result Number Goal Cloning Method DNA Host Strain Method

13b clone DNA fragments S. azureus, S. PÜC19 E. coli DH5e, Initial selection by No true encoding thiopep glaucogriseus, E. coll TBl ampF and lacZ blue- positives tide resistance and S. hawaii white selection. Trans found in methylases ensis DNA digeste formants with putative screening with BanHI, SphI, inserts were then with probes or SstI then size screened for desired selected for DNA DNA using colony hy Many false

fragments to whic , bridization to probes positives probes 2 and 5 2 and/or 5. found. could hybridize.

14® clone DNA fragment S. azureus DNA pUC19 E. coli DH5a Initial selection by No colonies encoding the nshA- digested with E. coli TBl amp"^ and lacZ blue- to which like gene from BamBI and DNA white selection. Trans probe 1 S. azureus. fragments > 9 kb formants with inserts hybridized selected for screened for desired were found. cloning. DNA using colony hy bridization to probe 1 15 clone DNA fragment S. azureus DNA PÜC19 E. coli DH5tt Initial selection by Cloned DNA encoding the nshA- digested with amp”- and lac blue- to which like gene from Bglll and PstI É i t e selection. Trans probe 1 S. azureus. and DNA fragments formants with inserts hybridized. of 2-4 kbp se screened for desired Putatively lected for DNA using colony hy the nshA- cloning. bridization to probe 1 like gene from S. azureus

% e genonic libraries in this experiment were constructed twice for both S. azureus and S. laurentii with the sane results each tine. The protocol used was that described by Pronega Biotech (Madison, MI) literature.

* ^ ese experiments were conducted separately more than ten tines each. % e s e experiments were conducted six times.

%iese experiments were conducted four times. % e s e experiments were conducted two times. ^See Table 4 for a list of the abbreviations. 140

The primary attempts were made by cloning DNA, either size-selected from the size of DNA fragments which hybridized to the probes (for example, using BamHI "DNA Windows" of 4 to

6 kbp and 6 to 9 kbp in order to clone the ca. 4.5 kbp and ca.

6.1 BamHI DNA fragments) or by "shotgun" methods, into various

Streptomyces vectors. These plasmid constructions were then introduced by transformation into the thiostrepton-sensitive organism, S. lividans TK24. After 12 to 20 hours incubation

(to allow for protoplast regeneration), the protoplasts were overlaid with thiostrepton to a final concentration of 50

/xg/mL (and other antibiotic markers on the plasmid, if applicable) to select for transformants containing plasmids carrying thiostrepton resistance genes. Special plasmids were obtained or constructed for this purpose, and both large, low- copy-number (e.g., pIJ915, pANT29, and pKC505) and small, high-copy-number (e.g., pSKN02, pSKN04, and pANT419) vectors were used.

Plasmid pIJlOl is a small (8.9 kbp), high-copy-number

(>40 copies per cell) first isolated from Streptomyces lividans ISP 5434 [91]. As will be discussed below, new DNA probes, specific to the replication region of plasmid pIJlOl

[89,91,161], were constructed. The restriction maps of these probes are shown in Figure 10. The plasmids, pANT425 and

PANT426. contain the 860 bp Pstl-BamUI fragment of the replication region of pIJlOl and the 530 bp BamHI-Sall fragment of the replication region of pIJlOl, respectively. Figure 10: The plJlOl-specific probes used in this study. The map shown is derived from [44,89,90,91]. The Pstl-Apal probe and the Sstll "minimal replicon" probe were derived from pIJ702, a pIJlOl derivative, since pIJ702 is identical to pIJlOl in this region of the plasmid. The Pstl- Apal probe was made double digesting pIJ702 with PstI and Apal and pooling the agarose gel- purified Pstl-Apal fragment (nt 8080 to nt 8594) with the large Apal fragment (nt 8594 to nt 1480). The Sstll "minimal replicon" probe was made by digesting pIJ702 with Sstll and isolating the correct fragment from an agarose gel, as described in the methods section. The plasmids, pANT425 and pANT426, were made by the double digestion of pIJlOl with BamHI and PstI (pANT425) and with BamHI and Sail (pANT426). The correct fragments were purified from an agarose gel and cloned into pUC19. For use as a probe, the inserts were excised from the pUC19-based plasmids by the appropriate restriction endonuclease digestion followed by gel purification.

141 1.0 kbp

7700 bp IÎ I 2490 bp rep o ip 6

Pstl-Aoal Probe I I 5tfII Minimal Replicon Probe

pANT425

PANT426

Figure 10

H to 143 cloned into pUC19. The replication region probes referred to as Pstl-Apal and Sstll were gel-purified from pIJ702 (a pIJl01-derivative which is identical to pIJlOl in this region of the plasmid).

B. Initial Observations

In the initial attempt to clone the genes encoding resistance to thiostrepton, the ca. 24 kbp, low-copy-number vector pIJ915 was used. This vector is derived from pSCPlOS

(an SCP2* vector) [Figure A20 in 67;114] and contains the viomycin resistance gene on a 2.0 kbp SamHI DNA fragment as a selectable marker [Table A9 in 67]. The cloning strategy employed was as follows: S. azureus DNA and pIJ915 were each digested with BamHI, ligated together and then used to transform S. lividans TK24. The transformants were selected by resistance to thiostrepton, and counterscreened for sensitivity to viomycin.

Eighteen tsr^, vio® transformants were obtained, 15 of which, when extracted for plasmid DNA, yielded 6.7 kbp, high- copy-number plasmids. The other three transformants contained

5.0 kbp, high-copy-number plasmids. Preliminary restriction analysis of one representative plasmid of each size, and the production of these aberrant plasmids when various different cloning vectors (e.g. pIJ915, pANT29, and pKC505, as well as other vectors described in Table 14) had been used, led me to suspect that these plasmids resulted from contamination of the 144

S. lividans protoplasts with the plasmid pIJ702, a high-copy- number plasmid derived from the naturally-occurring plasmid pIJlOl [91]. The preliminary restriction maps of these small, aberrant plasmids were compatible with a plJlOl-derived plasmid (unpublished observations), and the plasmid, pIJ702 (a streptomycete cloning vector derived from pIJlOl and containing the thiostrepton resistance gene as a selectable marker [91]) was being used by others in the laboratory at the time that these plasmids appeared.

To rule out the possibility of contamination from pIJ702, drastic steps were taken. All restriction enzyme buffers, all restriction enzymes used in these experiments, and T4 DNA

Ligase, were replaced. All buffers used in media, ligation, protoplast preparation, and protoplast transformation were discarded and then freshly prepared in acid-washed glass beakers. Magnetic stir bars used to aid in dissolving the buffers in water were cleaned with soap and water, and rinsed in double-distilled water just prior to use. Glass bottles used to store the buffers and glass pipets were acid-washed, rinsed in double-distilled water, and then baked dry. Plastic material (pipet tips, microcentrifuge tubes, etc.) was replaced where possible with fresh material, which was sterilized by extensive autoclaving and then baked dry. The protoplasts previously used were discarded and fresh protoplasts were carefully made. Finally, a fresh culture of

S. lividans TK24 was obtained from K.E. Kendrick, and fresh 145 protoplasts were made from this culture. Thus, the possibility of contamination by pIJ702 was essentially eliminated. Nevertheless, the 5.0 kbp and 6.7 kbp plasmids were still obtained as the result of various cloning strategies.

C. Subsequent Cloning Experiments

As seen in Table 14, repeated cloning experiments yielded consistent, unexpected results. When high-copy-number vectors were used to clone DNA from S. azureus containing the thiostrepton resistance genes into S. lividans TK24, tsr^ transformants of S. lividans TK24 were found, but no plasmid could be isolated from these transformants. When low-copy- number vectors were used, tsr^ transformants were also found, but plasmid extractions from these yielded either no apparent plasmid, a plasmid of approximately 6.7 kbp linear size, or a plasmid of approximately 5.0 kbp linear size. All of the isolated 5.0 kbp and 6.7 kbp plasmids could be linearized with the restriction endonuclease BamHI. Preliminary restriction analysis on representative isolates showed that all of the 6.7 kbp plasmids were alike and that all of the 5.0 kbp plasmids were alike. Furthermore, some similarity between the restriction maps of the 6.7 kbp and 5.0 kbp plasmids was noted. Figure 11 shows typical results from various plasmid preparations following digestion of the extracted DNA with

BamHI. Figure 11: Typical results from plasmid preparations of S. lividans protoplasts into which DNA from S. azureus cloned into various plasmids has been introduced by transformation. From left-to- right, the DNA in each lane, and the, restriction endonuclease with which it had been digested, is as follows: Top Row: lane 1, k, Hindlll; lane 2, pIJ702, BamHI; lanes 3, 4, and 5, three different preparations of DNA which had been digested with BamHI, "shotgun" cloned into shuttle cosmid, pKC505, introduced into S. lividans TK24 by transformation, and selected for apr^ and thio*"; lane 6 , A, Bindlll; lane 7, another preparation like those in lanes 3, 4, and 5; lanes 8 , 9, 10, and 11, S. azureus DNA digested with BamHI, selected for fragments in the size range of 4 to 6 kbp, cloned into pKC505, introduced into S. lividans TK24 by transformation, and selected for apr^ and thio^; and lane 12, A, Hlndlll. Bottom Row: lane 1 , A, Hindlll; lanes 2 and 3, S. azureus DNA digested with Bglll, cloned into pKC505, introduced into S. lividans TK24 by transformation, and selected for thio^ and apr^; lane 4, S. laurentii DNA digested with BamHI, "shotgun" cloned into pKC505, introduced into S. lividans TK24 by transformation, and selected for thio^ and apr^; lanes 5 to 10, S. actuosus DNA digested with BamHI, size-selected for DNA fragments in the range of 1 to 4 kbp, cloned into pKC505, introduced into S. lividans TK24 by transformation, and selected for nsh"^ and apr^; and lane 1 1 , A, Hindlll.

146 147

10 11 12

= »

4 5 6 7 8 9 10 11 =

#K>. '--S*

= s

Figure 11 148

Four hypotheses were advanced to explain the recurring isolation of the 6.7 kbp and 5.0 kbp plasmids (now named

PÀNT423 and pANT424, respectively). As previously stated, the first hypothesis put forward the idea that pIJ702 (5.8 kbp) was contaminating either the ligation or the transformation procedures. The second hypothesis was that the genes encoding resistance to thiostrepton were located in a physically unstable region of the S. azureus chromosome; thus, when the cloning of the region of DNA contiguous with the resistance genes was attempted, the instability resulted in the vector and insert DNA being rearranged. The third hypothesis was that pIJ702, pIJ350 (the mel~ parent of pIJ702 [91]), or some thiostrepton-resistant variant of pIJlOl (the 8.9 kbp parental plasmid of both pIJ702 and pIJ350 [44,67,90,91]) had contaminated the strain of S. azureus with which we worked and inserted itself into the chromosome adjacent to the thiostrepton resistance gene via homologous recombination between the vector tsrR gene and the chromosomal tsrJ? gene.

The fourth hypothesis was that a plJlOl-like replication region was located proximal to the resistance genes in the S. azureus chromosome, and during cloning experiments, it was released and expressed as a circularized plasmid in the recipient strain.

The first hypothesis, in which pIJ702 or pIJ350 contamination of the ligation and/or transformation was suspected was ruled out as already described. The second 149 hypothesis, that the DNA was unstable and thus difficult to clone, could not be directly ruled out. It should be noted that in repeated Southern hybridization analyses, no fragments of DNA were ever noted to stand out from the chromosomal digest which had been separated by electrophoresis on an agarose gel (data not shown). If such had been seen, it could be taken to indicate putative DNA amplification, a process known to occur commonly in streptomycetes [reviewed in 26,64].

Conversely, though stable mutants of S. azureus could be generated by MNNG treatment as already described, and though mobility shifts (possibly indicating deletions of DNA) in the fragments of DNA hybridizing to various probes putatively associated with thiostrepton biosynthesis DNA were observed in these mutants, no spontaneous mutants of S. azureus were ever observed to exist stably. This indicates that severe instability, such as might be reflected in spontaneous deletions of DNA as found in other streptomycetes

[26,27,51,64,65,66,79], is probably not the normal state of the region of DNA surrounding the nshA-like gene or the two tsrR genes of S. azureus. Thus, although deletions are the favored explanation for the mobility shifts in the thiostrepton-nonproducing mutants of S. azureus (see previous discussion), it is probable that the mutagenic treatment stimulated the putative instability and deletions in the DNA believed to be associated with thiostrepton biosynthesis; the 150 region is not likely to be normally unstable and prone to deletion.

To rule out the third hypothesis, in which contamination by a thiostrepton resistance gene-bearing plasmid variant of pIJlOl was suspected in S. azureus, two approaches were employed. To determine if a free plasmid (or plasmids) existed in S. azureus which were causing the unusual results, the first approach used to test this hypothesis was to transform S. lividans TK24 with DNA from S. azureus, using both DNA fragments from restriction endonuclease digestion and undigested DNA from whole chromosomal DNA preparations.

Aliquots of S. azureus chromosomal DNA were each digested separately with Bglll, BamHI (two separate digests), BcoRI,

SphI, and PstI. All digests, except one BamHI digest, were precipitated with ethanol. The DNA from one BamHI digest was separated by agarose gel electrophoresis, and fragments of 4.0 to 6.0 kbp size range and 6.0 to 10 kbp size range were gel- purified. Each DNA aliquot, including the size-fractionated

DNA, was further split into two aliquots, one which was self ligated and one which was left unligated. One aliquot of undigested chromosomal DNA from S. azureus was self-ligated.

Five aliquots of undigested S. azureus chromosomal DNA containing increasing amounts of DNA (5, 10, 20, 60, and 80

/ig) were left unligated. All of the DNA aliquots were then used to transform aliquots of freshly prepared S. lividans

TK24 protoplasts, after which tsr^ transformants were 151 selected. In none of the transformation experiments, each of which was repeated two to six times, did thiostrepton- resistant colonies of S. lividans TK24 arise. Moreover, a control experiment using tsr^ selection on untransformed S. lividans TK24 yielded no thiostrepton resistant transformants.

These results indicate that a vector was required as an intermediate in the apparent liberation of the 5.0 kbp and 6.7 kbp plasmids.

In the second approach to the study of the third hypothesis, representative isolates of pANT423 (the 6.7 kbp plasmid) and pANT424 (the 5.0 kbp plasmid) were digested in various ways (along with pIJ702) and Southern blotted after

DNA fragment separation by agarose gel electrophoresis. The blotted nitrocellulose filter containing the plasmid DNA digests was hybridized to a gel-purified fragment of DNA specific to the replication region of plJlOl-based plasmids

(the Pstl-Apal probe of Figure 10) as described earlier. The pIJlOl replicon DNA was used as a probe because, as noted earlier, the restriction maps of pANT423 and pANT424 appeared similar to the restriction map of the plasmid pIJ702, a derivative of pIJlOl. As shown in Figure 12, DNA fragments of pIJ702, and the two representative, aberrant, plasmid clones

(pANT423 and pANT424, the 6.7 kbp and 5.0 kbp plasmids, respectively) all hybridized to the plJlol-specific DNA.

Since no free plasmid bearing the thiostrepton resistance gene was found in DNA transformation experiments, the fourth Figure 12 : Hybridization of restriction endonuclease-digested DNA fragments from pIJ702 and two other plasmids to a 2,180 bp fragment of DNA internal to the essential DNA (replicon region) of plasmid pIJlOl. The other two plasmids are representative, aberrant plasmids isolated in attempts to clone large DNA fragments containing genes encoding thiostrepton resistance. The DNA of the plasmids was restriction endonuclease digested as described in the methods section and loaded into a 1.0% THE agarose gel (GIBCO/BRL Model H4 electrophoresis apparatus) for separation by electrophoresis and Southern blotting, as described in the methods section. The size standard used was DNA from bacteriophage k digested with Hindlll, and it did not hybridize to the pIJlOl DNA fragment used as a probe. The larger (6.7 kbp) aberrant plasmid, pANT423, is abbreviated 423. The smaller (5.0 kbp) aberrant plasmid, pANT424, is abbreviated 424. Plasmid pIJ702 is abbreviated 702. The lanes of hybridizing DNA are read left-to-right. A. Top Row: lane 1, 702, BamHI; lane 2, 423, BamHI; lane 3, 424; lane 4, 702, BamHI+MIuI; lane 5, 423, BamHI+MluI; lane 6 , 424, BamHI+Mlul; lane 7, 702, MIuI; lane 8 , 423, Mlul; lane 9, 424, Mlulj lane 10, 702, BamHI+Styl; lane 11, 423, BamHI+Styl; lane 12, 424, BamHI+Btyl; lane 13, 702, Mlul+Styl; lane 14, 423, Mlul+Styl; lane 15, 424, Mlul+Styl; lane 16, 702, Styl; lane 17, 423, Btyl; lane 18, 424, Styl; lane 19, 702, BamHI+Sall; lane 20, 423, BamHI+Sall; lane 21, 424, BamHI+Sall; lane 22, 702, Mlul+Sall; lane 23, 423, Mlul+Sall; lane 24, 424, Mlul+Sall; lane 25, 702, Sail; lane 26, 423, Sail; lane 27, 424, Sail. B. Bottom Row: lane 1, 702, BamHI+Bgrlll; lane 2, 423, BamHI+Bglll; lane 3, 424, BamHI+Bglll; lane 4, 702, Mlul+Bglll; lane 5, 423, Mlul+Bglll; lane 6 , 424, Mlul+Bglll; lane 7, 702, Bglll; lane 8 , 423, Bglll; lane 9, 424, Bglll; lane 10, 702, SphI; lane 11, 423, SphI; lane 12, 424, SphI; lane 13, 702, Sphl+PvuII; lane 14, 423, Sphl+PvuII; lane 15, 424, Sphl+PvuII; lane 16, 702, PvuII; lane 17, 423, PvuII; lane 18, 424, PvuII.

152 153

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 m

1 2 3 4 5 g 7 8 9 10 11 12 13 14 15 16 17 18

Figure 12 154 hypothesis, suggesting that a plJlOl-like plasmid had integrated into the S. azureus chromosome, was tested. This hypothesis could readily explain why no transformants were found when uncut chromosomal DNA was used to transform S. lividans TK24. Also, this hypothesis could conceivably explain why aberrant plasmids were found when cloning with pSCP2*-derived vectors (such as pKC505) and why no plasmids could be found when vectors using the pIJlOl replication mechanism were used. In the former case, once the integrated plasmid was cloned into the SCP2* vector, the high-copy-number replication mechanism of the inserted plasmid would conflict with and out-replicate the low-copy-number mechanism of the

SCP2* vector. Presumably, then, the high-copy-number vector would be under selective pressure to excise itself from the low-copy-number vector. This would result in the isolation of a plasmid of smaller size (e.g., «6.7kbp and «5.0 kbp) than the original cloning vector used (which were all >18 kbp in size) and bearing the thiostrepton resistance gene.

Furthermore, if the excision was not correctly carried out, or if the high-copy-number plasmid's replication mechanism was defective, a low-copy-number, very large, plasmid bearing thiostrepton resistance (and any other selectable marker the vector itself carried, such as apramycin for pKC505) would result. After plasmid isolation procedures, the plasmid might not be observed above the background of the chromosomal smear, since SCP2* vectors exist in l to 4 copies per cell [69,114]. 155

It is difficult to prepare such plasmids in S. lividans TK24, even using special procedures for large, low-copy-number vectors, as also reported by other investigators [140]. Thus, pKC505 is a quite attractive plasmid for these studies, because correct clones can be shuttled to apramycin-sensitive

E. coli and grown to large quantities [148].) Finally, if the plasmid DNA were integrated into the chromosome in the proximity of the thiostrepton resistance gene, fragments of

DNA should exist in the chromosome which hybridize to pIJlOl- specific probes. Based on the restriction map (Figure 7), some of these fragments should also hybridize to tsrR and nshR.

Thus, three specific tests of this hypothesis were available for experimental analysis. Moreover, if this hypothesis were true, then it appears (based on restriction analysis and

Southern blot hybridization analysis) that the 6.7 kbp plasmid

(pANT423) is derived from the ca. 6.1 kbp BamHI DNA fragment which hybridizes most strongly to tsrR, while the 5.0 kbp plasmid (pANT424) appears to be derived from the ca. 4.5 kbp

BamHI DNA fragment which hybridizes most strongly to nshR.

As already discussed and shown (see Table 14), repeated attempts to clone the DNA contiguous with the thiostrepton resistance gene from S. azureus resulted in repeated isolation of aberrant plasmids (when SCP2*-based vectors were used) or no apparent plasmids from phenotypically correct transformants

(typically when plJlOl-based plasmids were used). For example, DNA derived from the plasmid preparation seen in lane 156

7 of Figure 11 did not appear to contain any plasmid; however, the transformant from which this DNA was derived was resistant to 50 jug/mL of thiostrepton both on plate culture and in liquid culture. DNA from this plasmid preparation was digested in several, separate restriction endonuclease reactions. The DNA in each digest was then separated by electrophoresis on an agarose gel and blotted by the method of

Southern [157] to nitrocellulose. The blot was probed with tsrR, and hybridization was detected by autoradiography. As shown in Figure 13, the DNA of the plasmid preparation showed no apparent plasmid DNA above the "chromosomal smear" background. As shown in Figure 14, the DNA of the plasmid preparation did hybridize to tsrR. A rudimentary restriction map of this plasmid (pANT428) was derived (since there was considerable rearrangement of one end of the insert) and is also shown in Figure 14. This insert DNA appears to correspond to the 6.1 kbp BamHI DNA fragment containing the thiostrepton resistance gene. The considerable rearrangement of the DNA to the right of the resistance gene in pANT428

(which seems to correspond to the DNA which is to the left of the 6.1 kbp BamHI DNA fragment in Figure 7) is offered as proof of the occurrence of an incorrect excision of the plasmid. This cloned DNA also hybridized, quite weakly, to the Pstl-BamHI fragment of the plJlOl replication region probe

(data not shown). Figure 13: Photograph of the DNA of a plasmid preparation of S. azureus DNA which had been cloned into plasmid pKC505. Transformants which were thio^ and apr^ were selected and DNA isolated using a plasmid preparation technique for large plasmids. Note that the DNA in the photo appears to be chromosomal, and no plasmid DNA is readily apparent. From left-to-right, the restriction endonuclease digests are: bacteriophage A DNA digested with Hindlll); cloned DNA digested with BamHI; cloned DNA digested with BamHI+WIuI; clone digested with Mlul; clone digested with BamHI+Styl; clone digested with Mlul+Styl; clone digested with Styl; clone digested with SstI; clone digested with Bstl+Sstll; clone digested with Sstll; clone digested with SphI; clone digested with Sphl+PvuII; clone digested with PvuII; bacteriophage A DNA digested with HindiII.

157 : I e t m • V- I jBamHI

I BamHI+MluI { Mlul

J iBamHI+Styl IMV- •■ * \MluI+StyI, I { [Styl H W i (sstI

I \SstI+SstII

iSstll

. » • Sphi

Sphl+Pvull

Pvuii

' 1}I |z i I 111 I lo i A s I

ui 00 Figure 14: The 6.1 kbp tsrR-containing DNA cluster cloned from S. azureus into pKC505. A. Autoradiogram to Probe 5 of DNA containing thiostrepton resistance cloned into shuttle cosmid pKC505. From left-to-right, the lanes are as follows: bacteriophage A DNA digested with Hindlll (does not hybridize to the probe); cloned DNA digested with BamHI; cloned DNA digested with BamHI+MluIj clone digested with Mlul; clone digested with BamHI+Styl; clone digested with Mlul+Styl; clone digested with Styl; clone digested with SstI; clone digested with Sstl+Sstll; clone digested with Sstll; clone digested with Sphl; clone digested with Sphl+PvuII; clone digested with Pvull; bacteriophage A DNA digested with Hindlll (does not hybridize to the probe). B. Restriction map from the data in A. There is a rearrangement of the insert DNA which makes the map somewhat uninterpretable.

159 160

HH H 3 M H M -W 5 s 4 J s: CO •U % + 4* CO 5 HH + 4" I X W X H H H H B 3 a 3 >1 4 J ta (0 4 J %w » g. S' QQ S OQ s CO CO CO

EcoRl

Pvull

PANT428

Reananged Area of Insert /

Figure 14 161

D. Experiments with Authentic S. azureus ATCC 14921

It now appeared that a case could be made for the integration of a plJlOl-like plasmid into the chromosome of S. azureus. With proof that a plJlOl-like plasmid could be found in the chromosome of S. azureus and S. actuosus, it became necessary to establish whether the source of the DNA which hybridized to the pIJlOl replicon was artificial (i.e., if it had been accidentally introduced into the organism sometime after arriving at this laboratory) or if it had been there as a result of natural transfer of the plasmid during evolution of this strain.

To validate the presence in the S. azureus chromosome of sequences of DNA related to the replication region of pIJlOl, two new strains of S. azureus were obtained. Though all strains of S. azureus are believed to be derived from the ATCC strain, ATCC 14921, it is possible that contamination of a plJlOl-like plasmid had occurred in the course of manipulations of the organism. One strain of S. azureus from the stocks of the John Innes Institute (Norwich, U.K.) was obtained via Todd Smith. Another strain of S. azureus was newly purchased from the ATCC. The lyophilized culture obtained from the ATCC was from the original strain deposit of s. azureus ATCC 14921 made in 1965 [15]. Thus, if any plasmid were found in this strain, it would have had to originate with the organism and would not be the result of any contamination due to our manipulations of S. azureus. 162

The three strains of S. azureus are hereafter identified as follows; Saz-l/OSU, for the OSU strain of wild-type S. azureus in which the initial observations were made; Saz-l/JI, for the strain of wild-type S. azureus from the John Innes

Institute; and Saz-l/ATCC for the strain of wild-type S. azureus from the ATCC certified to be from the original stock.

Fresh, new stocks of chromosomal DNA were prepared separately from each of these S. azureus strains using methods described in Chapter IV.

Each preparation of chromosomal DNA was completely digested with BamHI or Mlul. All digests were carried out at conditions which allowed for eight-fold overdigestion of the

DNA. In addition to the three strains of S. azureus, the same kinds of restriction endonuclease digests were performed on

DNA from the following organisms: S. actuosus, S. hawaiiensis, S. laurentii, S. lividans TK24, S. sioyaensis, and Streptomyces species ATCC 31463. Southern hybridization was then performed as described in the methodology section.

The nitrocellulose blots were then probed (i) with the 2.3 kbp

Sstll DNA fragment of pIJlOl encoding the minimal replicon,

(ii) with the insert from pANT425 (which is the Pstl-BamH.1 fragment of pIJlOl containing part of the rep gene and some downstream DNA, as shown in Figure 10), and (iii) with the insert from pANT400 (the nshR probe), as shown in Figures 15,

16, and 17. 163

As can be seen in the figures, the nshR probe, the Sstll

"minimal replicon" probe, and the pANT425-derived probe all hybridized to the same fragments of DNA from Saz-l/OSU. Just as importantly, the pattern of hybridization is consistent among all three strains of S. azureus. This indicates that the pIJl01-1ike plasmid in S. azureus was present in that organism from the time of its original isolation and was not a laboratory contamination.

There are three BamHI and four Mlul fragments of DNA which hybridize to the plJlOl-specific probes. Two fragments of S. azureus DNA cut with BamHI and one fragment of DNA cut with Mlul hybridize to the nshR DNA probe in all three strains, which is consistent with previous results. Moreover, these fragments match the DNA fragments which hybridize to the plJlOl-specific probes, strongly suggesting that the pIJlOl- like plasmids carry the cloned thiostrepton resistance methylase genes. Hybridization of DNA fragments to the plJlOl-specific probes also occurred in S. actuosus, S. hawaiiensis (weakly), S. laurentii (weakly), S. lividans, and

S. species ATCC 31463. Additional experiments (not shown) showed that S. coelicolor A3(2) contains fragments of DNA hybridizing to plJlOl-specific probes.

Many species of Streptomyces are known to harbor a variety of plasmids [26,69], insertion sequences [26,62,107], and phage [26]. The most widely known plasmids in the streptomycetes are SLPl and its derivatives [12,68] (a plasmid Figure 15: Autoradiogram showing hybridization of the nshR DNA probe (derived from pANT400) to DNA isolated from several strains of Streptomyces spp. A. The chromosomal DNA samples digested with BamHI. From left-to-right, the DNA samples are as follows: S. actuosus, Saz-l/OSU, Saz-l/JI, Saz- l/ATCC, S. hawaiiensis, S. laurentii, S. lividans TK24, S. sioyaensis, S. species ATCC 31463, X DNA digested with Hindlll (did not hybridize to probe), and pANT34 size standard (the autoradiographic size standard for use with nshA and nshR). Also, a lane of X DNA digested with Hindlll which was to the left of the lane of S. actuosus DNA is not shown because it did not hybridize to the probe. B. The chromosomal DNA samples digested with Mlul. The order of the lanes is the same except that the rightmost lanes of pANT34 size standard and X DNA digested with Hindlll are reversed. The stringency of the wash was medium (0.4X SSC, 0.1% SDS, 42°C, two washes, 30 minutes per wash).

164 S. actuosus S. azureus/OSU

S. azureus/JI

S. azureus/ATCC

S. hawaiiensis

S. laurentii S. lividans T K 2 4

S. sioyaensis

ATCC 31463

X s i z e s t d . ►53 I oopo ro H Ü1 S , actuosus

S. azureus/OSU

S. azureus/JI

S. azureus/ATCC

S. hawaiiensis

S. laurentii

S. lividans T K 2 4

S. sioyaensis

ATCC 31463

s i z e s t d .

991 Figure 16: Autoradiogram showing hybridization of the Sstll "minimal replicon" DNA probe to DNA isolated from several strains of Streptomyces spp. A. The chromosomal DNA samples digested with BamHI. From left-to-right, the DNA samples are as follows: S. actuosus, Saz-l/OSU, Saz-l/JI, Saz- l/ATCC, S. hawaiiensis, S, laurentii, S. lividans TK24, S. sioyaensis, and S. species ATCC 31463. The lanes containing X DNA digested with Hindlll and the pANT34 size standard (the autoradiographic size standard for use with nshA and nshR) are not shown as they did not hybridize to the probe. B. The chromosomal DNA samples digested with Mlul. The order of the lanes is the same as in A. The stringency of the wash was medium (0.4X SSC, 0.1% SDS, 42“C, two washes, 30 minutes per wash).

166 H M S. actuosus S . azureus/OSU S. azureus/JI

S. azureus/ATCC S. hawaiiensis

S. laurentii

S. lividans T K 2 4 S . sioyaensis

I ATCC 31463

. s . actuosus # s. azureus/OSU

S . azureus/JI

^ C s, azureus/ATCC W f S. hawaiiensis I , S. laurentii S. lividans T K 2 4 S. sioyaensis

A T C C 3 1 4 6 3

Z.9T Figure 17: Autoradiogram showing hybridization of the pANT425-derived probe (containing the Pstl-BamHI DNA fragment of the pIJlOl replication region) to DNA isolated from various strains of Streptomyces spp. A. The chromosomal DNA samples digested with BamHI. From left-to-right, the DNA samples are as follows: S. actuosus, Saz-l/OSU, Saz- l/JI, Saz-l/ATCC, S. hawaiiensis, S. laurentii, S. lividans TK24, S. sioyaensis, S. species ATCC 31463, X DNA digested with Hindlll (did not hybridize to probe), and pANT34 size standard (the autoradiographic size standard for use with nshA and nshR). Also, a lane of X DNA digested with Hindlll which was to the left of the lane of S. actuosus DNA is not shown because it did not hybridize to the probe. B. The chromosomal DNA samples digested with Mlul. The order of the lanes is the same as in A. The stringency of the wash was medium (0.4X SSC, 0.1% SDS, 42°C, two washes, 30 minutes per wash).

168 K> K> C Lj S . actuosus a « &- • s. azureus/OSU ^ azureus/JI

^ S. azureus/ATCC

S. hawaiiensis

S. laurentii

'» - s., S. lividans TK24 S. sioyaensis

H- t ATCC 31463 A 1(0 s i z e s t d .

: k

S . actuosus

S. azureus/OSU

S. azureus/JI

S. azureus/ATCC » S. hawaiiensis

laurentii

lividans T K 2 4

S. sioyaensis

A T C C 3 1 4 6 3

691 170 which undergoes site-specific, stable, chromosomal integration

[56,139]), SCP2* and its derivatives [69,114,148], and pIJlOl and its derivatives [44,89,90,91,161]. In addition to SLPl, a new plasmid which also undergoes site-specific, stable, chromosomal integration, pSAM2, has been found in the streptomycete, S. ambofaciens [16,17,120].

Because of this diversity of plasmids, and because of the broad host range of some of the plasmids in streptomycetes

[26,69], the potential for genetic exchange among various species of Streptomyces is great. Indeed, Rafii and Crawford have recently shown that various derivatives of the conjugative plasmid plJlOl [26,69,91] can mobilize nontransferable pIJlOl derivatives in streptomycetes in soil

[145]. Specific to this work, Streptomyces azureus ATCC 14921 has been shown to harbor both a phage [134,136] and a unique,

8.8 kbp, pock-forming plasmid (pSAl) which undergoes both linear and circular phases in its life cycle [125,133]. There is no apparent relationship between the restriction map of pSAl [125] and the restriction maps of pANT423 and pANT424, nor does pock formation appear to increase above background levels in S. lividans TK24 when either pANT423 or pANT424 are introduced by transformation into protoplasts of S. lividans

TK24. (S. lividans TK24, though devoid of known plasmids, does form small, pock-like structures on plates [67].) Thus, it seems improbable that pSAl is related to pANT423 and

PANT424. 171

To better understand the arrangement of the plasmids in the S. azureus chromosome, and to compare Saz-l/ATCC with Saz- l/OSU, Southern hybridization of Saz-l/ATCC and Saz-l/OSU,

(cut in multiple, separate restriction endonuclease digests) to plJlOl-specific probes. This proved impossible to do, however, as there were far too many hybridizing fragments of

DNA, all possessing about the same level of hybridization, to construct a map. A rudimentary restriction endonuclease fragment map of the nshN/tsriî-hybridizing DNA fragments was also constructed (Fig. 18). There are several differences with the map constructed for Saz-l/OSU (see Fig. 7), but these do not appear to be significant. The differences, however, between the chromosomal restriction endonuclease maps (Fig. 7 and Fig. 18) and the maps of pANT423 and pANT424 (Fig. 19 and

Fig. 20, respectively) are significant. As has been shown in hybridization experiments, there are only two BamHI DNA fragments in S. azureus to which the nshR and tsrR DNA probes hybridize. Furthermore, two of the three BamHI DNA fragments of S. azureus to which the plJlOl-replicon DNA probe hybridizes correspond exactly to these two resistance genes.

As shown in cloning experiments, only two types of thiostrepton-resistant plasmids could be recovered from S. azureus. Thus, it is reasonable to assume that the two plasmids correspond to these two BamHI DNA fragments. The Figure 18: Restriction endonuclease map from Saz-l/ATCC of the chromosomal region hybridizing to tsrR and nshR. Restriction endonucleases are abbreviated as follows: Ba, BamHI; Cl, Clalj Ec, BcoRV; Ml, Mlul; Ps, PstI; Pv, PvuII; Sc, S a d ; and St, Styl.

172 Cl a Pv Sc S Pv Ec Ç l?t Cl Ba Ml Ba Sc St Ml Ml Ba Pv Sc Pa Cl Lu MU - 4 - nshR-like tsrR

Region of DNA Surrounding the Two Thiopeptide Antibiotic Resistance Genes in Saz-l/ATCC

1 . 0 k b p

Figure 18

H -v ] W Figure 19; Restriction endonuclease map of pANT423, the plasmid that appears to correspond to the 6.1 kbp BamHI fragment of DNA in the S. azureus chromosome to which tsrR and nshR hybridize.

174 175

Sstll

PANT423 «6750 bp

Styl EcoRV

P s t l Pvull

mu l

Sstn

Figure 19 Figure 20: Restriction endonuclease map of pANT424, the plasmid that appears to correspond to the 4.5 kbp BamHI fragment of DNA in the S. azureus chromosome to which tsrR and nshR hybridize.

176 177

BamlXL

PANT424 -5000 bp

-Sstll Styl P m U Sst\

Styl Gal Styl

Figure 20 178 differences in the chromosomal and plasmid maps are due partly to the different abilities to map chromosomal DNA and plasmid

DNA. For example, sites for Sstll could be placed on the plasmids, which were cultivated in S. lividans TK24, but not on the S. azureus chromosome, because S. azureus chromosomal

DNA cannot be cut with Sstll (from GIBCO/BRL), or the isoschizomers SacII (from New England Biolabs, Beverly, MA) and Kspl (from Boehringer Mannheim Corporation, Indianapolis,

IN), possibly due to méthylation of the restriction endonuclease recognition site (data not shown). Also, the differences in the chromosomal restriction map and the maps of the isolated plasmids are explainable if the plasmids integrate into the chromosome and then, upon subsequent cloning out of the chromosome, undergo some rearrangement to form a viable plasmid. This has been shown to be the case with other integrating plasmids of Streptomyces spp.

[12,16,26,56,68,69,120,139,145].

E. Discussion

This research has led to the significant finding that the genes encoding resistance to thiostrepton in S. azureus ATCC

14921 lie in tandem on the chromosome of S. azureus and that each resistance gene is contiguous with DNA which hybridizes to probes specific to the ■ replication region of pIJlOl.

Moreover, these plasmids were not introduced in our laboratory, but rather, they were present in S. azureus from 179 the time of its isolation (i.e., the ATCC strain had been deposited ca. 10 years before pIJlOl was discovered and derivatives from it were made).

This finding raises some perplexing questions about thiostrepton biosynthesis in S. azureus. One question is whether the thiopeptide resistance methylases are located anywhere near the genes encoding enzymes whose reactions are integral to the biosynthesis of thiostrepton. This assumption of the "clustering" of biosynthetic genes around the resistance genes was based on excellent evidence and precedence in other streptomycetes, but must now be called sharply into question for S. azureus. Furthermore, the role of the nsha-like gene in S. azureus, and of nshA itself in S. actuosus, is now more perplexing. As discussed earlier, fragments of DNA which hybridize to a probe made from nshA are widely distributed among the streptomycetes. In addition, the nshA-li'ke gene in S. azureus ATCC 14921 is not located near

(within 20 kbp) the tsrR and nshR-like thiopeptide resistance genes, whereas in S. actuosus it was linked on a polycistronic mRNA to nshR. It is tempting to speculate also that ribosomal méthylation may not have been the original mode of resistance to thiostrepton, but that it is rather fortuitous later occurrence brought about by the introduction of these pIJlOl- like plasmids into S. azureus.

The appearance in multiple species of Streptomyces of DNA fragments which hybridize to the plJlOl-specific probes is 180 also interesting. It may be that plJlOl-like plasmids form the basis of natural genetic exchange among the streptomycetes. Indeed, this speculation has some support from the work of Don Crawford and coworkers [e.g., 145] who found that pIJlOl and its derivatives are mobilized easily and naturally in soil.

The isolation of the two thiostrepton resistant plasmids, pANT423 and pANT424, from Saz-l/OSU was found to be highly reproducible as long as a low-copy-number vector (one based on

SCP2*, not pIJlOl) was involved as an intermediate step. That is, ligation of Saz-l/OSU DNA to the vector, followed by the subsequent introduction by transformation of that ligated DNA into S. lividans TK24 and selection for thiostrepton resistance, had to occur. This supports the idea that the plasmids are integrated into the chromosome. It may show also that the plJlOl-like sequences in S. azureus are insufficient by themselves to replicate, and that functions from the vector intermediate are subsumed by the plJlOl-like plasmids upon cloning. It was noted that transformations with uncut, unligated chromosomal DNA and with uncut, ligated chromosomal

DNA from Saz-l/OSU never yielded thiostrepton-resistant transformants in S. lividans TK24. This also strongly supports the idea that the plasmids are integrated into the chromosome. 181

Cloning of the nshA-Equivalent Gene from S. azureus

The nshA-like (orf699) gene of S. azureus was successfully cloned into E. coli DH5a using the vector pUC19

(Table 14). Using the DNA hybridization procedures discussed in a previous section, the gene was found to map to a 2.8 kbp

Pstl-Bglll fragment of the S. azureus chromosome. Cloning of the DNA was accomplished by the ligation of size-selected DNA into pUC19 and screening by colony hybridization as discussed in the methods section and shown in Table 14.

Using these procedures, a total of 240 E. coli DH5a transformants with inserts were obtained, and three of these hybridized to probe 1 (nshA). These three were taken along with seven randomly picked transformants to which the probe did not hybridize. The plasmids were extracted from cultures of the transformants, and restriction enzyme mapping and

Southern hybridization were performed on the plasmids. The plasmid DNA from the three colonies which the data indicated contained the gene of interest again hybridized to probe l, whereas the plasmid DNA from the seven randomly picked colonies did not hybridize. The three plasmids with hybridizing DNA were found to have identical inserts, as shown in the restriction map of pANT408 (Fig. 21).

The DNA of pANT408 contains three fragments which have a

PstI restriction site at one end and a Bglll site at the other end. In addition, two of the Pstl-Bglll fragments are separated by an intervening Bglll fragment. Hybridization of Figure 21: Restriction map of pANT408, the initial clone containing DNA which hybridized to probe l (nshA).

182 183

psti Bgm

Pvul

PA N T408 Bg/U

BglU Pvul Psfl pvul Sphl Mlul

Figure 21 184 the entire, isolated, radioactively-labelled insert of pANT408 to DNA from S. azureus and S. actuosus showed that the insert

DNA most likely was composed of non-contiguous DNA fragments from S. azureus. The Pstl-Bglll fragment of pANT408 which hybridized to probe 1 was located, using Southern hybridization techniques, and subcloned to make pANT409. The plasmid (pANT409) then was mapped further by restriction endonuclease digestion (Fig. 22).

Analysis of the Function of NshA

The structure and other features of the protein encoded by the DNA of nshA have been discussed previously (see Chapter

2). It is of great interest to establish a function for a gene which is apparently so widely distributed among the streptomycetes as was shown earlier in this chapter. To this end, I collaborated with Dr. D. C. Dosch to determine the function of this gene.

The plasmid, pANT407 (Fig. 23), was constructed from the plasmid, pANT403 (pIJ702 plus a 2.6 kbp Pstl-BamHI fragment of

DNA known to contain the nshA and nshR genes of S. actuosus, and also the 2.0 kbp viomycin resistance gene) by restriction endonuclease digestion at the single Mlul site contained within the nshA gene on the plasmid. Treatment with SI nuclease then destroyed the overhanging deoxynucleotides and deleted the Mlul site. Religation of the plasmid caused a missense mutation in the gene at the point of deletion, which Figure 22: Restriction map of pANT409, the subcloned construct of pANT408 containing only the 2.8 kbp Pstl-Bglll DNA fragment from S. azureus which hybridizes to probe 1 (nshA). Note that the BamHI site of the vector (pUCl9) and the Bglll site of the insert have been destroyed by the cloning procedure employed.

185 186

Sphl Mlul Pvul San

Smal ‘San PANT409

•Sphl Pvul

Pvul

Smal

Pvul

Smal

Figure 22 Figure 23: Restriction map of pANT407 used for the site- specific mutagenesis, via plasmid replacement, of nshA in S. actuosus. The construction of this plasmid is briefly described in the text.

187 188

BcanUi

Pstl

PANT407

Oal EcoRW

BamHl Pttl

Figure 23 189 subsequently resulted in a nonsense mutation downstream in the transcript. Since the mutation was a deletion, there was no chance of reversion of the DNA. The plasmid remained, nevertheless, selectable with both viomycin and thiostrepton and could be propagated in S.actuosus, which is viomycin- sensitive.

This plasmid was then introduced by transformation into

S. actuosus with selection for viomycin resistance. Isolated transformants were cultured and protoplasted as described in the methods section. The purpose of the protoplasting was to promote loss of the plasmid. The protoplasts were regenerated under nonselective conditions and regenerated, viomycin- sensitive (indicating probable plasmid loss) colonies were selected.

Theoretically, since the high-copy-number vector contains an insert highly homologous to a region of the chromosomal DNA

(in this case the nshA-nshR region), the insert (containing the mutation in the nshA gene) and the wild-type chromosomally-located gene will recombine. The mutated gene will then be in the chromosome, and the wild-type gene will be in the plasmid. By manipulating the transformants in ways to promote loss of the plasmid, some transformants will lose the plasmid after the recombination has occurred, and it will be the wild-type gene which is lost. By mutating the nshA gene at a convenient restriction endonuclease site (the single Mlul restriction site in the nshA gene), the presence or absence of 190 the wild-type nshA gene in any of the transformants can be readily ascertained by chromosomal DNA isolation from the transformant, restriction endonuclease digestion of the DNA, and Southern hybridization of the digested DNA to the nshA gene probe.

These colonies then were grown on R5 medium and assayed for bioactivity against Staphylococcus aureus, as described in the preceding chapter for the mutant analysis. A wide range of bioactivities, ranging from production of much larger zones of inhibition (12mm) than the wild-type zone (3mm) to almost no production at all (zone just greater than 1.0 mm), were obtained in this assay. The production of the melanin-like pigment secreted by S. actuosus was assessed also, since the production of the pigment varied among individual colonies.

There was no apparent correlation between the production of the melanin-like pigment and the size of the zone of inhibition. From these organisms, 10 strains were picked as representatives of the various inhibition zones. These were grown in YMB/YEME media, and chromosomal DNA was isolated from each strain. An aliquot of each sample of isolated DNA was digested separately with BamHI and Mlul. The presence of the mutation corresponded to the two strains with the highest zones of inhibition of S. aureus. The autoradiographic analysis of the mutants is shown in Figure 24. As measured by the zones of inhibition of S. aureus, the amount of nosiheptide produced in the two strains found to carry the Figure 24: Autoradiogram of Southern hybridization analysis of 10 strains of S. actuosus into which pANT407 had been introduced by transformation and subsequently lost. The DNA isolated from each strain was separately digested with Mlul (panel A) and BamHI (panel B). From left-to-right, in each panel, the lanes are: (1 ) autoradiographic size standard (Table 6 ); (2) S. actuosus wild type; (3) Sac-4; (4) Sac-11; (5) Sac-20; (6 ) Sac- 31; (7) Sac-38; (8 ) Sac-49; (9) Sac-53; (10) Sac- 74; (11) Sac-8 6 ; (12) Sac-94; and (13) autoradiographic size standard (Table 6 ).

191 1 2 3 4 5 6 7 8 9 10 11 12 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14

4 •1^

12.8

7 .4

# .S.O

... 2.6

Figure 24 W 193 mutation (Sac-11 and Sac-53) was not stable. In one strain, the production of nosiheptide returned to wild-type levels, while in the other strain, the production of nosiheptide ceased.

Biological assay, though it can be sensitive and accurate with certain organisms, is not the most reliable method to determine the amount of antibiotic produced. Houck [81] has developed a sensitive, high-pressure liquid chromatography

(HPLC) assay for nosiheptide production. This experiment is to be conducted again to confirm, if possible, these results.

These results, if confirmed by a complete repetition of the experiment, suggest that the function of nshA is that of a potential negative regulator of biosynthesis. These results would be in accord with the analysis of the putative protein product and the discovery of both putative DNA and RNA binding sites within the protein [109]. One can envisage a protein which acts to block transcription of the DNA encoding proteins used in nosiheptide biosynthesis until a point in the life cycle when some positive regulator turns off the effects of the negative regulator. Negative regulators are known to exist (see the discussion in Chapter II, page 16), but none have been described in the literature yet with respect to sequence and function. Thus, upon further analysis, this may be the first, clear example of a negative regulator of

Streptomyces antibiotic biosynthesis. 194

Time-Course Fermentation; Analysis of 5. azureus DNA

As described in the methods section, samples of S. azureus cells were taken from the fermentation vessel in order

(i) to analyze the DNA structure of the organism over time with respect to the stability of the putative antibiotic biosynthesis DNA, and (ii) to measure the increase in cell mass over time (by measuring the increase in the dry weight of cells). This experiment was conducted because, based on the known instability of other antibiotic biosynthesis DNA of

Streptomyces spp. and the possible instability of the mutants of S. azureus blocked in thiopeptide antibiotic biosynthesis, it was hypothesized that S. azureus antibiotic biosynthesis was intrinsically unstable.

The fermentation dry weights are shown in Figure 25.

These data show, when plotted over time, a normal, expected growth curve for S. azureus in the medium DPM3. The initial cell inoculum was 0.18 gm of cells/L of medium. The final concentration of cells in the fermentation medium, at 120 hr after inoculation, was 8.74 gm of cells/L of medium. The cells taken for DNA analysis were processed as described previously. Because of the potentially small amounts of DNA able to be recovered from cells in the early time points, it was decided to precipitate the DNA with isopropanol in a centrifuge tube, rather than spooling the DNA onto Pasteur pipets as described in the methods section. After DNA was isolated from cells at each of the time points taken, aliquots Figure 25: The Dry weights of fermentation of S. azureus (log scale). Samples for dry weights were taken approximately every 6 hr. through the 84th hr. then at 12 hr. intervals to the I20th hr. The samples were then processed to get the dry weights of the cells, as described in the methods section. Dry weights are shown on a log scale as gm of dry cells/L of media.

195 10.000

sS m "m O 1. 0 0 0 -

0.100 0 12 24 36 48 60 72 84 Time in Fermentation (hours)

H Figure 25 VO OV 197 were digested with either BamHI or SstI restriction endonucleases. The DNA was then separated by agarose gel electrophoresis and blotted to nitrocellulose as described.

Probe 2 (nshR) was used for hybridization. The washing of the filters was conducted at medium stringency (defined as 0.4X

SSC, 0.1% SDS, 42°C) in order to see both DNA fragments in 5. azureus which hybridize to the probe. These data are shown in

Figure 26. No apparent rearrangement of the DNA was observed over the time of the fermentation. This indicates that the mobility shifts (possibly deletions) of DNA observed in the mutants are not related to inherent instability of the DNA for thiostrepton production in S. azureus, insofar as probes for the antibiotic production are available, but are instead related either to the conditions of mutagenesis or to the mutagen itself. When coupled with the observation that no stable, spontaneous, thiostrepton-nonproducing mutants of S. azureus were found in the course of this work, the hypothesis of inherent instability in the region of DNA containing genes involved in thiostrepton biosynthesis is further weakened.

Study of the Mode of Biosynthesis of Thiostrepton and

Nosiheptide by Chloramphenicol Inhibition of Antibiotic

Production

One major, unanswered question has been whether thiostrepton and nosiheptide are synthesized via ribosomal mechanisms or from an enzymatic template. In the former case. Figure 26: Hybridization of S. azureus DNA digested with BamHI and SstI to probe 2 (nshR) at different time points in fermentation of the organism. In all four panels shown on the next two pages, the DNA digested with BamHI alternates sequentially with the DNA digested with SstI for each time point. The lanes are ordered as follows: A. lane 1, size standard (Table 6 ); lanes 2 & 3, time point 0; lanes 4 & 5, time point 1 (6 hr); lanes 6 & 7, time point 2, (12 hr); lanes 8 & 9, time point 3, (18 hr); lanes 10 & 11, time point 4, (24 hr); Icine 12, size standard (Table 6 ). B. lane 1, size standard (Table 6 ); lanes 2 & 3, time point 5 (30 hr); lanes 4 & 5, time point 6 (36 hr); lanes 6 & 7, time point 7 (42 hr); lanes 8 & 9, time point 8 (48 hr); lanes 10 & 11, time point 9 (54 hr); lane 12, size standard (Table 6 ). C. lanes 1 & 2, size standard (Table 6 ); lanes 3 & 4, time point 10 (60 hr); lanes 5 & 6, time point 11 (66 hr); lanes 7 & 8, time point 12 (72 hr); lanes 9 & 10, time point 13 (78 hr); lanes 11 & 12, time point 14 (84 hr; DNA did not appear on this exposure, but did show up at longer exposures, and showed the same pattern of hybridization as all other lanes of DNA); lane 13, size standard (Table 6 ). D. lane 1, size standard (Table 6 ); lanes 2 & 3, time point 15 (96 hr); lanes 4 & 5, time point 16 (108 hr); lanes 6 & 7, time point 17, (120 hr); lane 8, size standard (Table 6 ).

198 -J 00 00 N3 lu In O ôo I «41 * 4 « size std. 0 hr (BamHI] 0 hr (SstI)

6 hr (BamHI]

6 hr (SstI)

12 hr (BamHI)

12 hr (SstI)

18 hr (BamHI)

18 hr (SstI)

24 hr (BamHI)

24 hr (SstI)

size std ! w yt V î* !» s to é t lé é • « size std. 30 hr (BamHI) 30 hr (SstI) 36 hr (BamHI) 36 hr (SstI) 42 hr (BamHI)

A'.. 42 hr (SstI)

;.i^'48 hr (BamHI) If).' 48 hr (SstI) ! ^ ^ , 5 4 hr (BamHI) 54 hr (SstI)

size std. t I

661 4 H-

3 ro @%k'(size std. 50 hr (BamHI) 60 hr (SstI) I 66 hr (BamHI)

66 hr (SstI) I n sa: V W i» 72 hr (BamHI)

72 hr (SstI) 78 hr (BamHI) 78 hr (SstI) 84 hr (BamHI) 84 hr (SstI)

size std. SJ 00 00 M "fk LM \0 00

size std. 96 hr (BamHI)

96 hr (SstI) 108 hr (BamHI) 108 hr (SstI)

120 hr (BamHI)

120 hr (SstI) I M l j^size std.

ooz 201 a gene would contain DNA encoding the putative peptide precursor. This gene would then be transcribed and translated and the resulting peptide modified to the desired endproduct by enzymatic reactions. In the latter case, the peptide precursor would be synthesized from a complex synthetase, probably a thioesterase using the 4'-phosphopantetheine cofactor to transport the growing peptide chain from one domain to another domain within the enzyme or enzyme complex

(see the Chapter II for the complete discussion). Previous attempts [81,128] to determine the mode of synthesis of these thiopeptide antibiotics have given conflicting answers. Houck

[81] found that chloramphenicol temporarily inhibited the synthesis of nosiheptide, while Mocek and Floss [128] were unable to find such inhibition of biosynthesis. The work described here was an effort to resolve this conflict by using a more comprehensive analysis of the antibiotic biosynthesis.

A. Oligonucleotide Analvsis

The results of the study with oligonucleotides which could possibly hybridize with the consensus binding sequence of 4'-phosphopantetheine have been discussed previously in this chapter. To summarize, oligonucleotide 4PAS1 (Table 8 ) was found to hybridize to multiple DNA fragments in multiple strains of streptomycetes. This indicates, that multiple genes encoding proteins which putatively are able to bind the cofactor 4'-phosphopantetheine exist in these strains. The 202

functions of these genes are unknown, but it is not unrealistic to suppose that synthases and synthetases are

among these genes (especially fatty acid synthase), and that

in the thiopeptide producers, an enzyme (possibly a thioesterase) which constructs the peptide precursor is one of the hybridizing genes. Other genes might encode polyketide

synthases which also utilize 4'-phosphopantetheine

[6,13,154,163]. S. azureus, S. hawaliens is, and S. glaucogriseus may produce polyketide-derived compounds for

incorporation into their spores (personal observation), and S, coelicolor, and S. lividans TK24 certainly do [24,27,163].

Two other oligonucleotide probes, TSRl and NOSIl, were used in an attempt to find DNA fragments in S. azureus and S. actuosus which could encode the peptide precursors of thiostrepton and nosiheptide, respectively. The sequences of these oligonucleotides are shown in Table 8 . Hybridization of these oligonucleotides to DNA fragments from S. azureus and S. actuosus would suggest strongly that the thiopeptide antibiotics were synthesized via ribosomal mechanisms.

Several attempts were made to hybridize the ^^P-labelled oligonucleotides to BamHI-digested DNA fragments isolated from

S. azureus and S. actuosus. Even using low-stringency washing conditions (2 washes of 2X SSPE, 0.1% SDS at RT followed by exposure to film for autoradiography), and long exposures (21 days) of the nitrocellulose filters to film at -70°C, no hybridization of the oligonucleotides to the chromosomal DNA 203

fragments could be seen. The data suggest, then, that no sequences in the chromosomal DNA of either S. azureus or S. actuosus exist to which these oligonucleotides can hybridize.

Though this is a negative result and cannot stand by itself, the result may show that the ribosomal mechanism of synthesis of the thiopeptide antibiotics is not used.

B. End-Point Protein-Thiopeotide Antibiotic

Labelling Experiment

The end-point analysis experiment, studying the labelling of cellular proteins and thiopeptide antibiotics, was conducted as described in the methodology chapter. The dry weights of the cells from this experiment and the net amount of antibiotic recovered are shown in Table 15, while the radioactive counting data of TCA-precipitable materials are shown in Table 16, and the amount of incorporated into thiopeptide antibiotic is shown in Table 17.

The data from this first labelling experiment do not show conclusively that the production of thiostrepton and nosiheptide are unaffected by chloramphenicol, but the data do imply that chloramphenicol has no effect on the antibiotic synthesis, since the differences in incorporated radioisotope

(nCi/mg; Table 15) are small. The one exception, Saz-1 with

25 ng/mL added chloramphenicol, may be a case of sampling error. A 20 /xg sample of each of the antibiotics was separated using thin-layer chromatography as described in the 204

Table 15: Dry Weights of Cells and Net Weights of Antibiotic Recovered

Organism Amount of Added Dry Weight Amount of Tested® Chloramphenicol^ of Cells Antibiotic (gm/L) (mg)

Saz-1 0 pg/mL 8.158 13.1

Saz-1 25 /xg/mL 7.780 11.7

Saz-1 125 fig/xaL 8.059 11.6

Sac-1 0 /xg/itiL 8.640 15.5

Sac-1 25 /itg/roL 9.452 13.8

Sac-1 125 /xg/mL 9.245 12.5

“Saz-1 is wild-type S. azureus. Sac-1 is wild-type S. actuosus.

^All the chloramphenicol added was in total volume of 100 nL of 95% ethanol. For the 0 jug/mL added chloramphenicol, 100 juL of ethanol was added to the flask. 205

Table 16: Radioactivity Incorporated into TCA-insoluble Material

Organism Amount of Added avg. cpm of TCA- average Tested® Chloramphenicol precipitables nCi/mL^

Saz-1 0 ng/mL 98623.22 ± 16447.50 53

Saz-1 25 jug/mL 20642.68 ± 2974.06 11

Saz-1 125 jLtg/mL 18884.68 ± 864.41 10

Sac-1 0 ffg/mL 97591.16 ± 11441.89 52

Sac-1 25 jug/mL 29628.68 ± 298.69 16

Sac-1 125 pg/mL 32391.52 ± 9732.72 17

*^Saz-l is wild-type S. azureus. Sac-1 is wild-type S. actuosus.

^The cpm were the average of three 1.0 mL aliquots. The counter had an efficiency at this counting of 0.8500. The average cpm were converted to dpm by dividing the average cpm by the efficiency. The dpm were converted to nCi by the following relationship:

1.0 nCi = 2.2 10^ dpm 206

Table 17: Amount of Radioactivity Incorporated into Antibiotics

Organism Amount of Added cpm of 50% nCi/mg Tested* Chloramphenicol of sample’^’ Incorporated®

Saz-1 0 jug/mL 169 1.4 10-=

Saz-1 25 /ig/mL 46 4.2-10”=

Saz-1 125 /Ltg/mL 118 1 .1 -10“=

Sac-1 0 fjig/nL 423 2.9-10”=

Sac-1 25 /ng/mL 471 3.6-10”=

Sac-1 125 /ig/mL 379 3.2-10”=

“Saz-1 is wild-type S. azureus. Sac-1 is wild-type S. actuosus.

^The cpm were from an aliquot of 50% of the total recovered, purified sample of antibiotic. Total cpm in the sample was presumed to be twice the number shown. The counter had an efficiency at this counting of 0.8500. The average cpm were converted to dpm by dividing the average cpm by the efficiency. The dpm were converted to nCi by the following relationship:

1.0 nCi = 2.2 10= dpm

The nCi/mg figure was arrived at by dividing the total nCi of antibiotic incorporated by the total amount of antibiotic recovered. 207 methodology section. Autoradiography (exposure for 20 days) of the radiolabelled antibiotics revealed the presence of radioactively-labelled compounds comigrating with nonradioactive standards of the antibiotics (data not shown).

The major problems with the data are; (i) the small amount of antibiotic recovered; (ii) the quite low level of ==8 incorporation into the antibiotic; and (iii) the analysis of the data at the end of the fermentation rather than by sampling at defined intervals in the fermentation.

Furthermore, this experiment was only preliminary and was not considered to provide the final answer to the mode of antibiotic synthesis.

C. Time-Course Protein-Thiooeotide Antibiotic

Labelling Experiment

In order to correct for the deficiencies of the end-point analysis of ^®S-labelling of protein and thiopeptide antibiotics, a new experiment to study the potential inhibition by chloramphenicol of incorporation of into the thiopeptide antibiotics was designed, as described in the chapter on methods. The incorporation of radiolabelled cysteine into both total protein and thiopeptide antibiotic

(in the absence or presence of 125 pg/mL chloramphenicol, freshly dissolved in 95% ethanol) was studied over time for the streptomycetes, S. azureus, S. actuosus, and S. laurentii.

The method of rapidly extracting antibiotic from the cells. 208 described in the methods chapter, was used, without further purification, since nearly all of the extracted material appeared to be the thiopeptide antibiotics.

There were potential problems with this time-course analysis of the thiopeptide-producing streptomycetes. First, despite the use of equal amounts of cells from starter culture, the cells did not grow at equal rates. As seen in

Figure 27, the growth rate of the cells for S. actuosus and S. azureus, as measured by total protein analysis, are not equal prior to the addition of chloramphenicol. Second, the total protein analysis procedure itself had problems. The cells of

S. laurentii produce dispersed mycelia in almost every liquid media in which they will grow (personal observation). The cells of S. azureus and S. actuosus, however, tend to form tightly clumped mycelia in some media. In the medium DPM3, both S. azureus and S. actuosus form tightly clumped mycelial balls. When the cells were lysed by heating with 0.1 N NaOH, not all of the mycelial balls of these organisms ruptured. It is not known, however, if all of the protein inside these balls actually was extracted into the NaOH. It is doubtful that it did, as the mycelial balls could be stained to a faint bluish color when placed in Coomassie stain (not shown).

Nevertheless, the consistency of the method of protein analysis did yield curves which approximate the growth curve one would normally expect. Figure 27: Total protein analysis as an approximation of the growth curve of S. laurentii, S. actuosus, and S. azureus. Two l.O mL aliquots of these strains were taken from each culture flask and processed to analyze the total protein (as mg/mL) in the cultures. The processing for total protein was conducted as described in the methods section, and the actual concentration of the protein was obtained by comparing the absorbance at 595 nanometers of the 1.0 mL aliquots against an absorbance curve constructed by measuring the absorbance (at 595 nm) of proteins of known concentration, also as described in the methods section. Each data point of each curve represents the average value for total protein derived from the values obtained for each of the two aliquots. A. The total protein of cultures of S. laurentii over time. Filled diamonds represent samples drawn from the flask to which no chloramphenicol was added; Open diamonds represent samples drawn from the flask to which chloramphenicol was added at a final concentration of 125 /xg/mL. B. The total protein of cultures of S. actuosus (filled and open squares) and S. azureus (filled and open circles). Filled symbols represent samples from the flasks to which no chloramphenicol was added. Open symbols represent samples from the flasks to which chloramphenicol was added to a final concentration of 125 /xg/mL.

209 210

? 1.000::

0.010 -- 1.000 -:

0.010

Time in Hours

Figure 27 211

The data of radioactivity counting of glass fiber filters onto which TCA-precipitable materials from the cells were adsorbed (Fig. 28), show that the addition of chloramphenicol to the culture medium strongly depresses the protein synthesis of S. laurentii. Though the depression of protein synthesis in S. actuosus and S. azureus is not as strong as with the previous experiment, some depression of protein synthesis is seen in these strains. It is also known that protoplasts of

S. actuosus and S. azureus will not regenerate in the presence of 125 /xg/mL chloramphenicol; therefore, chloramphenicol is believed to be capable of inhibiting protein synthesis in these strains. When the radiolabelled antibiotic extracts were counted by scintillation originally, using ca. 80% of the antibiotic extract, the synthesis of thiostrepton and nosiheptide appeared noticeably depressed in the flasks to which chloramphenicol was added (Fig. 29). The counts per minute (cpm) of radioactivity were extremely high (on the order of lO'* to 1 0® cpm) for all samples, however, and as described in the methods chapter, correction for quenching of counts by the samples could not be performed. By using smaller aliquots of the antibiotic extracts (10% of S. actuosus and S. azureus and 1% to 2% of S. laurentii), it was possible to reduce the counts to an amount which could be assayed for the quenching of the signal by the sample.

Correction of quench was then performed as described in the methods section. The amount of radioactivity in the total Figure 28: Radioactivity, as cpm, in the TCA-precipitable material collected on glass fiber filters. Six hours after the addition of ^^S-L-cysteine to all of the culture flasks, two 1.0 mL aliquots of each flask were taken to assay the amount of radioactivity from the radiolabelled cysteine which could be incorporated into TCA-precipitable material. The aliquots were processed and adsorbed onto glass fiber filters as described in the methodology chapter. After baking the filters and adding scintillation cocktail, the filters were counted. The efficiency of the counter in the window of the counter was determined to be ca. 0.8712. Each data point of these plots represents the average counts of two filters of TCA precipitable material. A. The cpm of TCA precipitable material of cultures of S. laurentii over time. Filled diamonds represent aliquots drawn from the flask to which no chloramphenicol was added; Open diamonds represent samples drawn from the flask to which chloramphenicol was added at a final concentration of 125 ^g/mL. B. The cpm of TCA precipitable material of cultures of S. actuosus (filled and open squares) and S. azureus (filled and open circles). Filled symbols represent samples from the flasks to which no chloramphenicol was added. Open symbols represent samples from the flasks to which chloramphenicol was added to a final concentration of 125 /ug/mL.

212 213

5.000E5-- A •

2.000E5--

" 1.000E5--

0.000 m 2.000E5--

Ô 1.500E5-- L.

1.000E5--

5.000E4--

0.000 5466 72 7860 84 Time in Hours

Figure 28 Figure 29: Total cpm of the ^^S-L-cysteine incorporated into nosiheptide and thiostrepton extracted from 20 mL aliquots of cultures in medium DPM3 at specific time points after the addition of radiolabelled cysteine. The samples here represent ca. 80% of the total antibiotic extracted from the aliquots taken from the flasks. The procedure for this experiment is detailed in the methods section. A. Filled diamonds represent radiolabelled antibiotic extracted from the flask to which no chloramphenicol was added. Open diamonds represent radiolabelled antibiotic extracted from the flask to which 125 jitg/mL (final concentration) of chloramphenicol was added. B. Squares represent the radiolabelled nosiheptide extracted from S. actuosusj circles represent the radiolabelled thiostrepton extracted from S. azureus. Filled squares and circles represent extracts of antibiotic from flasks to which no chloramphenicol was added, while open squares and circles represent extracts of antibiotic from flasks to which a final concentration of 125 Hg/mL chloramphenicol was added.

214 cpm in Antibiotic Extract cpm in Antibiotic Extract o o o § o g o o o o o o o m m m

I to VO

m -

00 to H* oi 216 sample was calculated from this data, with corrections for both quench and counter efficiency. From the amount calculated, the nanocuries (nCi) incorporation of ^^S-L- cysteine into the antibiotic could be calculated. This amount of incorporation was then divided by the amount of protein measured in the sample to get a value of nCi of incorporation per mg of total protein. When these data are then plotted graphically (Fig. 30), the amount of radioactivity incorporated into the antibiotic per mg of protein is comparable for S. actuosus and S. azureus in the absence or presence of chloramphenicol.

These data weakly suggest that chloramphenicol does not inhibit the synthesis of antibiotic in these organisms. The same cannot be said, however, for the synthesis of thiostrepton in S. laurentii. In this case, graphical analysis of the data indicates that thiostrepton production in

S. laurentii is strongly depressed by the addition of chloramphenicol. It may be argued that this indicates that the antibiotic is synthesized by ribosomal mechanisms in S. laurentii and by enzymatic mechanisms in S. azureus. This would not make much sense from the standpoint of evolution.

One may also argue that the data in S. laurentii are flawed either with respect to total protein analysis or with respect to antibiotic synthesis. This is possible, but not directly provable without a complete repeat of the experiment on S. laurentii. It can be more easily argued, however, that the Figure 30; Total incorporated into thiostrepton and nosiheptide in the presence and absence of chloramphenicol. The data were generated as described in the text. The units of incorporation of radioactivity are given as nCi of incorporated, radiolabelled cysteine in the antibiotics per milligram of protein in the cultures, at each given time point sampled. A. Filled diamonds represent the total amount of radioactivity incorporated into thiostrepton per mg of protein in the culture of S. laurentii to which no chloramphenicol was added. Open diamonds represent the total amount of radioactivity incorporated into thiostrepton per mg of protein in the culture of S. laurentii to which chloramphenicol was added to a final concentration of 125 /xg/mL. B. Filled squares represent the total amount of radioactivity incorporated into nosiheptide per mg of protein in the culture of S. actuosus to which no chloramphenicol was added. The open circles represent the data from the culture of S. actuosus to which a final concentration of 125 jug/itiL of chloramphenicol was added. Filled and open circles represent the analogous data for the two cultures of S. azureus.

217 218

75.000 D»

■g 50.000-- e-

(/) 25.000 - -

0.000 ? 15.000-- s 12.000 X)

S 9.000

6.000 -

0.000 54 60 66 72 78 84 Time in Hours

Figure 30 219 effect observed in S. laurentii represents neither a different method of synthesis of the antibiotic nor seriously flawed data. One could reasonably speculate that the data represents the rapid loss of some critical, highly labile enzyme involved in thiostrepton biosynthesis. Indeed, it can be seen from thin-layer chromatography data (Figures 31, 32, and 33) that although the production of antibiotic in the flasks to which chloramphenicol was added is greatly depressed, some of the antibiotic does appear in some of the lanes of extracts with added chloramphenicol.

When all the data from the various experiments studying the method of biosynthesis of the thiopeptide antibiotics are considered together, a firm conclusion is difficult to obtain; however, the weight of the available evidence suggests, at least in S. actuosus and S. azureus, that an enzymatic mechanism is used to synthesize nosiheptide and thiostrepton, respectively in these organisms. The data generated by oligonucleotides probes 4PÀS1, TSRl, and NOSIl show: (i) that multiple sequences of DNA exist in the chromosomes of the thiopeptide antibiotic producing strains which putatively encode proteins which bind the enzyme cofactor 4'- phosphopantetheine; and (ii) no sequences of DNA can be found in these same strains which could encode the putative peptide precursors of the antibiotics. The data from the end-point analysis of radioactivity incorporated into antibiotics also suggest that chloramphenicol does not inhibit biosynthesis of Figure 31: Thin-layer chromatography of S. actuosus antibiotic extracts over time. The thin-layer chromatography was performed as described in the methods section. The solvent system was CHCl^zn- heptane: isopropanol (5:5:3). Each lane represents 1 0 % of the total antibiotic extracted. The indicated times are measured from the time in hours past the inoculation of the cultures. From left-to-right the lanes are as follows: 55 hours, chloramphenicol added; 55 hours, no chloramphenicol added; 60 hours, chloramphenicol added; 60 hours, no chloramphenicol added; 66 hours, chloramphenicol added; 66 hours, no chloramphenicol added; 72 hours, chloramphenicol added; 72 hours, no chloramphenicol added; 78 hours, chloramphenicol added; 78 hours, no chloramphenicol added; 83 hours, chloramphenicol added; 83 hours, no chloramphenicol added. After elution, the plate was air-dried for several hours. The spots in each lane of the plate then were visualized under UV light. Flanking the test samples on the plate, 1 ng of authentic nosiheptide had been spotted and eluted. These two spots were then used to find the spots of the extracts which were nosiheptide. The location of the nosiheptide, as determined from the authentic sample, is shown in the figure.

220 221

55 hrs. 60 hrs. 66 hrs. 72 hrs. 78 hrs 83 hrs

« I NSH i

Figure 31 Figure 32: Thin-layer chromatography of S. azureus antibiotic extracts over time. The thin-layer chromatography was performed as described in the methods section. The solvent system was CHClgZn- heptane: isopropanol (5:5:3). Each lane represents 10% of the total antibiotic extracted. The indicated times are measured from the time in hours past the inoculation of the cultures. From left-to-right the lanes are as follows: 55 hours, chloramphenicol added; 55 hours, no chloramphenicol added; 60 hours, chloramphenicol added; 60 hours, no chloramphenicol added; 66 hours, chloramphenicol added; 66 hours, no chloramphenicol added; 72 hours, chloramphenicol added; 72 hours, no chloramphenicol added; 78 hours, chloramphenicol added; 78 hours, no chloramphenicol added; 83 hours, chloramphenicol added; 83 hours, no chloramphenicol added. After elution, the plate was air-dried for several hours. The spots in each lane of the plate then were visualized under UV light. Flanking the test samples on the plate, 1 ng of authentic thiostrepton had been spotted and eluted. These two spots were then used to find the spots of the extracts which were thiostrepton. The location of the thiostrepton, as determined from the authentic sample, is shown in the figure.

222 223

+ hrs. 72 hrs. 78 hrs. 83 hrs. "^~ + — + — + —

-TSR t # f -■s

Figure 32 Figure 33: Thin-layer chromatography of S. laurentii antibiotic extracts over time. The thin-layer chromatography was performed as described in the methods section. The solvent system was CHClsZn- heptane: isopropanol (5:5:3). Each lane represents 10% of the total antibiotic extracted. The indicated times are measured from the time in hours past the inoculation of the cultures. From left-to-right the lanes are as follows: 55 hours, chloramphenicol added; 55 hours, no chloramphenicol added; 60 hours, chloramphenicol added; 60 hours, no chloramphenicol added; 66 hours, chloramphenicol added; 66 hours, no chloramphenicol added; 72 hours, chloramphenicol added; 72 hours, no chloramphenicol added; 78 hours, chloramphenicol added; 78 hours, no chloramphenicol added; 83 hours, chloramphenicol added; 83 hours, no chloramphenicol added. After elution, the plate was air-dried for several hours. The spots in each lane of the plate then were visualized under UV light. Flanking the test samples on the plate, 1 jug of authentic thiostrepton had been spotted and eluted. These two spots were then used to find the spots of the extracts which were thiostrepton. The location of the thiostrepton, as determined from the authentic sample, is shown in the figure.

224 225

55 hrs. 60 hrs. 66 hrs. 72 hrs. 78 hrs. 83 hrs. + — + + — + —

i-r

TSR

Figure 33 226 thiostrepton and nosiheptide. The data from the time-course experiment measuring incorporation of radioactivity into thiopeptide antibiotics make a weak case for an enzymatic mechanism of synthesis in S. actuosus and S. azureus, and make a case for a ribosomal mechanism of synthesis of thiostrepton in 5. laurentii. A clear answer to the question of the mechanism of the biosynthesis of these antibiotics may come from future experiments. CHAPTER VI: SUMMARY AND CONCLUSIONS

The thiopeptide antibiotics, thiostrepton and nosiheptide, are complex, highly modified antibiotics which are important in agriculture, in veterinary medicine, and in studies of the molecular biology of streptomycetes. These antibiotics have a complex array of enzymatic reactions associated with their biosynthesis. The use of molecular biology to study the genes encoding regulatory, resistance, and biosynthetic processes in thiostrepton- and nosiheptide- producing organisms may allow future researchers to improve the synthesis of the antibiotics or to make rational, structural modifications to the antibiotics which could allow them to become useful in human medicine. This work laid important foundations upon which future studies of the thiopeptide antibiotics can build.

First, stable mutants, blocked in the production of thiostrepton, were created. These mutants will be important in the cloning of genes believed to be involved in the biosynthesis of thiostrepton and nosiheptide, since they can serve as recipients of the cloned DNA.

Second, additional data was added to the study of the mechanism of synthesis of thiostrepton and nosiheptide.

227 228

Although the question of the mechanism of synthesis was not clearly answered, the data generated in this study added to data already acquired and suggest that the mechanism used in the synthesis of thiostrepton and nosiheptide is an enzymatic one (i.e., synthesis of the antibiotic from the template of a synthetase rather than translation of a mRNA followed by extensive modification of the translation product). This particular study suggests that a different inhibitor of protein synthesis needs to be used (e.g., lincomycin rather than chloramphenicol) and that the study should be conducted over a shorter time interval with more frequent sampling (such as, an 8 hr. time-course, with sample intervals of 30 to 60 min.).

Third, a gene which might be involved in regulation of synthesis of the thiopeptide antibiotics was studied, and its homolog from S. azureus was cloned. The gene, nshA, cloned from S. actuosus, was site-specifically mutated in the producing organism, and the effect on antibiotic production was observed. Although the results should be considered preliminary, the data are suggestive that nshA is a negative regulator of nosiheptide biosynthesis in S. actuosus. If this is confirmed, this gene would be the first clearly characterized example of a negative regulator. Interestingly, sequences of DNA which hybridize to a probe made from nshA were found to be widely distributed among the streptomycetes, suggesting a broader role than just regulation of thiopeptide 229 antibiotic synthesis. The homolog of nshA was cloned from S. azureus. Also, the location of this homolog in the S. azureus chromosome, with respect to the thiostrepton resistance genes, was determined. Contrary to the situation in S. actuosus

(where nshA has been found to be transcriptionally linked to the nosiheptide resistance gene, nshR [109]), the nshA homolog in S. azureus was found to be located more than 20 kbp away from the thiostrepton resistance genes.

The fourth, and most significant, finding of this work was the discovery that plJlOl-like sequences of DNA exist in

S. azureus, apparently integrated into the chromosome, and that these plasmid sequences carry the two ribosomal methylase genes which provide resistance to thiopeptide antibiotics.

Three plasmids containing both a gene encoding a resistance methylase and a gene (or genes) encoding the plJlOl-like replication functions were cloned from the OSU strain of S. azureus. Plasmids pANT423 and pANT424 are small (5.0 kbp to

6.7 kbp), high-copy-number plasmids which appear to originate from DNA containing the 6.1 kbp and 4.5 kbp BamHI DNA fragments which hybridize to the probes tsrR, nshR, and rep

(from pIJlOl). Plasmid pANT428 appears to result from cloning of the 6.1 kbp BamHI DNA fragment just described into pKC505, followed by the incorrect excision of a plJlOl-like plasmid from pKC505. These plasmid DNA sequences in S. azureus were found to be integrated into the chromosome as the result of natural genetic exchange rather than laboratory contamination. 230 since these same sequences were found in a strain of S. azureus freshly procured from the ATCC.

The finding of the plasmid sequences in S. azureus calls into question many widely-held ideas regarding antibiotic synthesis in streptomycetes. It is no longer possible to know whether the genes encoding proteins involved in the biosynthesis of thiostrepton remain tightly clustered on the chromosome, as is the case for all other cloned biosynthetic pathways [24,27]. It is also tempting to speculate that ribosomal méthylation may not be the original mode of resistance to thiostrepton, but rather, it might be a fortuitous, later occurrence. Finally, Southern hybridization experiments showed sequences of DNA in strains of streptomycetes other than S. azureus which hybridize to the plJlOl-specific probes. It is possible that this represents a case of broad, natural, genetic exchange of plJlOl-like sequences among the streptomycetes. This, also, if true, represents an important finding with powerful implications for the future of molecular biology in strains of Streptomyces spp. LITERATURE CITED

1. Adamidis, T., P. Riggle, and W. Champness. 1990. Mutations in a new Streptomyces coelicolor locus which globally block antibiotic biosynthesis but not sporulation. J. Bacteriol. 172: 2962-2969.

2. Anderson, B., D.C. Hodgkin, and M.A. Viswamitra. 1970. The structure of thiostrepton. Nature 225: 233-235.

3. Anke, T. and F. Lipmann. 1977. Studies on the biosynthesis of valinomycin. FEES Letters 82: 337-340.

4. Babcock, M. J. and K. E. Kendrick. 1988. Cloning of DNA involved in sporulation of Streptomyces griseus. J. Bacteriol. 170: 2802-2808.

5. Bartel, P.L. 1989. Analysis of natural and hybrid antibiotic formation in anthracycline-producinq Streptomvcetes. PhD Dissertation. Columbus, OH: The Ohio State University.

6 . Bëhal, V. 1982. Oligoketide-synthesizing enzymes, in Krumphanzl, V., B. Sikyta, and Z. Vanëk (Eds.), Overproduction of Microbial Products. New York: Academic Press, 301-309.

7. Benazet, F., M. Cartier, J. Florent, C. Godard, G. Jung, J. Lunel, D. Mancy, C. Pascal, J. Renaut, P. Tarridec, J. Theilleux, R. Tissier, M. Dubost, and L. Ninet. 1980. Nosiheptide, a sulfur-containing peptide antibiotic isolated from Streptomyces actuosus 40037. Experientia 36: 414-416.

8 . Bibb, M.J., M.J. Bibb, J.M. Ward, and S.N. Cohen. 1985. Nucleotide sequences encoding and promoting expression of three antibiotic resistance genes indigenous to Streptomyces. Mol. Gen. Genet. 199: 26-36.

9. Bibb, M.J., P.R. Findlay, and M.W. Johnson. 1984. The relationship between base composition and codon usage in bacterial genes and its use for the simple and reliable identification of protein-coding sequences. Gene 30: 157-160.

231 232

10. Bibb, M.J., G.R. Janssen, and J.M. Ward. 1985. Cloning and analysis of the promoter region of the erythromycin- resistance gene (ermE) of Streptomyces erythraeus. Gene 38: 215-226; and Erratum in 41 (1986) E357-E368.

11. Bibb, M.J. and G.R. Janssen. 1986. Unusual features of transcription and translation of antibiotic resistance genes in antibiotic-producing Streptomyces. in Alacevic, M., D. Hranueli, and Z. Toman (Eds), Genetics of Industrial Microorganisms. Proceedings of the Fifth International Symposium on the Genetics of Industrial Microorganisms, part B . Karlovac, Yugoslavia: Ognjen Prica Printing Works, pp 309-318.

12. Bibb, M.J., J.M. Ward, T. Kieser, S.N. Cohen, and D.A. Hopwood. 1981. Excision of chromosomal DNA sequences from Streptomyces coelicolor forms a novel family of plasmids detectable in Streptomyces lividans. Mol. Gen. Genet. 184: 230-240.

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

14. Birge, E.A. 1981. Chapter 3. Mutations and mutagenesis. in Bacterial and Bacteriophage Genetics. New York: Springer-Verlag, pp 40-66.

15. Birk, R. (ATCC Bacteriology Laboratory) 1991. personal communication.

16. Boccard, F., J. Pernodet, A. Friedmann, and M. Guérineau. 1988. Site-specific integration of plasmid pSAM2 in Streptomyces lividans and S. ambofaciens. Mol. Gen. Genet. 212: 432-439.

17. Boccard, F., T. Smokvina, J. Pernodet, A. Friedmann, and M. Guérineau. 1989. Structural analysis of loci involved in pSAM2 site-specific integration in Streptomyces. Plasmid 21: 59-70.

18. Buttner, M.J. 1989. RNA polymerase heterogeneity in Streptomyces coelicolor A3(2). Mol . Microbiol. 3 : 1653-1659.

19. Buttner, M.J., K.F. Chater, and M.J. Bibb. 1990. Cloning, disruption, and transcriptional analysis of three RNA polymerase sigma factor genes of Streptomyces coelicolor A3(2). J. Bacteriol. 172: 3367-3378.

20. Carlton, B.C. and B.J. Brown. 1981. Gene Mutation, in Gephardt, P. (Ed.), Manual of Methods for General 233

Bacteriology. Washington, DC: American Society for Microbiology, pp 222-242.

21. Chater, K.F. 1972. A morphological and genetic mapping study of white colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 169: 5715-5720.

22. Chater, K.F. 1989. Multilevel regulation of Streptomyces differentiation. Trends in Genetics 5(11): 372-377.

23. Chater, K.F. 1989. Sporulation in Streptomyces. in Regulation of Procaryotic Development. Smith, I., R.A. Slepecky, and P. Setlow (Eds). Washington, DC: American Society for Microbiology, pp 277-299.

24. Chater, K.F. 1990. The improving prospects for yield increase by genetic engineering in antibiotic-producing Streptomycetes. Bio/Technol. 8 : 115-121.

25. Chater, K.F., C.J. Bruton, K.A. Plaskitt, J.J. Buttner, C. Mendez, and J.D. Helmann. 1989. The developmental fate of Streptomyces coelicolor hyphae depends crucially on a gene product homologous with the motility sigma factor of Bacillus subtilus. Cell 59: 133-143.

26. Chater, K.F., D.J. Henderson, M.J. Bibb, and D.A. Hopwood. 1988. Genome flux in Streptomyces coelicolor and other streptomycetes and its possible relevance to the evolution of mobile antibiotic resistance determinants, in Kingsman, A.J., S.M. Kingsman, and K.F. Chater (Eds.), Transposition. Society for General Microbiology Symposium, 43. Cambridge: Cambridge University Press, pp 7-42.

27. Chater, K.F. and D.A. Hopwood. 1989. Antibiotic biosynthesis in Streptomyces. in Genetics of Bacterial Diversity. Hopwood, D.A., and K.F. Chater (Eds). London: Harcourt Brace Jovanovich, Publishers. pp 129-150.

28. Chater, K.F. and M.J. Merrick. 1979. Streptomycetes. in Developmental Biology of Prokaryotes, Parish, J.H. (Ed.). Oxford: Blackwell Scientific, pp 93-114.

29. Cho, H., T. Horisaka, H. Ohkishi, and H. Ogawara. 1989. in Koyama, Y . (Ed.), Trends in Actinomvcetology in Japan. Actinomycetologica Forum 1989. Tokyo: Soc. Actinomycetes, Japan, pp 85-86.

30. Connors, N.C. 1990. personal communication. 234

31. Cron, M.J., D.F. Whitehead, L.R. Hooper, B. Heinemann, and J. Lein. 1956. Bryamycin, a new antibiotic. Antibiotics and Chemotherapy 6 : 63-67.

32. Cundliffe, E. 1971. The mode of action of thiostrepton in vivo. Biochem. Biophys. Res. Commun. 44: 912-917.

33. Cundliffe, E. 1978. Mechanism of resistance to thiostrepton in the producing-organism Streptomyces azureus. Nature 272: 792-795.

34. Cundliffe, E. 1986. Mechanisms of self defence in antibiotic-producing organisms. in Alacevic, M., D. Hranueli, and Z. Toman (Eds), Genetics of Industrial Microorganisms. Proceedings of the Fifth International Symposium on the Genetics of Industrial Microorganisms, part B . Karlovac, Yugoslavia: Ognjen Prica Printing Works, pp 199-206.

35. Cundliffe, E., P. Dixon, M. Stark, G. Stoffler, R. Ehrlich, M. Stoffler-Meilicke, and M. Cannon. 1979. Ribosomes in thiostrepton-resistant mutants of Bacillus megaterium lacking a single 50 S subunit protein. J. Mol. Biol. 132: 235-252.

36. Cundliffe, E., R.H. Skinner, and J. Thompson. 1984. Resistance to antibiotics in antibiotic-producing organisms. in Biological. Biochemical. and Biomedical aspects of Actinomvcetes. Ortiz-Ortiz, L., L.F. Bojalil, and V. Yakoleff (Eds). Orlando: Academic Press, Inc., pp 303-314.

37. Danner, D.B. 1982. Recovery of DNA fragments from gels by transfer to DEAE-paper in an electrophoresis chamber. Anal. Biochem. 125: 139-142.

38. Dary, A., B. Simonet, J.-M. Simonet, and B. Decaris. 1989. Clonage d/un gène de résistance au nosiheptide de Streptomyces actuosus. C.R. Acad. Soi. Paris 308: 35-41.

39. Davies, J. 1986. A new look at antibiotic resistance. FEMS Microbiol. Reviews 39: 363-371.

40. Davis, B.D. and W.K. Maas. 1952. Analysis of the biochemical mechanism of drug resistance in certain bacterial mutants. Proc. Nat'l. Acad. Sci. USA 38: 775-785.

41. Dekleva, M.L., J.A. Titus, and W.R. Strohl. 1985. Nutrient effect on anthracycline production by Streptomyces peucetius in a defined medium. Can. J. Microbiol. 31: 287-294. 235

42. Delîc, L., D.A. Hopwood, and E.J. Friend. 1970. Mutagenesis by N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) in Streptomyces coelicolor. Mutation Res. 9: 167-182.

43. Demain, A.L. 1981. Industrial Microbiology. Science 214: 987-995.

44. Deng, Z., T. Kieser, and D.A. Hopwood. 1988. "Strong incompatibility" between derivatives of the Streptomyces multi-copy plasmid pIJlOl. Mol. Gen. Genet. 214: 286-294.

45. Dosch, D.C., W.R. Strohl, and H.G. Floss. 1988. Molecular cloning of the nosiheptide resistance gene from Streptomyces actuosus ATCC 25421. Biochem. Biophys. Res. Commun. 156: 517-523.

46. Dowson, C.G., A. Hutchison, J.A. Brannigan, R.C. George, D. Hansman, J. Linares, A. Tomasz, J.M. Smith, and B.G. Spratt. 1989. Horizontal transfer of penicillin-binding protein genes in penicillin-resistant clinical isolates of Streptococcus pneumoniae. Proc. Nat'l. Acad. Sci. USA 8 6 : 8842-8846.

47. Endo, T. and H. Yonehara. 1978. Identity of multhiomycin with nosiheptide. J. Antibiot. 31: 623-625.

48. Feinberg, A.P. and B. Vogelstein. 1983. A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132; 6-13.

49. Feitelson, J.S., F. Malpartida, and D.A. Hopwood. 1985. Genetic and biochemical characterization of the red gene cluster of Streptomyces coelicolor A3(2). J . Gen. Microbiol. 131: 2431-2441.

50. Floss, H.G., unpublished.

51. Floss, H.G. 1987. Hybrid antibiotics — the contribution of the new gene combinations. Trends in Biotechnology 5: 111-115 =

52. Floss, H.G. 1988. Biosynthesis of the antibiotics nosiheptide and thiostrepton. in Okami, Y., T. Beppu, and H. Ogawara (Eds.), Biology of Actinomycetes '8 8 . Tokyo: Japan Scientific Societies Press, pp 401-405.

53. Floss, H.G., and J.M. Beale. 1989. Biosynthetic studies on antibiotics. Anaewandte Chemie (International Edition in English) 28: 146-177. 236

54. Floss, H.G., J.M. Beale, D.C. Dosch, T. Frenzel, D.R. Houck, J.J. Lee, Y. Li, T. Kamigauchi, U. Mocek, D. O'Kagan, R. Tsuchiya, R. Woodman, L.-D. Yuen, Z. Zeng, P. Zhou, and W.R. Strohl. 1990. Biosynthesis of natural products — nosiheptide and thiostrepton. Proc. 5th FAOB Meet.. Seoul. South Korea fAua. 1989 . IN PRESS.

55. Floss, H.G., S.P. Cole, X.-G. He, B.A.M. Rudd, J. Duncan, J. Fujii, C.-j. Chang, and P.J. Keller. 1985. Biosynthesis of polyketide antibiotics. in Kleinkauf, H ., H , von Dohren, H . Dornauer, and G . Nesemann (Eds.), Regulation of Secondary Metabolite Formation. Weinheim, Germany; VCH Verlagsgesellschaft mbH, pp. 283-304.

56. Grant, S.R., S.C. Lee, K. Kendall, and S.N. Cohen. 1989. Identification and characterization of a locus inhibiting extrachromosomal maintenance of the Streptomyces plasmid SLPl. Mol. Gen. Genet. 217: 324-331.

57. Hale, R.S., K.N. Jordan, and P.F. Leadlay. 1987. A small, discrete acyl carrier protein is involved in de novo fatty acid biosynthesis in Streptomyces erythraeus. FEBS Letters. 224: 133-136.

58. Hale, R.S. and P.F. Leadlay. 1985. Oligonucleotide probes for bacterial acylcarrier protein genes. Biochimie 67: 835-839.

59. Hanahan, D. 1985. Techniques for transformation of E. coli. in Glover, D.M. (Ed.), DNA Cloning, a practical approach, volume I. Oxford: IRL Press, pp 109-135.

60. Hara, O. and T. Beppu. 1982. Induction of a streptomycin-inactivating enzyme by A-factor in Streptomyces griseus. J. Antibiot. 35: 1208-1215.

61. Hasegawa, M . , G. Hintermann, J.-M. Simonet, R. Crameri, J. Piret, and R. Butter. 1985. Certain chromosomal regions in Streptomyces glaucescens tend to carry amplifications and deletions. Mol. Gen. Genet. 200: 375-384.

62. Henderson, D.J., D.F. Brolle, T. Kieser, R.E. Melton, and D.A. Hopwood. 1990. Transposition of IS117 (the Streptomyces coelicolor A3(2) mini-circle) to and from a cloned target site and into secondary chromosomal sites. Mol. Gen. Genet. 224: 65-71.

63. Hensens, O.D. and G. Albers-Schonberg. 1983. Total structure of the highly modified peptide antibiotic components of thiopeptin. J . Antibiot. 36: 814-831. 237

64. Hershberger, C.L, 1988. Genetic structure and instability in antibiotic producing streptomycetes. in Okami, Y., T. Beppu, and H. Ogawara. Biology of Actinomycetes '8 8 . Tokyo; Japan Scientific Societies Press, pp 133-138.

65. Hintermann, G., R. Crameri, M. Vogtli, and R. Hütter. 1984. Streptomycin-sensitivity in Streptomyces glaucescens is due to deletions comprising the structural gene coding for a specific phosphotransferase. Mol. Gen. Genet. 196: 513-520.

6 6 . Hopwood, D.A. 1988. Understanding the genetic control of antibiotic biosynthesis and sporulation in Streptomyces. in Biology of Actinomvcetes '88. Okami, Y., T. Beppu, and H. Ogawara (Eds.). Tokyo: Japan Scientific Societies Press, pp 3-10.

67. Hopwood, D.A., M.J. Bibb, K.F. Chater, T. Kieser, C.J. Bruton, H.M. Kieser, D.J. Lydiate, C.P. Smith, J.M. Ward, and H. Schrempf. 1985. Genetic Manipulations of Streptomyces — A Laboratory Manual. Norwich, UK: The John Innes Foundation.

6 8 . Hopwood, D.A., G. Hintermann, T. Kieser, and H.M. Wright. 1984. Integrated DNA sequences in three streptomycetes form related autonomous plasmids after transfer to Streptomyces lividans. Plasmid 11: 1-16.

69. Hopwood, D.A., D.J. Lydiate, F. Malpartida, and H.M. Wright. 1985. Conjugative sex plasmids of Streptomyces. in Donald R. Helinski (Ed.), Plasmids in Bacteria. New York: Plenum Press, 615-631.

70. Hopwood, D.A., F. Malpartida, and K.F. Chater. 1986. Gene cloning to analyze the organization and expression of antibiotic biosynthesis genes in Streptomyces. in Kleinkauf, H., H. von Dohren, H. Dornauer, and G. Nesemann (Eds.), Regulation of Secondary Metabolite Formation. Weinheim, Germany: VCH Verlagsgesellschaft mbH, pp 23-33.

71. Horinouchi, S. and T. Beppu. 1986. A-factor and regulatory network that links secondary metabolism with cell differentiation in Streptomyces. in Alacevic, M . , D. Hranueli, and Z. Toman (Eds.), Genetics of Industrial Microorganisms. Proceedings of the Fifth International Symposium on the Genetics of Industrial Microorganisms. Part B ■ Karlovac, Yugoslavia: Ognjen Prica Printing Works, pp 41-48. 238

72. Horinouchi, S. and T. Beppu. 1988. Regulation of secondary metabolism in Streptomyces. in Biolocry of Actinomycetes '8 8 . Okami, Y., T. Beppu, and S. Ogawara (Eds). Tokyo: Japan Scientific Societies Press, pp 71-75.

73. Horinouchi, S., K. Furuya, M. Nishiyama, H. Suzuki, and T. Beppu. 1987. Nucleotide sequence of the streptothricin acetyltransferase gene from Streptomyces lavendulae and its expression in heterologous hosts. Bacteriol. 169: 1929-1937.

74. Horinouchi, S., 0. Hara, and T. Beppu. 1983. Cloning of a pleiotropic gene that positively controls biosynthesis of A-factor, actinorhodin and prodigiosin in Streptomyces coelicolor A3(2) and Streptomyces lividans. Bacteriol. 155: 1238-1248.

75. Horinouchi, S., Y. Kumada, and T. Beppu. 1984. Unstable genetic determinant of A-factor biosynthesis in streptomycin-producing organisms: cloning and characterization. J. Bacteriol. 158: 481-487.

76. Horinouchi, S., M. Nishiyama, H. Suzuki, Y. Kumada, and T. Beppu. 1985. The cloned Streptomyces bikiniensis A- factor determinant. J. Antibiot. 38: 636-641.

77. Horinouchi, S., H. Suzuki, and T. Beppu. 1986. Nucleotide sequence of afsB, a pleiotropic gene involved in secondary metabolism in Streptomyces coelicolor A3(2) and "Streptomyces lividans.” J. Bacteriol. 168: 257-269.

78. Horinouchi, S., H. Suzuki, M. Nishiyama, and T. Beppu. 1989. Nucleotide sequence and transcriptional analysis of the Streptomyces griseus gene {afsA) responsible for A-factor biosynthesis. J. Bacteriol. 171: 1206-1210.

79. Hornemann, U., C.J. Otto, G.G. Hoffman, and A.C. Bertinuson. 1987. Spectinomycin resistance and associated DNA amplification in Streptomyces achromogenes subsp. rubradiris. J. Bacteriol. 169: 2360-2366.

80. Hoshiko, S., c. Nojiri, K. Matsunaga, K. Katsumata, E. Satoh, and K. Nagaoka. 1988. Nucleotide sequence of the ribostamycin phosphotransferase gene and of its control region in Streptomyces ribosidificus. Gene 6 8 : 285-286.

81. Houck, D.R. 1986. Studies on the biosvnthesis of the modified-peptide antibiotic. nosiheptide. PhD Dissertation. Columbus, OH: The Ohio State University. 239

82. Houck, D.R., L.-C. Chen, P.J. Keller, J.M. Beale, and H.G. Floss. 1988. Biosynthesis of the modified peptide antibiotic nosiheptide in Streptomyces actuosus, J. Am. Chem. Soc. 110: 5800-5806.

83. Hunter, I.S. 1985. Gene cloning in Streptomyces. in Glover, D.M. (Ed.), DNA Cloning, a practical approach, volume II, Oxford: IRL Press, pp 19-44.

84. Ingram, L. 1970. A ribosomal mechanism for synthesis of peptides related to nisin. Biochim. Biophys. Acta 224: 263-265.

85. Kaiser, K. and N.E. Murray. 1985. in Glover, D.M. (Ed.), DNA Cloning, a practical approach, volume I. Oxford: IRL Press, pp 1-48.

8 6 . Kaletta, C., and K.-D. Entian. 1989. Nisin, a peptide antibiotic: cloning and sequencing of the nisA gene and posttranslational processing of its peptide product. J\_ Bacteriol. 171: 1597-1601.

87. Kaletta, C., K.-D. Entian, U. Schneider, F. Gotz, H. Zahner, R. Kellner, and G. Jung. 1988. Prepeptide sequence of epidermin, a ribosomally synthesized antibiotic with four sulphide-rings. Nature 333: 276-278.

8 8 . Kamagauchi, T., and H.G. Floss. 1989. unpublished results.

89. Kendall, K.J. and S.N. Cohen. 1988. Complete nucleotide sequence of the Streptomyces lividans plasmid pIJlOl and correlation of the sequence with genetic properties. Bacteriol. 170: 4634-4651.

90. Kieser, T., Z. Deng, and D.A. Hopwood. 1989. "Strong incompatibility" and co-integrate formation between derivatives of the Streptomyces multi-copy plasmid, pIJlOl. in Okami, Y., T. Beppu, and H- Ogawara (Eds.), Biology of Actinomycetes '8 8 . Tokyo: Japan Scientific Societies Press, pp 97-102.

91. Kieser, T., D.A. Hopwood, H.M. Wright, and C.J. Thompson. 1982. pIJlOl, a multi-copy broad host-range Streptomyces plasmid: functional analysis and development of DNA cloning vectors. Mol. Gen. Genet. 185: 223-228.

92. Kinashi, H. and M. Shimaji. 1987. Detection of giant linear plasmids in antibiotic producing strains of Streptomyces by the OFAGE technique. J. Antibiot. 40: 913-916. 240

93. Kinashi, H., M. Shimaji, and A. Sakai. 1987. Giant linear plasmids in Streptomyces which code for antibiotic biosynthesis genes. Nature. 328: 454-456.

94. Kirst, H.A., E.F. Szymanski, D.E. Dorman, J.L. Occolowitz, N.D. Jones, M.O. Chaney, R.L. Hamill, and M.M. Hoehn. 1975. Structure of althiomycin. J. Antibiot. 28: 286-291.

95. Kleinkauf, H. and H. von Dohren. 1982. A survey of enzymatic peptide formation, in Kleinkauf, H. and H. von Dohren (Eds), Peptide Antibiotics: Biosynthesis and Functions. Berlin: Walter de Gruyter, pp 3-21.

96. Kleinkauf, H., and H. von Dohren. 1983. Peptides, in Vining, L.C. (Ed.), Biochemistry and Genetic Regulation of Commercially Important Antibiotics. Reading, MA: Addison-Wesley Publishing Company, pp 95-145.

97. Kleinkauf, H. and H. von Dohren. 1986. Enzyme systems synthesizing peptide antibiotics, in Kleinkauf, H., H. von Dohren, H. Dornauer, and G. Nesemann (Eds.), Regulation of Secondary Metabolite Formation. Weinheim, Germany: VCH Verlagsgesellschaft mbH, pp 173-207.

98. Kleinkauf, H. and H. von Dohren. 1987. Biosynthesis of peptide antibiotics. Ann. Rev. Microbiol. 41: 259-289.

99. Komori, T., M. Ezaki, E. Kino, M. Kohsaka, H. Aoki, and H. Imanaka. 1983. Lavendomycin, a new antibiotic. I. Taxonomy, isolation and characterization. J. Antibiot. 38: 691-698.

100. Kratzschmar, J., M. Krause, and M.A. Marahiel. 1989. Gramicidin S biosynthesis operon containing the structural genes grsA and grsB has an open reading frame encoding a protein homologous to fatty acid thioesterases. J. Bacteriol. 171: 5422-5429.

101. Krause, M. and M.A. Marahiel. 1988. Organization of the biosynthesis genes for the peptide antibiotic gramicidin S. J. Bacteriol. 170: 4669-4674.

102. Kutscher, A.H., L. Seguin, R.M. Rankow, and J.D. Piro. 1958. Thiostrepton, a new antibiotic. Antibiot. Chem. Rev. 8 : 576-583.

103. Lawen, A. and R. Zocher. 1990. Cyclosporin synthetase. The most complex peptide synthesizing multienzyme polypeptide so far described. J. Biol. Chem. 265: 11355-11360. 241

104. Lawlor, E.J., H.A. Baylis, and K.F. Chater. 1987. Pleiotropic morphological and antibiotic deficiencies result from mutations in a gene encoding a tRNA-like product in Streptomyces coelicolor A3(2). Genes Develop. 1: 1305-1310.

105. Lawlor, E.J., B.K. Leskiew, B.L. Daly, and K.F. Chater, in preparation, cited in [2 2 ].

106. Lee, Y.H.W., Z.Y. Tzecheng, S.C. Wang, W.L. Cheng, and C.W. Chen. 1986. Structural stability of heterologous genes cloned in Streptomyces plasmid pIJ702. Biochem. Biophys. Res. Commun. 140: 372-378.

107. Leskiew, E.K., M. Mevarech, L.S. Barritt, S.E. Jensen, D.J. Henderson, D.A. Hopwood, C.J. Bruton, and K.F. Chater. 1990. Discovery of an insertion sequence, IS116, from Streptomyces clavuligerus and its relatedness to other transposable elements from actinomycetes. Gen. Microbiol. 136: 1251-1258.

108. Li, Y. 1989. unpublished work.

109. Li, Y., D.C. Dosch, W.R. Strohl, and H.G. Floss. 1990. Nucleotide sequence and transcriptional analysis of the nosiheptide-resistance gene from Streptomyces actuosus. Gene 91: 9-17.

110. Li, Y., D.C. Dosch, R.H. Woodman, H.G. Floss, and W.R. Strohl. 1991. Transcriptional organization and regulation of the nosiheptide resistance gene in Streptomyces actuosus, Dev. Industr. Microbiol. Submitted 7/24/90.

111. Li, Y. and W.R. Strohl. 1990. personal communication.

112. Lipmann, F. 1982. On biosynthesis of the cyclic antibiotics gramicidin S and tyrocidine, and of linear gramicidin, in Kleinkauf, H. and H. von Dohren (Eds), Peptide Antibiotics:____ Biosynthesis and Functions. Berlin: Walter de Gruyter, pp 23-45.

113. Lbpez-Cabrera, M., J.A. Pèrez-Gonzalez, P. Heinzel, W. Piepersberg, and A. Jimènez. 1989. Isolation and nucleotide sequencing of an aminocyclitol acetyltransferase gene from Streptomyces rimosus forma paromomycinus. J. Bacteriol. 171: 321-328.

114. Lydiate, D.J., F. Malpartida, and D.A. Hopwood. 1985. The Streptomyces plasmid SCP2*: its functional analysis and development into useful cloning vectors. Gene 35: 223-235. 242

115. Malik, V.S. March 1980. Microbial secondary metabolism. TIBS, pp. 68-72.

116. Malpartida, F. and D.A. Hopwood. 1984. Molecular cloning of the whole biosynthetic pathway of a Streptomyces antibiotic and its expression in a heterologous host. Nature 309; 462-464.

117. Maniatis, T., E.F. Fritsch, and J. Sambrook. 1982. Molecular Cloning — A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratories.

118. Marahiel, M.A., M. Karuse, and J.J. Skarpeid. 1979. Biological role of gramicidin S in spore function: studies on gramicidin S-negative mutants of Bacillus brevis ATCC 9999. Eur. J. Biochem. 99: 49-55.

119. Martin, J.F., and J.A. Gil. 1984. Cloning and expression of antibiotic production genes. Bio /Technology 2: 63-72.

120. Mazodier, P., C. Thompson, and F. Boccard. 1990. The chromosomal integration site of the Streptomyces element pSAM2 overlaps a putative tRNA gene conserved among actinomycetes. Mol. Gen. Genet. 222: 431-434.

121. Mendez, C. and K.F. Chater. 1987. Cloning of whiG, a gene critical for sporulation of Streptomyces coelicolor A3(2). J. Bacteriol. 169: 5715-5720.

122. Mikulik, K., I. Janda, J. Weiser, J. Stastna, and A. Jiranova. 1984. RNA and ribosomal protein patterns during aerial spore germination in Streptomyces granaticolor. Eur. J. Biochem. 145: 381-388.

123. Miller, J.H. 1972. Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, pp 132, 432.

124. Miyake, K., T. Kuzuyama, S. Horinouchi, and T. Beppu. 1990. The A-factor-binding protein of Streptomyces griseus negatively controls streptomycin production and sporulation. J. Bacteriol. 172: 3003-3008.

125. Miyoshi, Y.K., S. Ogata, and S. Hayashida. 1986. Multicopy derivative of pock-forming plasmid pSAl in Streptomyces azureus. J. Bacteriol. 168: 452-454.

126. Mocek, U., J.M. Beale, and H.G. Floss. 1989. Reexamination of the ^H and ^^C nmr spectral assignments of thiostrepton. J. Antibiot. 42: 1649-1652. 243

127. Mocek, U., L.-C. Chen, P.J. Keller, D.R. Houck, J.M. Beale, and H.G. Floss. 1989. ^H and ^^C nmr assignments of the thiopeptide antibiotic nosiheptide. J. Antibiot. 42: 1643-1648.

128. Mocek, U. and H.G. Floss. 1988. unpublished results.

129. Narva, K.E. and J.S. Feitelson. 1990. Nucleotide sequence and transcriptional analysis of the redD locus of Streptomyces coelicolor A3(2). J. Bacteriol. 172: 326-333.

130. Neesen, K. and G. Volckaert. 1989. Construction and shuttling of novel bifunctional vectors for Streptomyces spp. and Escherichia coli. J. Bacteriol. 171: 1569-1573.

131. Neil, R. 1987. Structure and regulation of the methylenomycin A resistance gene of Streptomyces coelicolor A3(2). PhD Thesis, University of East Anglia, Norwich, England.

132. Ochi, K. 1989. The tsr gene-coding plasmid pIJ702 prevents thiopeptin from inhibiting ppGpp synthesis in Streptomyces lividans. FEMS Microbiol. Lett. 61: 219-223.

133. Ogata, S., Y. Koyama, Y. Sakaki, and S. Hayashida. 1983. Isolation of a linear DNA associated with pock formation in Streptomyces azureus. Aoric. Biol. Chem. 47: 2127-2129.

134. Ogata, S., H. Suenaga, and S. Hayashida. 1985. A temperate phage of Streptomyces azureus. Appl. Environ. Microbiol. 49: 201-204.

135. Ogata, S., S. Yamada, and S. Hayashida. 1985. Genetic recombination in Streptomyces azureus by protoplast fusion and high production of thiostrepton by the recombinants. J. Gen. Aopl. Microbiol. 31: 187-191.

136. Ogata, S., S. Yoshino, H. Suenaga, K. Aoyama, N. Kitajima, and S. Hayashida. 1981. Specific lysogenicity in Streptomyces azureus. Appl. Environ. Microbiol. 42: 135-141.

137. Ohnuki, T., T. Imanaka, and S. Aiba. 1985. Self-cloning in Streptomyces griseus of a str gene cluster for streptomycin biosynthesis and streptomycin resistance. J. Bacteriol. 164: 85-94.

138. Olesker, A., L. Valente, L. Barata, G. Lukacs, W.E. Hull, K. Tori, K. Takura, K. Okabe, M. Ebata, and H. Otsuka. 1978. Natural abundance nuclear magnetic resonance spectroscopic evidence for the structural relationship between the peptide antibiotics thiostrepton and siomycin-A. J. Chem. Soc.. Chem. Commun. 14: 577-578.

139. Orner, C.A. and S.N. Cohen. 1984. Plasmid formation in Streptomycesi Excision and integration of the SLPl replicon at a specific chromosomal site. Mol. Gen. Genet. 196: 429-438.

140. Otten, S.L., K.J. Stutzman-Engwall, and C.R. Hutchinson. 1990. Cloning and expression of daunorubicin biosynthesis genes from Streptomyces peucetius and S. peucetius subsp. caesius. J. Bacteriol. 172: 3427-3434.

141. Pagano, J.F., J.J. Weinstein, H.A. Stout, and R. Donovick. 1956. Thiostrepton, a new antibiotic. I. In vitro studies. Antibiot. Ann. 1955/1956: 554-559.

142. Pascard, C., A. Ducroix, J. Lunel, and T. Prange. 1977. Highly modified cysteine-containing antibiotics. Chemical structure and configuration of nosiheptide. J. Am. Chem. Soc. 99: 6418-6423.

143. Pei, Z. 1990. unpublished results.

144. Perkins, J.B., S.K. Guterman, C.L. Howitt, V.E. Williams II, and J. Pero. 1990. Streptomyces genes involved in biosynthesis of the peptide antibiotic valinomycin. J. Bacteriol. 172: 3108-3116.

145. Rafii, F. and D.L. Crawford. 1989. Donor/recipient interactions affecting plasmid transfer among Streptomyces species: a conjugative plasmid will mobilize nontransferable plasmids in soil. Current Microbiology 19: 115-121.

146. Rhône-Poulênc, S.A., U.S. Patent 3,155,581, dep. 16 February 1962, publ. 3 November 1964. French Patent 1,392,453, dep. 24 February 1961, publ. 19 March 1965.

147. Richardson, M.A., S. Kuhstoss, M.L.B. Huber, L. Ford, 0. Godfrey, J.R. Turner, and R.N. Rao. 1990. Cloning of spiramycin biosynthetic genes and their use in constructing Streptomyces ambofaciens mutants defective in spiramycin biosynthesis. J. Bacteriol. 172: 3790-3798.

148. Richardson, M.A., S. Kuhstoss, P. Solenberg, N.A. Schaus, and R. N. Rao. 1987. A new shuttle cosmid vector, pKC505, for streptomycetes: its use in the cloning of 245

three different spiramycin-resistance genes from a Streptomyces ambofaciens library. Gene 61: 231-241.

149. Ristow, H. , J. Salnikow, and H. Kleinkauf. 1974. Biosynthesis of valinomycin. FEBS Letters 42: 127-130.

150. Schaffer, P. 1969. Sporulation and the production of antibiotic, exoenzymes and exotoxins. Bacteriol. Rev. 33: 48-71.

151. Schauer, A., M. Ranes, R. Santamaria, J. Guijarro, E. Lawlor, c. Mendez, K. Chater, and R. Losick. 1988. Visualizing gene expression in time and space in the filamentous bacterium Streptomyces coelicolor. Science 240: 768-772.

152. Schrempf, H. 1983. Deletion and amplification of DNA sequences in melanin-negative variants of Streptomyces reticuli. Mol. Gen. Genet. 189: 501-505.

153. Selden, R.F. 1989. Southern Blotting and Hybridization. in Ausubel, F.M. (Ed.), Current Protocols in Molecular Biology. New York: John Wiley and Sons, pp 2.9.5-2.9.10.

154. Sherman, D.H., F. Malpartida, M.J. Bibb, H.M. Kieser, M.J. Bibb, and D.A. Hopwood. 1989. Structure and deduced function of the granaticin-producing polyketide synthase gene cluster of Streptomyces violaceoruber Tu 22. EMBO J. 9: 2717-2725.

155. Shoji, J., T. Kato, Y. Yoshimura, and K. Tori. 1981. Structural studies on thiocillins I, II, and III. (Studies on antibiotics from the genus Bacillus XXIX). J. Antibiot. 34: 1126-1136.

156. Sinkawa, H., M. Sugiyama, and O. Nimi. 1988. Two genes for streptomycin phosphotransferase in Streptomyces griseus. in Biology of Actinomycetes '88. Okami, Y., T. Beppu, and H. Ogawara (Eds). Tokyo: Japan Scientific Societies Press, pp 353-358.

157. Southern, E. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503-517.

158. Speedie, M.K., and D.L. Hartley. 1984. Demonstration of a metabolic grid at an early step in the streptonigrin biosynthetic pathway in Streptomyces flocculus. Journal of Antibiotics 37: 159.

159. Steigerwald, V. 1989. personal communication. 246

160. Stein, D. and S. Cohen. 1989. A cloned regulatory gene of Streptomyces lividans can suppress the pigment deficiency phenotype of different developmental mutants. J. Bacteriol. 171: 2258-2261.

161. Stein, D.S., K.J. Kendall, and S.N. Cohen. 1989. Identification and analysis of transcriptional regulatory signals for the kil and kor loci of Streptomyces plasmid pIJlOl. J. Bacteriol. 171: 5768-5775.

162. Strohl, W.R. 1989. unpublished analysis.

163. Strohl, W.R., P.L. Bartel, N.C. Connors, C.-B. Zhu, D.C. Dosch, J.M. Beale, Jr., H.G. Floss, K. Stutzmann-Engwall, S.L. Otten, and C.R. Hutchinson. 1989. Biosynthesis of natural and hybrid polyketides by anthracycline-producing streptomycetes. in Hershberger, C.L., S.W. Queener, and G. Hegeman (Eds.), Genetics and Molecular Biology of Industrial Microorganisms. Washington, DC: American Society for Microbiology, pp 68-84.

164. Thiemann, J.E., C. Coronelli, H. Pagani, G. Beretta, G. Tamoni, and V. Arioli. 1968. Sporangiomycin, an antibacterial agent isolated from Planomonospora parontospora var. antibiotica var. nov. J. Antibiot... 31: 525-531.

165. Thompson, C.J., T. Kieser, J.M. Ward, and D.A. Hopwood. 1982. Physical analysis of antibiotic-resistance genes from Streptomyces and their use in vector construction. Gene 20: 51-62.

166. Thompson, C.J., J.M. Ward, and D.A. Hopwood. 1980. DNA cloning in Streptomyces: resistance genes from antibiotic-producing species. Nature 286: 525=527.

167. Thompson, J. and E. Cundliffe. 1980. Resistance to thiostrepton, siomycin, and sporangiomycin in actinomycetes that produce them. J. Bacteriol. 142: 455-461.

168. Thompson, J. and E. Cundliffe. 1981. Purification and properties of an RNA methylase produced by Streptomyces azureus and involved in resistance to thiostrepton. J. Gen. Microbiol. 126: 185-192.

169. Thompson, J., F. Schmidt, and E. Cundliffe. 1982. Site of action of a ribosomal RNA methylase conferring resistance to thiostrepton. J. Biol. Chem. 257: 7915-7917. 247

170. Vinci, V.A. 1988. Serine proteases in Streptomyces. PhD Dissertation. Columbus, OH; The Ohio State University.

171. Walker, J., À. Olesker, L. Valente, R. Ranabal, and G. Lukacs. 1977. Total structure of the polythiazole- containing antibiotic micrococcin P. A nuclear magnetic resonance study. J. Chem. Soc.. Chem. Commun. 20: 706-708.

172. Yoshimoto, A., T. Oki, T. Takeuchi, and H. Umezawa. 1980. Biosynthesis of daunomycinone from aklavinone and e-rhodomycinone. J. Antibiotics 33: 1199-1201.

173. Zhou, P., D. O'Hagan, U. Mocek, Z. Zeng, L.-D. Yuen, T. Frenzel, C.J. Unkefer, J.M. Beale, and H.G. Floss. 1989. Biosynthesis of the antibiotic thiostrepton. Méthylation of tryptophan in the formation of the quinaldic acid moiety by transfer of the methionine methyl group with net retention of configuration. J. Am. Chem. Soc. Ill: 7274-7276.

174. Zocher, R., Madry, N., Peeters, H., and Kleinkauf, H. 1984. Biosynthesis of Cyclosporin A. Phytochemistry, 23: 549-551.