LIBRARY Michigan State University PLACE iii RETURN BOXto romavothb chockwtflom your record. TO AVOID FINES Mum on or before date duo. DATE DUE DATE DUE DATE DUE ll LJ__] F—T—lf—j MSU IoAn AMnnaflvo Action/Equal Opportunity Instituion MOLECULAR GENETIC ANALYSIS OF REGULATION OF ANrIBIo'rIc BIOSYNTHESIS IN STREPTOMYCES COELICOLOR By Trifon Adamidis A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Program in Genetics 1994 ABSTRACT MOLECULAR GENETIC ANALYSIS OF REGULATION OF ANTIBIOTIC BIOSYNTHESIS IN STREPTOMYCES COELICOLOR By Trifon Adamidis The filamentous soil bacterium Streptomyces coelicolor is known to produce four antibiotics. An extensive search for mutants which are blocked in antibiotic production but not in sporulation led us to the discovery of the abs mutants (for antibiotic synthesis deficient. Genetic analysis of these mutants defined two loci, absA and absB, which map at 10 o’clock and 5 o’clock, respectively. Each of these loci is defined by several independent mutants. We also have obtained preliminary evidence for the existence of several additional abs loci. Based on cloning experiments, two putative absB“ clones capable of restoring all four antibiotics in absB mutants, were isolated. In addition,during the course of cloning the absB gene, several clones were isolated which were able to "bypass" the absB mutation including actII-orf4, redD and ast. The actII-orf4 and redD genes are the positive regulators of actinorhodin and undecylprodigiosin biosynthesis, respectively. The ast gene activates the production of both actinorhodin and undecylprodigiosin. Furthermore, several novel clones which stimulate antibiotic production were isolated. Based on our results, we conclude that the absA and absB genes are involved in global regulation of antibiotic biosynthesis. To my parents, Evagelos and Anastasia iv ACKNOWLEDGMENTS First and formost, I would like to thank Dr. Wendy Champness. Her love of, and dedication to, science was highly contagious and she created a very stimulating environment in which to learn and work. Wendy believed in me and gave me encouragement and support all throughout my graduate studies. She was always there, willing to share her knowledge and time when scientific or personal problems arose. For all that you have taught me about science and about life and for being a boundless source of guidance, encouragement, and care, I thank you, Wendy. I feel very fortunate to have been associated with an exceptional group of faculty and students and I would like to thank all of them who helped me through interesting discussions, sharing of their materials and friendship. I wish to thank the members of my guidance committee, namely Dr. Lee Kroos, Dr. Michael Thomashow and Dr. Michael Bagdasarian who have given beneficial advice and direction. I also acknowledge the Bouyoucos Fellowship for supporting me during the first four years of my graduate studies. I want to thank all members of the lab, most notably Perry Riggle, for their time, shared ideas and companionship. I am deeply indebted to Dr. Ronald Patterson for his guidance, patience and friendship. I would like to acknowledge Dr. Sue Conrand for many interesting discussions and Dr. Tom Comer for letting me use his lab equipment continously. A special word of thanks must be given to Sue Dagher, for her heroic efforts to understand my project and help me any way she could. Thank you, Sue, for being such a devoted friend. I thank all my friends who helped me during these years; without them, this thesis would have been finished earlier! Special thanks to Maria, a loyal friend who typed a large portion of this manuscript, old Zoy, Sophia and Stacy, for being such a faithful friends, young Zoe, George, Flora, Dimitris, Laurie, Sotiris, Filippos, Dia,Marina, Christos, Dimitra, Eftychia, my uncle Terry and my aunt Stella and my cousins, espessially Kelly for providing happiness through their friendship. I particularly wish to acknowledge Brendan and Sophia for providing me with support, a home and their friendship throughout. Additionally, I wish to thank Nikos, in Greece, a true friend who provided many years of support. I would like to express my gratitude to my teachers in Aristotelian University of Thessaloniki, Greece; special thanks to Drs. E. Alichanidis, A. Alichanidou, E. Litopoulou-Tzanetaki, N. Tzanetakis and G. Mourkidis. I am deeply indebted to my family - my father, my mother, and my sister - whose encouragement, help me to transform my dreams into reality. I especially thank my mother and father for supporting me, emotionally and financially, at any COSI. TABLE OF CONTENTS List of tables List of figures Chapter 1: Streptomyces coelicolor, a model organism for studing differentiation A review Introduction Antibiotics and secondary metabolites Physiological regulation Developmental genetics Bibliography Chapter 2: Mutations in a new Streptomyces coelicolor locus which globally block antibiotic biosynthesis but not sporulation Abstract Introduction Materials and methods Results Discussion Literature cited vii xii 13 21 37 50 51 51 52 53 56 57 Chapter 3: Genetic analysis of absB, a Streptomyces coelicolor locus involved in global antibiotic regulation Abstract Introduction Materials and methods Results Discussion Literature cited Chapter 4: At least four loci regulate globally the antibiotic production in Streptomyces coelicolor Introduction Materials and methods Results Discussion References Chapter 5: Further studies on phenotypic and genetic characterization of abs and bid mutants of Streptomyces coelicolor Introduction Materials and methods Results Discussion References viii 59 60 6O 60 61 63 65 67 68 69 72 82 83 86 87 90 70 95 99 Chapter 6: Cloning of the putative absB“ allele and other DNA sequences that" bypass" the absB mutation Introduction Materials and methods Results Discussion References Chapter 7: Suppression of antibiotic synthesis deficiences inStreptomyces coelicolor developmental mutants by cloned ast alleles Abstract Introduction Results References Chapter 8: Summary and conclusion Addendum 102 103 103 107 127 130 133 135 136 137 153 157 166 LIST OF TABLES Chapter 2 1. Strains of S. coelicolor, phage, and plasmids used in this study Chapter 3 1. Strains, phages, and plasmids used 2. Observation of actinorhodin and undecylprodigiosin pigments in absB mutant strains 3. Measurements of actinorhodin and undecylprodigiosin production in absB mutant and actIl-Orf4-stimulated strains Chapter 4 1. Strains and plasmids used Chapter 5 1. Strains, plasmids, and phages used in this study 2. Resistance to erythromycin, grth at 40° C, and melanin production of S. coelicolor bid and abs mutants 3. Antibiotic production is medium depedent in S. coelicolor bid and abs mutants 4. C249(bldl) carries a bldA mutation 61 62 62 70 89 91 94 96 Chapter 6 1. Strains and plasmids used 2. Characterization of clones that "bypass" the absB mutation 3. Production of actinorhodin and undecylprodigiosin in redD stimulated strains 4. Undecylprodigiosin production by bldA mutant harboring the ast and redD clones 5. Act, Red and CDA production in different developmental mutants in the presence of either the plasmids pTA108 and pTA128 Chapter 7 1. Strains of S. coelicolor and plasmids used in this study 2. Measurements of actinorhodin and undecylprodigiosin production in ast-stimulated strains 104 112 120 121 123 138 144 LIST OF FIGURES Chapter 2 1. CDA assay of Abs' mutant strain 2. Methylenomycin assay of an Abs' mutant strain 3. Time course of actinorhodin and undecylprodigiosin production 4. Mapping of abs-542 5. Backcross of C5422 strain to 11501 parental strain Chapter 3 1. Calsium dependent antibiotic assay of AbsB' mutant strains 2. Methylenomycin assays of AbsB' mutant strain 3. Mapping of abs-5 76 Chapter 4 1. Mapping of abs-8752 2. Mapping of abs-95 3. Mapping of abs-155 Chapter 6 1. Restriction map of a fragment of the act region and its relationship to the pTA107 clone xii 54 54 54 55 55 62 63 74 77 80 109 2. Clone pTA227 carries the entire whiE gene cluster 3. Clone pTA963 carries the redD and redC genes 4. Restriction map of pTA108 clone Chapter 7 1. Restriction maps for pWCS and pTA140 2. Effect of the ast clone pWCS on CDA production in Bld‘ and Abs’ mutant strains 3. Methylenomycin assay in Bld' and Abs' strains carrying pWCS xiii 114 118 125 141 146 149 Chapter 1: Streptomyca coelicolor, a model organism for studying differentiation, A review 2 INTRODUCTION Streptomycetes are gram positive soil bacteria which have an interesting life cycle and produce a plethora of secondary metabolites. The secondary metabolites represent a vast array of structural diverse low molecular weight compounds many of which have medical or agricultural value. Products of Streptomycetes include the common antibiotics, streptomycin, erythromycin, tylosin, chloramphenicol, tetracycline, the antitumor drug adriamycin, the immunosuppressant cyclosporin, and the herbicide bialaphos. Besides their commercial value Streptomyces exhibit a fascinating life cycle. After germination of a Streptomyces spore, vegetative grth occurs by hypha] extension and branching. Three days later, in laboratory conditions, aerial hyphae rise out of the substrate mycelium. Aerial hyphae formation coincides with the commencement of secondary metabolism. Eventually the aerial hypha which grows as a multinucleoid filament, subdivides to create haploid spores. A large portion of the substrate mycelium is lysed to provide nutrients for the maturation of the spores. Over the next several weeks , the colony functions as a differentiated multicellular organism, since the edges of the colony continue to grow vegetatively, aerial hyphae rise, the spores maturate, while secondary metabolites are produced. Studies on Streptomyces coelicolor were initiated in early 19705 in D. Hopwood’s laboratory. Since then many scientists have been attracted by the unique features of Streptomyces coelicolor. First S. coelicolor produces at least four antibiotics, two of Which are pigments; thus their presence or absence can be 3 detected visually. Even though these antibiotics are not used commercially, actinorhodin is a polyketide and biosynthetically is related to important antibiotics. Second, a plasmid mediated mapping system has been well characterized and at least 120 loci have been mapped (63). Recently a physical map has been established. Third, molecular genetic techniques have been developed and several plasmids, temperate phage vectors and transposons exist. Fourth, S. coelicolor is less subject to unstable DNA rearrangements common in many other species. This combination of features established S. coelicolor as a model organism to study the genetic regulation of morphological differentiation and antibiotic biosynthesis. ANTIBIOTICS AND SECONDARY METABOLITES Secondary metabolites are compounds with a wide range of chemical structures and biological activities while they are not essential for the growth of the organism which produces them. For Streptomycetes specifically the formation of secondary metabolites is directed by gene clusters which are usually controlled temporally and spatially. Streptomyces coelicolor produces four antibiotics, namely actinorhodin, undecylprodigiosin, methylenomycin and calcium dependent antibiotic, and it is a suitable host for production of secondary metabolites derived from other Streptomycetes such as melanin. Actinorhodin Actinorhodin (9) is a polyketide antibiotic. Polyketides comprise a family of compounds synthesized by both prokaryotic and eukaryotic cells where they play a 4 variety of roles (66). Their structures are diverse but they share a common biosynthetic route: the successive condensation of small organic acid units into a chain structure containing several keto groups. The primary structure of actinorhodin derives from 16 acetate units (40). Actinorhodin is a pigment; it is blue in alkaline conditions and red in acidic conditions, thus its presence can be detected visually. Genetic studies revealed that the biosynthetic genes of actinorhodin are clustered at 5 o’clock on Streptomyces coelicolor chromosome (146,117). Cosynthesis tests defined six classes of act mutants (117). The complete set of actinorhodin biosynthetic and resistance genes has been cloned and expressed in an actinorhodin non-producing strain, S. parvulus (93). The entire act cluster extends for 25kb and at least 20 ORFs transcribed in at least six mRN As have been defined (93). Streptomyces coelicolor actII mutants do not produce actinorhodin and do not cosynthesize it with mutants blocked in actinorhodin biosynthesis suggesting that the act]! locus plays a regulatory role in actinorhodin production (117). The "actII region" was cloned by Fernadez-Moreno et al., 1991. It contains 4 open reading frames; namely ORFl, ORF2, ORF3 and ORF4. actII-oer and actII-orf3 resemble trans-membrane proteins that confer resistance by export of various agents in other microorganisms (29). The actII-orfl encodes a product that resembles the repressors of tetracycline resistance genes (29). Disruption of actII-orfl, actII-on? or actII-orj‘3 does not prevent actinorhodin biosynthesis but the ORF2/ORF3 mutations interferes with actinorhodin export, leading to accumulation of actinorhodin inside the cells (29). The fourth ORF, actII-orjf4 encodes a regulator which stimulates actinorhodin 5 production (29). Transcription of actII-orf4 is growth-phase dependent, reaching a maximum during the transition from exponential to stationary phase (43). The transcription of the biosynthetic genes, act!!! and actVI-orf], occurs after the act”- orf4 transcript reaches maximal levels of accumulation (43); transcription of the act!!! gene requires the actII-orf4 product (48). Many Streptomycetes produce polyketide antibiotics which appear to have similar modes of formation (66). The positive activator Dan of daunorubicin biosynthesis in S. peucetius is capable of stimulating actinorhodin production in act]!- 0114 mutants. Similarly actII-orf4 is able to activate daunorubicin production in dan mutants in S. peucetius. These results imply that ActII-orf4 and Dan proteins are able to recognize common DNA sequences (127). Hopwood et al.,1985 reported the production of novel "hybrid" polyketides resulting from introduction of actinorhodin biosynthetic genes into the medermycin producer Streptomyces sp.AM-7161 resulting in production of mederrhodins A and B. Undecylprodigiosin Undecylprodigiosin is the second antibiotic pigment found to be produced by S. coelicolor. The pigment is structurally very similar to prodigiosin, the red antibiotic produced by Senatia marcescens and is a member of a closely related family of molecular species, the prodigiosenes. Prodigiosin and prodigiosin-like pigments have a wide distribution among bacteria and are produced under quite different environmental conditions. Serratia marcescens (142), Vibrio psychroerythms (23), Pseudomonas magnesiorubra (33), and at least two families of actinomycetales, Actinomycetaceae and Streptomycetaceae 6 (Gerber and Lechevalier, 1976), are representatives of bacteria capable of prodigiosin production. Prodigiosin and prodigiosin-homologues are soluble in many organic solvents such as chloroform and methanol and completely insoluble in water over the entire pH range (54). All pigments have a characteristic red color in acidic pH solutions and yellow in alkaline ones. This phenomenon makes studies on these compounds easy since the amount of antibiotic produced can be determined colorimetrically. When red, prodigiosin solutions have a maximum absorbance at about 525um and when yellow at 460nm. The instability of yellow solutions of prodigiosin at room temperature (54) and in sunlight (35), make the acidified solutions most suitable for quantitation of the red antibiotic. The antibiotic is bactericidal /bacteriostatic to a wide range of Gram-positive bacteria (e.g. genus Micrococcus, Bacillus etc). Also the antibiotic can act against different parasites (54). Although prodigiosin was shown in mice to have a definite activity against the parasite Trypanosome concolege (malaria) and it was used in humans for treatment of coccidioidomycosis, it is considered too toxic for therapeutic use (13). Studies of biosynthesis of prodigiosin in Serratia marcescens showed that proline, acetate, glycine and methionine are direct precursors of the pigment (134). The formation of the antibiotic was not induced only by the direct precursors but also by ornithine, glutamic acid, alanine, aspartic acid, serine and thiamine (38). A supply of dissolved oxygen is required for prodigiosin production (56) while streptomycin (143), ATP, inorganic phosphate and ribose (11) cause inhibition of the antibiotic production in S. marcescens. Even though undecylprodigiosin is a pigment, its discovery in S. coelicolor was delayed because of the existance of the blue actinorhodin which masks the red color. Only when mutants defected in actinorhodin production were isolated, did the presence of undecylprodigiosin became apparent (118). Later it was found that the red pigment is not just one compound but a mixture of mainly undecylprodigiosin and butylcycloheptylprodigiosin (136). Undecylprodigiosin, like prodigiosin in Sematia marascens, is sensitive to phosphate (60) . Cosynthesis experiments of red mutants isolated by UV. mutagenesis ordered them in six cosynthesis groups (A-F) (118,28). Genetic analysis of these mutants revealed that all genes required for the biosynthesis of undecylprodigiosin are clustered in the S. coelicolor genome, in a locus distinct from that of the actinorhodin biosynthetic genes (118). On the assumption that the biosynthetic pathways of undecylprodigiosin in S. coelicolor and prodigiosin in S. marascens might be similar, Feitelson and Hopwood performed cosynthesis experiments which confirmed the hypothesis (27). The redE gene was cloned by a shot-gun cloning system and was showed to encode an O- methyltransferase enzyme (27). Later studies showed that both the redE and redF gene products are required for O-methyltransferase activity (28). Kinetic studies showed that O-methyltransferase activity coincided with the red pigment production suggesting that O-methyltransferase activity may be the rate-limiting step for the biosynthesis of the red antibiotic (28). The whole biosynthetic gene cluster for red production was cloned by in viva 8 recombination between a clone carrying two contiguous segments derived from opposite ends of the red cluster and the wild type chromosome of S. coelicolor (94). This 36 kb sequence carries all the information for the red antibiotic biosynthesis and resistance since it is sufficient to produce undecylprodigiosin when it is introduced into an heterologous host, S. parvalus. As in the actinorhodin pathway, the undecylprodigiosin biosynthetic genes are activated by a regulator. Evidence that the redD gene plays the role of an activator of the red antibiotic production came first from the redD' mutants that cannot act either as converters or as secretors in cosynthesis experiments (118) and second from the isolation of the redD clone which even in low copy number caused overproduction of the pigment in the wild type (104). Also, redE mRNA is lacking in a redD mutant (104). Dot blot analysis revealed that wild-type transcription reached a peak at 60 h after inoculation, in liquid culture, and that the redE and redBF messages require the presence of the redD product to produce the wild type level of antibiotic (104). There is no AT-rich -10 promoter region in the putative redD promoter but there is homology between promoter regions of redD and strR, the streptomycin regulatory gene (104) and actII-orf4 the actinorhodin regulatory gene (29). The transcription of redD is growth-phase-dependent and even though a low level of transcription can be seen during exponential growth, transcription of redD is highly induced during the transition to stationary phase (131). DNA sequences homologous to red sequences occur in other Streptomyces also. S. lividans, S. violaceolatus and S. scabies were shown to possess sequences homologous to red sequences (28). Methylenomycin Methylenomycin has antibacterial activity against both Gram-positive and Gram-negative organisms (49). The antibiotic was first detected in S. coelicolor by Wright and Hopwood, 1976. Genetic studies have shown that the genes for the biosynthesis of and resistance to methylenomycin are present on the SCP1 plasmid (83). Methylenomycin is the only Streptomycete antibiotic which is known to be plasmid encoded. Plasmid SCP1 has been physically characterized as a linear plasmid of 350kb (82). Calcium dependent antibiotic (CDA) Calcium dependent antibiotic requires calcium ions in order to exert its antibacterial activity (85). Mutants deficient in CDA production have been isolated and their mutations mapped at 10 o’clock on the S. coelicolor chromosome (65). The genes responsible for CDA production have not been cloned yet. Melanin Tyrosinase (E.C. 1.14.18.1) is an enzyme that catalyzes the conversion of tyrosine to the black pigment melanin. The whole process involves the oxidation of L-tyrosine via L-dihydroxyphenylalanine to dopaquinone which after spontaneous oxidization and polymerization forms melanin (57). Tyrosinase is a metalloprotein so in order to be active it requires two atoms of copper per molecule (88). Many organisms (e.g. bacteria, fungi, plants, insects and mammalian cells) are capable of producing melanin or melanin-like pigments. Several Streptomyces exhibit the same ability. S. glaucescens tyrosinase was purified and characterized, and was found to contain two copper atoms per molecule. Both copper atoms are necessary 10 for the activity of the tyrosinase. Removal of the copper atoms from the native enzyme results in catalytically inactive apoenzyme (apotyrosinase) (88). These results were confirmed by reactivation of the purified apotyrosinase with a supply of copper (87). The level of tyrosinase production which leads to the production of melanin is regulated in Streptomyces. The amino acids L-leucine, L-methionine and L- phenylalanine can act as specific inducers of tyrosinase production in S. glaucescens (5), while in S. antibioticus induction of melanin production can be accomplished only with L-methionine or methionine-analoques (76). Experiments in S. antibioticus with S-labelled methionine showed that methionine causes de novo synthesis of tyrosinase and it does not simply activate an inactive form of apotyrosinase (8). Addition of actinomycin D, rifampicin or chloramphenicol just before induction with methionine caused absence of tyrosinase synthesis suggesting that both transcriptional and translational events are required (8). Several other factors seem to play a positive role in tyrosinase production, namely in S. antibioticus small amounts of inoculum and young cultures seem to be more responsive to methionine induction, while the carbon source seems to play no role (76). Tyrosinase activity have been found to be present both intracellularly and extracellularly (5). In S. glaucescens tyrosinase synthesis in liquid cultures starts 24 hours after inoculation and only a low extracellular activity can be observed under non-induced conditions. When L-methionine was used as an inducer, a marked increase in tyrosinase activity was observed. Under induced conditions, the rate of the enzyme synthesis is greater than the secretion rate (which has a maximum 11 constant) leading to the accumulation of the enzyme intracellularly. Both intracellular and extracellular enzymes are identical in molecular weight, copper content, enzyme kinetics and amino-terminal amino acid sequence (19). Three groups of mutants have been isolated from S. glaucescens defective in the production or regulation of tyrosinase synthesis. The first group of mutants consists of constitutive mutants which do not require specific inducers for tyrosinase synthesis (80). The second group contains mutants able to produce functional tyrosinase intracellularly but not extracellularly (5,20). In the third group are mutants which are unable to produce any intracellular or extracellular tyrosinase activity (5,128,129). The third group of mutants can be divided into three genetically independent classes: melA, melB, and melC (20,59). The melC gene from both S. antibioticus (77) and S. glaucescens (59) was cloned in a melanin-nonproducer strain S. lividans; transformed S. lividans colonies became melanin producers suggesting that the melC gene encodes the tyrosinase structural gene. The regulation of secretion of the enzyme though seems to be regulated differently in S. lividans since tyrosinase activity occurs only intracellularly (77). Nucleotide sequence analysis of the melC gene of S. antibioticus revealed the existence of an operon consisting of the melCI and melC2 genes. A promoter lying at the 5’ end of the melCI gene drives the transcription of a polycistronic mRNA. melC2 encodes the structural gene of the apotyrosinase while melCI encodes a protein with an amino-terminal signal peptide characteristic of exported proteins (7). Insertional inactivation of the melCI gene inactivates the expression of the tyrosinase 12 structural gene even when some of these insertions contain promoters that should transcribe the melC2 gene (77). To rule out the possibility that the melCI gene disruption causes a Mel’ phenotype due to a polar effect causing premature transcription termination, complementation tests were carried out. Plasmids containing the melCI * melC2 and in inverted orientation the meICI’ melC2+ genes were generated and transformed into S. lividans JT46, which is defective in intraplasmid recombination. All the IT 46 transformants displayed a Mel” phenotype indicating that melCI acts in trans on the expression of the tyrosinase gene (87). Interestingly, although S. lividans containing the melCI' melC2+ construct showed a Mel’ phenotype, RNA was synthesized from this region. Thus melCI affected the tyrosinase gene expression but did not act at the transcriptional level (87). Since the apotyrosinase produced by a melCI' melC2+ mutant could be activated in vitro with copper sulfate, Lee et. al.,1988, proposed that the melCI gene encodes a trans-acting protein which seems to facilitate the incorporation of copper to apotyrosinase (87). Similarly the nucleotide sequence of the meIC transcribed region of S. glaucescens revealed the presence of an operon containing three open reading frames (ORF). ORFl gene product was postulated to be involved in copper transfer, ORF2 encodes for apotyrosinase and ORF3 has an unknown function (73,74). The DNA sequences encoding the apotyrosinases of S. antibioticus (7) and of S. glaucescens (73) show a 85.8% homology and the comparison of the aminoacid sequence between the two tyrosinases reveale 86.4% identity. Northern blot analysis revealed that the induction of tyrosinase expression by different amino acids is regulated at the 13 transcriptional level and it was proposed that melA and melB genes play a positive regulatory role in the transcriptional regulation of the tyrosinase gene (59). Further investigation by Geistlish et. al., 1989 localized three DNA sequences responsible for the promoter regulation of the meIC operon. One of the regulatory DNA sequences is responsible for the induction of tyrosinase by different amino acids, the second is recognized by a putative protein which is postulated to be a repressor and the third acts like an upstream activator site, although there is no evidence that a protein binds in this region. Several lines of evidence suggest that the melC operon of both S. antibioticus and S. glaucescens is regulated in a similar manner; namely an exact match between the two transcriptional start sites, a 70% sequence similarity in the -10 region, a 65- 68% sequence similarity in the upstream regulatory region and inducibility by L- methionine. The melC promoter is not recognized by E. coli RNA polymerase since substitution of the melC native promoter with the E. coli tac promoter led to melanin production in E. coli (92). Studies in other Streptomyces species revealed that S. scabies tyrosinase gene is present extrachromosomally (45). In S. michiganesis, an actinomycin and melanin producer, melanin plays no role in tyrosinase induction but copper does. Expression of both actinomycin and tyrosinase is repressed by ammonium at the transcriptional level (57). PHYSIOLOGICAL REGULATION Many scientific studies have described the effects of nutrition and 14 environment on secondary metabolism. In general when nutrients are available, secondary metabolism is suppressed, while under nutrition limitation, secondary metabolism commences. In S. coelicolor different growth conditions have been described which inhibit or repress certain antibiotics and favors others. Actinorhodin production, for example, can be inhibited by either ammonium or high levels of phosphate (26). The nature of the nutritional imbalance and the intracellular regulatory factors which leads to the activation or derepression of antibiotic biosynthesis genes is not well understood. The stringent response and chemical signals such as A-factor may link nutritional stress to molecular control mechanisms. Investigation of the role of the stringent response in antibiotic regulation In E. coli, amino acid limitation produces the "stringent response" whereby complex changes in the pattern of gene expression occur and transcription of stable RNA genes is immediately reduced (12). Bacteria] cells exhibiting a stringent response rapidly accumulate mM levels of ppGpp, a process triggered by the binding of uncharged tRNA to ribosomes, and the synthesis of GTP is reduced severely (32). rel mutants show a "relaxed" phenotype since neither ppGpp nor pppGpp is accumulated after nutritional shift-down and RNA synthesis continues normally. Two kinds of mutations were found, namely relA and relC. relA mutants carry mutations of the gene which encodes a ribosome-associated enzyme (p)ppGpp synthetase I, which synthesizes ppGpp when uncharged tRNA molecules bind to the ribosomal A site. relC mutants carry mutations of the gene which encodes the ribosomal protein L11. Ribosomes lacking L11 protein were found among thiostrepton - resistant mutants in E. coli (31). Thiostrepton, a modified peptide 15 antibiotic, is a protein synthesis inhibitor. The drug blocks the functional domain that apparently contains at least part of the aminoacyl-tRNA binding site (A site) and is involved in some ribosome-associated GTPase events (31). The binding of thiostrepton to a single site in 23 S rRNA is greatly enhanced in the presence of protein L11, although L11 itself does not bind to thiostrepton (21). In Bacillus subtilis, a Gram-positive bacterium which produces endospores, the decrease of intracellular GTP caused by the stringent response, results in the initiation of sporulation (111,112). The stringent response was studied in many Streptomyces species and all of them showed that pppGp and ppGpp were increased significantly after nutritional shift-down. Two main methods were used for induction of the stringent response: a) actual nutrient shift-down by transferring colonies growing in casamino acid supplied medium to a synthetic "poor" medium b) using chemical compounds that can mimic nutritional shift-down. 1) One such compound, DL-serine hydroxamate (SHX), a structural analogue of L-serine, provokes the stringent response in E. coli by acting as a competitive inhibitor of seryl-tRNA synthetase, thus leading to the accumulation of uncharged tRNAset (135). 2) methyl-a-D- glucopyranoside, an analogue of glucose which inhibits glucose uptake and accumulates as non-metabolized aMG-6-phosphate (50). 3) decoyinine, a specific inhibitor of GMP synthetase (107). The work of Ochi, 1986, revealed that there is a correlation between stringent control and physiological (secondary metabolites) and morphological (aerial hyphae formation) 16 differentiation. S. lavedulae MA406-A-1 produces a formycin (a nucleotide antibiotic) and melanin (pigment) as secondary metabolites. In the presence of casamino acids (1%), formycin started to be produced at the end of exponential growth. ppGpp concentration in the presence of casamino acids was low during exponential growth and increased significantly at the end of it. rel mutants, isolated as thi0peptin (a thiostrepton homolog) resistant mutants, exhibited a relaxed phenotype, accumulating much less ppGpp after nutritional shift-down. rel mutants produced strikingly less formycin in both liquid and agar-plate medium and did not produce melanin. Although the mutants retained the ability to form aerial mycelium and spores, the amount of these was lower than the wild-type and the onset of formation of aerial mycelium was delayed (106). The interesting observation that the mutants retained the ability to accumulate ppGpp and formycin during glucose starvation indicated that formycin production was tightly coupled with the accumulation of ppGpp (106). Ochi continued this work in several different species (S. antibioticus, S. lavedulae, S. griseoflavus, S. gnlseus) and found rel mutants in several different species. All these mutants were found to have an altered or missing ribosomal protein designated tentatively ST-Lll, so all these mutants were classified as relC mutants (109). All relC mutants showed a reduction in accumulation of ppGpp, antibiotic production and aerial mycelia formation (109). All relC mutants (except relC mutants of S. antibioticus), were sensitive to erythromycin, an antibiotic inhibitor of translation (109). Gram-positive bacteria are generally more resistant to erythromycin (21). The same phenomenon was observed with the relC mutants of B. subtilis (39). relC mutants (except those of S. 17 antibioticus) were also unable to grow at high temperature (40°C), whereas the parental strains could grow (109). In S. griseus addition of decoyinine induced sporulation under growth conditions that normally suppress sporulation, . Concomitant addition of guanosine with decoyinine inhibited sporulation completely (107). ppGpp inhibited the enzyme 1MPdehydrosenase, the first enzyme of the pathway leading to GTP (107). These observations led Ochi to propose a relation between GTP pool size and initiation of sporulation. S. griseus relC mutants gave rise at high frequency (5-30%) to revertants which were capable of producing streptomycin and aerial mycelia but not ppGpp. These suppressors retain the ability to be thiopectin resistant and lack the protein L11. Genetic analysis showed that the sup (suppressor) mutation occured at a chromosomal locus distinctive from relC (109). relC mutants of S. griseus were shown to produce A factor normally, while afs mutants were capable of accumulating ppGpp like the parental strain upon nutritional shift-down (109) (A-factor and the afs genes are discussed in another section). Triple mutants of S. gn'seus (relC, sup, afs) were constructed and the mutants were blocked for production of streptomycin and aerial mycelia formation. Addition of exogenous A-factor restored streptomycin production and aerial mycelia formation to wild-type levels (109). These observations indicated that the synthetic processes of ppGpp and A factor proceeded independently. More evidence of a correlation between production of secondary metabolites and stringent response came from the observation that two enzymatic activities 18 necessary for actinomycin production in S. antibioticus were absent or very much decreased in a relC mutant (78). Different results were obtained in S. clavuligerus, a cephalosporin producer. RNA synthesis was under stringent control after nutritional shift-down from complex medium but thiostrepton resistant colonies which had a diminished capacity to produce ppGpp upon nutritional shift-down were shown to belong to three different types. These mutants produce little, the same, or higher amounts of antibiotics as compared to the wild-type strain. So mutations conferring thiostrepton resistance can influence cephalosporin production in opposite ways, depressing or stimulating the process (4). In S. coelicolor, relC mutants isolated by Ochi (1 10) exhibited the same phenotype as the relC mutants isolated from other Streptomyces species by the same researcher (109). So the relC mutants did not accumulate ppGpp and the synthesis of RNA continued normally upon nutritional shift-down. The relC mutants were sensitive to erythromycin and high temperature (110). Aerial mycelia were reduced and delayed, also. The antibiotic production followed a different pattern in relC mutants than the wild-type. No pigmented antibiotics were produced for 10 days, but at the tenth day actinorhodin started to accumulate and soon reached levels higher than wild type, whereas undecylprodigiosin never appeared (110). Similar results, namely reduction of RNA synthesis in S. coelicolor upon accumulation of ppGpp were found by other researchers also (126). All four rrnD promoters of the operons for rRNA transcription were shown to be subject to negative stringent control (126). The results though for actinorhodin and undecylprodogiosin 19 production were different. The act!!! gene (a structural gene for actinorhodin biosynthesis) was shown to be transcribed shortly after entry into stationary phase. After nutritional shift-down, a switch to secondary metabolism was shown by the appearance of the act!!! transcript one hour later. Suppression of ppGpp accumulation by addition of chloramphenicol before shift-down, did not prevent antibiotic production indicating that ppGpp is not necessary to promote this transition (126). relC mutants isolated by the same researchers, appeared to be normal in antibiotic production (126). For undecylprodigiosin production, ppGpp was shown to play no role. Undecylprodigiosin appeared upon entry into stationary phase when ppGpp was detected at normal levels. After nutritional shift-down, colonies of S. coelicolor accumulated high amounts of ppGpp but the levels of redD and redX transcripts did not increase. So for both, actinorhodin and undecylprodigiosin showed that the antibiotic production is not specifically induced by the increased levels of ppGpp (131). All these results indicate that no correlation between secondary metabolism and functioning of the stringent response could be established. Cellular signals in Streptomycm spp. One may presume that actinomycetes require effective chemical signals for communication between different parts of their mycelia as in the case in other filamentous organisms such as fungi (6). Several autoregulators having a y-lactone ring in their structures have been reported to stimulate secondary metabolism in Streptomyces: namely A-factor in S. griseus which controls streptomycin production and aerial mycelia formation (51); factor F1 which controls anthracycline biosynthesis 20 in S. Vitidochromogenes (42); a factor that controls cytodifferentiation and anthracycline production in S. bikiniensis and in S. cyanofuscatus (42); and "inducing factors" which control virginiamycin production in S. virginiae (105,147). Azfaszm The best characterized autoregulator is A-factor (autoregulating factor) from S. gn'seus. A-factor was first detected as a factor stimulating streptomycin production in S. griseus (79). A-factor deficient mutants of S. gn'seus are unable either to produce streptomycin or to form aerial mycelia (51) and they become sensitive to streptomycin (52). The only promoter found to respond directly to A-factor from the streptomycin gene cluster belonged to the strR gene, the positive regulator of the streptomycin biosynthetic and resistance genes. In the presence of A-factor, transcription of the strR gene is activated, which in turn stimulates transcription of the other genes in the cluster (137). A protein with high affinity to A-factor was isolated (Miyake et al., 1989); this protein seems to act as a repressor of streptomycin production, since mutants deficient in A-factor binding protein are able to form aerial mycelia and to produce streptomycin even in the absence of A-factor (101). N 0 site in the vicinity of strR promoter was found that A-factor binding protein was able to bind, suggesting that A-factor binding protein exerts its function in streptomycin biosynthesis indirectly. A model for the function of A-factor has been proposed by Miyake et al., 1990. When A-factor is absent, the A-factor binding protein binds to specific regions of some regulatory genes and represses their expression. In the presence of A-factor, the binding protein binds to A-factor, relieving the repression of the regulatory genes, 21 which activate the expression of other regulatory genes such as strR. According to this model, the timing of the expression of the regulatory genes depend on the molecular ratio of A—factor: A-factor binding protein. This hypothesis correlates well with the finding that A-factor is produced just before the cells begin to differentiate (51). The extremely low effective concentration of A-factor (52) and the complex regulatory cascade for transmission and amplification of the A-factor signal (137) suggests that A-factor acts as an ancestral hormone in procaryotes. Interestingly, even though A-factor mutants cannot form aerial mycelia, they are able to sporulate suggesting that aerial mycelia deficiency does not necessarily mean that spore formation is blocked (130). afsA (A-factor deficient) mutants of S. coelicolor and S. lividans have also been isolated but these mutants showed no apparent phenotypic change suggesting that A- factor does not play a role in morphological and physiological differentiation of S. coelicolor (69). Later results showed that the A-factor binding protein was absent in both strains, thus according to the current model they do not require the presence of A-factor for their differentiation (100). Two loci have been identified to regulate A-factor production, namely afsB and ast (69,70). The role of these genes in S. coelicolor and S. lividans differentiation has not yet been well defined. DEVELOPMENTAL GENETICS Three major classes of genes have been defined to affect the developmental differentiation of S. coelicolor. Mutations of the genes of the first class result in bld mutants which are deficient in both sporulation and antibiotic production. Genes 22 that are involved in aerial mycelia formation, the sap genes, are also discussed in this class of genes. The second class of genes regulate the antibiotic production but not sporulation. Finally, the third class contains genes involved in sporulation but not in antibiotic production. The bld genes Genetic evidence that the onset of sporulation is coupled to antibiotic production comes from the isolation of bld mutants. Wild-type S. coelicolor colonies appear to have a fuzzy, velvety surface due to the presence of aerial mycelia. bld (bald) mutants are so named because of their inability to form aerial mycelia, so that their colonies remain smooth. Most of the bld mutants that have been isolated also fail to produce antibiotics. Many bld mutants have been isolated from S. coelicolor but only a few of them have been genetically and phenotypically characterized; namely bldA, bldB, bldC, bldD (99), bldG, bldH (14), bld] (14,55), bld-221 (139), bld-5M5, bld-5M1 (120). In most bld mutants, the Bld' phenotype is variable depending on the carbon source provided. For bldA, bldB, bldC, bldD, bldG, bIdH, and bld-221 mutants, aerial mycelium formation is blocked on glucose minimal medium but does occur on other carbon sources such as maltose or mannitol. Even though the mutants are able to sporulate on different carbon sources, only bldH mutants regain the antibiotic production (14). The carbon source effects on sporulation and antibiotic production are not understood. W bldA mutants fail to produce aerial mycelia on glucose minimal medium or 23 on the complex R5 medium but are capable of doing so when maltose or mannitol are the carbon sources. bldA mutants are blocked for all four antibiotics, namely actinorhodin (Act), undecylprodigiosin (Red), methylenomycin (Mmy) and calcium dependent antibiotic (Cda), known to be produced by S. coelicolor. The carbon source seems to play no role on the antibiotic synthesis since bldA mutants fail to produce any on different kinds of media (99,14,115). Interestingly, on low-phosphate medium, bldA mutants produce undecylprodigiosin but not actinorhodin (47). The observed phenotypes were confirmed when a bldA deletion mutant was created by gene replacement (91). The bldA mutations map at 10 o’clock on the S. coelicolor genetic (99,14) and physical (81) maps. Phage-mediated cloning of the bldA gene showed that a single copy of the bldA gene was capable of inducing the production of actinorhodin and the formation of aerial mycelium (115). Sequence analysis revealed that the bldA gene encodes a tRNA-like transcript which presumably recognizes the UUA codon for leucine (86). The TTA codon appears to be extremely rare in the S. coelicolor chromosome (73% GC) and the fact that the bldA deletion mutation is not lethal demonstrates that the bldA gene product is not essential for vegetative growth. A comparison of the occurrence of the three most rare codons, namely TI‘A, CTA and 'ITI‘ in 100 Streptomyces genes, revealed that TI‘A codons are present only in genes involved in regulation of morphological and / or physiological differentation. Both the two other codons, CT A and TIT, occur in genes involved in both primary and secondary metabolism (89). 24 The nonrandom occurrence of the 'ITA codon (89) and the presence of only one copy of the bldA gene in the S. coelicolor chromosome (86) suggest that bldA plays a role in translational regulation of some gene(s) required for initiation of differentiation in Streptomyces. Genetic evidence that bldA specifies a tRN A recognizing the UUA codon came by testing the expression of genes containing 'ITA codons. The ampC gene of E. coli which encodes for delactamase, the lacZ gene from E. coli which encodes for 6-galactosidase and the carB gene from S. thermotolerans which encodes a product that confers resistance to lincomycin were tested. All three genes, which contain at least two TI‘A codons, failed to be expressed in bldA mutants of S. coelicolor and / or S. lividans (a close relative of S. coelicolor), but were expressed in bldA+ strains. When the two TI‘A codons of the carB gene were changeded to the CT C codon, both bldA+ strains and bldA mutants became resistant to lincomycin. The latter result suggests that the expression of the carB gene depends on the expression of the bldA gene due to the presence of the two TTA codons in carB (90). Two other genes containing TI‘A codons, namely the aad and hyg genes whose products confer resistance to spectinomycin and hygromycin respectively, were expressed poorly in bIdA mutants on minimal media but were expressed equally as well as the bldA+ strains on complex medium (90). TI‘A codons were found in two genes in the act cluster, namely actII-orf4 which encodes a putative transcriptional activator for actinorhodin biosynthetic genes and actII-orfl which encodes a trans-membrane protein that confers resistance by exporting actinorhodin out of the cell (29). Upon site-directed mutagenesis of the single TI‘A codon to a TI‘G codon in a cloned copy of the actII-orf4 gene, the bldA 25 mutants were capable of overproducing actinorhodin which accumulated in the cell. The actinorhodin did not diffuse into the medium probably because the actII-orfl gene was not expressed due to TI‘A codons that it contains (29). When the actII-orf4 gene is present in high copy number it is capable of inducing actinorhodin production in bldA mutants (114). The expression of aad, hyg and aCtII-Oif4 in high copy numbers has been attributed to a low level of translation of the UUA codons by noncognate tRNA species (90). Even though the UUA codon is unsuited for reading via "wobble" rules, it is possible that wobble occurs in the first position of the codon (89). Since bldA mutants do not produce any of the known antibiotics the role of the bldA gene in regulation of production of Act and Red was examined. Transcriptional fusions which placed the promoterless xylE gene under the chromosomal promoter of act] (structural gene for undecylprodigiosin) were constructed (10,47). The xylE gene of Pseudomonas putida encodes for the catechol 2,3-dioxygenase which converts the colorless catechol to a yellow diffusable pigment (148). Although xylE was not expressed in bldA mutants, it was expressed in bldA” strains suggesting that the bldA gene exerts its function in the transcription level of the structural genes of the act (10) and red (47) cluster. These results are explained for Act since the actII-orf4 gene, the transcriptional activator of the act cluster, contains a TI‘A codon (29). RedD, though, the transcriptional activator of the red cluster, does not contain any TI‘A codon (104) suggesting that the bldA gene product translates a mRNA whose protein participates in regulation of the red genes. In low-phosphate media, the redX-xylE fusion is expressed in a bldA mutant which 26 suggests an alternative regulatory pathway for red gene expression (47). , Evidence for temporal regulation of bldA expression would support the hypothesis of a regulatory role for bldA in Streptomyces development. In one study only the 5’ end of the premature tRNAUUA was present in young cultures (91). The mature tRN A increased in abundance late in growth whereas a major lysyl-tRNA did not show any induction (91). In another report there was only a very small increase of the mature tRNA during the time of physiological differentiation (43). The presence of an antisense tRNA transcript is possibly implicated in the maturation of the tRNA (91). The temporal control of the translatability of the UUA codons could also support the regulatory role of the bldA gene. The mRNA of the ampC gene which contains seven UUA codons was shown to accumulate early in growth while 6-lactamase, the ampC gene product, appeared to accumulate later (91). However, three actII-orf4::ennE constructs containing none, one or two TI‘A codons were shown to be transcribed and translated even in young cultures (43). The different results obtained by these experiments were attributed to different strains, growth conditions and protocols used. The regulatory role of the bldA gene remains to be established. Examples in other bacteria have shown that tRNAs may play a regulatory role. A tRNA for arginine and a tRNA for serine seem to play a role in DNA replication and cell division respectively in E. coli (102,132). Additionally a tRNA for leucine is accumulated when E. coli bacteria are starved for methionine (22). In Clostridium acetobutylicum, mutants of the threoninyl-tRNA gene neither sporulate nor produce secondary metabolites. The tRNA recognizes the ACG codon which is very rare in 27 Clostndium spp (30-40% GC) (119). Both the latter case and the fact that bldA mutations exert the same phenotypes in S. lividans and S. gn'seus suggest a wide occurrence of regulation of morphological and physiological differentiation through tRNAs that recognize rare codons. W The second bld locus that has been cloned and sequenced is bldB. bldB mutants do not show dependence on the carbon source used for antibiotic production and sporulation, remaining bald and nonpigmented in any media tested (99,55). Recent results show that the bldB mutants are competent to produce undecylprodigiosin selectively on SY medium (1% starch, 0.3% yeast extract) (2). The bldB locus was mapped to 5 o’clock on the genetic map of S. coelicolor (99,14). The bldB gene was cloned using a phage-mediated shotgun cloning system (55). Surprisingly, the bldB locus appears to be unique to S. coelicolor since the bldB sequence does not hybridize to DNA from other Streptomyces (55). Sequence analysis did not reveal anything about the function of the bldB gene product. 81 nuclease protection assays and Northern blot analysis showed that the transcription of bldB gene is not temporally regulated (Piret, personal communication). Thus, the function of bldB in development remains obscure. The sap genes The Sap proteins were isolated from the surface of the spores of S. coelicolor. The sapA gene was cloned using reverse genetics and was found to encode a preprotein in which the 13 kD polypeptide is preceded by a signal sequence. The sapA promoter appeared to be temporally and spatially regulated. Unexpectedly, the 28 transcription of the sapA gene was found to occur in all bld mutants tested and in some of them the level of transcription of the sapA gene was comparable to wild-type levels. Furthermore, sapA was found to be transcribed even in liquid cultures (S. coelicolor does not undergo morphological differentiation in liquid cultures) (46). The latter results show that even though the SapA protein is associated with spore formation, the activation of the sapA gene does not require morphological differentiation of S. coelicolor cultures. sapC-E genes were mapped on the giant linear plasmid SCP1, the presence of which is not a prerequisite for the developmental cycle of S. coelocolor (McCormick and Lossick, personal communication). Thus these proteins are nonessential for sporulation. W The SapB protein consists of 18 amino acids and contains a moiety carrying a hydroxyl group such as a sugar residue. Antibodies raised against proteins purified in different developmental stages from S. coelicolor showed that the appearance of the SapB protein coincides with the onset of morphological differentiation (139). SapB protein is suggested to be synthesized by a nonribosomal mechanism since its expression was not inhibited by addition of chloramphenicol in the medium , (chloramphenicol is known to inhibit translation) (140). This finding suggests that one or more peptide synthetases are involved in SapB protein synthesis; as it has been shown to be the case for the synthesis of some peptide antibiotics (95). All bld mutants tested, namely bldA, bldB, bldC, bldD, bldH, bldI, bld221 and bld261 were found to be impaired in SapB production. Other developmental 29 mutants impaired in either antibiotic biosynthesis such as abs (1) or in sporulation but not in aerial mycelia formation such as whi (15) were found to be competent to produce SapB protein (139,140). These findings supported the idea that SapB protein is specifically correlated with aerial mycelia formation. When bld mutants were provided with purified SapB protein, they acquired a fuzzy appearance which is an indication of the presence of aerial mycelia. Interestingly, the effect of SapB protein on sporulation was transient. After seven days after of incubation, the SapB- treated bld mutants lost their fuzzy appearance and microscopic analysis revealed that the aerial mycelia had been collapsed (139,140). SapB protein diffuses into the medium as shown by in situ immunoblot assays and bld mutants could be restored for aerial mycelia formation when were grown near to SapB producing colonies. Indeed, all bld mutants acquired the characteristic fuzzy appearance when were grown next to wild-type colonies. The bld colonies did not loose their fuzzy appearance after 7 days as happened when purified SapB was provided. Instead they sporulated normally, suggesting that the bld mutants need a constant supply of SapB protein in order to sporulate (139,140). Extracellular complementation that restores SapB production in bld mutants was shown by growing bld mutants in pairs close to each other. Every bld mutant (except bld261) was able to be complemented for both SapB production and aerial mycelia formation by one or more bld mutants. Since bld mutants are impaired in SapB synthesis, some signal-molecules must diffuse in the medium to induce aerial mycelia restoration. From the pattern of bld mutants that complement each other, an hierarchical cascade is proposed to exist: 30 bld261 < bldA/bldH< bldG < bIdC < bldD. Assuming that each complementation group produces one signal-molecule, at least 4 distinct signals may exist which facilitate SapB synthesis and subsequently aerial mycelia formation (140). To test the latter hypothesis, the medium on which a bldD mutant (bldD mutant complements all other blds) had been grown previously was either boiled at 100°C for 10 min, or treated with proteinase K. The bld261 mutant, grown on the resolidified, treated medium, was able to sporulate, suggesting that one of the signals is a molecule resistant to heat and to proteinase K (140). A hypothesis was proposed involving the SpaB protein. SapB was proposed to be either a protein that allows aerial hyphae to break the surface tension and protrude vertically out of the vegetative mycelium, or to act as a factor that activates the genes responsible for aerial mycelia formation (139). Both the fact that the bld mutants are capable of sporulating on some minimal media and the fact that the SapB producer strains do not express SapB on any minimal medium tested, indicate the existence of at least two different pathways for aerial mycelia formation. This hypothesis was supported by the absence of SapB protein in some Streptomyces species (139). GENES INVOLVED IN ANTIBIOTIC PRODUCTION The genes involved in antibiotic production can be divided into two major subclasses; the genes which regulate globally the antibiotic biosynthesis such as the abs genes (which are discussed later) and the genes that regulate some but not all four antibiotics. 31 The abaA gene A gene from S. coelicolor was found to be capable of stimulating actinorhodin production in S. lividans when it was present in high copy number (30). S. lividans is a close relative of S. coelicolor which carries the whole set of genes for actinorhodin biosynthesis but does not express it. The gene was named abaA for antibiotic biosynthesis activator and was mapped on the 2 o’clock in S. coelicolor chromosomal map. Disruption of the abaA gene caused total deficiency in actinorhodin production and great reduction in undecylprodigiosin and calcium dependent antibiotic but did not affect the methylenomycin biosynthesis (30). The abaA gene cannot stimulate actinorhodin production in bldA mutants (30) indicating a different mode of action than that of actII-orf4 which in high copy munber can stimulate actinorhodin overproduction in bldA mutants (114). DNA sequences of other Streptomyces species were found to hybridize with abaA suggesting that abaA plays a role in the common regulatory network of antibiotic production in Streptomyces genus (30). Magma: afsQI gene was cloned as being capable of causing actinorhodin overproduction in S. lividans, in a low copy number. afsQI was also able to stimulate undecylprodigiosin and A-factor production [A-factor is a pheromone-like molecule, the presence of which is obligatory for sporulation and antibiotic production in S. gn'seus but not for either S. coelicolor or S. lividans (53)]. The afsQI gene was shown to be capable of restoring antibiotic production to absA but not absB mutants (75). 32 Sequence analysis showed that the afsQl gene encodes a protein that resembles two regulatory proteins of the two-component regulatory systems (75 ). Two component regulatory systems have been found in procaryotes to be involved in adaptive responses by controlling gene expression and are composed of a sensory histidine kinase and a response regulator. A family of response regulators contain an aspartate residue which is phosphorylated allowing the regulatory protein to exert its function (125). The aspartate residue of the M501 protein was changed by site directed mutagenesis at the DNA level. The mutagenized afisQI gene was unable to stimulate actinorhodin production in S. lividans, suggesting that afsQl needs to be phosphorylated in order to be active. Since the genes encoding the two-component regulatory systems are usually linked, a search for the histidine kinase revealed the existence of the afsQZ gene. The afsQZ gene was shown to encode a protein that resembles the sensor-member of the two-component system. The afsQ2 gene was found to not stimulate antibiotic production in S. lividans (75). The genes were mapped at 7 o’clock on the S. coelicolor physical map (75). The transcription of both genes was shown to be temporally controlled since the afsQI-afsQZ transcript appeared just before actinorhodin production starts. me1 and afsQZ gene disruption did not reveal any phenotype other than the normal wild- type which suggests that afsQ1,2 genes do not play an obligatory role in antibiotic biosynthesis in S. coelicolor (75). Mam afsB mutants of S. coelicolor were isolated on the basis of lacking the ability 33 to produce A-factor and the two pigmented antibiotics, actinorhodin and undecylprodigiosin. The afsB gene was named after its capacity to stimulate A- factor synthesis. Even though afsB mutants are impaired in the two pigmented antibiotics and A-factor synthesis, they are able to sporulate normally which suggests that A-factor does not play any role in S. coelicolor and S. lividans morphological differentiation (69). The afsB mutation was localized on the S. coelicolor genetic map to 5 o’clock (69). A clone able to stimulate pigment prduction and A-factor synthesis in afsB mutants was isolated (53). Genetic evidence showed that the cloned gene did not correspond to the afsB“ allele and the cloned allele was renamed ast for A-factor synthesis regulator (124). The ast gene was originally isolated as an open reading frame encoding a 243 aminoacid protein (53) but further analysis revealed that the ast gene encodes a 993 amino acid protein (70). So the ast gene is composed of the originally cloned afsB region which encodes the carboxy-terminal region of the Ast protein and a region, first called the "afsC region", encoding the amino-terminal region of the Ast protein (70). The ast gene was mapped on the genetic (1) and on the physical (81) chromosomal map of S. coelicolor at 7 o’clock. The Ast protein was shown to induce excessive amounts of actinorhodin and undecylprodigiosin, not only in the afsB mutants but also in the wild type (ast*) strains. The inability to clone the ast gene in high copy number either in S. coelicolor or in S. lividans was attributed to a such overproduction of antibiotics that led to the death of the host (70). Interestingly, when either the afsC (70) or the afsB (68) part of the ast gene was present in a high or low copy number , 34 overproduction of the antibiotics in wild-type S. coelicolor strains occured, but to a lesser extent than with the whole ast gene. The Ast protein was shown to activate the transcription of the act gene cluster. Of particular interest is the transcriptional activation of the nod] region, which is known to contain the actII-orf4 gene, the positive activator for actinorhodin biosynthesis. These experiments though did not show if the actII-orf4 gene specifically was activated in the presence of the ast gene (71). Sequence analysis revealed that the afsC region of the ast gene contains two ATP binding sites. When these sites were mutagenized by site-directed mutagenesis, the ast gene was still able to induce antibiotic production but in reduced levels compared to the ajEsR+ allele (70). The afsB part of the ast gene, contains two potential DNA-binding domains which suggest that the Ast protein may activate transcription of developmental genes by binding to the promoters of these genes (53). The afsB part of the ast gene was shown to be able to "bypass" bld mutants of S. lividans (124). Later experiments showed that the Ast protein is phosphorylated by the phosphate of an ATP molecule. The same experiments showed that the Ast protein is not autophosphorylated but requires a specific kinase named Asz (67). The specific kinase (Asz) was localized in the membrane fraction of the cell (67). These results suggest that the Ast-Asz proteins act as a two-component regulatory system; Asz acts as a sensor protein and activates Ast which subsequently acts as a regulatory protein. Interestingly, the ast DNA sequence did not reveal any homology to any regulatory protein of a two-component system (67). However, the 35 N -terminus of the protein showed homology to both RedD and ActII-orf4 proteins, the transcriptional activators of undecylprodigiosin and actinorhodin biosynthesis respectively (67). Whether or not Ast interacts with RedD and ActII-ofr4 remains to be elucidated. The whi genes The whi genes comprise the third major class of genes which are developmentally regulated in S. coelicolor. Based on the observation that a Streptomyces wild-type colony turns from white to grey as its spores mature, it was reasoned that colonies remaining white on prolonged incubation might be unable to make mature spores; this phenotype was designated Whi for white (64). Genetic and morphological characterization of one hundred whi mutants defined eight distinct loci; whiA,B,C,D,E,G,H and whiI (15). Construction of double whi mutants and phenotypic examination revealed that an hierarchy occurs: whiG < whiH < whiA, whiB < whil ; where whiG is epistatic to whiH,A,B,I, whiH to Whi/1,8,1, and whiA and whiB to Mail (16). Only whiB and whiG have been sequenced and their sequences suggest that these genes play a regulatory role. W The whiG gene encodes a sigma factor which resembles the sigma factor D of Bacillus subtilis and the sigma factor F of Salmonella typhimurium and Pseudomonas aeruginosa (17); both of which interact with core RNA polymerase to activate promoters of genes concerned with motility and chemotaxis (58). Overexpression of the whiG gene caused hypersporulation and at the same time 36 inhibition of growth and actinorhodin production was observed (97). Transcriptional studies of whiG revealed that transcription is very low until the onset of aerial mycelium formation (18). Two cloned S. coelicolor promoters have been described that are dependent on whiG (133). These promoters are activated when aerial hyphae become visible. At a high copy number, these promoters and a sigma factor D-dependent promoter of Bacillus subtillis are capable of causing a partial sequestering of the whiG gene product (133,17). W The whiB gene product is an unusually small protein which while not obviously related to any known proteins, bears some resemblance to various transcription factors (24). Two promoters have been defined for whiB. The more upstream, whiBPl, is very weak and constitutive while whiBPZ is strongly expressed only when aerial mycelia are forming (122). Both promoters were active in bldA, bldB, whiA, whiB, whiG, and whiH mutants. It is interesting that whiG mutant can express both whiB promoters even though whiG was morphologically epistatic to whiB in a double whiG whiB mutant (16). These results suggest that there is more than one hierarchy at play during Streptomyces development. The thE' gene The whiE gene cluster was also cloned and shown to be responsible for the pigmentation of the spores (25). There is only one promoter responsible for the transcription of the whiE cluster which is developmentally controlled since transcription takes place only in the aerial mycelia (25). BIBLIOGRAPHY 10. 37 BIBLIOGRAPHY Adamidis, T., P. Riggle, and W. Champness. 1990. Mutations in a new Streptomyces coelicolor locus which globally block antibiotic synthesis but not sporulation. J. Bacteriol. 172: 2962-2969. Adamidis, T., and W. Champness. Unpublished results. An, 6., and L. C. Vining. 1978. Intracellular levels of guanosine 5’-diphosphate 3’-diphosphate(ppGpp) and guanosine 5’-triphosphate 3’-diphosphate(pppGpp) in cultures of Streptomyces griseus producing streptomycin. Can. J. Microbiol. 24: 502-511. Bascaran, V., L. Sanchez, C. Hardisson and A. F. Brana. 1991. Stringent response and initiation of secondary metabolism in Streptomyces clavuligems. J. Gen. Microbiol. 137: 1625-1634. Baumann, R., and H. P. Kocher. 1976. Genetics of Streptomyces glaucescens and regulation of melanin production. In: Genetics of industrial microorganisms, 2nd International Symposium. Macdonald, K.D. (ed.). London, Academic Press, pp. 535-551. Beppu, T. 1992. Secondary metabolites as chemical signals for cellular differentiation. Gene 115: 159-165. Bernan, V., D. Filpula, W. Berber, M. Bibb and E. Katz. 1985. The nucleotide sequence of the tyrosinase gene from Streptomyces antibioticus and characterization of the gene product. Gene 37: 101-110. Betancourt, A. M., V. Bernan, W. Herber and E. Katz. 1992. Analysis of tyrosinase synthesis in Streptomyces antibioticus. J. Gen. Microbiol. 138: 787- 794. Brockman, H., A. Zeeck, K. Van der Merve, and W. Muler. 1966. Die konstitution des actinorhodins. Justus Liebigs Annln Chem. 698: 3575-3579. Burton, C. J ., E. P. Guthrie, and K. F. Chater. 1991. Phage vectors that allow monitoring of transcription of secondary metabolism genes in Streptomyces. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 38 Bio/Technology 9: 652-656. Bunting, M. I., C. F. Robinow, and H. Bunting. 1949. Factors affecting the elaboration of pigment and polysaccharide by Serratia marcescens. J. Bacteriol. 58: 114-115. Cashel, M., and J. Gallant. 1969. Two compounds implicated in the function of the RC gene of Escherichia coli. Nature 221: 838-841. Castro, A. J. 1967. Antimalarial activity of prodigiosin. Nature. 213: 903. Champness, W. C. 1988. New loci required for Streptomyces coelicolor morphological and physiological differentiation. J. Bacteriol. 170: 1168-1174. Chater, K. F. 1972. A morphological and genetic mapping study of white colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 72: 9-28. Chater, K. F. 1975. Construction and phenotypes of double sporulation deficient mutants of Streptomyces coelicolor. J. Gen. Microbiol. 72: 9-28. Chater, K., C. J. Burton, K. A. Plaskitt, M. J. Buttner, C. Mendez, and J. D. Helmann. 1989. The developmental fate of S. coelocolor hyphae depends upon a gene product homologous with the motility sigma factor of B. subtilis. Cell 59: 133- 143. Chater, K. F. Unpublished results. Crameri, R., L. Ettinger, R. Hutter, K. Lerch, M. A. Suter and J. A. Vetterli. 1982. Secretion of tyrosinase in Streptomyces glaucescens. J. Gen. Microbiol. 128: 371-379. Crameri, R., G. Hintermann, R. Butter and T. Kieser. 1984. Tyrosinase activity in Streptomyces glaucescens is controlled by three chromosomal loci. Can. J. Microbiol. 30: 1058-1067. Cundliffe, E. Involvement of specific portions of ribosomal RNA in defined ribosomal functions: a study utilizing antibiotics. In: Structure, function, and genetics of ribosomes, Hardesty B. and G. Kramer (eds.), Springer-Verlag, pp. 586-604. Danchin, A., and L. Dondon. 1979. Regulatory features of tRNA“" expression in Escherichia coli K12. Biochem. Biophys. Res. Commun. 90: 1280-1286. D’Aoust, J. Y. and N. N. Gerber. 1974. Isolation and purification of prodigiosin from Vibn'a psychraerythreus. J. Bacteriol. 118: 756-757. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 39 Davis, N. K., and K. F. Chater. 1992. The Streptomyces coelicolor whiB gene encodes a small transcription factor-like protein dispensable for growth but essential for sporulation. Mol. Gen. Genet. 232: 351-358. Davis, N. K., and K. F. Chater. 1990. Spore colour in Streptomyces coelicolor A3(2) involves the developmentally regulated synthesis of a compound biosynthetically related to polyketide antibiotics. Mol. Microbiol. 4: 1679-1691. Dan], J. L., and L. C. Vining. 1990. Nutritinal control of actinorhodin production by Streptomyces coelicolor A3(2): suppressive effects of nitrgen and phosphate. Appl. Microbiol. 33: 449-454. Feitelson, J. S. and Hopwood, D. A. 1983. Cloning a Streptomyces gene for an O-methyltransferase involved in antibiotic biosynthesis. Mol. Gen. Genet. 190: 394-398. Feitelson, J. S., F. Malpartida and D. A. Hopwood. 1985. Genetic and biochemical characterization of the red gene cluster of Streptomyces coelicolor A(3)2. J. Gen. Microbiol. 131: 2431-2441. Fernandez-Moreno, M. A., J. L. Caballero, D. A. Hopwood, and F. Malpartida. 1991. The act cluster contains regulatory and antibiotic export genes, direct targets for translational control by the bldA transfer RNA gene of Streptomyces. Cell 66: 769-780. Fernandez-Moreno, M. A., A. J. Martin-Triana, E. Martina, J. Niemi, H. M. Kieser, D. A. Hopwood, and F. Malpartida. 1992. abaA, a new pleiotropic regulatory locus for antibiotic production in Streptomyces coelicolor. J. Bacteriol. 174: 2958-2967. Gale, E. F., E. Cundliffe, E. P. Reynolds, H. M. Richmond, J. M. Waring. 1981. The molecular basis of antibiotic action. John Wiley and Sons, London and New York. Gallant, J. A. 1979. Stringent control in E. coli. Annu. Rev. Genet. 13: 393- 415. Gandhi, N. M., J. Nazareth, P. V. Divekar, H. Kohl and N. J. daSouza. 1973. Magnesidin, a novel magnesium-containing antibiotic. J. Antibiot. 26: 797-798. Geistlieh, M., S. Irniger and R. Hutter. 1989. Localization and functional analysis of the regulated promoter from the Streptomyces glaucescens mel operon. Mol.Microbiol. 3: 1061-1069. Gerber, N. N. 1975 . A new prodiginine (prodigiosin-like) pigment from 36. 37. 38. 39. 40. 41. 42. 43. 45. 46. 47. 40 streptomyces. Antimalarial activity of several prodiginines. J. Antibiot. 28: 194- 199. Gerber, N. N. 1975. Prodigiosin-like pigments. Crit. Rev. Microbiol. 3: 469-485. Gerber, N. N. and M. P. Lechevalier. 1976. Prodigine (prodigiosin-like) pigments from Streptomyces and other aerobic Actinomycetes. Can J. Microbial. 22: 658-667. Goldschmidt, M. C. and R. P. Williams. 1968. Thiamine-induced formation of the monapyrrale moiety of prodigiosin. J. Bacteriol. 96: 609-616. Goldthwaite, C., and 1. Smith. 1972. Physiological characterization of antibiotic resistant mutants of Bacillus subtilis. Mol. Gen. Genet. 114: 190-204. Gorst-Alhnan, C. P., B. A. M. Rudd, C-J Chang, H. G. Floss. 1981. Biasynthesis of actinorhodin. Determination of the point of dimerizatian. J. Org. Chem. 46: 455-456. Grate, U., G. Reinhardt, W. Schade, I. Eritt, W. F. Fleck, and L. Radics. 1983. Interspecific inducers of cytodifferentiatian and anthracycline biosynthesis from Streptomyces bikiniensis and S. cyaneafitscatus. Biatechnal. Lett. 5: 591- 596. Grate, U., W. Schade, I. Eritt, and W. F. Fleck. 1982. A new inducer of anthracycline biosynthesis from Streptomyces viridochromagenes. J. Antibiot. 35: 1722-1723. Gramaja, H. C., E. Takana, and M. J. Bibb. 1993. Stationary-phase production of the antibiotic actinorhodin in Streptomyces coelicolor A3(2) is transcriptionally regulated. Mol. Microbiol. 7: 837-845. Gregory, K. F., and J. C. C. Huang. 1964a. Tyrosinase inheritance in Streptomyces scabies. 1. Genetic recombination. J. Bacteriol. 87: 1284- 1286. Gregory, K. F., and J. C. C. Huang. 1964b. Tyrosinase inheritance in Streptomyces scabies. 11. Induction of tyrosinase deficiency by acridine dyes. J. Bacteriol. 87 : 1287-1294. Guijarro, J., R. Santamaria, A. Schauer, and R. Losick. 1988. Promoter determining the timing and spatial localization of transcription of a cloned Streptomyces coelicolor gene encoding a spore-associated polypeptide. J. Bacterial. 170: 1895-1901. Guthrie, E. F., and K. F. Chater. 1990. The level of a transcript required for 48. 49. 50. 51. 52. 53. ' 54. 55. 56. 57. 58. 59. 41 the production of a Streptomyces coelicolor antibiotic is conditionally dependent on a transfer RNA gene. J. Bacteriol. 172: 6189-6193. Hallam, S. E., F. Malpartida, and D. A. Hopwood. 1988. Nucleotide sequence, transcription and deduced function of a gene involved in polyketide antibiotic synthesis in Streptomyces coelicolor. Gene 74: 305-320. Haneishi, T., N. Kitahara, K. Takiguchi, M. Arai, and S. Sugawara. 1974. New antibiotics, methylenomycin A and B. Producing organism, fermentation and isolation, biological activities and physical and chemical properties. J. Antibiot. 27 : 386-392. Hansen, M. T., L. M. Pato, S. Malin, P. N. Fill, and K. Van Meyenburg. 1975. Simple downshift and resulting lack of correlation between ppGpp pool size and ribonucleic acid accumulation. J. Bacterial. 122: 585-591. Hara, 0., and T. Beppu. 1982a. Mutants blocked in streptomyc in production in Streptomyces gn'seus-the role of A-factar. J. Antibiot. 35: 349-358. Hara, 0., and T. Beppu. 1982b. Induction of streptomycin inactivating enzyme by A-factar in Streptomyces gn'seus. J. Antibiot. 35: 1208-1215. Hara, 0. S., T. Horinouchi, T. Uozumi, and T. Beppu. 1983. Genetic analysis of A-factor synthesis in Streptomyces coelicolor A3(2) and Streptomyces griseus. J. Gen. Microbiol. 129: 2939-2914. Harashima, K., N. Tsuchida, T. Tanaka and J. Nagatsu. 1967. Progigiosin- 25C. Isolation and the chemical structure. Agric. Biol. Chem. 31: 481-489. Harasym, M., and J. Piret. 1990. The Streptomyces coelicolor A3(2) bldB region contains at least two genes involved in morphological development. J. Gen. Microbiol. 136: 1543-1550. Heinemann, B., A. J. Howard and H. J. Palocz. 1970. Influence of dissolved oxygen levels on production of L-asparaginase and prodigiosin by Serratia marcescens. Appl. Micriabiol. 19: 800-804. Held, T., and H. J. Kutzner. 1990. Transcription of the tyrosinase gene in Streptomyces michiganensis DSM 40015 is induced by copper and repressed by ammonium. J. Gen. Microbial. 136: 2413-2419. Helmann, J. D. 1991. Alternative sigma factors and the regulation of flagellar gene expression. Mol. Microbiol. 5: 2875-2882. Hintermann, G., M. Zatchej and R. Hutter. 1985. Cloning and expression of 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 42 the genetically unstable tyrosinase structural gene from Streptomyces glaucescens. Mol.Gen. Genet. 200: 422-432. Hobbs, G., C. M. Frazer, D. C. J. Gardner, F. Flett and S. G. Oliver. 1990. Pigmented antibiotic production by Streptomyces coelicolor A3(2): kinetics and the influence of nutrients. J. Gen. Microbiol. 136: 2291-2296. Hang, S-K., M. Kito, T. Beppu, and S. Horinouchi. 1991. Phosphorylation of the Ast product, a global regulatory protein for secondary metabolite formation in Streptomyces coelicolor A3(2). J. Bacteriol. 173: 2311-2318. Hopwood, D. A., M. J. Bibb, K. F. Chater, T. Kieser, C. J. Brunton, H. M. Kieser, D. J. Lydiate, C. P. Smith, J. M. Ward, and H. Schrempf. 1985. Genetic manipulation of Streptomyces. A laboratory manual. The John Innes Foundation. Norwich, UK Hopwood, D. A., F. Malpartida, H. M. Kieser, H. Ikeda, J. Duncan, 1. Fujii, B.A. M. Rudd, H. G. Floss, and S. Omura. 1985. Production of "hybrid" antibiotics by genetic engineering. Nature 314: 642-644. Hopwood, D. A., H. Wildermuth, and H. M. Palmer. 1970. Mutants of Streptomyces coelicolor defective in sporulation. J. Gen. Microbiol. 61: 397-408. Hopwood, D. A. and H. M. Wright. 1983. CDA is a new chromasomally determined antibiotic in Streptomyces coelicolor A3(2). J. Gen. Microbiol. 129: 3575- 3579. Hopwood, D. A., and D. H. Sherman. 1990. Molecular genetics of polyketides and its comparison to fatty acid biosynthesis. Annu. Rev. Genet. 24: 37-66. Horinouchi, S. and T. Beppu. 1992. Regulation of secondary metabolism and cell differentiation in Streptomyces: A-factar as a microbial hormone and the Ast protein as a component of a two-component regulatory system. Gene 115: 167-172. Horinouchi, S., and T. Beppu. 1984. Production in large quantities of actinorhodin and undecylprodigiosin induced by afisB in Streptomyces lividans. Agric. Biol. Chem. 48: 2131-2133. Horinouchi, S., O. Hara, and T. Beppu. 1983. Cloning of a pleitropic gene that positively controls biosynthesis of A-factor, actinorhodin, and prodigiosin in Streptomyces coelicolor A3(2) and Streptomyces lividans. J. Bacteriol. 155: 1238- 1248. Horinouchi, S., M. Kito, K. Furuya, S. K. Hang, K. Miyake, and T. Beppu. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 43 1990. Primary structure of Ast, a global regulatory protein for secondary metabolite formation in Streptomyces coelicolor A3(2). Gene 95: 49-56. Horinouchi, S., F. Malpartida, D. A. Hopwood, and T. Beppu. 1989. afsB stimulates transcription of the actinorhodin biosynthetic pathway in Streptomyces coelicolor A3(2) and Streptomyces lividans. Mol. Gen. Genet. 215: 355-357 Horinouchi, 8., H. Suzuki, M. Nishiyama, and T. Beppu. 1989. Nucleotide sequence and transcriptional analysis of the Streptomyces gn'seus gene (afsA) responsible for A-factor biosynthesis. J. Bacterial. 171: 1206- 1210. Huber, M., G. Hintermann and K. Lerch. 1985 . Primary structure of tyrosinase from Streptomyces glaucescens. Biochem. 24: 6038-6044. Huber, M., R. Hutter and K. Ierch. 1987. The promoter of the Streptomyces glaucescens mel aperon. Nucl. Acids Res. 15: 8106. Ishizuka, H., S. Horinouchi, H. M. Kieser, D. A. Hopwood, and T. Beppu. 1992. A putative two-component regulatory system involved in secondary metabolism in Streptomyces spp. J. Bacterial. 174: 7585-7594. Katz, E., and A. Betancourt. 1988. Induction of tyrosinase by L-methionine in Streptomyces antibioticus. Can. J. Microbiol. 34: 1297-1303. Katz, E., C. J. Thompson and D. A. Hopwood. 1983. Cloning and expression of the tyrosinase gene from Streptomyces antibioticus in Streptomyces lividans. J. Gen. Microbiol. 129: 2703-2714. Kelly, K., K. Ochi and G. H. Jones. 1991. Pleitrapic effects of a relC mutation in Streptomyces antibioticus. J. Bacterial. 173: 2297-2300. Khakhlov, A. S. 1980. Problems of studies of specific cell autoregulators (on the example of substances produces by some actinomycetes). p.201-210. In S. N. Ananchenko (ed.). Frontiers of bioorganic chemistry and molecular biology. Pergamon Press, Oxford and New York. Kieser, T., L. Ettilinger and R. Hutter. 1976. Mutants of Streptomyces glaucescens constitutive far tyrosinase synthesis. In: Genetics of Actinomycetales, Freerksen, E., I. Tamok and J .H. Thumin (eds.). Stuttgart & New York: Gustav Fischer. pp. 59-60. Kieser, H. M., T. Kieser, and D. A. Hopwood. 1992. A combined genetic and physical map of the chromosome of Streptomyces coelicolor A3(2). J. Bacteriol. (In press). 82. 83. 85. 86. 87. 88. 89. 90. 91. 92. 93. 44 Kinashi, H., and M. Shimaji-Murayama. 1991. Physical characterization of SCP1, a giant linear plasmid from Streptomyces coelicolor A3(2). J. Bacterial. 173: 1523-1529. Kirby, R. and D. A. Hopwood. 1977. Genetic determination of methylenomycin by the SCP1 plasmid of Streptomyces ceolicalor A3(2). J. Gen. Microbiol. 98: 239-252. Kleiner, E. M., V. V. Onoprienko, S. A. Pliner, V. S. Soifer and A. S. Khokhlav. 1977. The synthesis of racemic A-factar. A biological regulator from Streptomyces gn'seus (Actinomyces streptomycini). Bioarg. Khim. 3: 424- 426. Lakey, J. H., E. J. A. Lea, B. A. M. Rudd, H. M. Wright, and D. A. Hopwood. 1983. A new channel-forming antibiotic from Streptomyces coelicolor A3(2) which requires calcium for its activity. J. Gen. Microbial. 129: 3565-3573. 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 Dev. 1: 1305-1310. Lee, Y.-H.W., B.-F. Chen, S.-Y. Wu, W.-M. Leu, J.-J. Lin, C. W. Chen and S. J. Lo. 1988. A trans-acting gene is required for the phenotypic expression of a tyrosinase in Streptomyces. Gene 65: 71-81. Lerch, K., and L. Ettlinger. (1972). Purification and characterization of a tyrosinase from Streptomyces glaucescens. Eur. J. Biochem. 31: 427-437. Leskiw, B. K., M. J. Bibb, and K. F. Chater. 1991a. The use of a rare cadon specifically during development? Mol. Microbiol. 5: 2861-2867. Leskiw, B. K., E. J. Lawlor, J. M. Fernandez-Abalos, and K. F. Chater. 1991b. TI‘A codons in some genes prevent their expression in a class of developmental, antibiotic-negative, Streptomyces mutants. Proc. Natl. Acad Sci. USA 88: 2461-2465. Leskiw, B. K., R. May, E. J. Lawlor, and K. F. Chater. 1993. Accumulation of bldA -specified tRNAis temporally regulated in Streptomyces coelicolor A3(2). J. Bacteriol. 175: 1995-2005. Leu, W.-M., S.-Y. Wu, J.-J. Lin, S. J. Lo and Y.-H.W. Lee. 1989. Analysis of the promoter region of the melanin locus from Streptomyces antibioticus. Gene 84:267-277. Malpartida, F, and D. A. Hopwood. 1984. Molecular cloning of the whole 94. 95. 96. 97. 99. 100. 101. 102. 103. 104. 105. 106. 45 biosynthetic pathway of a Streptomyces antibiotic and its expression in an heterologous hast. Nature. 309: 462- 464. Malpartida F., J. Niemi, R. Navarrete and D. A. Hopwood. 1990. Cloning and expression in a heterologous host of the complete set of genes for biosynthesis of the Streptomyces coelicolor antibiotic undecylprodigiosin. Gene. 93: 91-99. Marahiel, M. A. 1992. Multidomain enzyme involved in peptide synthesis. FEBS Lett. 307: 40-43. McVittie, A. 1974. Ultrastructural studies on sporulation in wild-type and white colony mutants of Streptomyces coelicolor. J. Gen. Microbial. 81: 291-302. Mendez, C., and K. F. Chater. 1987. Cloning of whiG, a gene critical for sporulation of Streptomyces coelicolor A(3)2. J. Bacteriol. 169: 5715-5720. Merrick, M. J. 1976. A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 96: 299-315. Miyake, K., S. Horinouchi, M. Yoshida, N. Chiba, K. Mori, N. Nogawa, N. Morikawa, and T. Beppu. 1989. Detection and properties of A-factor-binding protein from Streptomyces gn'seus. J. Bacteriol. 171: 4298-4302. Miyake, K., T. Kuzuyama, S. Horinouchi, and T. Beppu. 1990. The A-factar- binding protein of Streptomyces griseus negatively controls streptomycin production and sporulation. 172: 3003-3008. Mullin, D. A., G. M. Garcia, and J. R. Walker. 1984. An E. coli DNA fragment 118 base pairs in length provides complementing activity. Cell 37: 669-674. Nagase, T., S. Ishii and F. Imamato. 1988. Differential transcriptional control of the two tRNAM“ genes of Escherichia coli K-12. Gene 678: 49-57. 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. Nihara, T., Y. Shimizu, H. S. Kim, and Y. Yamada. 1988. Structure-activity relationship of virginiae butanolide C, an inducer of virgiamycin production in Streptomyces virginiae. J. Antibiot. 42: 1828-1837. Ochi, K. 1986. Occurrence of the stringent response in Streptomyces sp. and its significance for the initiation of morphological and physiological differentiation. J. Gen. Microbiol. 132: 2621-2631. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 46 Ochi, K. 1987. Changes in nucleotide pools during sporulation of Streptomyces griseus in submerged culture. J. Gen. Microbiol. 133: 2787-2795. Ochi, K. 1989. The tsr gene-coding plasmid pIJ702 prevents thiopeptin from inhibiting ppGpp synthesis in Streptomyces lividans. FEMS Microbiol. Lett. 61:219-224. Ochi, K. 1990a. Streptomyces relC mutants with an altered ribosomal protein ST -L11 and genetic analysis of a Streptomyces griseus relC mutant. J. Bacteriol. 172: 4008-4016. Ochi, K. 1990b. A relaxed (rel) mutant of Streptomyces coelicolor A3(2) with a missing ribosomal protein lacks the ability to accumulate ppGpp, A-factor and prodigiosin. J. Gen. Microbiol. 136: 2405-2412. Ochi, K., J. C. Kandala, and E. Freese. 1981. Initiation of Bacillus subtilis sporulation by the stringent response to partial amino acid deprivation. J. Biol. Chem. 256: 6866-6875. Ochi, K., J. C. Kandala, and E. Freese. 1982. Evidence that Bacillus subtilis sporulation induced by the stringent response is caused by the decrease in GTP or GDP. J. Bacteriol. 151: 1062-1065. Ochi, K., A. Penyige and G. Barabas. 1992. The possible role of ADP- ribosylatian in sporulation and streptomycin production by Streptomyces griseus. J. Gen. Microbiol. 138: 1745-1750. Passantino, R., A. M. Puglia, and K. F. Chater. 1991. Additional copies of the act]! regulatory gene induce actinorhodin production in pleiotropic bld mutants of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 137: 2059-2064. Piret, J .M., and K. F. Chater. 1985. Phage-mediated cloning of bldA, a region involved in Streptomyces coelicolor morphological development and its analysis by genetic complementation. J. Bacteriol. 163: 965-972. Riesenberg, D., F. Bergter and C. Kari. 1984. Effect of serine hydroxamate and methyl a-D-glucopyranoside treatment on nucleoside polyphosphate pools, RNA and protein accumulation in Streptomyces hygroscopicus. J. Gen. Microbiol. 130: 2549-2558. Rudd , B. A. M., and D. A. Hopwood. 1979. Genetics of actinorhodin biosynthesis by Streptomyces coelicolor A3(2). J. Gen. Microbial. 114: 35-43. Rudd, B. A. M. and D. A. Hopwood. 1980. A pigmented mycelial antibiotic in Streptomyces coelicolor: control by a chromosomal gene cluster. J. Gen. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 47 Microbial. 119: 333-340. Sauer, U., and P. Durre. 1992. Possible function of tRNAT'" in regulation of solvent formation in Clastridium acetabutylicum. FEMS Microbiol. Lett. 100: 147-154. Shauer Smith, 1., P. Paress and S. Pestka. 1978. Thiostrepton-resistant mutants exhibit relaxed synthesis of RNA. Proc. Natl. Acad. Sci. USA. 75: 5993-5997. Soliveri, J ., K. L. Brown, M. J. Buttner, and K. F. Chater. 1992. Two promoters for whiB sporulation gene of Streptomyces coelicolor A3(2) and their activities in relation to development. J. Bacteriol. 174: 6215-6220. Spadara, A., A. Spena, V. Santonastaso and P. Donini. 1981. Stringency without ppGpp accumulation. Nature 291: 256-258. Stein, D., and S. N. Cohen. 1989. A cloned regulatory gene of Streptomyces lividans can suppress the pigment deficiency phenotype of different developmental mutants. J. Bacterial. 171: 2258-2261. Stock, J. B., A. J. Ninfa, and A. M. Stock. 1989. Protein phaspharylation and regulation of adaptive responses in bacteria. Microbiol. Rev. 53: 450-490. Strauch, E., E. Takano, H. A. Baylis and M. J. Bibb. 1991. The stringent response in Streptomyces coelicolor A3(2). Mol. Microbiol. 5: 289-298. Stutzman-Engwall, K., S. Otten, and R. Hutchinson. 1992. Regulation of secondary metabolism in Streptomyces spp. and overproduction of Daunarubicin in Streptomyces peacetius. J. Bacteriol. 174: 144-154. Suter, M. A. , R. Hutter and T. Leisinger. 1978a. Mutants of Streptomyces glaucescens affected in the production of extracellular enzymes. In: Genetics of the Actinomycetales. Freerksen E., I. Tamok and J.H. Thumim (eds). Gustav Fischer Verlag, pp.61-64. Suter, M. A., R. Butter and T. Leisinger. 1978b. Induction of albino mutants in Streptomyces glaucescens. Experientia 34: 1669. Szaba, G., and S. Vitalis. 1992. Sporulation without aerial mycelium formation on agar medium by Streptomyces bikiniensis HHl, an A-factor-deficient mutant. J. Gen. Microbiol. 138: 1887-1892. Takano, E., H. C. Gramajo, E. Strauch, N. Andres, J. White and M. .I. Bibb. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141 142. 48 1992. Transcriptional regulation of the redD transcriptional activator gene accounts for growth-phase-dependent production of the antibiotic undecylprodigiosin in Streptomyces coelicolor A3(2). Mol.Microbiol. 6: 2797- 2804. Tamura, F., S. Nishimura, and M. Ocki. 1984. The E. coli divE mutation, which differentially inhibits synthesis of certain proteins, is in tRNAs“. EMBO J. 3: 1103-1107. Tan, H., and K. F. Chater. 1993. Two developmentally controlled promoters of Streptomyces coelicolor A3(2) that resemble the major class of motility- related promoters in other bacteria. J. Bacteriol. 175: 933-940. Tanaka, W. K., L. B. deMedina and W. R. Hearn. 1972. Labelling patterns in prodigiosin biosynthesis. Biochem. Biophys. Res. Commun. 46: 731-737. Tosa, T., and I. L. Pizer. 1971. Biochemical bases for the antimetabolic action of L-serine hydraxymate. J. Bacteriol. 106: 972-982. Tsao, S.-W., B. A. M. Rudd, X.-G. He, C.-J. Chang and H. G. Floss. 1985. Identification of a red pigment from Streptomyces coelicolor A(3)2 as a mixture of prodigiosin derivatives. J. Antibiot. 38: 128-131. Vujakliia, D., K. Ueda, S. Hang, T. Beppu, and S. Horinouchi. 1991. Identification of an A-factor-dependent promoter in the streptomycin biosynthetic gene cluster of Streptomyces gn'seus. Mol. Gen. Genet. 229: 119- 128. Wildermuth, H., and D. A. Hopwood. 1970. Septation during sporulation in Streptomyces coelicolor. J. Gen. Microbiol. 60: 57-59. Willey, J. W., R. Santamaria, J. Guijarro, M. Geislich, and R. Losick. 1991. Extracellular complementation of a developmental mutation implicates a small sporulation protein in aerial mycelium formation by S. coelicolor. Cell 65: 641- 650. Willey, J. W., J. Schwedock, and R. Losick. 1993. Multiple extracellular signals govern the production of a marphogenetic protein involved in aerial mycelium formation by Streptomyces coelicolor. Genes & Dev. 7: 895-903. Williams, R. P., C. L. Gott and S. M. H. Qadri. 1971. Induction of pigmentation in nonproliferating cells of Serratia marcences by addition of single amino acids. J. Bacteriol. 106: 444-448. Williams, R. P. and W. R. Hearn. 1967. Prodigiosin. In: D. Gottlieb and PD. 143. 144. 145. 146. 147. 148. 49 Shaw (eds.), Antibiotics, vol. 2. Springer-Verlag, Berlin, pp. 410—432. Williams, R. P. and C. L. Gott. 1964. Inhibition by streptomycin of the biosynthesis of prodigiosin. Biochem. Biophys. Res. Commun. 16: 47-50. Witney, F. R., M. L. Failla and E. D. Weinberg. 1977. Phosphate inhibition of secondary metabolism in Serratia marcescens. Appl. Env. Microbiol. 33: 1042- 1046. Wright, H. M., and D. A. Hopwood. 1976. Actinorhodin is a chromasomally determined antibiotic in Streptomyces coelicolor A3(2). J. Gen. Microbiol. 96: 289-297. Wright, H. M., and D. A. Hopwood. 1976. Identification of the antibiotic determined by the SCP1 plasmid of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 95: 96-106. Yamada, Y., K. Sugamura, 1K. Kondo, M. Yanagimoto, and H. Okada. 1987. The structure of inducing factors for virgiamycin production in Streptomyces virginiae. J. Antibiot. 40: 496-504. Zukawski, M. M., D. F. Gaffney, D. Speck, M. Kauffmann, A. Findeli, A. Wisewp, and J. Lecocq. 1993. Chromogenic identification of genetic regulatory signals in Bacillus subtilis based on expression of a cloned Pseudamanas gene. Proc. Natl. Acad. Sci. USA 80: 1101-1105. Chapter 2: Mutations in a new Streptomyca coelicolor locus which globally block antibiotic biosynthesis but not sporulation 50 louaNAL or BACTERIOLOGY. June 1990. p. 2962-2969 0021-9193/90/062962-0850200/0 Copyright 0 1990. American Society for Microbiology 51 Vol. 172. No. 6 Mutations in a New Streptomyces coelicolor Locus Which Globally Block Antibiotic Biasynthesis but Not Sporulationi TRIFON ADAMIDIS.1 PERRY RIGGLE.2 AND WENDY CHAMPNESSM“ Genetics Programl and Department of Microbiology and Public Health.2 Michigan State University, East Lansing. Michigan 48824-1101 Received 11 October 1989/Accepted 1 March 1990 Streptomyces coelicolor produces four known antibiotics. To define genetic elements that reguhte antibiotic synthesis, we screened for mutations that visibly blocked synthesis of the two pigmented antibiotics and found that the mutant strains which we recovered were of two classes—double mutants and mutants in which all four antibiotics were blocked. The mutations in these multiply blocked strains define a new locus of S. coelicolor which we have named 063.4. The genetic location of absA. at 10 o’clock. is distinct from the locations of the antibiotic gene clusters and from other known mutations that affect antibiotic synthesis. The phenotype of thcabsA mutantssuggeststhatalls. coelicolorantibloticsynthesisgenesaresubjecttoacommonglobal regubfluthmhathauhpandkthafmmmhmwthatabukagenedcmmpuentdthe mechanism maul-wry Streptomyces spp. are well known for their capacity to synthesize an enormous variety of antibiotics as secondary metabolites. The complex life cycle of these bacteria in- cludes differentiation into a sporulating. antibiotic-producing multicellular organism. In colonies. difi'erentiation is first visible after several days of growth. when prespore aerial hyphae grow vertically out of the vegetative mycelium. Aerial hypha formation is temporally coordinated with anti- biotic production, and there is genetic evidence that they are coordinately regulated (26). The control of Streptomyces coelicolor antibiotic synthe- sis is an aspect of development that is especially amenable to genetic analysis, since two of its four antibiotics are pig- mented and easily seen. Also. the four antibiotics have been genetically characterized (15. 20. 24. 30, 31. 36). and three antibiotic gene clusters have been cloned (6. 9. 24). Genetic evidence that the onset of sporulation and antibiotic produc- tion are subject to common control has come from the isolation of single mutations that block both processes—the bld mutations. Several bld loci have been defined by such mutations (3. 5, 26). One interpretation of the Bid phenotype is that bld genes function in the vegetative colony to initiate expression of genes involved in sporulation and antibiotic synthesis. A partial. nutritional uncoupling of antibiotic synthesis from sporulation has been observed in mutants of the bldA. -C. -D. and -G loci (3. 26). In these bld mutants. the Bld‘ phenotype varies with nutritional conditions: although Bld‘ on glucose media. the mutants are capable of some sporula- tion on poor carbon sources. However. antibiotic production is not normally restored under the nutritional conditions that restore sporulation. Defective antibiotic synthesis has also been reported in isolates with mutations of the afsB locus, which was identi- fied during studies on the synthesis and role of A-factor in S. coelicolor (10). A-factor. a small. difi'usible molecule. has been implicated in both antibiotic production (streptomycin) ‘ Corresponding author. t Journal article no. 13112 of the Michigan Agricultural Experi- ment Station. 2962 and sporulation in Streptomyces griseus (reviewed in refer- ence 10). In this streptomycete. A-factor-deficient mutants were found to be defective in streptomycin synthesis and sporulation. but the mutants could be phenotypically cor- rected by growth in the presence of added A-factor.'ln S. coelicolor. afsB mutants were isolated on the criterion of lacking A-factor and were found to be also defective in production of actinorhodin and undecylprodigiosin. How- ever, since exogenous A-factor could not correct the pig- mentation defect in afsB mutants, and since additional mutants could be isolated that lacked A-factor but that were not pigmentation defective (afsA mutants). the role. if any, for A-factor in S. coelicolor antibiotic production has not been determined. An additional locus. ast, has been reported to be in- volved in global regulation of Streptomyces antibiotic syn- thesis. This locus was first identified in attempts to clone the wild-type qfsB gene (17). Although afsB‘ was not cloned. a clone capable of suppressing the afsB phenotype was ob- tained (27. 32). The gene responsible has been named qst (R for regulatory) (32) because it is capable. when on a multicopy plasmid. of increasing pigmented antibiotics not only in afsB mutants but also in wild-type S. coelicolor and Streptomyces lividans and even in a variety of S. lividans mutants blocked early in development (17. 32). This stimu- latory effect. at least in the case of actinorhodin synthesis. has been shown to be due to increased transcription of the actinorhodin genes ( 18). The null phenotype of ast has not been determined. so it is not known whether ast has an obligatory role in normal antibiotic synthesis. We sought to isolate a new class of Streptomyces devel- opmental mutants—mutants that are globally but specifically blocked in antibiotic synthesis so that no antibiotics are synthesized but sporulation proceeds normally. If such mutants could be found. they would provide evidence for the existence of a class of regulators that act directly and specifically on the antibiotic genes. Such mutants would also indicate that sporulation can proceed independently of anti- biotic synthesis. suggesting that sporulation and antibiotic synthesis may be regulated by independent. parallel genetic 52 VOL. 172. 1990 S. COELICOLOR ANTIBIOTIC SYNTHESIS MUTANTS 2963 TABLE 1. Strains of S. coelicolor. phage. and plasmids used in this study Strain|. phaglc. 0' Relevant genotype“ Source S. coelicolor 1650 cyleS rnthB2 agaAl NF SCP2’ K. Chater 1514 proAl cysAl5 argAI uraAl nicAl agaAl NF SCPZ' K. Chater 1668 cyle8 mthBZ bldA39 NF K. Chater (28) 11501 ham] :4er strAl SCPl‘ SCP2' Pgl' K. Chater (7). RES cyle8 proAl argAl afsB SCP‘Z‘ S. Horinouchi (10) TK18 argAl uraAl strAl act-l4] redEoo D. Hopwood C505 hisAl uraAl strAl abs-505 SCPI' SCP2‘ Pgl' This work C542 hisAl uraAl strAl abs-542 SCP1“ SCP2’ Pgl’ This work C554 hisAl uraAl strAl abs-554 SCPI' SCP2‘ Pgl' This work C577 hisAl uraAl strAl abs-577 SCPl‘ SCPZ' Pgl’ This work C5421 hisAl strAl abs-542 NF This work’ C5422 argAl cysAl5 proAl strAl abs-542 NF This work‘ C5423 him] 14er MM! abs-542 NF This work" ¢C31 KC603 c’ AattP and bldA‘ insert. Vio' K. Chater (28) ¢C31 KC516 c‘ AattP. Vio' Tsr’. Sstl and Psrl cloning sites K. Chater (29) ¢C3l PR106 c’ AattP and 2.1-kp SacI-Psti (1st insert. Tsr’ This work Plasmid p11702-AP22 2.1-kp Sacl-Pstl ast insert S. Horinouchi (l9) ‘ Abheviations: SCP1. S. coelicolor plasmid 1 (33). SC P2. S. coelicolor plasmid 2 (2); NF. SC Pi is integrated into the chromosome at 9 o'clock (see text): Pgl‘ . 0C3] sensitive (7). Vio'. viomycin resistant: Tsr'. thiostrepton resistant. ‘ Recombinant from 1650 x C542. ‘ Recombinant from C5421 x 1514. pathways, once the decision to difi‘erentiate has been made and executed in the streptomyces colony. in this report we describe the isolation and characteriza- tion of a new class of mutations that globally block antibiotic synthesis without blocking sporulation. These mutations have no nutritionally conditional phenotype and define a new S. coelicolor locus. absA. (A portion of this work was presented at the American Society for Microbiology Conference on the Genetics and Molecular Biology of Industrial Microorganisms. University of Indiana. Bloomington, October 1988.) MATERIAISANDMETHODS Bacterial strain and phages. The strains used for genetic analysis were derivatives of S. coelicolor A3 (2). S. lividans 1326 was used for phage propagation (23). The phage ¢C3l KC603 contains the bldA* allele on the 5.6-kilobase insert. KC603 lysogens for complementation studies were con- structed as described previously (3). Media and culture techniques. Minimal plate media for genetic analysis. nutrient agar. R5. and YEME were as described by Hopwood et al. (14). YMPG was described previously (16) YEG contained 1% yeast extract and 1% glucose. Liquid minimal medium had the same composition as plate minimal medium. without agar. For assay of the calcium-dependent antibiotic (CDA), low-calcium nutrient agar (Oxoid. Inc.) with or without added calcium [as Ca(NO,)2 to 12 mM] was used (21). Culture growth. genetic crossing techniques. and spare preparations were as de- scribed by Hopwood et al. (14). Antibiotic assays. Methylenomycin production and resis- tance genes are located on the plasmid SCP1. so SCP1- carrying abs strains were obtained by crossing the abs isolates with 1650 (Table l) and picking NF abs recombi- nants (see below). Methylenomycin production by these strains was tested by cross-streaking them against the meth- ylenomycin-sensitive S. coelicolor 11501 SCPI‘. Methyle- nomycin activity was assessed at 3 days of growth at 28°C. since methylenomycin is most active against late mycelia (33). Assay conditions for the calcium-dependent antibiotic were adapted from Lakey et al. (21). Briefly. agar plugs of 2-day-old cultures of the test strains C542. 11501. and TK18. which were grown on Oxoid nutrient agar (ONA), were placed onto plates of ONA with and without added calcium. Soft ONA or ONA plus Ca was seeded with CDA-sensitive Staphylococcus aureus (21) and was overlaid around the plugs. After overnight refrigeration. the plates were incu- bated overnight at 37°C. For actinorhodin. undecylprodigiosin. and growth assays. ZOO-ml cultures of YEG. in l-liter bafiled flasks, were inoc- ulated to a density of 5 x 10‘ spores per ml with 11501 or an abs mutant strain and incubated with shaking at 30°C. At daily intervals 5-ml samples were taken. For growth mea- surements. mycelium from a S-mi sample was collected on Whatman paper and weighed. For actinorhodin measure- ment, NaOH was added to a 5-ml sample to achieve a pH of 12. After 1 h at room temperature. the culture was filtered through a 0.2-um-pore-size filter unit (Nalgene). Absorbance of the filtrate was measured over a range of wavelengths from 400 to 900 nm. From YEG-grown cultures. the absorb- ance maximum was at 580 nm (27. 35). After adjustment to pH 2 with HCi. the actinorhodin maximum shifted to a broad peak of 510 to 530 nm. as reported by Horinouchi and Beppu (16). For undecylprodigiosin. a 5-ml sample was first ex- tracted with NaOH to solubilize actinorhodin as above and then centrifuged. and the mycelial pellet was washed. The mycelium was then extracted with methanol (pH 2) over- night at room temperature. After filtration through a 0.2- um-pore-size unit. the absorbance maximum of the filtrate was measured over a range of wavelengths from 450 to 650 nm: the absorbance maximum was at 530 nm but shifted to 468 nm after the addition of NaOI-I to adjust the pH to 12, as 2964 ADAMIDIS ET AL. reported for undecylprodigiosin by Horinouchi and Beppu (16). Mutageneais and mutant isolation. Spores of 11501 were treated with N-methyl-N'-nitro-N-nitrosoguanidine (14) or irradiated with UV light (254 nm) to a survival rate of 0.1 to 1% (14) and then plated at about 400 colonies per plate on glucose minimal medium. Incubation was at 35°C for 4 to 5 days. Under these growth conditions. 11501 colonies sporu- lated and were red due to production of the two pigmented antibiotics. actinorhodin and undecylprodigiosin. Actinorho- din, which normally is blue and difiusible at alkaline pH. remained cell bound and red under these conditions. because 11501 produces relatively large amounts of acid when grown on glucose at 35°C (22). Thus. diffusing actinorhodin did not obscure unpigmented mutants. Genetic mapping techniques. Crosses and data analyses were carried out as described previously (3). Chromosomal recombination was mediated primarily by the plasmid SCP1 integrated at 9 o‘clock on the genetic map to give the NF fertility type (12). Several phenotypes are associated with the NF state. (i) NF strains are Aga“ because they fail to produce an agarase for which the gene is deleted when SCP1 integrates (ll). Aga* strains sink down into the agar; Aga“ strains do not. (ii) NF strains are normally methylenomycin producers and are methylenomycin resistant. since the pro- duction and resistance genes for this antibiotic are encoded by SCP1. Production of and resistance to methylenomycin were assayed as discussed above. In an NF x SCP1“ cross. close to 100% of the progeny will be NF (13). The frequency of other markers donated by the NF parent decreases in both directions from the insertion region at 9 o’clock. In this work we have exploited the observation (13) that in a cross between an NF strain and an SCP1“ strain such as a 11501 derivative. greater than 95% of the spore progeny will be his.“ strAl when plated nonse- lectively (our unpublished results). The biased recombina- tion of an NF x SCP1“ cross was exploited to obtain backcrossed recombinants in which the chromosomal region from 10 o‘clock clockwise to eight o‘clock originated from the SCP1“ parent and the region from 8 o’clock to 10 o'clock originated from the NF parent. For mapping the abs muta- tions. biased recombination was avoided in some crosses by using NF Abs“ strains derived from primary crosses in subsequent crosses with NF strains carrying standard mark- ers (34). NF Abs“ strains were picked based on their Aga“ phenotype. Construction ofa ¢C31 phage strain, P8106,eontainingthe 4st sequence. The plasmid AP22 carries the ast sequence on a 2.1-kilobase-pair (kb) SacI-PstI fragment (19) (Table 1). After digestion of AP22 by SacI and P311. the 2.1-kb frag- ment was purified from an agarose gel and ligated into $57] (an isoschizomer of SacI) and PstI-cut KC516 (29). The ligation mixture was used in a liposome-assisted transfection of S. lividans 1326 protoplast. Several plaques were purified. and phage minilysates were spotted onto a lawn of 11501 spores spread on RZYE: the 11501 lawn was replica plated onto RZYE medium containing thiostrepton (Sigma Chemi- cal Co.) at 50 ug/ml. Phage strains containing the ast sequence were detected by their ability to form Tsr’ lysogens on 11501: these phage strains were further analyzed by agarase gel electrophoresis to confirm they contained the 2.1-kb ast sequence. Methods for the assay. propagation. transfection. and in vitro DNA manipulation of KC516 were as described by Hopwood et al. (14). 53 J. Bxcrsator. RESULTS Isolation of Abs“ mutants. A mutant screen was carried out with the goal of obtaining mutants with a sporulation- competent. antibiotic-deficient phenotype. We have named this phenotype Abs“. for antibiotic synthesis deficient. To identify mutants with a global block in antibiotic synthesis. we used a visual screening procedure that took advantage of the fact that two of the four S. coelicolor antibiotics are pigmented. By requiring that a single mutation simulta- neously prevent the synthesis of both pigmented antibiotics. we thought we could identify “control" mutations rather than mutations that blocked either the actinorhodin or the undecylprodigiosin pathway. i.e.. act (30) or red (31) muta- tions. respectively. We expected that mutations causing a global block could be relatively rare. so we devised condi- tions that would allow us to screen very large numbers of colonies (see Materials and Methods). The colonies resulting from mutagenized spores were visually screened. and colonies that sporulated normally but that produced no pigment were isolated for further analysis. We isolated eight unpigmented mutants after examination of about 800.000 colonies; 700.000 of these were from UV mutageneses. Half of the unpigmented isolates proved to carry double mutations of the act and red loci (4). Calculated from the results of previous mutant screens. the predicted frequency for unpigmented double mutants would be ap- proximately 2 x 10“". since act and red mutants were reportedly obtained at frequencies of about 1 x 10“3 and 2 X 10 “3. respectively (30. 31). from UV mutageneses. Thus. the frequency at which we found double mutants. about 5 x 10““. is in accordance with results from earlier screens by other investigators. After the double mutants were elimi- nated from consideration. the frequency of the remaining unpigmented mutants was also about 5 x 10““. These unpigmented isolates were all obtained from UV mutagene- sis. They were named strains C505. C542. C554. and C577. In this mutant screen we also isolated an additional group of partially pigment-deficient mutants that will be described elsewhere. characterization of Abs“ mutants. If the unpig- mented isolates were the globally blocked mutants sought in this study. they might also be defective in production of the two unpigmented S. coelicolor antibiotics. methylenomycin and CDA. The Abs“ isolates were tested for production of these antibiotics in comparison with their Abs+ parent strain. 11501. Figure 1 shows the results of an assay for CDA. The parent strain produced an antics. aureus killing activity that was calcium dependent. In contrast. the Abs“ isolate pro- duced no detectable CDA activity. The abs mutation also blocked the synthesis of methyle- nomycin (Fig. 2). In this experiment we compared SCP1+ and SCP1“ derivatives of an Abs“ strain constructed as described in Materials and Methods. The SCP1“ parent strain was strongly inhibited by methylenomycin produced by its SCP1-carrying derivative (Fig. 2. intersection of streaks C and E). In contrast. the SCP1+ Abs“ strain produced no inhibitory activity (intersection of streaks A and E). When the SCP1-carrying Abs+ parent was streaked against itself. no inhibition occurred (intersection of streaks C and D). Interestingly. the SCP1+ Abs“ strain also ex- pressed methylenomycin resistance (intersection of streaks D and A). although it did not produce methylenomycin. All four Abs“ isolates exhibited similar behavior in both the CDA and methylenomycin assays. Both the methylenomy- VOL. 172. 1990 FIG. 1. CDA assay of Abs“ mutant strain. Growth conditions are described in Materials and Methods. Plug A is 11501 (Abs‘ parent). plug 1: is C542 (Abs ). plug C Is TK18 (act red). Plugs were taken from 2—day—old plates of ONA. With longer incubation on 0NA111501produces actinorhodin. which also kills S auteur: actinorhodin killing activity was apparent on plates without calcium after 3 days of incubation (data not shown). Therefore. straI rain TKIS. an act red double mutant (Table 1). was included to show that. under these assay conditions. we were observing the CDA activity and not actinorhodin actI ivity. cin and CDA assays were repeated many times with no detection of either antibiotic activity The e and mycelial accumulation of an Abs“ isolate were normal compared with those of the Abs parent. Figure 3A shows the results for mycelial accumula- tion. The Abs“ isolate C505 was normal in comparison with the parent strain when grown in YEG (the small difference at day 2 was not reproducibly observed in other experiments). Similar dry weight measurements were also made for cul- tures grown in liquid R5. YMPG. and glucose minimal media. and no growth defect in the Abs“ mutants was observed in any medium. Thus. the ab: mutation prevents the synthesis of antibiotics without significantly affecting growth rate or mycelial accumulation. Production of actinorhodin and undecylprodigiosin was quantitated as described in Materials and Methods. Figure 3B compares actinorhodin synthesis in YEG for the parent and an Abs“ isolated. C505. over a 9~day period. The mutant FIG. Methylenomycin assay of an Abs“ mutant strain. Strain construction and grwo wth conditions are desc nbed In Materials and Me thod od.s Streaks: C542 SCP1 (A) C542 (B). 11501 SCP1 (C and D). 11501 (E) 54 S. COELICOLOR ANTIBIOTIC SYNTHESIS MUTANTS 2965 FIG. 3. Time course of actinorhodin and undecylprodigiosin production. in liquid culture. by 11501 and the Abs“ mutant C505. Cultures of YEG were inoculated with spores to a cell density of 5 X10‘ per ml and were incubated at 30‘C withae on. At daily Ie.rvals 5. ‘ " nundecylpro- dligiosin. and dry weight determinations as described in Materials M.ethods Symbols. 0. 11501; A. iC505. (A) Dry weight mea- surements. (B) Actinorh rhodin produc TheA S”at pH 12 is indicated (C) Undecylprodigiosin production. Them A,” at pH 2 Is indie cate ted. C505 was completely deficient in actinorhodin production over the entire course of the fermentation. The failure to produce actinorhodin was also observed in cultures grown in liquid R5. YMPG. YEME. and liquid glucose minimal me- dia. Undecylprodigiosin was also assessed in the same YEG cultures; the C505 mutant produced no detectable antibiotic (Fig. 3C). in contrast to the parent. The other three Abs“ strains. C542. C554. and C577. were also completely defi- cient in both actinorhodin and undecylprodigiosin (data not shown). In summary. the Abs“ mutants were isolated on the basis of simultaneous loss of the ability to synthesize two pig- mented antibiotics and were shown to be severely defective in the synthesis of both of the other known antibiotics. Nevertheless. the Abs“ mutants sporulated normally on all plated media tested (see below). with aerial hyphae appear- ing and maturing to spore chains at the same time as in the parent strain. Figure 2 illustrates the sporulation capability of C542 (streaks A and B) in comparison with that of the parent (streaks C. D. and E). The Abs“ phenotype did not vary with any nutritional condition which we tested. Plate-grown cultures on R5. nutrient agar. or minimal medium with glucose or maltose failed to produce actinorhodin undecylprodigiosin. or CDA. Methylenomycin was not produced on R5 or nutrient agar. it could not be assayed on minimal medium. Genetic characterization of ab: mutants. In a preliminary cross. an Abs“ mutant strain. C542. was crossed with an antibiotic-producing strain. 1514. The recovery of only two colony phenotypes among the recombinant progeny—either 2966 ADAMIDIS ET AL. .0010 1 514 an‘ aura In" aw av 17 m 112 11a nan-ac as so an as “0.001 ”.5 FIG. 4. Mapping of abs-542. Strain C5421 was crossed with strain 1514. with selection as indicated by triangles: 35% of the hisA‘ srrAl recombinants were abs-542. Allele frequencies among the recombinants are indicated. and segregation of abs-542 with cysA and argA is tabulated. normal pigmentation or complete lack of pigmentation like the Abs“ parent—suggested the Abs“ phenotype was due to mutation at a single locus. The frequency at which unpig- mented colonies occurred among recombinants suggested a position for the abs-542 mutation near cysA (data not shown). For more definitive crosses. an NF (SCP1-con- taining) derivative of C542 was obtained from a cross with 1650 (see Materials and Methods). The Abs“ NF derivative, C5421, was then used in the cross shown in Fig. 4. The frequency at which the abs-542 allele was recovered among recombinants indicated that it was located either between the cysA and uraA loci or between the MM and argA loci. Segregation of abs-542 was independent with respect to the argA+ allele but not with respect to the cysA+ allele. Therefore the position counterclockwise to cysA was pre- ferred. In crosses comparable to that shown in Fig. 4. strains C505. C554. and C577 were also analyzed genetically and were shown to carry mutations that also mapped to a position near to and counterclockwise of cysA. An additional cross was performed to confirm the location of the abs mutations. This cross. which was a backcross of the abs mutant strain with the parent from which it was derived. was also used to replace most of the chromosome of the abs strain with unmutagenized chromosome from the parent strain. We were able to perform such a backcross because the position of the abs-542 mutation allowed us to take advantage of the unusual recombinant genotypes that arise after an NF X SCP1“ cross (see Materials and Meth- ods). This cross allowed us to move the abs mutation to an unmutagenized strain. For this cross, C5422. an abs-542-carrying recombinant from the cross shown above in Fig. 4. was crossed with 11501. The allele frequencies among recombinants indicated two possible positions for the abs-542 mutation. either clockwise of the SCP1 integration site and before cysA or counterclockwise before uraA (Fig. 5). Because segregation of abs-542 was independent with respect to the aunt ' allele but not with respect to the cysA15 allele. the position between cysA and the 9 o'clock region was preferred. Abs“ 55 1. BxcrI-zruou on" wars am“ an"! a." as 0 1a 41 ass-an 14 to so 14 pecans '80.! FIG. 5. Backcross of C5422 strain to 11501 parental strain. Recombinants were obtained from nonselective plating conditions (see Materials and Methods). Allele frequencies among the progeny are indicated. and segregation of abs-542 with cysA and mat is tabulated. cys+ pro+ hisAl arg+ srr'uraAl recombinants were obtained from this cross. These recombinants carried a chromosome originating from the 11501 parent for the entire region from cys clockwise to am. The region clockwise of am. including SCP1. originated from either 1650 or 1514. Thus these strains carried C542 DNA only from the region between cysA and the 9 o‘clock region. These recombinants had a phenotype identical to that of the C542 isolate. eliminating the possibil- ity that any mutations outside of this region contributed to the Abs“ phenotype. An assessment of whether all of the abs isolates carried mutations of the same locus was made by crossing the mutants against each other. For these crosses, NF uraAl pro‘ and NF um" proAl strains were obtained for each abs mutation. since the uraA and proA loci fiank all the abs mutations. All possible pairwise combinations of abs mu- tants were crossed. with selection for ura+ pro+ recombi- nants. In these crosses. pigmented colonies arose at frequen- cies of less than 0.05 to 0.3%. In no case did reciprocal crosses between a given pair of mutants (for example. abs-542 um+ proAl crossed with abs-554 uraAl pro+ and abs-542 uraAI pro+ crossed with abs-554 ura‘ proAl) yield significantly more recombination with one selection than with the other. As mentioned below. spontaneous pigmented apparent revertants arose readily in the abs mutants; we attribute the pigmented colonies obtained in these crosses to apparent reversion rather than recombination. Thus. we conclude that all of the abs mutations were closely linked. and we have named this locus absA. We compared absA with other loci to determine whether it was identical to three other previously described 5. coeli- color loci. The first of these was the bldA locus. since the map position of the abs/t locus was near bldA (26). As mentioned above. the phenotype of bldA mutants is different from that of absA mutants; although both are blocked in production of the four S. coelicolor antibiotics. bldA mutants are also blocked in formation of a sporulating aerial mycelium. Nevertheless. to rule out the possibility that the absA mutations were alleles of bldA that resulted in a different phenotype. they were tested for complementation by a bldA+ allele as described in Materials and Methods. VOL. 172. 1990 Phage strain KC603 (containing the bldA“ locus) failed to restore pigmentation after lysogen formation (data not shown). Therefore. the absA mutations were not recessive alleles of either bldA or of any gene neighboring bldA on the 5.6-kb cloned insert. As further confirmation of the noniden- tity of absA and bldA. the bldA strain 1668 was crossed with the absA strain C5423 (T able 1). With selection for strAl um“. 14% of the recombinants were pigmented. In a recip- rocal cross, with selection for sin“ his”. only 0.2% of the recombinants were pigmented. These results clearly indi- cated that the absA mutation was not an allele of bldA and positioned the absA locus counterclockwise of bldA. We also tested the afsB locus. since mutants of this locus had been reported to sporulate but to be pigment deficient (10) (see Discussion). The map position of the afsB locus had been reported to be at 5 o'clock. and we verified that the absA locus. which we had mapped to 10 o‘clock, was distinct from afsB by crossing an absA mutant strain. C5423. with an afsB mutant strain. BH5 (Table I). With selection for the uni hisA“ alleles. 41% of the recombinants were nor- mally pigmented. indicating that the dirt? and absA loci were clearly distinct. The third locus to test was qst. Since its genetic location had not been determined. it was possible that absA muta- tions were in the qst locus. To test this possibility. we had to determine the genetic origin of the cloned qst sequence. A 2.1-kb SacI-Psrl fragment from pI1702-AP22 (19) was subcloned into KC516 to form the phage strain PR106 (see Materials and Methods). This phage. which has its attach- ment site deleted, can form lysogens by Campbell-type integration at the chromosomal region that is homologous to the cloned ast insert. Lysogens of 11501 were selected as described in Materials and Methods. The thiostrepton resis- tance marker carried on the phage vector was then used as a mapping marker to locate ast. An NF strain (1668). lysog- enized with a ¢C31 strain (KC603) to prevent zygotic induction. was crossed with 11501::PR106 SCP1“ . (The bldA genotype of 1668 was irrelevant in this context.) With selection for srrAI his“. 85% of 480 recombinants were thiostrepton resistant. suggesting a location for Tsr' close to strA. Segregation was observed with the mat locus at 8 o‘clock (P = 0.001) but not with the mthB locus at 5 o'clock (P a 0.2). No segregation with Vior (carried by KC603, which was integrated at the bldA locus at 10 o'clock) was observed (P a 0.3). Because 63% of the ant! his“ progeny were uraA, the favored location for Tsr' was between 31M and uraA. at about 7 o'clock. These results demonstrated that ast and absA were difi‘erent loci. DISCUSSION The result of an extensive search for S. coelicolor muta- tions that uncouple sporulation from antibiotic synthesis and. in so doing. globally block antibiotic synthesis has been the discovery of the absA locus. Although they were isolated for failing to produce two antibiotics. absA mutant strains failed to produce any of the four known S. coelicolor antibiotics. strongly indicating that they globally blocked antibiotic synthesis. Nevertheless. they sporulated nor- mally. The map location of absA. at 10 o'clock, is distinct from the locations of at least three of the four affected antibiotics. undecylprodigiosin. actinorhodin. and methyle- nomycin. which are. respectively. at about 5. 6. and 9 o'clock (20. 30. 31). The data presented here did not locate absA with respect to the CDA locus. which also lies in the 56 S. COELICOLOR ANTIBIOTIC SYNTHESIS MUTANTS 2967 region between 9 and 11 o‘clock (15). Thus the possibility remains that absA may be close to or associated with the CDA locus. However. none of the mutants isolated for being CDA defective. with the exception of some bld mutants. was reported to be multiply defective for antibiotic synthesis (15). The location of the absA mutations. together with the observed structural dissimilarity of the affected antibiotics. indicates that absA mutations do not block a step in antibi- otic synthesis but rather penurb a regulatory element impor- tant in global regulation of antibiotic synthesis. The low frequency at which absA mutations were found is curious. One possibility is that the rarity of ab: mutations was due to the difficulty of visual detection. However. our detection of double acr red strains at about the same frequency as abs strains argues against this explanation. A second possibility is that the abs mutant phenotype is due to multiple mutations. The mapping data and the backcross data show that if multiple mutations are responsible. they must all be clustered in the region to which we have assigned the absA locus. Arguing against multiple mutations is the observation that all of the absA mutant strains spontane- ously acquired apparent revertants in which all four antibi- otics were normally produced (P. Riggle and W. Champness. unpublished data). If the absA strains are multiple mutants. all the mutants must be suppressed by a single mutation. A third possibility is that the abs mutations do not cause a loss of gene activity. since loss-of-function mutations for many other S. coelicolor loci occur at much higher frequencies. Perhaps the abs mutations result in an inhibitory activity or deregulate a repressor; such mutations would be likely to be relatively rare. The antibiotic-deficient phenotype of absA mutants seems to be more severe than that of other Streptomyces mutants with defects in antibiotic synthesis. Although afsB mutants. isolated for lacking A factor. also have a pigment-deficient. sporulation-competent phenotype (10). in our hands they are less pigment deficient than are absA mutants and produce some methylenomycin and CDA. The lack of phenotypic variation of the absA mutants under difl'erent growth condi- tions is especially noteworthy. since many Streptomyces developmental mutants of the bld class. in not only S. coelicolor but also S. gn'seus (1), are phenotypically affected by nutritional conditions. Two implications of the results of the mutant screen are noteworthy. F irst. in screening for simultaneous loss of two antibiotics. all mutations that tightly blocked both actinorho- din and undecylprodigiosin also blocked CDA and methyle- nomycin, suggesting that all antibiotics may be regulated by a common control mechanism. Second. all mutations that very tightly blocked all antibiotic synthesis and cleanly uncoupled sporulation from antibiotic synthesis mapped to the absA locus. This result suggests that absA may be the only locus that can mutate to a tight Abs“ phenotype. In this mutant screen. we did not recover any mutants that were candidates for ast mutants. based on the genetic location of ast. Determination of the null phenotype of ast may clarify whether ast can mutate to confer an Abs“ phenotype. The effect of absA mutations on the development of antibiotic resistance is interesting. Generally. the level of resistance to a given antibiotic increases at roughly the time of onset of production of the antibiotic (25). In many cases. the cluster of antibiotic biosynthetic genes includes at least one gene that encodes a resistance function (6. 8): in the case of streptomycin. Distler et al. have observed that the strep- tomycin resistance gene is cotranscribed with a putative 2968 ADAMIDIS ET AL. positive regulatory gene (8). We observed that absA mu- tants. although failing to produce methylenomycin. were as resistant as the wild type to methylenomycin (Fig. 2). This result suggests that resistance and production genes may not all be commonly regulated or that production genes have additional requirements for expression. The phenotypes of the bld and abs mutants invite specu- lation about the order of gene activity in the genetic pathway or pathways that mediate the transition from a vegetative to a differentiated colony. In the simplest model. abs activity could be placed downstream from bld activity. with abs involved more directly than are the bld genes in turning on the antibiotic operons. An independent pathway to sporula- tion, diverging from the common developmental pathway downstream of the bld activities. is suggested by the ability of sporulation to proceed independently of antibiotic produc- tion in the abs mutants. ACKNOWLEDGMENTS We thank Keith Chater. David Hopwood. and Sueharu Horinou- chi for providing strains used in this work. T.A. was supported by a George 1. Bouyoucos Graduate Fellow- ship. This work was supported by the National Science Foundation grant OMB-8811338 (to WC.) and by the Biotechnology Research Center at Michigan State University. LITERATURECITED 1. Baheaek, M., nd K. Kendrick. 1989. Cloning of DNA involved in sporulation of Streptomyces griseus. 1. Bacterial. 170:2802- 2808. 2. Bibb, M. 1., and D. A. Hopwood. 1981. Genetic studies of the fertility plasmid SCP2 and its SCP2' variants in Streptomyces coelicolor A3(2). 1. Gen. Microbiol. 126:427-442. 3. Champness. W. C. 1988. New loci required for Streptomyces coelicolor morphological and physiological differentiation. 1. Bacteriol. 170:1168—1174. 4. Chlnpnem. W. C., T. Adamidis. and P. Riggle. 1989. Develop- mental genes and antibiotic synthesis. p. 108-112. In C. L. Hershberger, S. W. Queener. and G. Hegeman (ed.). Genetics and molecular biology of industrial microorganisms. American Society for Microbiology. Washington. D.C. 5. Chater. K. 1984. Morphological and physiological difl'erentia- tion in Streptomyces. p. 89. In R. Losick and L. Shapiro (ed.). Microbial development. Cold Spring Harbor Laboratory. Cold Spring Harbor. NY. 6. Chater. K. F ., and C. 1. Brain. 1985. Resistance. regulatory and production genes for the antibiotic methylenomycin are clustered. EMBO J. 4:1893-189‘7. 7. Cheer. K. F., C. J. Baton. A. A. King. and 1. E. Suarez. I982. The expression of Streptomyces and Escherichia drug resiso tance determinants cloned into the Streptomyces phage 6C31. Gene 19:21-32. 8. Dhtier. 1.. E. Andrea. K. Marinas-I. K. Pbowotaki. M. Stock- mn. and W. Pipersherg. 1987. Gene cluster for streptomycin biosynthesis in Streptomyces griseus: nucleotide sequence of three genes and analysis of transcriptional activity. Nucleic Acids Res. 15:8041-8056. 9. Feitelson. J. S.. F. Mflpartida. and D. A. Hopwood. 1986. Genetic and biochemical characterization of the red gene cluster of Streptomyces coelicolor A3(2). 1. Gen. Microbiol. 1111:2431- 2441. 10. Hara. 0.. S. Horinouchi, T. Uozumi, and T. Beppu. 1983. Genetic analysis of A-factor synthesis in Streptomyces coeli- color A3(2) and Streptomyces griseus. 1. Gen. Microbiol. 129: 2939-2944. 11. Hodpon. D., Ill K. Chater. 1981. A chromosomal locus controlling extracellular agarase produced by S. coelicolor A3(2) and its inactivation by chromosomal integration of plas- mid SCPI. 1. Gen. Microbiol. 124:339—348. 12. Hopwood, D., III K. F. Chater. 1974. Streptomyces coelicolor. 57 13. 14. I5. 16. 17. 18. 19. 21. 24. 26. 27. 1. BACTERIOL. p. 237-255. In R. C. King (ed.). Handbook of genetics. vol. 1. Plenum Publishing Corp.. New York. Hopwood. D., R. Harold. A. Vivian. and H. Ferguson. 1969. A new kind of fertility variant in Streptomyces coelicolor. Genet- ics 62:461-477. Hopwood. D. A., M. J. Bibb. K. F. Chater. T. Keher, C. 1. Bruton. H. M. Kieser. D. J. Lydiate, C. P. Smith. J. M. Ward, and H. Schrempf. 1985. Genetic manipulation of Streptomyces. a laboratory manual. John Innes Foundation. Norwich. United Kingdom. Hopwood. D. A., and H. M. Wright. 1983. CDA is a new chromosomally determined antibiotic in Streptomyces coeli- color A3(2). 1. Gen. Microbiol. 129:3575—3579. Horinouchi. S.. and T. Beppu. 1984. Production in large quan- tities of actinorhodin and undecylprodigiosin induced by 0138 in Streptomyces lividans. Agric. Biol. Chem. 48:2131-2133. Horinouchi, S.. O. 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. 1. Bacteriol. 155:1238-1248. Horinouchi. S.. F. Malpartida. D. A. Hopwood. ad T. Beppu. 1989. afsB stimulates transcription of the actinorhodin biosyn- thetic pathway in Streptomyces coelicolor A3(2) and Strepto- myccs lividans. Mol. Gen. Genet. 215:355—357. liaisonehi. 8.. H. Suzuki. and T. Beppu. 1986. Nucleotide sequence of afsB. a pleiotropic gene involved in secondary metabolism in S (repromyces coelicolor A3(2) and Streptomyces lividans. J. Bacteriol. 168:257-269. . Kirby. R.. and D. A. Hopwood. 1977. Genetic determination of methylenomycin synthesis by the SCP1 plasmid of Streptomyo ces coelicolor A3(2). 1. Gen. Microbiol. 98:239—252. Lakey. 1. 11.. E. 1. A. Lea. B. A. M. Rudd, H. M. Wright. .d D. A. Hopwood. 1983. A new channel-forming antibiotic from Streptomyces coelicolor A3(2) which requires calcium for its activity. 1. Gen. Microbiol. [29:3565-3573. . Lin. C.-1., B. Pogell. and P. Redshaw. 1985. Regulation of "conditional" aerial mycelium mutants of streptomycetes. 1. Gen. Microbiol. 131:1015-1025. . Innovskaya. N. D., N. M. Mkrtumhn. N. L. WI”. and V. B. Danlienko. 1972. Characterization of temperate actino- phage 0C3] isolated from Streptomyces coelicolor A3(2). 1. Virol. 9:259—262. Mahartida, 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 (London) ”9:462-464. . Martin. 1. F., and A. L. Demain. 1980. Control of antibiotic biosynthesis. Microbiol. Rev. 44:230-251. Merrick. M. 1. 1976. A morpholoigcal and genetic mapping study of bald colony mutants of Streptomyces coelicolor. 1. Gen. Microbiol. 96:299—315. Outer. C. A., D. Stein. ad S. N. Cohen. 1988. Site-specific insertion of biologically functional adventitious genes into the Streptomyces lividans chromosome. 1. Bacteriol. 110:2174- 2184 . Piret.. r. M., and K. r. cum. 1985. Phage-mediated cloning of HIM. a region involved in Streptomyces coelicolor morpholog- ical development. and its analysis by genetic complementation. 1. Bacteriol. 163:965-972. . Rodicio. M. R.. C. 1. Briton. and K. F. Chater. 1985. New derivatives of the Streptomyces phage oC31 useful for the cloning and functional analysis of Streptomyces DNA. Gene 34:283-292. . Rudd. B. A. M., and D. A. Hopwood. 1979. Genetics of actinorhodin biosynthesis by Streptomyces coelicolor A3(2). 1. Gen. Microbiol. 114:35-43. . Rudd. B. A. M., and D. A. Hopwood. 1980. A pigmented mycelial antibiotic in Streptomyces coelicolor: control by a chromosomal gene cluster. 1. Gen. Microbiol. 119:333—340. . Stein. D., and S. Cohen. 1989. A cloned regulatory gene of Streptomyces liridans can suppress the pigment deficiency phenotype of diflerent developmental mutants. 1. Bacteriol. 171:2258—2261. 58 VOL. 172. 1990 S. C OELICOLOR ANTIBIOTIC SYNTHESIS MUTANTS 2969 33. Vivian, A. 1971. Genetic control of fertility in Streptomyces coelicolor A3(2): plasmid involvement in the interconversion of UF and IF strains. 1. Gen. Microbiol. 95:96-106. 34. Vivian. A., and D. Hopwood. 1970. Genetic control of fertility in Streptomyces coelicolor A3(2): the IF fertility type. 1. Gen. Microbial. 64:101—117. 35. 36. Wright. H. M., and D. A. Hopwood. 1976. Identification of the antibiotic determined by the SCP1 plasmid of Streptomyces coelicolor A3(2). 1. Gen. Microbial. 95:96—106. Wright. H. M., and D. A. Hopwood. 1976. Actinorhodin is a chromasomally-determined antibiotic in Streptomyces coeli- color A3(2). 1. Gen. Microbiol. 96:289—297. Chapter 3: Genetic analysis of absB, a Streptomyces coelicolor locus involved in global antibiotic regulation 59 JOURNAL or Bacnsaiowcv. July 1992, p. 46224628 0021-9193/92/144622-07502.00/0 Copyright 0 1992. American Society for Microbiology 60 Vol. 174. No. 14 Genetic Analysis of absB, a Streptomyces coelicolor Locus Involved in Global Antibiotic Regulation TRIFON ADAMIDISl AND WENDY CHAMPNESS‘J‘ Genetics Program1 and Department of Microbiology. 2 Michigan State University, East Laming. Michigan 48824-1101 Received 3 February 1992Accepted 6 May 1992 The filamentous soil bacterium W coelicolor is known to produce four antibiotics which are genetically and structurally distinct. An extensive search for antibiotic regulatory mutants led to the discovery of absB mutants, which are antibiotic deficient but sporulation proficient. Genetic analysis of the absB mutants has resulted in definition of the absB locus at 5 o’clock on the genetic map. Multiple cloned copies of the call-ORR! gene, an activator of synthesis of the antibiotic actinorhodin, restore actinorhodin biosynthetic capability to the absB mutants. These results are interpreted to mean that the failure of M mutants to produce antibiotics results from decreased expression of the antibiotic genes. The absB gene is proposed to be involved in global regulation of antibiotic synthesis. Bacteria of the order Actinomycetales are capable of producing a diverse array of medically useful secondary metabolites. The streptomycetes have been the major focus of antibiotic research, and the genetically best-characterized species. Streptomyces coelicolor, is known to produce four structurally and genetically distinct antibiotics: actinorho- din, undecylprodigiosin. methylenomycin, and calcium-de- pendent antibiotic. Genetic studies on S. coelicolor’s four antibiotics have progressed to the point where all of the biosynthetic gene clusters have been mutationally defined and mapped (18. 22, 32, 33). Three gene clusters have been cloned: act (26, 27), red (9. 28). and my (7). Regulatory studies have progressed farthest with the act and red genes. The 22-kbp act gene cluster encodes many open reading frames, expressed in at least six transcripts (10. 27). The actIl-ORF4 gene encodes a gene product required for transcription of the act biosyn- thetic genes (reviewed in reference 10). The red cluster (9, 28) is also large (35 kbp) and complex and also includes a gene. redD, which plays a positive role in expression of the red biosynthetic genes (30). In a plate-grown Streptomyces culture. growing hyphae initially form a dense. matted mycelium. Later, serially directed hyphae form on the colony surface and develop into chains of spores. Antibiotic synthesis is developmentally regulated and is generally found to coincide with sporulation (6). In liquid culture, most streptomycetes do not sporulate. but antibiotic production is delayed until the culture enters the stationary phase. Evidence that sporulation and production of the four S. coelicolor antibiotics are subject to a common genetic con- tral derives from the isolation of single mutations that block both processes—the bld mutations (5. 29, 34. 37). The best-characterized bld gene, bldA, encodes a Ieucyl-tRNA capable of translating the codon UUA (23). UUA has thus far been found primarily in genes involved in development. including several antibiotic resistance genes (reviewed in reference 24). a gene involved in Streptomyces griseus sporulation (2), and the actll-ORF4 gene (10). Therefore, ‘ Corresponding author. 4622 one aspect of actinorhodin regulation involves a translational requirement for the bldA tRNA (10). Single mutations that completely block production of the four S. coelicolor antibiotics but allow abundant sporulation (Abs' phenotype [1]) define the previously discovered absA locus (l). The Abs" phenotype suggests the existence of a global regulatory mechanism that is, at least in part, specific to antibiotic synthesis and distinct from sporulation control. Here, we report on the discovery and characterization of an additional locus. absB. which can also cause an Abs” phenotype by mutation. MATERIALS AND METHODS Bacterial strains and phages. The strains used for genetic analysis were derivatives of S. coelicolor A3(2) (Table 1). Streptomyces lividans 1326 was used for phage propagation (25). Procedures for phage propagation were as described by Hopwood et al. (15). Lysogen formation was accomplished as described previously (1. 5). Media and culture techniques. Minimal plate medium for genetic analysis. nutrient agar, and media R5 and R2 were as described by Hopwood et al. (15). YEG (30) contained 1% yeast extract and 1% glucose; SY (12) contained 0.3% yeast extract and 1% starch. R5. YEG, and SY were supplemented with histidine and uracil at 50 and 7.5 uyml, respectively. Thiastrepton (gift of S. 1. Lucania. E. R. Squibb and Sons. lnc., or Sigma Chemical Co.) was used at a concentration of 50 ug/ml. Antibiotic assays. The assays used for methylenomycin and calcium-dependent antibiotic were those described pre- viously (1). For actinorhodin and undecylprodigiosin quan- titations. the tested strains were streaked onto cellulose- acetate filters an R5 medium. After 5 to 6 days, the mycelia were scraped off and weighed. Approximately 20 mg was extracted with 0.5 ml of chloroform for 30 min at room temperature. with shaking. Then. 0.5 ml of l N NaOH was added. and the tubes were vortexcd and then spun in a microcentrifuge for 15 s. The aqueous phase contained actinorhodin, which is blue at alkaline pH. The A5... of the aqueous phase was determined. The chloroform phase con- tained the undecylprodigiosin, which was yellow. For absorbance measurements of undecylprodigiosin. the chlo- VOL. 174, 1992 61 S. COELICOLOR ANTIBIOTIC SYNTHESIS MUTANTS 4623 TABLE 1. Strains. phages, and plasmids used Strain. Pm" Relevant characteristics“ Source or reference or plasmid S. coelicolor M124 cyle8 tug/ll pmAI SCPl' SCPZ‘ D. Hopwood (15) 1650 cyle8 mthBZ agaAI NF SCPZ‘ K. Chater (29) 1514 praAl cysAIS my!) arm" nicA7agaAl NF SCPZ‘ K. Chater (15) 11501 W] uraAI srrAl SCPl‘ SCP2' Pgl' K. Chater (7, 15) BH5 cyle8 pro/ll org/ll afsB SCP2+ S. Horinouchi (14) BHSI‘ Same as BH5 but NF This work C120 his.“ «MA! strAI abs-120 SCPl" SCP2‘ Pgl' This work C1201” Same as C120 but NF This work C170 hiLsAl urnAI strAI abs-I70 SCPI‘ SCP2‘ Pgl" This work C1701’ Same as C170 but NF This work C175 hisAI urnAI strAl abs-I75 SCPI' SCPZ" Pgl’ This work C1751‘ pmAl cysAIS argAI abs-I75 NF This work C1752“ hisAl umAI strAI abs-I75 NF This work C246 his/i] uraAl strAI abs-246 SCPI‘ SCP2' Pgl‘ This work C252 his/ll urn/ll srrAI abs-252 SCPl‘ SCP2‘ Pgl’ This work C576 hisAI umAI strAI abs-5 76 SCPl' SCP2' Pgl’ This work C5761" his/ll umAI strAI abs-576 NF This work C5762” hrLsAI srrAI abs-5 76 NF This work C604‘ hisAI umAl strAI act red SCPI' SCPZ' Pgl’ This work es ¢C31 KC9m c+ AattP act! insert, Thio' Vio' K. Chater (4) ¢C31 KC902 c+ AattP red insert, Thio' Vio' K. Chater (12) Plasmids p11702 Thio' Mel‘ 21 pAT107 p11702 (21) and 2.7-kb Sphl actll-ORF4 insert. Thio' This work ‘ Abbreviations: SCP1. S. coelicolor plasmid 1 (35); SCP2, S. coelicolor plasmid 2 (3): NF. $01 is integrated into the chromosome at 9 o'clock (reviewed in reference 16): Pgl’. “31 sensitive (8); Vio'. viomyein resistant; Thio'. thiostrepton resistant: Mel’. melanin production. NF derivatives were obtained as deserted previously (1. 5). ‘ Recombinant from a cross with 1650. ‘ Recombinant from a cross with 1514. ‘ The unpigmented strain C604 was dcrnrmstrated to carry act and red mutations by genetic mapping (data not shown). roforrn layer was acidified with HCl. TheAuo of the now-red chloroform phase was then determined. For visual assessment of actinorhodin and undecylprodi- giosin production on plate medium, strains with disruptions of either the act or red genetic pathway were used to facilitate observation of each pigment. The act and red clusters were disrupted by insert-directed integration of the previously described phage KC900 (4) or KC902 (12), re- spectively. Both phages also carry the xylE gene (catechol 2,3-dioxygenase [20]), and transcriptional fusions to the act! or red promoter are formed in the integrants, but this feature of the phages is not relevant to their use in these experi- ments. Mutagenesis and mutant isolation. Mutagenesis and mutant isolation were carried out as described previously (1). Genetic mapping techniques. Crosses and data analysis were done as described previously (1. 5). Chromosomal recombination was mediated primarily by the plasmid SCPl, integrated at 9 o’clock on the genetic map to give the NF fertility type (17). For the crosses shown in Fig. 3B and C, NF absB derivatives were obtained from NF x SCPI‘ crosses; these were identified by their Dag‘ and methyle- nomycin-resistant phenotypes. as described previously (1). Recombinant DNA techniques. DNA isolations for plasmid and chromosomal DNA were done as described by Hap- wood et al. (15). Sphl-digested total DNA from M124 was ligated with SphI-digested, dephasphorylated pI1702 DNA and used to transform A120 protoplasts, with selection in thiostrepton (50 ugjml). Protoplast manipulations and trans- formations were dane as described by Hopwood et al. (15). RESULTS isolation and characterisation ofAbsB‘ mutants. A two- step screen for sporulation-proficient. antibiotic-deficient mutants was devised; well-sporulating colonies visibly lack- ing the antibiotic pigments actinorhodin and undecylprodi- giosin were isolated and then tested for loss of calcium- dependent antibiotic and methylenomycin. An initial screen for mutants lacking any detectable antibiotic led to the isolation of absA mutants (1). Double act red mutants such as C604 were also isolated (Table 1). In the course of that screen. mutants with a leaky Abs’ phenotype were also observed, but only one was isolated. Such mutants could define additional genes involved in antibiotic regulation, so a second mutant hunt was undertaken with the goal of isolat- ing mutants with a strong but not absolute Abs' phenotype. In this study. six AbsB’ mutants were isolated, C576 by UV mutagenesis. and the others (Table 1) by N—methyI-N'- nitro-N-nitrosoguanidine mutagenesis. The frequency was about 1 in 10.000 survivors of mutagenesis. The phenotypes of all six were very similar. This screen of approximately 120,000 colonies did not yield any additional abaA mutants. a result that was not surprising in light of the previously observed abs/i isolation frequency of l in 200.000 colonies. although that was obtained by UV mutagenesis (1). The mutant colonies sporulated as well as the parent strain on media such as R5 and glucose minimal medium, R2, SY. YEG, and maltose minimal medium. Table 2 shows that the representative mutant strain C120 was unable to produce actinorhodin or undecylprodigiosin on a variety of complex 62 4624 ADAMIDIS AN’D CHAMPNESS 1. BACI‘EIIIOL. TABLE 2. Observation of actinorhodin and undecylprodigiosin pigments in absB mutant strains‘ Pigment” in medium: Strain Lysogen Antibiotic observed MM +5, R2 K5 SY YEG MM+G MMeM low phosphate 11501 KC902 Actinorhodin + + + + + + + + + + + + + + KC900 Undecylprodigiosin + + + + + + z + + C120 KC902 Actinorhodin — z — — — — _ KC9m Undeq'lprodigiasin - z — - — _ - ‘ All media are described In Materials and Mthods; MM+G and MM+M are mtnimal medium with glucose or maltose. respectively. To facilitate observation of the actinorhodin and undecylprodigiosin pigments. strains with red undecy incubation at . +.pigment tnmedium; +. pigmentin mycelia:z or act gene disruptions. respectively. were used. Phage KC90 Iprodiposin pathway. and phage KC‘XX) inactivated the actinorhodin pathway (see Materials and Methods). Obscrv 30’C. 2 insetmted the ations were made after 6 days of + + +. extensive blue pigment in medium; .faint pigment in mycelia; - .no detectable pigment. Undecylprodigiosin remains cell-bound; it turns yellow at alkaline pH. ++ dark red mycelia: 4. red mycelia; z. taint pink mycelia; - .no detectable pigment and minimal media. Actinorhodin and undecylprodigiosin production was quantitated on K5 medium (Table 3). as described in Materials and Methods. The AbsB phenotype (i.e.. uncoupling of sporulation from antibiotic synthesis) ntwo conditions: (i) on thin R5 plates. the mutant sswell than 11501, and (ii) produced detectable but greatly reduced amounts of actinorhodin and undecylprodigiosin. Guthrie and Chater (12) have proposed the existence of a phosphate-sensitive, bldA-independent mechanism for acti- vation of undecylprodigiosin synthesis, since they had ob- served that a bldA mutant could produce undecylprodigiosin an low-phosphate (0.04 mM) glucose minimal medium (or R2, which lacks yeast extract). Accordingly. the Abs' strains were tested on these media. Undecylprodigiosin production in the 11501 parental strain responded to lower- ing of the phosphate concentration in glucose minimal me- um. However, the AbsB' mutant strain produced no more tmd prodigiosin on the low-phosphate medium than an other media (Table The level of the third known antibiotic, calcium-dependent antibiotic was also greatly reduced In the absB isolates, as shown for the representative strain C120 in Fig. 1. Because the methylenomycin production and resistance genes are carried on the SCPl plasmid, SCPI‘ derivatives of absB mutants were constructed as described in Materials and Methods. The SCP1" parent strain, 11501, was strongly inhibited by methylenomycin produced by its SCP1-carrying derivative (Fig. 2. intersection of A and B). However. the SCP1 abs-B mutant strain produced much less antibiotic (Fig. 2. intersection of A and D). Although the C1201 strain was impaired in methylenomycin production. it did exhibit TABLE 3. Measurements of actinorhodin and undecylprodigiosin production in absB mutant and actll-ORF4-stimulated strains" Strain A9... A“, 11501 0.46 1.1 C604 0.03 0.02 C576 0.03 0.09 C120 0.01 0.17 C120(pAT107) 2.68 0.03 'Values represent the average of duplicates. extracted from 20 mg of mycelia grown an R5 plates as described In Materials and Method. Absorb- Iihml' the maximum for undecylprodigiosin methylenomycin resistance comparable to that of 11501 NF (data not 5 own . Genetic analysis of AbsB mutants. An initial cross between an AbsB~ isolate, C576. and a mapping strain, 1650, is shown in Fig. 3A. The recombinant progeny sorted into only two phenotypic classes, Abs‘ and Abs‘. suggesting that a single mutant locus was responsible for the Abs' phenotype. The frequency at which Abs' progeny occurred suggested that the mutant locus was either In the umA-strA interval or in the mthB- hisA interval. The frequency of casegregatian of the putative absB mutant allele was higher with the nlthB‘ allele than with the uraA allele. Because the AbsB' mutant strain was SCPl', this type of cross with an SCPI’ (NF) strain resulted in complex allele gradients (36; reviewed in reference 16) among the progeny. Therefore, for more de- finitive mapping, SCPI‘ NF abs-B mutant derivatives were isolated from crosses with strains 1514 and 1650 (see Mate- rials and Methods). The absB NF strain C5761 was then crossed against 1650 (Fig. 38) to confirm absB segregation with the MM]? locus. In this cross, the gradient of allele frequencies and clear segregation of the absB allele with the mthB‘ allele ed us to assign a position for the absB mutant ocus at approximately 5 o c’lock, count terelockwise of mthB Segregation of the absB allele with the hisAI allele in this FIG. 1. Calcium-dependent antibiotic assay of AbsB' mutant strains. Plugs were cut out of 2—day-ald Oxoid nutrient agar plates nd placed onto plates of Oxoid nutrient agar with (right plate) and without (left plate) 12 mM Ca(NO,)3. Soft agar with and without calcium was seeded with S. aurora. Because actinorhodin also kills S. aureus. an act red double mutant. C604. was included to ensure that the assay conditions were testing calcium-dependent antibiotic activity. Plug A. C604 (Abs' act red); plug B. 11501 (Abs’ parent): plug C. C120 (AbsB'). VOL. 174. 1992 FIG. 2. Methylenomycin assays of AbsB‘ mutant strain. See Materials and Methods for strain construction and growth condi- tions. Streaks: A, 11501: B. 11501 NF; C. C120; D. C1201 NF. cross was also consistent with the assigned map location. In an additional cross. shownin nFig. 3C, an NF absB mutant strain, C5762, was crossed with the mapping strain 1514. The two alternative positions for abs-B, based on the allele frequencies. would be between WM and cysA or between .rrrA and argA. Segregation of abs-576 was independent with respect to the um’ and cys‘ alleles but not with respect to arg’. These results are all consistent with a location for absB at approximately 5 o’clock. The AbsB' mutant strains C120. C170, C175, C246. and C252 were all analyzed in crosses similar to those used for C576. For each strain, the genetic data indicated that a single mutant locus. with a location similar to that of abs-576. was responsible for the mutant as.phenotype An additional erossw aspcrfonned to locate the abs-B mutation abs-120 with respect to the cyle ocus, counter- clockwise of mthB. Strain C1201 (Table 1) was crossed with strain 1650. Selection for 262 his“ strAl recombinants yielded 235 mth’, 215 abs-120. and 187 eys+ recombinants. These data indicate that the relative order is mtIIB-absB- cysD, in the 5 o’clock map location. Many of the AbsB‘ mutant strains were also crossed against each other. in a series of pairwise crosses. For example. C1701 (NF abs-I70 uraAI hisAl srrAl) was crossed with C1751 (abs-I75 argAl cysAIS pmAl strAl). Two selective conditions were used to effect reciprocal selections: selection for um mg recombinants (cross I) and selection for is“ pro recombinants (cross II). Abs recombinants occurred at a frequency of 1% for cross I and 0.4% for cross 11. Similar crosses yielding similar results were performed with strains carrying the abs-120 and ab:— 576 mutations. The very low recombination frequencies between the abs mutations strongly suggested that they all afiected the same locus. Accordingly, we have named the locus absB. The position of the abs-B locus was similar to that of another, possibly related. locus. afsB (14). Mutants carrying mutations of the afsB locus were reported to sporulate but produce reduced amounts of the actinorhodin and undecyl- prodigiosin pigments (14). We compared an afsB mutant strain, BH5, wit th two absB mutant strains, C120 and C576. S. COELICOLOR ANTIBIOTIC SYNTHESIS MUTANTS 63 4625 BH5 and an SCPl‘ BH5 derivative, BH51 (Table 1). were tested for calcium-dependent antibiotic and methylenomycin production. respectively. in plate tests and were found to produce as much antibiotic as J 1501 or J 1501 NF (data not shown). Although BH5 was unpigmented on minimal me- dium with glucose, it produced abundant actinorhodin an R5; undecylprodigiosin production was not tested. In addi- tion. genetic crosses between BH5 and C1201, C5761. and C1752 (Table 1) were performed. With selection for um+ strAl recombinants, the progeny were 67, 42. and 45% Pgm’ (pigmented). respectively. These substantial recombination frequencies indicated that abs-B and afsB were different loci. Because of the plasmid status (Table 1) of the strains in these crosses. they yielded complex allele gradients and were not useful for ordering absB and afsB with respect to each other. Eflect of multiple copies of the regulatory gene null-0k}? on actinorhodin synthesis. In the course of experiments intended to clone the absB’ allele (unpublished data). we isolated a plasmid. pAT107, that stimulated copious aeti- norhodin but not undecylprodigiosin production in strain C120 (T able 3). The plasmid carried a 2.7-kbp Sphl insert. The same insert was found in 17 of approximately 20. 000 transformants and occurred In both orientations. Restriction mapping of the Sphl fragment with Fall and BamHI indi- cated that it correspone aportion of the actinorhodin gene cluster (13, 27) and included the actll-ORF4 promoter and open reading frame (10). The plasmid vector pU702 (21) replicates to a copy number of 30 to 100 copies per genome. Thus, the multiple copies of actIl-ORF4 present in C120(pAT107) were able to bypass the abs-B block to acti- norhodin synthesis. DISCUSSION An extensive search for S. coelicolor mutants that are sporulation proficient and antibiotic deficient has resulted in identification of a new abs locus at 5 o’clock, absB. Muta- tions at the absB locus cause a phenotype similar to the previously described absA mutant phenotype (1). Like absA mutations, which map to 10 o‘clock (1), the abs-B mutations affect production of the four known S. coelicolor antibiotics. These antibiotics, undecylprodigiosin. actinorhodin, methyl- enomycin, and calciu umdependent antibiotic, are produced gene products encoded In gene clusters located at 5, 6. 9, and 11 o'clock, respectively (15). The absA and absB mutants can be phenotypically distin- guished in two ways: absB mutants are somewhat leakier for antibiotic production on same complex media and do not accumulate spontaneous suppressive mutations, as do abaA mutants (31a). Evidence that the absB mutants fail to produce antibiotics because they are defective in antibiotic gene expression comes from the observation that multiple cloned copies of the acrll-ORF4 gene bypass the block to actinorhodin syn- thesis. Because acIIl-ORF4 is a regulatory gene for ac! expression (reviewed in reference 10), this result argues that the chili mutants are metabolically and biosynthetically competent to produce actinorhodin but fail to do so because the abs-B mutation prevents adequate expression of the acIII-ORF4 gene. A similar bypass effect of multicopy acrll-ORF4 in bldA. bldB. bldD, bldG, and bldH mutants has recently been described (31). The frequency at which abs-B mutations occurred is simi- lar to the frequency for occurrence of loss—of-funetion mu- tations in other S. coelicolor genes. In addition, the six mutant strains all showed very similar phenotypes. These J. BACTERIOL. ADAMIDIS AND CHAMPNESS 4626 .3: .5. v0.58 .0: 9.03 2:35:53. 93,—. 62.2.5.2 m_ we»... can .540 is: 5.3 03.36 .0 5:39:96 63:22. 803 $9 ”55:350.. :5». $3 .2 e380?» 5.? .32 5.3 03.88 ma? 3th 526 AU. .25. one. mace. Q90 2: c0393. $90 .22 a £85 m2. 5 3.3. .0: ma? EQQC gaze-=3. 682.58 a. 9.2: can VH5 .53 033.36 .0 ectawocwom 63:26 203 secs. 325.5832 :55. t E: .8 5:838 .EB 69: 5.3 commode as? 35.0 525 Am: 622.58 m. as... E:— VE: :33 cause no :33?ch tea 623:2: 2a «Eu—.5882 2: mecca «06:252. o_o=< 69336 293 2:23:53. ~33. . Mi 2: he «can “moi—.3... .3 “623:2: as 8:038 .EB 69.: £23 .23 panacea a; 2.90 5.2% 2. .9333: .c meanes— .... .0...— udnn wooévn vain pooévn 3°61. Sodvo 96...... a 8' as 3 .o o. .3-vo a .9 S. no :93- . co 9 an {n.3- n on nu u S 8 + can no «u 3 no + cos 8 on u on + no. :2: +2: :Eo +Ea 91.50 +30 «hie. +5.: .32 +3.. «.55 +5.: pa.2 Figure 1 76 using different markers and the abs-8752 allele was always mapped at around 10 o’clock in the S. coelecolor chromosome. The bldA gene also maps counterclockwise of the cysA locus; thus, it was interesting to see where the abs-8752 allele mapped in comparison to the bldA allele. The abs-87522 strain was crossed against J 668, which carried a bldA mutation. When selecting for strAI and ura+ alleles, 8% of the recombinants were pigmented; whereas when selecting for strAI and his" alleles, only 0.65% of the recombinants were pigmented. These results indicated that the abs8752 allele mapped counterclockwise of the bldA allele. The latter finding gave rise to the possibility that the abs-8752 mutation is another allele of the absA gene, since the absA gene maps in the same general location. The C8752 mutant might just be leakier than other absA mutants. Crosses of abs8752 against absA mutants were performed and wild type colonies were recovered at very low frequencies, always less than 1%. absA mutants are known to acquire spontaneous second site mutations, such that the double mutants show a wild type phenotype with respect to antibiotics (17). It was probable that these second site mutations contributed to the number of colonies that arose from the cross between abs-8752 and absA mutants. Recently several clones of the abs/l” allele were isolated (18). Upon transformation of C8752 protoplasts with the plasmid pPRlSl carrying an absA” allele, the C8752 transformants remained unpigmented; thus the absA * clone could not complement the abs-8752 mutation, suggesting that the abs-8752 mutation defines a new locus which maps close to the absA locus. Genetic characterization of stain C95. A cross between the C95 mutant and 77 Figure 2: Mapping of abs-95. Strain C95 was crossed with strain 1514, with selection as indicated by triangles. Allele frequencies among the recombinants are indicated, and segregation of abs-95 with proA and argA is tabulated. 78 Strain 095 Siraln 151+ 4. 0 PW" eroAl _9_°" ’ m1 abs“ 82 137 185 94 abs-95 9 0 4 S p<0.0002 p>03 79 the mapping strain 1514 was performed. The appearance of only two different phenotypes, namely Abs‘ and Abs”, among the recombinant progeny suggested the presence of only one mutation causing pigment deficiency. The frequency at which the abs-95 allele was recovered among the recombinants indicated that it was located either between the proA * and hisAI loci or between the hisA and argA+ loci. Since the abs95 allele segregates with the proA” allele but not with the argA+ allele, the position between proA * and hisAI was preferred (Figure 2). Crosses of C95 against absA542 were performed. The appearance of a substantial number of Abs“ progeny suggested that A95 does not map close to the absA locus but rather it maps clockwise of it. When plasmid pPR151 was transformed in C95 protoplasts, all the transformants remained Abs' confirming the result that abs-95 and absA define two different loci. Genetic characterization of stain C155. The C155 mutant was crossed against the abs” strain J650. Two phenotypes were recovered among the recombinants, suggesting that only one mutation dictates the production of Act, Red and CDA in the C155 mutant (data not shown). The strain C1551, isolated from the previous cross was crossed against 1514, an abs” strain. From the recombination frequency of the progeny, the abs-155 mutation appeared to map either between the cysA " and uraA+ loci or between the hisAI and argA‘ loci. Tabulation of the segregation of abs-155 mutation mapped it counterclockwise of the cysA * allele since it segregates more frequently with the cysA“ allele than with the argA+ allele (Fig.3). Upon transformation of C155 protoplasts with the plasmid pPR151, the C155 transformants remained unpigmented suggesting that the abs-155 mutation define a 80 Figure 3: Mapping of abs-155. Strain C1551 was crossed with strain 1514, with selection as indicated by triangles. Allele frequencies among the recombinants are indicated, and segregation of abs-155 with cysA and argA is tabulated. 81 SCP1 Strain 1514 a 2+ 1 nrg+ argAl abs * 2 43 23 22 abs-155 11 4 9 6 p<6.6061 p>a3 Figure 3 sug ch 82 ea aP eXl.‘ 303 82 locus other than absA. Strains C821, C108 and C584. All remaining mutants,namely C821, C108 and C584 when crossed against J650 showed only two phenotypes, Abs' and Abs+, suggesting the presence of a single mutation in both strains. Further genetic characterization remains to be done, in order to have a definite location of the abs- 821, abs-108 and abs-584 mutations on the S. coelicolor map. None of these mutants were able to produce antibiotics when transformed with the plasmid pPRlSl. DISCUSSION Six mutants of S. coelicolor have been isolated that are deficient in antibiotic production but competent for sporulation. There are two lines of evidence that the six mutants do not carry mutations in the biosynthetic genes of the four antibiotics. Act, Red, Mmy and CDA biosynthetic gene clusters are located in quite different loci in S. coelicolor chromosome, namely at 5, 6, 9 and 11 o’clock respectively (11,12,20,13). The parental strain J1501 does not contain the plasmid SCP1, so the biosynthetic genes for methylenomycin were not present during the mutagenesis and five out of the six mutants were not able to produce methylenomycin upon receiving the SCP1 plasmid. Genetic analysis indicated that only one mutation is present in each mutant, since in crosses with the wild type, only two different phenotypes appeared; namely, wild type and the respected mutant phenotype. absA (2) and absB (1) mutants have a similar phenotype. Thus the possibility existed that the six mutants carried either an absA or an absB mutation. Genetic analysis showed that all six mutations, namely abs-95, abs-155, abs-584, abs-8752, abs- 83 108 and abs-821, are located far from 5 o’clock where the absB allele was mapped (1). A plasmid containing the absA " allele was introduced into the six mutants but was not able to complement the mutation(s) they carry, ruling out the possibility of the existence of an absA mutation (only if these mutations are not dominant alleles over the absA+ allele). Since two of the mutations, abs-584 and abs-95, were mapped at the same locus at 11 o’clock, a possibility exists that abs-584 and abs-95 are alleles of the same gene. The same possibility exists for the mutations abs-8752 and abs-155 which were mapped at 10 o’clock, even though the C155 mutant was able to produce methylenomycin whereas C8752 was not. Further crosses between the mutants will clarify the situation. Another possibility exists that the six mutants carry multiple mutations in different regulatory genes. This possibility is unlikely since the mutations were found at a frequency of 5x104 whereas double mutations are usually found at a rarer frequency. This evidence is consistent with the fact that genetic analysis indicates that all six mutants carry one mutation each. Therefore, at least two different loci previously unidentified appear to play a regulatory role in antibiotic production. REFERENCES 1. Adamidis, T., and W. Champness. 1992. Genetic analysis of absB, a Streptomyces coelicolor locus involved in global antibiotic regulation. J. Bacteriol. 174: 4622-4628. 2. 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. 10. 11. 12. 13. 14. 84 Bibb, M. J ., and D. A. Hopwood. 1981. Genetic studies of the fertility plasmid SCP2 and its SCP2. variants in Streptomyces coelicolor A3(2). J. Gen. Microbiol. 126: 427-442. Champness. W. C. 1988. New loci required for Streptomyces coelicolor morphological and physiological differentiation. J. Bacteriol. 170: 1168-1174. Chater, K. 1984. Morphological and physiological differentiation in Streptomyces. In: R. Losick and L. Sharipo (eds.), Microbial development. Cold Spring Harbor laboratory, Cold Spring Harbor, N .Y., pp. 89. Chater, K. F., and C. J. Bruton. 1985. Resistance, regulatory and production genes for the antibiotic methylenomycin are clustered. EMBO J. 4: 1893-1897. Fernandez-Moreno, M. A., A. J. Martin-Triana, E. Martinez, J. Niemi, H. M. Kieser, D. A. Hopwood, and F. Malpartida. 1992. abaA, a new pleiotropic regulatory locus for antibiotic production in Streptomyces coelicolor. J. Bacteriol. 174: 2958-2967. Hara, 0., S. Horinouchi, T. Uozumi, and T. Beppu. 1983. Genetic analysis of A-factor synthesis in Streptomyces coelicolor A3(2) and Streptomyces griseus. J. Gen. Microbiol. 129: 2939-2944. Horinouchi, S., M. Kito, K. Furuya, S. K. Hang, K. Miyake, and T. Beppu. Primary structure of Ast, a global regulatory protein for secondary metabolite formation in Streptomyces coelicolor A(3). Gene 95: 49-56. Hopwood, D. A., M. J. Bibb, K. F. Chater, T. Keiser, C. J. Bruton, H. M. Keiser, D. J. Lydiate, C. P. Smith, J. M. Ward, and H. Schrempf. 1985. Genetic manipulation of Streptomyces, a laboratory manual. J ahn Innes Foundation, Norwich, United Kingdom. Hopwood, D. A., and H. M. Wright. 1983. CDA is a new chromasomally determined antibiotic in Streptomyces coelicolor A3(2). J. Gen. Microbiol. 129: 3575-3579. Kirby, R., and D. A. Hopwood. 1977. Genetic determination of methylenomycin synthesis by the SCP1 plasmid of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 98: 239-252. Lakey, J. H., E. J. A. Lea, B. A. M. Rudd, H. M. Wright, and D. A. Hopwood. 1983. A new channel-forming antibiotic from Streptomyces coelicolor A3(2) which requires calcium for its activity. J. Gen. Microbiol. 129: 3565-3573. Malpartida, F., and D. A. Hopwood. 1984. Molecular cloning of the whole 15. 16. 17. 18. 19. 20. 21. 85 biosynthetic pathway of a Streptomyces antibiotic and its expression in a heterologous host. Nature (London) 309: 462-464. Malpartida, F., J. Niemi, R. Navarrete, and D. A. Hopwood. 1990. Cloning and expression in a heterologous host of the complete set of genes for biosynthesis of the Streptomyces coelicolor antibiotic undecylprodigiosin. Gene 93: 91-99. Merrick, M. J. 1976. A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 96: 299-315. Riggle, P., and W. Champness. Unpublished results. Riggle, P., P. Brian, and W. Champness. Unpublished results. Rudd, B. A. M., and D. A. Hopwood. 1979. Genetics of actinorhodin biosynthesis by Streptomyces coelicolor A3(2). J. Gen. Microbiol. 114: 35-43. Rudd, B. A. M., and D. A. Hopwood. 1980. A pigmented mycelial antibiotic in Streptomyces coelicolor: control by a chromosomal gene cluster. J. Gen. Microbiol. 119: 333-340. Vivian, A. 1971. Genetic control of fertility in Streptomyces coelicolor A3(2): plasmid involvement in the interconversion of UP and IF strains. J. Gen. Microbiol. 69: 353-364. Chapter 5: Further studies on phenotypic and genetic charactirization of abs and bld mutants of Streptomyces coelicolor. 86 87 INTRODUCTION In Escherichia coli, amino acid limitation produces the "stringent response" whereby the RNA synthesis and numerous other cellular functions are severely reduced (9). Under amino acid deprivation, uncharged tRNAs accumulate and upon entering the A site of ribosomes, the relA gene product catalyzes the synthesis of ppGpp (guanosine 5’diphosphate-3’diphosphate) which is proposed to be responsible for the "stringent response" effect (5). Two classes of mutants that do not accumulate ppGpp and subsequently display a relaxed (Rel) phenotype have been isolated; relA and relC. In relC mutants the complete absence or alteration of the ribosomal protein L-11 prevents ppGpp synthesis (5). relC mutants from several Streptomyces species have been isolated, namely S. lavendulae the producer of formycin, S. antibioticus the producer of actinomycin, S. griseus the producer of streptomycin, S. griseoflavus the producer of bicozamycin and S. coelicolor (17,21). All Streptomyces relC mutants have an altered or absent ST- L11 protein (immunologically similar to L11 protein of E. coli) and are unable to produce the relevant antibiotics but can sporulate, albeit less abundantly. A contradiction exists for S. coelicolor relC mutants. A relC mutant isolated by Ochi, 1990b, showed an Act'Red‘ phenotype and only after 10 days incubation actinorhodin started to be produced. Sporulation was also late and aerial hyphae appeared after 5 days of growth. Interestingly the relC mutant isolated by Strauch et al.1991, is not different from the wild-type with respect to antibiotic production and sporulation. All Streptomyces isolated relC mutants (with the exception of S. antibioticus relC mutants) are erythromycin hypersensitive and cannot grow at elevated 88 temperatures at which wild-types normally are able to grow (17). S. coelicolor is naturally resistant to erythromycin at a concentration up to long/ml (18), while the relC mutants isolated by Ochi (18) cannot grow even in concentrations as low as 0.1ug/ ml. Furthermore, while wild type strains of S. coelicolor are able to grow at 40°C, the relC mutant was unable to do so (17). Tyrosinase is an enzyme that catalyzes the conversion of L-tyrosine to the black pigment melanin. Tyrosinase is encoded as apotyrosinase and in order to be fully activated requires the presence of two atoms of copper (12). Several Streptomyces produce melanin as a secondary metabolite and two melanin operons have been cloned, one from S. antibioticus (melC operon) (13) and one from S. glaucescens (10). Both operons contain the gene that encodes the apotyrosinase which is preceded by an open reading frame that has been proposed to act as a copper transferase to apotyrosinase. The exression of both genes is necessary for melanin formation (3,8). Both S. glaucescens and S. antibioticus exhibit tyrosinase activity intracellularly and extracellularly but when the me! operon of S. antibioticus was introduced to the melanin nonproducing strain S. lividans only intracellular accumulation of melanin formation (13). In this study, preliminary results of further phenotypic characterization of the bld and abs mutants are presented, namely the possibility of these mutants to be ppGpp deficient, their ability to produce melanin and their ability to produce antibiotics on different media. 89 Table 1: Strains, plasmids, and phages used in this study 4“ Strains, Relevant genotype' Reference plasmids phages Streptomyces coelicolor 11501 hisAI uraAI strAI SCPI‘ SCPZ' Pgl‘ Hopwood et al., 1985 I C301 his/11 uraAI strAI bldA SCPI‘ SCP2‘ Champness, 1988 Pgl‘ C112 hisAI uraAI strAI bldB SCP1” SCP2 Champness, 1988 Pgl“ 774 Merrick, 1976 I C536 hisAI uraAl strAI bldG SCPI' SCPZ‘ Champness, 1988 I Pgl' C109 hisAI uraAI strAI bldH SCPI' SCP2' Champness, 1988 Pgl‘ C249 hisAI uraAI strAI bld] SCP1 ‘ SCPZ' Pgl‘ Champness, 1988 C542 hisAI uraAI strAI absA SCPI' SCP2“ Adamidis et al., Pgl' 1990 C120 hisAI uraAI strAI absB SCP1 ' SCPZ' Adamidis and Pgl' Champness, 1992 C95 hisAI uraAI strAI abs-95 SCP1 ’ SCP2 Adamidis and Pgr Champness, unpublished results Phage phage ¢C31 and a bIdA“ insert Piret and ¢KC603 Chater,1985 Plasmid Mel‘Thio’ Katz et al., 1983 pl] 702 Plasmid ast+ allele in plasmid pIJ922 Adamidis et al., pWC5 unpublished results " Abbreviations: SCP1, Str. coelicolor plasmid 1 (23); SCP2, Str. coelicolor plasmid 2 (4); Pgl', ¢C31 sensitive; Mel+ melanin producer; Thio', thiostrepton resistance. 90 MATERIALS AND METHODS Bacterial strains and phages. All strains used in this study were derivatives of Streptmyces coelicolor A3(2) (Table 1). Steptomyces lividans 66 was used for phage propagation. oKC603 carries a cloned 5.6-kb insert which includes the bldA " allele (19). Media and culture techiques. R5 medium, a complex buffered medium and minimal medium was used for phenotypic analysis of bld and abs mutants (11). SY medium has been described by Tsao et al.,1985. R5 medium was also used for protoplast regeneration and phage infection. Recombinant DNA methods. Preparation of Streptomyces coelicolor protopasts, transformation,and phage infection have been described elsewhere (11). RESULTS Resistance of bld and abs mutants to erythromycin and their ability to grow at 40° C. Several bld and abs mutants were tested for their ability to grow at 40°C and in the presence of erythromycin. As shown in Table 2, all bld and abs mutants tested were able to grow at 40°C. At this temperature S. coelicolor wild-type strains cannot sporulate and consequently all the abs mutants remained asporogenous. The high temperature, though, did not alter the phenotype of the strains tested with respect to production of the two pigmented antibiotics. Similarly, all abs mutants grew normally in the presence of erythromycin. The results for bld mutants were more complex; while bldA, bldB and bld! mutants grew 91 Table 2: Resistance to erythromycin, growth at 40°C, and melanin production of Str. coelicolor bld and abs mutants. Strain Resistance to Growth at 40°C erythromycina production” 11501 (wt) + + + C301(bldA) + + - C112 (bldB) + + - 774 (bldD) - + + C536 (bldG) NT + + C109 (bldH) + /- + - C249 (bld!) + + - C542 (absA) + + + C120 (absB) + + + C95 (abs-95) + + NT ‘ Mutants were grown an R5 medium supplemented with 8ug/ml erythromycin. +, resistant; -, sensitive; + /-, not sure; NT, non tested ° Melanin was observed visually as black pigment +, melanin production; -, no melanin; NT; non tested 92 normally, bldD was unable to grow, while bldH grew in a smaller number of colonies than the expected number (number of colonies without selection for erythromycin). Melanin production by Str. coelicolor abs and bld mutants. The high copy number plasmid PIJ702, which carries the melC operon from S. antibioticus, was utilized to transform a melanin nonproducing strain of S. coelicolor. Streptomyces coelicolor transformed colonies produced copious amounts of melanin which was found not only intracellularly but extracellularly also. As shown in Table 2, absA and absB mutants are capable of producing melanin although blocked in antibiotic production. Only bldG and bldD mutants were found to produce melanin, while bldA, bldB, bldH and bld] do not produce any or produce extremely small quantities. Media dependence of antibiotic production of Str. coelicolor abs and bld mutants. The phenotype of bldA,B,G,H,I as well of absA,B and abs-95 was scored on different media, namely R5, ONA, YEG, YEMA and SY. In all media (except SY) the strains tested exhibited the relevant phenotype as if they were grown on glucose minimal medium e.g. all bld mutants did not produce the pigmented antibiotics and did not sporulate; abs mutants did not produce any pigmented antibiotic but sporulated normally (with exception of the ONA medium on which even the parental strain of abs mutants, J1501, does not sporulate). SY medium has been observed previously to cause strong production of undecylprodigiosin in S. coelicolor strains (22). Since proline has been shown to enhance undecylprodigiosin production, SY medium supplemented with proline was also used. The results that are summarized (Table 3) show that bldA and bid] 93 mutants were the only bld strains tested that did not produce any of the pigmented antibiotics on these media. Surprisingly, bldB, which shows the ”tightest” bld phenotype of any bld mutant on other media, was able to produce a vast amount of undecylprodigiosin on SY medium. Since actinorhodin production generally masks undecylprodigiosin production, the presence of undecylprodigiosin can be visually detected only if actinorhodin is not present. Both bldG and bldH mutants produced actinorhodin on SY medium but on some media supplemented with proline bldG mutants produced only undecylprodigiosin; perhaps the presence of proline enhanced the undecylprodigiosin production so much that the mutant was incapable of producing actinorhodin. The results were also surprising for the abs mutants. Since absB mutants are leakier in antibiotic production an R5 medium than absA mutants (1), one would expect production of antibiotics from absB rather than absA mutants. Interestingly, on SY medium absA mutants showed a small production of undecylprodigiosin while absB mutants did not produce any. abs-95 produced large amounts of actinorhodin confirming that abs-95 is not only genetically different than absA and absB mutants but is phenotypically different also. bld! is a double mutant containing a bldA mutation. Except for the fact that bldA mutants sporulate on maltose minimal medium while the bld] mutant strain does not, their phenotypes are strikingly similar. Neither produces melanin, they do not produce antibiotics on SY medium and in the presence of ast alleles are unable to produce actinorhodin while all the other bld mutants tested can (see Chapter 7). In order to elucidate further the difference between bldA and bldI mutants, it was 94 Table 3: Antibiotic production is medium depedent in Str. coelicolor abs and bld mutants _— |{ Strain SY SY + proline C301 (bldA) Act' Red’ Act’ Red' C112 (bldB) Act” Red+ Act’ Red” C536 (bldG) Act+ Act“ Red“ c109 (bldH) Act: + Act" + + C249 (bld!) Act‘ Red‘ Act‘ Red‘ C542 (absA) Act’ Red” Act‘ Red”: C120 (absB) Act‘ Red‘ Act' Red‘ C95 (abs-95) Act“ Act” Mutants were grown on SY medium with or without proline. Act, actinorhodin production; Red, undecylprodigiosin production. + + +, overproduction; + +, wild type level production; + or + /-, very small production. 95 necessary to see if the bld] mutant carries a bldA mutation. The phage ¢KC603 which carries a bldA + allele was used to infect bldA and bld] strains (C301 and C249 respectively) and their phenotypes were scored for production of the two pigmented antibiotics and calcium dependent antibiotic. In addition, ¢>KC603 was used to infect bldA and bldI mutants carrying the plasmid pWC5 which carries an ast allele; the ast allele is known to enhance the production of the two pigmented antibiotics. As shown in Table 4, strain C249 (bld!) produced actinorhodin in the presence of a bldA+ allele with or without the presence of the aer allele (lines 6, 8). These results suggest that the C249 mutant is a double mutant carrying a bldA mutation and the block in actinorhodin production in the C249 strain was due merely to the presence of the bldA mutation. Sequence analysis of the bldA gene in C249 has confirmed the presence of a bldA mutation (14a). Nevertheless, C249 is a bld mutant since it is unable to sporulate even when the bldA” allele was present (data not shown). As shown in line 6, C249 produced actinorhodin but did not produce calcium dependent antibiotic even a bldA“ allele was present; this phenotype has never been observed before among the bld mutants isolated, suggesting that C249 carries a mutation in a previously unidentified allele along with a bldA mutation. DISCUSSION relC mutants of Streptomyces carry a mutation that causes alteration or absence of the ST-Lll ribosomal protein which subsequently causes failure of these mutants to accumulate ppGpp, a compound proposed to be responsible for Table 4: 96 C249 (bld!) carries a bldA mutation Strain Actinorhodini Calcium dependent 1 production“ antibiotic production C301 (bldA) - - C249(bld1) - - l C301 + pWC5b - - C249 + pWC5 - - I l C301 + ¢KC603° + + + I C249 + ¢KC603 + - I C301 + ¢KC603 + + + + + pWC5 I C249 + ¢KC603 + + + + nt l pWC5 ' Actinorhodin and calcium dependent antibiotic (CDA) were assayed as described previously (2); For actinorhodin: + + + , overproduction; + +, production at wild type levels; +, smallproduction; F or CDA: + , production at wild type levels; ° Plasmid pWC5 carries the entire ast gene; ° Phage ¢KC603 carries the bldA * gene; it was introduced into the mutants by infection. -, no production. -, no production; at, non tested 97 stringent" response". Interestingly, relC mutants isolated from Bacillus subtilis (20) and Streptomyces species were defective in antibiotic production and while B. subtilis relC mutants cannot sporulate, Streptomyces relC mutants are capable of doing so even though not as well as wild type strains (17). These results suggest a possible role for the "stringent response" in determining the onset of secondary metabolism. The RelC phenotype resembles the Abs phenotype and thus it is possible that the abs mutants carry a relC mutation. To test this hypothesis, we utilized the observation that all relC mutants isolated either from B. subtilis~ or from Streptomyces spp are hypersensitive in erythromycin and cannot grow in elevated temperatures (17). abs mutants were able to grow at high temperature and in the presence of erythromycin; and the same is true for some of the bld mutants. So preliminary results showed that the abs and some of the bld mutants tested do not carry a relC mutation. Another explanation of these results could be that the sensitivity to erythromycin and the failure to grow at elevated temperatures of the relC mutants are coincidental due to the way that these mutants were isolated (see Chapter 1) and are irrelevant to the RelC phenotype. Further experiments that show directly the absence or alteration of the ST-Lll protein by two dimensionsal PAGE of total ribosomal protein isolated from abs and bld mutants, are remaining to be done. Another piece of evidence that supports the hypothesis that abs mutants do not carry a relC mutation comes from the observation that abs mutants are able to produce melanin while relC mutants from S. lavendulae (16), a melanin producer, fail to do so. It is interesting that bldA mutants cannot produce melanin. The bldA gene is known to encode a tRNA that recognizes the UUA codon for leucine. The UUA 98 codons are extremely rare in S. coelicolor mRNAs which are high GC content organisms) and furthermore 'ITA codons were found only in genes involved in development (14b). Surprisingly, the melC operon does not contain a ’ITA codon in its translated sequences, indicating that the bldA gene product does not act directly on the translation of the relC genes but rather another factor which depends on the bldA gene regulates melanin formation. bldB and bldH mutants also do not produce melanin and it will be interesting to see if these mutants along with the bldA mutant are blocked at the transcriptional level in melanin formation. The pattern of the melanin production in different developmental mutants suggests an hierarchy: bldA, bldB, bldH < bldG,bldD < absA, absB. bldA,B,H genes are involved in secondary metabolism generally and in sporulation, while bldG and bldD genes are more specific and blocked only in antibiotic production and in sporulation. absA and absB mutants are even more specific and blocked only in antibiotic production. The results on SY medium suggest an alternative pathway for antibiotic production since bldA and absB were the only mutants tested, that were not able to produce either of the two pigmented antibiotics. Surprisingly, even absA mutants produce small quantities of undecylprodigiosin. Recent results show that absA mutants produce actinorhodim on low phosphate medium (7). Since SY medium contains phosphate at least three pathways for undecylprodigiosin production should exist. Finally, the similarity of phenotypes of bldA and bld] mutants led us to further 99 investigate and the possibillity that the bldI mutant represented by strain C249 carries a bldA mutation in addition to a previously undefined bld mutation. 10. REFERENCES Adamidis, T., and W. Champness. 1992. Genetic analysis of absB, a Streptomyces coelicolor locus involved in global antibiotic regulation. J. Bacteriol. 174: 4622-4628. Adamidis, T., P. Riggle, and W. Champness. 1990. Mutations in a new Streptomyces coelicolor locus which globally block antibiotic synthesis but not sporulation. J. Bacteriol. 172: 2962-2969. Bernan, V., D. Filpula, W. Herber, M. Bibb and E. Katz. 1985. The nucleotide sequence of the tyrosinase gene from Streptomyces antibioticus and characterization of the gene product. Gene 37: 101-110. Bibb, M. J ., and D. A. Hopwood. 1981. Genetic studies of the fertility plasmid SCP2 and its SCPZ' variants in Streptomyces coelicolor A3(2). J. Gen. Microbiol. 126: 427-442. Cashel, M., and K. E. Rudd. 1987. The stringent response. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology. p1410-1438. American Society for Microbiology, Washington, D.C. Champness, W. C. 1988. New loci required for Streptomyces coelicolor morphological and physiological differentiation. J. Bacteriol. 170: 1168-1174. Champness, W. Unpublished results. Crameri, R., G. Hintermann, R. Hutter and T. Kieser. 1984. Tyrosinase activity in Streptomyces glaucescens is controlled by three chromosomal loci. Can. J. Microbiol. 30: 1058-1067. Gallant, J. A. 1979. Stringent control in E. coli. Annu. Rev. Genet. 13: 393- 415. Hintermann, G., M. Zatchej and R. Hutter. 1985. Cloning and expression of the genetically unstable tyrosinase structural gene from Streptomyces glaucescens. Mol. Gen. Genet. 200: 422-432. 11. 12. 13. 14a. 14b. 15. 16. 17. 18. 19. 20. 21. 22. 100 Hopwood, D. A., M. J. Bibb, K. F. Chater, T. Keiser, C. J. Bruton, H. M. Keiser, D. J. Lydiate, C. P. Smith, J. M. Ward, and H. Schrempf. 1985. Genetic manipulation of Streptomyces, a laboratory manual. John Innes Foundation, Norwich, United Kingdom. Huber, M., G. Hintermann and K. Lerch. 1985. Primary structure of tyrosinase from Streptomyces glaucescens. Biochem. 24: 6038-6044. Katz, E., C. J. Thompson and D. A. Hopwood. 1983. Cloning and expression of the tyrosinase gene from Streptomyces antibioticus in Streptomyces lividans. J. Gen. Microbiol. 129: 2703-2714. Leskiw, B. K. Unpublished results. Leskiw, B. K., E. J. Lawlor, J. M. Fernandez-Abalos, and K. F. Chater. 1991b. 'ITA codans in some genes prevent their expression in a class of developmental, antibiotic-negative, Streptomyces mutants. Proc. Natl. Acad Sci. USA 88: 2461-2465. Merrick, M. J. 1976. A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 96: 299-315. Ochi, K. 1986. Occurrence of the stringent response in Streptomyces sp. and its significance for the initiation of morphological and physiological differentiation. J. Gen. Microbiol. 132: 2621-2631. Ochi, K. 1990a. Streptomyces relC mutants with an altered ribosomal protein ST-Lll and genetic analysis of a Streptomyces griseus relC mutant. J. Bacteriol. 172: 4008-4016. Ochi, K. 1990b. A relaxed (rel) mutant of Streptomyces coelicolor A3(2) with a missing ribosomal protein lacks the ability to accumulate ppGpp, A-factor and prodigiosin. J. Gen. Microbiol. 136: 2405-2412. Piret, J. M., and K. F. Chater. 1985. Phage-mediated cloning of bldA, a region involved in Streptomyces coelicolor morphological development and its analysis by genetic complementation. J. Bacteriol. 163: 965-972. Smith, I., P. Paress, and S. Pestka. 1978. Thiostrepton-resistant mutants exhibit relaxed synthesis of RNA. Proc. Natl. Acad. Sci. USA. 75: 5993-5997. Strauch, E., E. Takano, H. A. Baylis and M. J. Bibb. 1991. The stringent response in Streptomyces coelicolor A3(2). Mol. Microbiol. 5: 289-298. Tsao, S.-W., B. A. M. Rudd, X.-G. He, C.-J. Chang and H. G. Floss. 1985. 23. 101 Identification of a red pigment from Streptomyces coelicolor A(3)2 as a mixture of prodigiosin derivatives. J. Antibiot. 38: 128-131. Vivian, A. 1971. Genetic control of fertility in Streptomyces coelicolor A3(2): plasmid involvement in the interconversion of UF and IF strains. J. Gen. Microbiol. 69: 353-364. Chapter 6: Cloning of the putative absB" allele and other DNA sequences that "bypass" the absB mutation. 102 103 INTRODUCTION absA and absB mutants of S. coelicolor do not produce any of S. coelicolor’s four antibiotics (1,2). Thus a putative absB" clone would be expected to restore production of all four antibiotic. In this report, we present the cloning of two putative absB" clones as well as clones that are able to "bypass" the absB mutation. Several DNA sequences have been identified by their ability to enhance production of two antibiotics in S. coelicolor. Genes, which are capable of inducing actinorhodin (Act) and undecylprodigiosin (Red), are: abaA (9), ast (12), afsQ (14), and asaA (23). Thus none of these genes are capable of globally regulating all four antibiotics. MATERIALS AND METHODS Bacterial strains. All S. coelicolor strains used in this report were derivatives of S. coelicolor A3 (2) (Table 1). Streptomyces lividans 1326 is a wild-type strain (11). Media and culture techniques. Minimal plate medium, nutrient agar and medium R5 were as described by Hopwood et al. Thiostrepton was used at a concentration of 50ug/ n11. Chloramphenicol was used at a concentration of 15ug/ ml. Antibiotic assays. The assays used for methylenomycin and calcium dependent antibiotic were those described previously (1). Quantitation of actinorhodin and undecylprodigiosin was performed as described by Adamidis and Champness, 1992. Table 1: Strains and plasmids used 104 Strain Relevant genotypeII Reference Str. coelicolor I —=- M124 ng1 pmAI cysDIS SCPl' SCP2 D. Hopwood et al,1985 J1501 hisAI uraAl strAI SCPl' SCP2“ D. Hopwood et al,1985 J650 cysD18 mthBZ agaAI NF SCPZ" M. J. Merrick, 1976 C542 hisAI uraAI strAI absA542 Adamidis, T. et a1, SCPl' SCPZ‘ 1990 C120 hisAl uraAI strAI absB]20 Adamidis, T. and W. SCPl' SCPZ' Champness, 1992 C1201” Same as C120 but NF Adamidis, T. and W. Champness, 1992 C170 hisAI uraAI strAI absBI70 Adamidis, T. and W. SCPl' SCPZ' Champness, 1992 C175 hisAI uraAI strAI absB] 75 Adamidis, T. and W. SCPl' SCPZ' Champness, 1992 C301 hisAI uraAI strAI bldA301 W. C. Champness, SCPl' SCP2’ 1988 C112 hisAI uraAI strAI bldBIIZ W. C. Champness, SCPl' SCP2' 1988 C536 hisAI uraAI strAI bldGS36 W. C. Champness, SCPl' SCPZ' 1988 C109 hisAI uraAl strAI bldH109 W. C. Champness, SCPl' SCP2' 1988 C181 hisAI uraAI strAI bldH181 W. C. Champness, SCPl' SCP2‘ 1988 C604 hisAI uraAI strAI act red SCPl' Adamidis, T. and W. SCPZ' Champness, 1992 105 Table 1 (cont’d) Plasmids Relevant characteristics" Reference pIJ702 Tsrr Mel” Katz, E. et a1, 1983 pIJ922 Tsrr Ltz’" Lydiate, D. J. et a1, 1985 pTA140 pIJ702 and 3.5-kb SphI ast insert This work pTA120 pIJ 702 and 2.7-kb SphI actII-orf4 This work insert pTA107 pIJ 702 and 2.7-kb SphI actII-orf4 This work insert pTA265 pIJ702 and 2.1-kb BglII insert This work pTA276 pIJ702 and and 5.8-kb BglII insert This work pTA310 pIJ702 and 3.5-kb Bng insert This work pTA364 pIJ702 and 2.5-kb BglII insert This work pTA201 pU702 and unknown size Bng This work insert pTA274 pIJ702 and 5.8-kb BglII insert This work pTA227 pIJ922 and 11.5-kb BglII whiE This work insert pTA963 pIJ922 and 20-kb Sau3A redD This work insert pTA404 pIJ922 and unknown size Sau3A This work insert pTA405 pIJ 922 and unknown size Sau3A This work insert pTA410 pIJ 922 and unknown size Sau3A This work insert pTA942 pIJ922 and unknown size Sau3A This work insert pTA108 pIJ 922 and 12-kb Sau3A absB: This work insert pTA128 pIJ922 and 13-kb Sau3A absB: This work insert 106 ' Abbreviations: SCPl, Str. coelicolor plasmid 1; SCP2, Str. coelicolor plasmid 2; NF, SCPl intergrated into the chromosome at 9 o’clock; Tsr‘, thiostrepton resistant; Mel”, melanin production; Ltz*, conjugative functions. ° Recombinant from a cross with J 650 (2). Recombinant DNA techniques. DNA isolations for plasmid and chromosomal DNA were done as described by Hopwood et al., 1985. SphI or Bng digested total DNA from M124 was ligated with SphI or Bng digested, dephasphorylated pIJ702 DNA and used to transform C120 protoplasts with selection in thiostrepton. Protoplast manipulations and transformations were done as described by Hopwood et al., 1985. For the low copy number library, partially Sau3A digested total DNA from J 1501 was ligated to BamHI digested pIJ922 DNA and used to transform protoplasts of M124 (21). PIJ922 is a self transmissible plasmid and exists in 1-2 copies per genome (16). Thiostrepton resistant transformants were replicated onto lawns of C120 (absB) mutant an R5 medium. After sporulation, the mating plates were replicated onto media selective for absB recipient colonies which had acquired the thiostrepton resistance marker from pIJ922. This medium was glucose minimal medium supplemented with uracil, histidine and thiostrepton to support growth of C120 transconjugants and streptomycin to kill the M124 strain. Absence of proline, cystine and arginine was also used to select against M124. The transconjugants were tested for complementation of the Abs' mutant phenotype by observing Act and Red and testing for CDA and Mmy. Southern blotting was done as described by Sambrook et al., 1989. Hybridization experiments were done with nonradioactive probes using the Genius Kit ( Boehringer 107 and Mannheim ) according to manufacturer instructions. RESULTS Cloning genes that bypass the absB mutation in high copy number. Two libraries were constructed using as a vector the high copy number plasmid pIJ702. The libraries were introduced into absB protoplasts and the transformants were screened visually for pigment production indicating the presence of at least one antibiotic. The pigmented colonies were isolated and they were further tested for the production of the calcium dependent antibiotic (CDA). Mm SphI restriction enzyme was used to construct the first library due to its ability to restrict the ast (12) and redD (19) genes; thus, we thought that we would avoid cloning of these genes which we knew could "bypass" the absB mutation. Twenty thousand transformants, carrying clones which represented more than 95% of the S. coelicolor wild-type genome, were screened. Forty-five pigmented colonies were isolated and were further tested for CDA production. None of the transformants were able to produce CDA, indicating that the absB‘ allele was apparently not represented in any of these clones. Nevertheless plasmid DNA was isolated from all these transformants and was retransformed into absB protoplasts. Only twenty-five clones were still able to stimulate pigment production, suggesting that the presence of pigment in the other twenty transformants was not due to the plasmids they carried. Streptomyces coelicolor is known to undergo spontaneous deletions and amplifications of its chromosome and these events are often accompanied both by loss of the natural resistance to the antibiotic chloramphenicol 108 and by overproduction of undecylprodigiosin; this phenotype is called "scarlet" (4,13). All of the forty-five original transformants were tested for chloramphenicol resistance and the twenty that were unable to induce pigment production were sensitive to chloramphenicol. This result suggests that these transformants had undergone chromosome deletions and/or amplifications. The twenty-five clones that induced pigment production in absB mutants, were divided in two classes according to their size and their restriction map. The first class contained eight with an insert size of 3.5 Kb. All of them were able to induce actinorhodin and undecylprodigiosin production in absB mutants. Since the ast was known to enhance the production of both pigmented antibiotics, we compared the published restriction map of the ast gene (12) with the restriction map of one of the isolated clones, namely pTA140. The comparison revealed that pTA140 carried part of the ast gene which encodes the carboxyl-terminus of the Ast protein. Further hybridization studies on these clones are analyzed in Chapter 7. All the remaining seventeen clones constituted the second class of pigment inducing clones. All of them had an insert size of about 2.6 Kb and were able to induce only actinorhodin production in absB mutants. The restriction map of two of these clones, namely pTA120 and pTA107, was compared with the published restriction map of the actII-orf4 gene (8), the positive activator of actinorhodin biosynthetic gene. Hybridization studies have shown that both pTA120 and pTA107 carry the actII-orf4 gene (Figure 1). Plasmid pTA107 was transformed into different developmental mutants, including bldA, bldB, bldG, bldH and absA mutants and it was capable of inducing actinorhodin production in all of them. Similar results for 109 Figure 1: Restriction map of a fragment of the act region and its relationship to the pTA107 clone. pTA107 carries the entire actIl-orf4 gene. The restriction sites of actII-orfli, actII-orf4 and actIII were described previously (10,8). The arrows at the top of the diagram indicate the direction of transcription for the corresponding act genes; for act!!! this was deduced from Hallam et al., 1988, and for actII-orfi and actII-orf4 from the results of Fernandez-Moreno et al., 1991. Sp, Sphl; 8, Sad; P, Pstl; B, BamHI; X, Xhol; A, ApaI. 110 on 4:10an m-- ao 4:10.13 111 the Bld mutants were obtained by Passantino et al.,1991. BgLILlibrary Twelve thousand absB transformants were visually screened and sixty-four pigmented colonies were isolated. All sixty-four transformants were tested for their ability to induce calcium dependent antibiotic production but none of them was able to do so. Only-thirty six of the transformants were able to grow in the presence of chloramphenicol, suggesting that the remaining twenty-eight transformants had the "scarlet" phenotype. Thirty-five clones were further characterized by retransforming them into absB protoplasts and by transforming them into the bldA and absA mutants. Nineteen of these were not able to induce pigment production either in absA or bldA transformants, while sixteen of them were able to bypass only the absA mutation. Six of the transformants that overproduce actinorhodin or produce it at wild-type levels were further studied. The results are summarized in Table 2. All six of them were tested for the possibility that these clones represent the ast or the actII-orf4 gene by hybridization experiments. None of the clones (except pTA201 which was found to be difficult to isolate) hybridized either with the ast or with the actII-orf4 clones. The inserts in pTA265, pTA276, pTA310 and pTA364 were subcloned to the low copy number plasmid pIJ 922 and were transformed into absB protoplasts. None of them was able to "bypass" the absB mutation, confirming our speculation that these plasmids did not carry the absB+ allele. Since all clones had a different size and a different restriction pattern (data not shown), we conclude that they represent different DNA sequences that are able to "bypass" the absB mutation when they are present in high copy number. 112 Table 2: Characterization of clones that "bypass" the absB mutation . actmorhodin‘I production in : ‘, Actinorhodin was assayed as described by Adamidis and Champness, 1992. + + + , actinorhodin over production; + + , actinorhodin produced at wild type levels; + or + /-, a small stimulation of actinorhodin production. b, pAT201 was found to be difficult to isolate, thus the size of the clone remains unknown. pTA227 was isolated from the Bng library by its ability to induce green pigment in both absA and absB mutants but not in bldA mutants. Plasmid pTA227 was found to carry a 11.5 Kb DNA fragment which was digested by the Bng restriction enzyme, creating a 3.0 Kb and a 8.5 Kb fragment. The presence of a Bng site suggested that the 11.5 Kb fragment resulted either from a partial digestion of the wild-type chromosome or the two fragments derived from different parts of the chromosome and were ligated together prior to ligation to the plasmid. The green pigment was found to turn blue under alkaline conditions, as actinorhodin does, 113 suggesting a relationship between the green pigment and actinorhodin. Restriction analysis of the clone revealed that the 8.5 Kb fragment had an identical restriction pattern with the whiE gene cluster which is responsible for spare pigmentation (7) (Figure 2). Even though the mature spores of S. coelicolor appear grey, it is possible that the overexpression of the whiE gene cluster resulted in the green color of the absB and absA transformants. Interestingly, even though the expression of the whiE gene cluster is temporally (late in growth) and spatially (only in spores) controlled (7), plasmid pTA227 induced green pigment production early in development in the substrate mycelia of absA and absB mutants. Early green pigment production was difficult to be detected in wild-type strain because the presence of the antibiotics masked the production of the green pigment. This result suggests that in plasmid pTA227, the whiE gene cluster is expressed by another promoter provided either by the 3.0 Kb fragment or by the plasmid. In order prove that actinorhodin was not induced by pTA227, hybridization experiments were performed and revealed that the 3.0 Kb fragment does not carry the actII-orf4 gene. The simplest explanation for these results is that since the whiE gene cluster product is a polyketide, as actinorhodin is, (7) the green pigment is the one that becomes blue at high pH. An alternative explanation is that the 3.0 Kb fragment is responsible for the production of another pigmented compound e.g. an antibiotic precursor. It is interesting that the actII-orf4 and redD genes were not represented in the BglII library. afirR is already known to be nonclonable at high copy number (12). The most probable explanation is that the BglII fragments carrying these genes are bigger than 10 Kb and it is known that plasmid pIJ702 becomes unstable when it 114 Figure 2: Clone pTA227 carries the entire whiE gene cluster. Partial restriction map of the whiE gene cluster was adopted from Davis and Chater, 1990. Only the restriction sites used in this work are shown. The arrows above the restriction map show the locations of open reading frames of whiE biosynthetic genes. Clone pTA227 carries the entire whiE gene cluster in a 8.5 kb BglII DNA fragment plus an uncharacterized 3.0 kb BglII fragment. Bg, BglII; s, SphI; P, Pstl; B, BamHI. 115 iam no.8 089.0... Umpmmw .3 g T _. m? E “'1 (Sea +1.7 ”'01 pro h‘fl 116 carries clones that big (15). Cloning genes that partially complement absB mutants for antibiotic production. The absB mutants fail to produce all four antibiotics, thus the putative absB“ clone was expected to be able to induce all four of them when introduced into absB mutants. Since none of the clones isolated with either of the high copy number libraries were able to cause calcium dependent antibiotic production in absB mutants, a low copy number library was used as described in Materials and Methods. One hundred pigmented absB transconjugants were isolated from the low copy number library. The presence of the pigment was an indication of antibiotic production so all these colonies were further tested for calcium dependant antibiotic production. Forty of the colonies were able to produce CDA. Plasmid isolation and retransformation of the absB mutants with these forty clones revealed that none of them were able to induce CDA production and actually only five of them were still able to induce pigmentation in absB mutants; pTA404, pTA405, pTA410, pTA942 and pTA963. It is interesting that the majority of the colonies were false-positive; we assume that integration of the plasmid via homologous recombination into the chromosome gave a high frequency donor state. Thus these colonies were recombinants derived from the cross of the absB mutants against the absB" strain M124 (see also Materials and Methods) Plasmids pTA404, pTA405 and pTA942 were able to cause overproduction of both actinorhodin and undecylprodigiosin in absB mutants. Furthermore pAT942 was introduced into bldA, bldB, bldG, bldH and absA mutants and showed the ability to induce pigment production in all of them. The phenotype of absB mutants 117 carrying these plasmids was indistiguisable from the phenotype of the same mutants harboring plasmid pWC5 which carries the ast” allele (see also Chapter 7) suggesting that all three plasmids carry the ast” allele. Hybridization experiments using as a probe the ajErR DNA sequence will prove if this hypothesis is true. Alternatively if these plasmids do not contain the ast gene, they carry novel regulatory DNA sequences and further characterization is required to elucidate their function. Plasmid pTA410 also caused actinorhodin and undecylprodigiosin overproduction in absB mutants, but the antibiotic appeared later in growth, not until the 5th day of development. Transformation of pTA410 into bldA, bldB, and bldH also caused production of the antibiotics but the phenotypes of the mutants transformed with pTA410 were quite different than the phenotypes of the same mutants carrying a cloned ast+ allele suggesting that pTA410 does not carry an ast+ allele. Another possibility is that pTA410 carries a truncated ast allele since it’s known that parts of the gene are able to cause overproduction of Act and Red, in lesser amounts, th‘ough, than the whole ast gene. Finally, plasmid pTA963 was isolated for its ability to induce red pigment accumulation in absB mutants. Since the red pigment became yellow-orange when the pH of the medium was increased, we concluded that the red pigment was the antibiotic undecylprodigiosin. The redD gene is known to encode a positively acting regulator of the red biosynthetic cluster (19). Hybridization experiments using the red gene as a probe showed that pTA963 actually carries the redD gene and at least one biosynthetic gene, redC (Figure 3). Plasmid pTA963 was transformed into 118 Figure 3: Clone pTA963 carries the redD and redC genes. Restriction sites of the chromosomal red sequences carrying the redD and redC genes were described by Malpartida et al., 1990. The solid line of the pTA963 clone indicates the common DNA sequences with the chromosomal red DNA sequences as shown by hybridization experiments. The dotted line indicates cloned DNA sequences, yet uncharacterized while the dashed line indicates chromosomal DNA sequences non relevant to this work. X, Xhol; Bg, BglII; B, BamHI; E, EcoRI. 119 l"III —cn 3.6mm .................... ~_6 120 Table 3: Production of actinorhordin and undecylprodigiosin production in redD stimulated strains" plasmid content none pTA963 relevant mutant J1501 none + + + + + + + + C604 act red - - . - C542 absA - - .. + + C120 absB - - - + + + C301 bldA - - - + + + C 1 12 bldB - - - + C536 bldG - - - + + 4- C109 bldH - - - - C181 bldH - - - + + + ", The presence of the antibiotics were detected as previously described (Adamidis and Champness, 1992); b, Complete genotypes are in Table 1; C , + +, wild type levels of production; + + +, over production; +, production of small amount; -, no production 121 Table 4: Undecylprodigiosin production by bldA mutant harboring the ast and redD clones. Red production in minimal mediab _ bldA plasmid content" glucose maltose arabinose strain C301 - - - - I C301 pTA140(ajEsR¢) - - + C301 pTA963(redD*) + + + + + + of the Ast protein. pTA963 carries the redD+ allele. + +, over production; +, production at wild type levels; -. no production. pTA140 carries part of the ast gene which encodes for the carboxyterminus The presence of undecylprodigiosin was inspected visually,after 5 days of growth. 122 several developmental mutants and, as shown in Table 3, all of them (except, strain C109) were able to overproduce undecylprodigiosin. Strain C109 which carries a bldH mutation probably carries a second mutation in the red gene cluster since cloned ast alleles were found also to be incapable of inducing Red production in this strain (3). bldA mutants have been shown not to be able to produce undecylprodigiosin on minimal glucose media (18) even when extra copies of the ast gene are present (3). As shown in Table 4, bldA mutants harboring the plasmid pTA963 are able to overproduce undecylprodigiosin in all minimal media tested with alternative carbon sources. Thus, just an extra copy of the red gene is capable of relieving the glucose suppression of Red biosynthesis. Another possibility is that the pTA963 clone contains extra DNA sequences which are responsible for these results. Subcloning of the redD gene in a smaller fragment will elucidate further these results. Cloning of the absB' allele. Since none of the clones isolated from the low copy number library was able to induce CDA production, the same library was reintroduced into the absB mutant. This time one hundred seventy-two pigmented colonies were isolated but only twenty-three of these were able to produce CDA. Upon isolation and retransformation of these twenty-three clones into absB mutants, only two clones, namely pTA108 and pTA128, were able to induce CDA production in absB mutants. Strain C120 (absB mutant), harboring either of these two plasmids showed a wild-type phenotype with respect to actinorhodin, undecylprodigiosin and calcium dependent antibiotic production. C120 does not contain the SCP1 plasmid that carries the methylenomycin biosynthetic and resistance genes. The strain C1201 was isolated from the progeny 123 Table 5: Act, Red and CDA production in different developmental mutants in the presence of either the plasmids pTA108 or pTA128 I pigment production” CDA (Act/Red) I C542( absA ) - - I C301( bldA ) - - I C112( bldB) - - C109( 11de ) - - C120( absB ) Act" /Red” + C170( absB ) Act” /Red” + c175( absB ) Act+ /Red* + 11501 Act" /Red* .- Str. lividans Red+ nt : Act, Red, CDA were detected as described by Adamidis and Champness, 1992; : +, wild type levels of the relevant antibiotic; -, no relevant antibiotic is present; nt, non tested 124 of the cross of C120 against wild type strain J650 and has been shown to carry the SCP1 plasmid (2). After transformation of pTA108 into C1201 protoplasts, the mutant was capable of producing methylenomycin. Thus plasmid pTA108 was capable of inducing all four antibiotics in an absB mutant. To test the ability of both plasmids to induce antibiotic production in other developmental mutants we transformed absA, bldA, bldB, bldH mutants as well as a S. lividans strain a close relative of S. coelicolor. S. lividans produces only undecylprodigiosin even though it contains all the biosynthetic and resistance genes for actinorhodin. As shown in Table 5, none of the transformed mutants produce Act, Red or CDA. Furthermore the parental strain of absB mutants, J1501, and S. lividans transformed by either pAT108 or pAT128 did not show any change in their phenotypes. All regulatory genes cloned at this time now; abaA (9), ast (12), afsQ (14) and asaA (23) were capable of activating actinorhodin biosynthesis in S. lividans; thus it is interesting that both plasmids pTA108 and pTA128 were not able to confer an Act" phenotype in S. lividans. Both plasmids were transformed into other absB mutants, namely C170 and C175, and all transformants showed CDA and Act production even though actinorhodin appeared later in development and in smaller amounts than the wild type. Thus, both plasmids were capable of inducing all four antibiotics in absB mutants but were not capable of doing so to other developmental mutants, suggesting that they carry an absB" allele. Since we have not yet obtained genetic proof that the clones carry the absB" allele, we designated these clones as absB’. A restriction map of the clone pAT108 was established (Figure 4). Southern 125 Figure 4: Restriction map of pTA108 clone. The solid line of pTA128 indicates the DNA fragment that shares with pTA108, the dashed lines indicate unknown DNA sequences. The arrows indicate where pIJ922 DNA sequences start. E,EcoRl; B, BglII; P, PstI. 126 "-1! 1 B .BNM 127 blotting and hybridization experiments showed that both pTA108 and pTA128 share the two PstI internal fragments, thus pTA128 and pTA108 carry common DNA sequences. DISCUSSION Two clones, pTA108 and pTA128, have been isolated from the low copy number library, that can restore the production of all four antibiotics in absB mutants. The two clones share a common homologous region and we designated them absB', pending proof that these" clones correspond to the absB locus. It is possible that the absB' clones simply bypass the absB mutation. All known regulatory genes which control antibiotic production in S. coelicolor can bypass one or more than one classes of developmental mutants, but none of them have the ability to activate all four antibiotics in any mutant tested. In contrast, the absB' clones can activate all four antibiotics in absB mutants and cannot activate any antibiotic production in any developmental mutant tested except absB. Thus, even if absB' clones do not correspond to absB gene, the clones are still very interesting, since they would represent a novel class of regulatory genes capable of globally controlling the antibiotic production in a single class of developmental mutants. Proof that absB' clones are complementing and are not suppressing clones will come from two experiments. First we can predict that the absB' clones should hybridize to a specific band of the physical map, since the location of the absB mutation on the chromosome of S. coelicolor is known (2). In the second experiment, marker rescue of absB alleles by the absB' clones will be performed. This is 128 conveniently accomplished using the phage ¢KC516. Phage ¢KC5 16 carrying the absB' clone will be infected into an absB mutant and will be lysogenized via recombination between the cloned segment and the homologous chromosomal region, since the phage attachment site has been deleted. Phage released from these lysogens will be a mixture of absB" and absB‘ only if the cloned insert was absB". The released phage will be used to inject C120. Phage which carry the absB‘ allele form Abs' lysogens. It is interesting that absB" gene was not found among the clones derived from the high copy number libraries. Many possible hypothesis may explain these results; the restriction enzymes we used to construct the libraries may restrict the absB gene or the absB gene (or a gene in this locus) is lethal in high copy number as is the case for ast and afsQ genes (12,14) or the absB" gene in high copy number does not display the expected phenotype. The redD and actlI-orf4 genes were isolated by their ability to activate undecylprodigiosin and actinorhodin, respectively, in absB mutants. Both genes were found to be capable of activating the relevant antibiotic in every developmental mutant tested, thus all these mutants are metabolically able to produce these antibiotics but fail to do so because their regulatory program has been disturbed. It is interesting that these genes, even in the presence of only one extra copy per cell, are able to stimulate antibiotic production in the tested developmental mutants; suggesting that the regulatory system which governs antibiotic production is delicately balanced. Cloning their promoters in front of a promoterless reporter gene like xylE (24) or the luciferase gene will enable us to monitor their temporal and spatial 129 expression. Plasmid pTA227 was found to carry the whiE gene cluster which is responsible for the color of the spores of S. coelicolor (7). The product of this gene cluster did not show any killing activity against Staphylococcus aureus since both absA and absB mutants harboring pTA227 were not able to kill Staphylococcus aureus. Even though the whiE gene cluster has been cloned (7), its product was unable to be purified since it is found to be strongly associated with the spores (6). Our clone has the ability to express the whiE cluster even in the vegetative mycelia, thus it will be easier to isolate the spare pigment and test for a possible antibiotic activity. Four clones have been isolated from the low copy number library able to bypass the absB mutation with respect to actinorhodin and undecylprodigiosin production. Hybridization experiments with known regulatory genes will identify the ones, if any, that are novel. Several clones have been isolated for being able to bypass the absB mutation in high copy number. Five of these clones do not represent the ast or the actII-orf4 gene. It is possible that they represent other known regulatory genes like abaA (9) or asaA (23). Another explanation for the abundance of these genes is that they do not represent genes that specifically regulate antibiotic production but can activate antibiotic genes by "cross-talk" when present in high copy number. Determination of the null phenotype of these genes will establish if any are essential for antibiotic synthesis or if they are accessory genes with stimulatory capabilities. 10. 11. 130 REFERENCES Adamidis, T., P. Riggle, and W. Champness. 1990. Mutations in a new Streptomyces coelicolor locus which globally block antibiotic synthesis but not sporulation. J. Bacteriol. 172: 2962-2969. Adamidis, T., and W. Champness. 1992. Genetic analysis of absB, a Streptomyces coelicolor locus involved in global antibiotic regulation. J. Bacteriol. 174: 4622-4628. Adamidis, T, and W. Champness. Unpublished results. Betzler, M., P. Dyson, and H. Schrempf. 1987. Relationship of an unstable ArgG gene to a 5.7-kilobase amplifiable DNA sequence in Streptomyces lividans 66. J. Bacteriol. 169: 4804-4810. Champness, W. C. 1988. New loci required for Streptomyces coelicolor morphological and physiological differentiation. J. Bacteriol. 170: 1168-1174. Chater, K. F. Unpublished results. Davis, K. N., and K. F. Chater. 1990. Spore colour in Streptomyces coelicolor A3(2) involves the developmentally regulated synthesis of a compound biosynthetically related to polycetide antibiotics. Molecular Microbiol. 4: 1679- 1691. Femandee-Moreno, M. A., J. L. Caballero, D. A. Hopwood, and F. Malpartida. 1991. The act cluster contains regulatory and antibiotic export genes, direct targets for translational control by the bldA transfer RNA gene of Streptomyces. Cell 66: 769-780. Fernandez-Moreno, M. A., A. J. Martin-Triana, E. Martinez, J. Niemi, H. M. Kieser, D. A. Hopwood, and F. Malpartida. 1992. abaA, a new pleiotropic regulatory locus for antibiotic production in Streptomyces coelicolor. J. Bacteriol. 174: 2958-2967. Hallam, E. S., F. Malpartida, and D. Hopwwod. 1988. Nucleotide sequence, transcription and deduced function of a gene involved in polyketide antibiotic synthesis in Streptomyces coelicolor. Gene 74: 305-320. 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 manipulation of Streptomyces: a laboratory manual. The John Innes Foundation, Norwich, United Kingdom. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 131 Horinouchi, S., M. Kita, K. Furuya, S. K. Hang, K. Miyake, and T. Beppu. 1990. Primary structure of Ast, a global regulatory protein for secondary metabolite formation in Streptomyces coelicolor A3(2). Gene 95: 49-56. Hutter, R., and T. Eckhardt. 1988. Genetic manipulation. In: Goodfellow M. Williams ST, Mordarski, M. (eds). Actinomycetes in biotechnology, Academic Press, London, pp 89-184. Ishizuka, H., S. Horinouchi, H. M. Kieser, D. A. Hopwood, and T. Beppu. 1992. A putative two-component regulatory system involved in secondary metabolism in Streptomyces spp. J. Bacteriol. 174: 7585-7594. Katz, E., C. Thompson, and D. Hopwood. 1983. Cloning and expression of the tyrosinase gene from Streptomyces antibioticus in Streptomyces lividans. J. Gen. Microbiol. 129: 2703-2714. 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. Malpartida, F., J. Niemi, R. Navarrete, and D. A. Hopwood. 1990. Cloning and expression in a heterologous host of the complete set of genes for biosynthesis of the Streptomyces coelicolor antibiotic undecylprodigiosin. Gene 93: 91-99. Merrick, M. J. 1976. A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 96: 299-315. 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. Passantino, R., A. M. Puglia, and K. Chater. 1991. Additional copies of the act” regulatory gene induce actinorhodin production in pleiotropic bld mutants of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 137: 2059-2064. Riggle, P., and W. Champness. Unpublished results. Sambrook, J., E. F. Fritsch, T. Maniatis. 1989. Molecular Cloning: a laboratory manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Strahl. Unpublished results. 24. 132 Zukawski, M. M., D. F. Gaffney, D. Speck, M. Kauffinann, A. Findeli, A. Wisewp, and J. Lecocq. 1993. Chromogenic identification of genetic regulatory signals in Bacillus subtilis based on expression of a cloned Pseudomonas gene. Proc. Natl. Acad. Sci. USA 80: 1101-1105. Chapter 7: Suppression of antibiotic synthesis deficiencies in Streptomycm coelicolor developmental mutants by cloned ast alleles 133 134 Suppression of antibiotic synthesis deficiencies in Streptomyces coelicolor developmental mutants by cloned afiR alleles Trifon Adamidis,1 Perry Riggle,2 and Wendy Champnessl'z' Genetics Program", Department of Microbiology, and NSF Center for Microbial Ecology,2 Michigan State University, East Lansing, Michigan 48824 'Corresponding author: Dr. Wendy Champness 369 Giltner Hall Department of Microbiology Michigan State University East Lansing, MI 48824-1101 Telephone: 517-353-9770 Fax: 517-353-8957 (submitted for publication) 135 Abstract Cloned alleles of the Streptomyces coelicolor ast locus were isolated on the basis of their ability to restore production of the antibiotic pigments actinorhodin and undecylprodigiosin to an antibiotic-nonproducing absB mutant strain. These cloned alleles also restored antibiotic production to other S. coelicolor mutant strains affected in antibiotic regulation including absA and several blds. Although the abs and bld mutations globally prevent production of all four of S. coelicolor’s known antibiotics, the cloned ajEsR alleles apparently affect regulation of only two of the antibiotics. 136 The spore-forming, filamentous bacterium Streptomyces coelicolor is known to produce four antibiotics as secondary metabolites. These are actinorhodin, undecylprodigiosin, calcium-dependent antibiotic (CDA) and methylenomycin A. Because they are pigments, actinorhodin and undecylprodigiosin are easily visualized on plate grown cultures and their production is seen to coincide with formation of pre-spore aerial hyphae. In liquid culture, production of ‘ actinorhodin (8,15), undecylprodigiosin (12,34), and methylenomycin (16) is generally associated with stationary phase. Evidence for common genetic control of the four antibiotics comes from isolation of bld mutations in several different loci which block sporulation and all of the four antibiotics (5,30, reviewed in 7), and abs mutations which block all four antibiotics but allow abundant sporulation (1,3). In addition, a disruption mutation of the abaA gene reduces production of three antibiotics (actinorhodin, undecylprodigiosin and CDA) and, like abs mutations, does not affect sporulation (10). An additional gene involved in antibiotic production but not sporulation is afsB. Mutations of afsB (13) were first isolated as S. coelicolor strains that were defective in production of A-factor, a butyrolactone involved in both streptomycin production and sporulation in Streptomyces griseus (reviewed in 21). The two afsB strains isolated were also affected in production of actinorhodin and undecylprodigiosin but sporulated normally (13). In our conditions, the afsB mutant phenotype is media dependent and does not cause a strong reduction in all of the four antibiotics (1). Unpigmented afsB mutants of S. lividans, which normally 137 produces undecylprodigiosin, have also been isolated (19). A cloned S. coelicolor gene that restored pigment production to a S. lividans afsB mutant strain was isolated (19). First named afsB (19), this cloned gene was later renamed ast [32 (see below)]; accordingly, it will be referred to as ast in this manuscript. The cloned ast gene also stimulated copious overproduction of actinorhodin in afs" S. lividans and S. coelicolor (20). The ast homolog of S. lividans was isolated as a clone which stimulated pigment (actinorhodin) production in a number of bld mutants of S. lividans (32). These bld mutants were shown to carry mutations of at least two different loci (32) but whether the mutant genes corresponded to any of the known (7) S. coelicolor or S. griseus bld genes was not determined. Genetic evidence that the cloned gene did not correspond to the afsB" allele led to its renaming as ast (32). The map location of ast was determined by genetic (3) and physical mapping (26). The open reading frame first reported for ast was 243 amino acids (22). However, later work indicated that the open reading frame extended upstream to include the "afsC" gene, and the complete afiR open reading frame was reported to be 993 amino acids (24). Ast has been suggested to function as a transcription activator of antibiotic genes (21,24). In this communication we report on characterization of the phenomenon of ast-mediated restoration of two of S. coelicolor’s four antibiotics to well- characterized S. coelicolor developmental mutants. Isolation of ast clones. In the course of experiments intended to isolate a clone of absB" (which will be described in more detail elsewhere), SphI-digested M124 DNA was ligated to SphI-digested (phosphatased) pIJ702 (copy number 40 138 Table 1. Strains of S.coelicolor and plasmids used in this study Strain or plasmid reference S. coelicolor M124 argAI proAI cysD18 SCPl' SCP2 D. Hopewood (18) M138 aIgAI proAI cysDI8 SCP1" scrz- D. Hopewood (18) M146 hisAI uraAI strAI SCP1" SCP2“ D. Hopewood (18) I 11501 hisAI uraAI strAI SCPl‘ scpz- pgl' D. Hopewood (18) I C109 hisAI uraAI strAI bldH109 SCPl' SCP2' pgl' (5) I C112 hisAI uraAI strAI bldBIIZ SCPl‘ SCP2‘ pgl' (5) C120 hisAI uraAI strAI absBIZO SCPl‘ SCP2‘ pgl' (1) C181 hisAI uraAI strAI bldH18I SCPl' SCP2' pgl' (5) C186 hisAI uraAI strAI bldBI86 SCP1°SCP2’pgl’ (5) i 1 C249 hisAI uraAI strAI bld1249 SCPl' SCP2' pgl' (5,14) I C536 hisAI uraAI strAI bld0536 SCPl‘ scrz- pgl' (Sb) C301 hisAI uraAI strAI bldA301 scpr SCPZ’ pg]' (5) C3015 hisAI uraAI strAI bldA301 SCP1"SCP2' pgl' (this work) C542 hisAI uraAI strAI absA542 SCP1" SCP2'pgl' (3) C5425 hisAI uraAI strAI absA542 SCP1" SCP2'pgl' (this work) hisAI uraAI strAI act red SCPl' SCP2' pgl' 139 Table 1 (cont’d.) Strain or Genotype" Source or plasmid reference plasmids pIJ702 Tsrr Mel" (25) pIJ922 Tsr' Ltz" (28) I pIJ702-AP22 2.1-kb PstI-BclI ast insert S. Horinouchi (22) pTA140 3.5-kb SphI ast insert (this work) I pWC5 18.5-kb Sau3A ast insert (this work) I 'Abbreviations (18): SCPl, S. coelicolor plasmid 1; SCP2, S. coelicolor plasmid 2; Pgl', ¢C31 sensitive; Tsr‘, thiostrepton resistant; Mel", melanin production; Ltz", fertility functions. to 300). Protoplasts of the absB strain C120 (Table 1) were transformed by the library, with selection for thiostrepton resistance. Transformants which produced substantial blue or red pigment were isolated. Five colonies contained plasmids with a 3.5 kb insert. Restriction mapping showed a correspondence with the reported location of SphI and NsiI sites of ast (24). Southern hybridization with the plasmid pIJ702-AP22 (22), which contains a PstI-BclI fragment of ast (Fig. 1), confirmed the identity of these clones with ast (data not shown). These clones carried only 233 amino acids of the carboxyl end afiR open reading frame and may include an open reading frame reported to be 3’ of afiR (24) (the ORF has not been delimited). One 140 such plasmid was named pTA140 (Figure 1). None of the identified afisR promoters (22,24) were present, so expression possibly occurred from a vector promoter. The SphI insert was found in both orientations. A clone of ast was also isolated from a library of J1501 Sau3A partially digested DNA ligated to BamHI-digested pIJ922 (copy number 1-2). This plasmid, which was named pWC5, restored production of both of the antibiotic pigments to the absB strain C120 (Table 1). This clone included the entire coding region for ast (Figure 1). Restoration of actinorhodin and undecylprodigiosin to S. coelicolor developmental mutants. The plasmids pTA140 and pWC5 were introduced by transformation (18) into bld and abs strains. Actinorhodin and undecylprodigiosin were measured after extraction, as described previously (1), from cultures grown an R5 plates for 8 days; representative results are shown in Table 2. Both the pWC5 and pTA140 cloned alleles dramatically increased actinorhodin production and pTA140 increased undecylprodigiosin in J 1501. Antibiotic production in C604 was not restored because the act and red mutations affect biosynthetic genes (1). The abs and bld mutants’ responses could be grouped into three categories. For class I, consisting of absA, absB and bldG both actinorhodin and undecylprodigiosin were restored, with pTAl40 stimulating undecylprodigiosin more than pWC5, as in J 1501. For class 11, consisting of bldA and bldI, the clones strongly stimulated undecylprodigiosin but not actinorhodin (colonies were deep pink-red). It is known that transcription of at least some red and act biosynthetic genes depends on a functional bldA gene (4,12). In the case of actinorhodin the bldA requirement 141 Figure 1. Restriction maps for pWC5 and pTA140. Sites to the right of Kpnl are from references 22 and 24. Sites leftward of the Kpnl site were not determined. CD! it: ZZdV-ZOLI‘Id (MflVld J] // SOMd P .———w%———— 142 rBamHl as: ' P \BamHl —Sacl - Pstl ‘--Sphl -BcH -Sphl -Xhol. —EcoRl —Kpn| 143 has been at least partially defined: bldA mutations, which alter the cell’s only leucyl tRNAUUA, prevent translation of the UUA-containing actII-ORF4 gene (11), whose product is required for expression of the act biosynthetic genes (11). A gene product of the red cluster, RedD, has substantial amino acid similarity to ActII-ORF4 (11) and is similarly required for red gene expression (29). The redD gene (29), however, does not contain any UUA codons so the nature of the bldA requirement in red gene expression is as yet unknown. Because the bldl response was similar to bldA’s, it will be interesting to assess the UUA-translation capacity of bld! mutants to determine if they are defective in a bldA pathway. Finally, class III consisted of bldB and bldH, which showed idiosyncratic responses: C112 (bldBIIZ) did not respond to pTA140 (Table 11) although C186 (bld186) responded to pTA140 but not pWC5 (data not shown); and C109 failed to produce undecylprodigiosin. A second bldH mutant strain, C181, produced both undecylprodigiosin and actinorhodin in response to pWC5 (data not shown); C109 may carry an uncharacterized red mutation (both bldH109 and red map to 5 o’clock). For all these results, the differences seen for pWC5 and pTA140 could be due to the differing coding regions, gene expression or copy number of the two plasmids. Thus, the cloned ast alleles: 1) bypassed the requirements for the absA, absB, bldA, bldB, bldG and bld] genes in undecylprodigiosin gene expression; 2) bypassed the requirements for the absA, absB, bldB, bldG and bldH genes in actinorhodin gene expression; and 3) failed to bypass the requirement for the bldA function in actII-ORF4 translation. Failure of the afisR clone to affect methylenomycin or calcium-dependent antibiotic. The abs and bld strains containing pTA140 or pWC5 were assayed for 144 Table 2. Measurements of actinorhodin and undecylprodigiosin production in ast-stimulated strains" Plasmid content and antibiotic tested: None pWC5 pTA140 C536 bldG 0.3 0.4 394.5 0 2602 53 I Class 11 C301 bldA 0 0 0.3 1.4 0 11.2 C249 bld! 0 0 0 1.9 0 6.2 Class III C112 bldB 0 0 506.5 1.7 0 0 C109 bldH 0.2 0 1529.5 0 1426 0 —7 _- "Antibiotic extractions are described in the text. The values represent the average of duplicate extractions from 20 mg of mycelia grown an R5 plates. Similar results were obtained from three replicate experiments. l’Complete genotypes are in Table 1. cAct, actinorhodin; values represent mg actinorhodin per mg mycelia. The absorbance at 640 nm was the maximum for actinorhodin; one OD unit = 120 _g/ml (20). °Red, undecylprodigiosin; values represent mg undecylprodigiosin per gm mycelia. The absorbance at 530 nm, in acidified methanol, was the maximum; one OD unit = 3.91 _g/ml (35). 145 CDA production, using a method as previously described (1). As shown in Figure 2, the abs " strain C604 showed a strong calcium-dependent inhibition of the indicator strain. In contrast, the bldA, bldB, absA and absB strains showed no CDA activity. The clone pWC5 failed to restore CDA to any of the bld or abs strains tested. A similar failure of pWC5 to restore CDA was observed with bldG and bld! strains (data not shown). In bldH, the very early, strong stimulation of actinorhodin, which also kills Staphylococcus aureus, precluded assessment of CDA (data not shown). The high-copy ast clones, pTA140 and pIJ702-AP22, also failed to stimulate CDA in any of the same set of bld or abs strains (results not shown). For tests of methylenomycin activity, strains carrying the plasmid SCP1, which encodes the methylenomycin biosynthetic and resistance genes (27), were constructed. The strain M138 (argAI cysDIS proAI SCPl") was crossed with abs and bld strains (all hisAI uraAI strAI SCPl' pWC5). The extrachromosomal SCP1 is transferred at extremely high frequency (36); close to 100% of progeny are expected to carry it. Transconjugants with the genotype arg" cys" pro" strAI tsrr were selected and tested for acquisition of SCP1 by testing for methylenomycin resistance as previously described (3). All transconjugants tested were mmy', as expected. Strains that were abs or bld (but pigmented due to the presence of pWC5) were cross-streaked against the methylenomycin-sensitive strain J 1501 (Table 1) an R5 media and incubated for 5 days at 28 C. The plasmid pWC5 did not restore methylenomycin activity to either the bldA or absA strains (Figure 3) or to absB or the other bld strains (not shown). Conclusions. These results show that the cloned extra copies of ast alleles were able to restored actinorhodin and undecylprodigiosin biosynthesis, but not CDA 146 Figure 2. Effect of the ast clone pWC5 on CDA production in Bld‘ and Abs' mutant strains. Strains were grown on Oxoid nutrient agar for 42 hours, when 1 cm square plugs were cut out and transferred to Oxoid nutrient agar plates with or without calcium added to 12 mM (as CaNO3). Soft oxoid nutrient agar (with or without 12 mM calcium) was seeded with Staphylococcus aureus and overlaid onto the plates. Zones of inhibition were observed after overnight 370 C incubation. Plug A is C542, plug B is C120, plug C is C604, plug D is C301, plug E is C186. Panel A: strains lack an ast clone; panel B: strains carry pWC5. Plates on the left lack calcium; plates on the right have calcium. 147 148 or methylenomycin to a wide range of well-characterized S. coelicolor developmental mutants. It is noteworthy that extra cloned copies of the actII-ORF4 and redD genes also restore antibiotic production - actinorhodin or undecylprodigiosin, respectively - to most of the developmental mutants studied here (1,2,31). The actII-ORF4 and redD mode of action has not been established in this phenomenon but such observations argue strongly that the absA, absB, bldA, bldB, bldG, bldH and bld! mutants are all metabolically and biosynthetically competent to produce antibiotics, but fail to do so because the regulatory program is perturbed. One hypothesis is that the abs and bld genes are normally required for expression of sufficient, active ActII- ORF4 and RedD but extra copies of actII-ORF4 or redD, or deregulated expression of these genes from a plasmid promoter, can bypass the abs or bld requirements (1,31). The ability of cloned alleles of ast to restore antibiotic production to abs and bld mutants is consistent with the hypothesis that all of these are regulatory mutants. One possible explanation for the phenomenon of ast actinorhodin and undecylprodigiosin activation is that Ast can substitute for ActII-ORF4 and RedD, but this is inconsistent with the ast clones’ inability to restore actinorhodin to the bldA mutant [which fails to translate the actII-orf4 gene (11)]. Another possibility is that ast acts by increasing actII-ORF4 and redD expression. An observation consistent with this hypothesis was made by Horinouchi et al. (23) who observed that the abundance of two transcripts from the act]! region increased in strains containing cloned extra copies of ast. This result must be qualified by the information that the act]! region was later shown to include not two but three distinct transcripts and the 149 Figure 3. Methylenomycin assay in Bld' and Abs' strains carrying pWC5. Strain construction is described in the text. The strains were cross-streaked on thin R5 plates and incubated at 30 C for 5 days. Streaks: A, J1501; B, M146; C, C3015; D, C3015 (pWC5); E, C5425 (pWC5); F, C5425. 150 A . 8“: \I {<1 .____ “\/ "L I...) 8"" l [l \K F:_ 7’ 151 Horinouchi et al. report did not specifically identify the actII-ORF4 transcript. Passantino et al. (31) observed that multicopy actII-ORF4 can restore actinorhodin to bldA mutants, possibly because a low level of bldA-independent translation of actII- ORF 4 can produce adequate gene product. Thus, if Ast does stimulate actII- ORF 4 transcription, the mRN A levels may be lower than in such plasmid-carrying strains. It is intriguing that the N-terminus of Ast shows substantial amino acid similarity (11,33) to the ActIl-ORF4 and RedD proteins over a distance of 90 amino acids. F ernandez-Moreno et al. have suggested (11) that this similarity might indicate an association between ActII-ORF4 and Ast via their similar domains; such an association might be involved in the ast stimulation observed here. Yet another possibility is that the bld and/or abs mutants lack an additional, as yet undescribed, factor that interacts with ActII-ORF4 and RedD; Ast might increase expression of, or substitute for, such a factor. It is noteworthy that the truncated ast alleles encoded by the pTA140 and pIJ702-AP22 clones are as capable of producing as strong an effect an antibiotic production in many of the mutant strains as is the entire ast clone (Table 2; data not shown for pIJ702-AP22). Two putative DNA-binding domains have been suggested to be in the C-terrninal 200 amino acids (22) of Ast; these would occur in all three clones. However, the mechanism for expression of ast in pTA140 is unclear: it lacks both of the reported promoters (22,24), so transcription would have to occur from a plasmid promoter and translation from an internal start site. Perhaps an ORF downstream from ast (24), contributes to the stimulatory effect in pTA14O and le702-AP22. Alternatively an RNA product or DNA site associated 152 with the clones may be responsible for the observed effects. Whether or not ast is required for all antibiotic production in S. coelicolor or S. lividans remains to be established. Disruptions of the ast open reading frame (24) were reported to reduce and delay actinorhodin production, but effects on undecylprodigiosin, calcium-dependent antibiotic or methylenomycin were not discussed. Deletion of the ast gene has not been reported. Interestingly, an extensive search for S. coelicolor mutants severely defective in actinorhodin and undecylprodigiosin defective mutants, which identified the absA and absB mutants and the act red strain C604 (1,3) did not identify any strains with mutations mapping to the ast locus. The inability of cloned Ast alleles to restore CDA or methylenomycin to the developmental mutants is interesting in light of the coordinate genetic control of all antibiotics observed in abs and many bld mutants. Other cloned sequences, which are unrelated to afiR, show similar behavior in that they also restore actinorhodin and undecylprodigiosin, but not CDA, to bld and abs developmental mutants (2,6). It is well established that physiological conditions can be manipulated to strongly favor production of one secondary metabolite over the others that a strain is capable of producing. Defined media that allow production of either methylenomycin, actinorhodin or undecylprodigiosin in S. coelicolor strain 1147 have been described (15,16). Differential responses to physiological conditions have also been observed in the S. coelicolor developmental mutants. For example, low phosphate media allows production of undecylprodigiosin in bldA (12) but not absB mutants (1). The extra copies or deregulated expression of the plasmid-borne ast alleles may obscure 153 ast’s primary role in antibiotic regulation; perhaps its major role is in activation of actinorhodin and/ or undecylprodigiosin in response to a specific, as yet undefined, nutritional signal. Alternatively, ast could be part of a regulatory system that is specific to actinorhodin and undecylprodigiosin and that acts downstream of the bld and abs genes in a regulatory hierarchy. We thank David Hopwood for providing S. coelicolor strains and Sueharu Horinouchi for providing pIJ702-AP22. T.A. was supported by a George J. Bouyoucos Graduate Fellowship. This work was supported by the National Science Foundation grants DMD-8811338 and MCB-9206068 to W.C. and by the Biotechnology Research Center and the NSF Center for Microbial Ecology (DIR 8809640) at Michigan State University. References 1. Adamidis, T., and W. Champness. 1992. Genetic analysis of absB, a Streptomyces coelicolor locus involved in global antibiotic regulation. J. Bacteriol. 174:4622-4628. 2. Adamidis, T., and W. Champness. Unpublished data. 3. 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. 17212962-2969. 4. Bruton, C., E. Guthrie, and K. Chater. 1991. Phage vectors that allow monitoring of transcription of secondary metabolism genes in Streptomyces. Bio/Technology 9(7):652-656. 5. Champness, W. 1988. New loci required for Streptomyces coelicolor morphological and physiological differentiation. J. Bacteriol. 170:1168-1174. 5b. Champness, W. Unpublished data. 10. 11. 12. 13. 14. 15. 16. 154 Champness, W., P. Riggle, T. Adamidis, and P. Vandervene. 1992. Identification of Streptomyces coelicolor genes involved in regulation of antibiotic synthesis. Gene 115:55-60. Chater, K. F. 1989. Sporulation in Streptomyces. In 1. Smith, R. A. Slepecky and P. Setlow (eds.), Regulation of procaryotic development. Amer. Soc. Microbial, Washington, DC. Daull, J. L., and L. C. Vining. 1990. Nutritional control of actinorhodin production by Streptomyces coelicolor A3(2): suppressive effects of nitrogen and phosplate. Appl. Microbiol. Biotechnol. 32:449-454. Feitelson, J., F. Malpartida, and D. Hopwood. 1986. Genetic and biochemical characterization of the red gene cluster of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 131:2431-2441. Fernandez-Moreno, M. A., A. J. Martin-Trisha, E. Martinez, J. Niemi, H. M. Kieser, D. A. Hopwood, and F. Malpartida. 1992. abaA, a new pleiotropic regulatory locus for antibiotic production in Streptomyces coelicolor. J. Bacteriol. 174:2958-2967. Femandee-Moreno, M. A., J. L. Caballero, D. A. Hopwood, and F. Malpartida. 1991. The act gene cluster contains regulatory and antibiotic export genes: direct targets for translational control by the bldA tRNA gene of Streptomyces coelicolor. Cell 66:769-780. Guthrie, E. P., and K. F. Chater. 1990. The level of a transcript required for production of a Streptomyces coelicolor antibiotic is conditionally dependent on a tRNA gene. J. Bacteriol. 172:6189-6193. Hara, 0. S., T. Horinouchi, T. Uozumi, and T. Beppu. 1983. Genetic analysis of A-factor synthesis in Streptomyces coelicolor A3(2) and Streptomyces gn'seus. J. Gen. Microbiol. 129:2939-2914. Harasym, M., and J. Piret. 1990. The Streptomyces coelicolor A3(2) bldB region contains at least two genes involved in morphological development. J. Gen. Microbiol. 136:1543-1550. Hobbs, G., C. M. Frazer, D.C.J. Gardner, J. A. Cullum, and S. G. Oliver. 1990. Pigmented antibiotic production by Streptomyces coelicolor A3(2): kinetics and the influence of nutrients. J. Gen. Microbiol. 136:2291-2296. Hobbs, G., A.I.C. Obanye, J. Petty, J. C. Mason, E. Barratt, D.C.J. Gardner, F. Flett, C. P. Smith, P. Broda, and S. G. Oliver. 1992. An integrated approach to studying regulation of production of the antibiotic 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 155 methylenomycin byStreptomyces coelicolor A3(2). J. Bacteriol. 174: 1487-1494. Hang, S.-K., M. Kito, T. Beppu, and S. Horinouchi. 1991. Phosphorylation of the afiR product, a global regulatory protein for secondary metabolite formation in Streptomyces coelicolor A3(2). J. Bacteriol. 173:2311-2318. Hopwood, D. A., J. J. Bibb, K. F. Chater, T. Kieser, C. J. Brunton, H. M. Kieser, D. J. Lydiate, C. P. Smith, J. M. Ward, and H. Schrempf. 1985. Genetic manipulation of Streptomyces. A laboratory manual. The John Innes Foundation. Norwich, UK. Horinouchi, S., O. 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. J. Bacteriol. 155:1238-1248. Horinouchi, S., and T. Beppu. 1984. Production in large quantities of actinorhodin and undecylprodigiosin induced by afsB in Streptomyces lividans. Agric. Biol. Chem. 48(8):2131-2133. Horinouchi, S., and T. Beppu. 1992. Regulation of secondary metabolism and cell differentiation in Streptomyces: A-factor as a microbial hormone and the Ast protein as a component of a two-component regulatory system. Gene 115:167-172. 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(1):257-269. Horinouchi, S., F. Malpartida, D. A. Hopwood, and T. Beppu. 1989. afsB stimulates transcription of the actinorhodin biosynthetic pathway in Streptomyces coelicolor A3(2) and Streptomyces lividans. Mol. Gen. Genet. 215:355-357. Horinouchi, 8., M. Kito, M. Nishiyama, K. Furuya, S.-K. Hang, K. Miyake, and T. Beppu. 1990. Primary structure of Ast, a global regulatory protein for secondary metabolite formation in Streptomyces coelicolor A3(2). Gene 95(1):49-56. Katz, E., C. J. Thompson, and D. A. Hopwood. 1983. Cloning and expression of the tyrodinase gene from Streptomyces antibioticus in Streptomyces lividans. J. Gen. Microbiol. 129:2703-2714. Kieser, H. M., T. Kieser, and D. A. Hopwood. 1992. A combined genetic and physical map of the chromosome of Streptomyces coelicolor A3(2). J. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 156 Bacterial. 174:5496-5507. Kirby, R., and D. Hopwood. 1977. Genetic determination of methylenomycin synthesis by the SCP1 plasmid of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 98:239-252. Lydiate, D. J ., F. Malpartida, and D. A. Hopwood. 1985. The Streptomyces plasmid SCPZ': its functional analysis and development into useful cloning vectors. Gene 35(3):223-235. Narva, K. E., and J. S. Feitelson. 1990. Nucleotide sequence and transcriptional analysis Of the redD locus of Streptomyces coelicolor A3(2). J. Bacteriol. l72(l):326-333. Merrick M. J. 1976. A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 96:299- 315. Passantino, R., A.-M. Puglia, and K. Chater. 1991. Additional copies of the act]! regulatory gene induce actinorhodin production in pleiotropic bld mutants of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 137:2059-2064. Stein, D., and S. N. Cohen. 1989. A cloned regulatory gene of Streptomyces lividans can suppress the pigment deficiency phenotype of different developmental mutants. J. Bacteriol. 17112258-2261. Stutzman-Engwall, K. J ., S. L. Otten, and C. Richard Hutchinson. 1992. Regulation of secondary metabolism in Streptomyces spp. and the overproduction of daunombicin in Streptomyces peucetius. J. Bacteriol. 174:144-154. Takano, E., H. C. Gramajo, E. Strauch, N. Andres, J. White, and M. J. Bibb. 1992. Transcriptional regulation of the redD transcriptional activator gene accounts for growth-phase-dependent production of the antibiotic undecylprodigiosin in Streptomyces coelicolor A3(2). Mol. Microbiol. 6(19):2797-2804. Tsao, S. W., B.A.M. Rudd, X. He, C. Chang, and H. G. Floss. 1985. Identification of a red pigment from Streptomyces coelicolor A3(2) as a mixture of prodigiosin derivatives. J. Antibiot. 38:128-130. Vivian, A. 1971. Genetic control of fertility in Streptomyces coelicolor A3(2): Plasmid involvement in the interconversion of UF and IF strains. J. Gen. Microbiol. 69:353-364. Chapter 8: Summary and conclusions 157 158 The Streptomycetes are common filamentous soil bacteria which attracted the attention of scientists not only because they have an unusual morphological complexity but also because they produce a plethora of secondary metabolites with commercial value such as antibiotics, insecticides, and immunosuppressors. In addition to their commercial importance, Streptomycetes exhibit a primitive developmental cycle. While vegetative growth occurs as hypha] extension, continuous branching creates a substrate mycelium. The third day of their growth, aerial hypha rise out of the substrate mycelium which in turn subdivides to create spores. The secondary metabolism commences at the onset of aerial hyphae formation. Many mutants have been isolated which are deficient in both antibiotic production and aerial mycelia formation, suggesting that both processes are coordinately controlled. Streptomyces coelicolor produces four known antibiotics actinorhodin (Act), undecylprodigiosin (Red), calcium dependent antibiotic (CDA), and methylenomycin (Mmy). The biosynthetic and resistance genes for these antibiotics are clustered. All four antibiotics have been genetically characterized and the red, act, and mmy gene clusters have been cloned. Production of each of S. coelicolor’s antibiotics, and the antibiotics of other Streptomycetes, has been thought to be regulated by physiological factors acting differently on each individual pathway. In our laboratory we sought to identify genetic elements involved in temporal regulation of antibiotics. We searched for two classes of mutants: a) mutants that produce antibiotics but do not sporulate, or b) mutants that do not produce any antibiotic but sporulate normally. Such mutants would uncouple sporulation from antibiotic production and would suggest: 1) that 159 sporulation and antibiotic synthesis may be regulated by independent pathways, once the decision for differentiation has been made, and 2) the antibiotics are subject to coordinate regulation. We took advantage of the fact that two of the four antibiotics produced in S. coelicolor are pigmented, actinorhodin which is blue and undecylprodigiosin which is red. Thus we searched for mutants that abolished simultaneously both pigments while sporulation was still occurring. After UN. and NTG mutagenesis, 800,000 colonies were screened for a non-pigmented phenotype. Three classes of mutants were isolated. Phenotypic and genetic characterization showed that one class of mutants consisted of double mutants for the two pigmented antibiotics (i.e, act red). The other two classes of mutants defined two new loci which we named absA and absB, for antibiotic synthesis deficient. Spectrophotometric quantitation of actinorhodin and undecylprodigiosin showed that the absA and absB mutants were completely deficient for production of both antibiotics. Production of the colorless antibiotics methylenomycin and calcium dependent antibiotic was assessed by killing assays which showed both absA and absB mutants were deficient in production of these antibiotics. Both absA and absB mutants appeared to produce melanin, a secondary metabolite with no antibiotic activity. Thus, our results suggest that the absA and absB genes may not regulate secondary metabolism generally, but rather, they regulate specifically antibiotic production in S. coelicolor. Both absA and absB mutations were mapped by plasmid mediated crosse against the wild-type. The presence of only two phenotypes, Abs‘ and wild-type, among the recombinants indicated the presence of only one mutation in each mutant tested. The absA locus maps at 10 o’clock and the absB locus maps at 5 o’clock in 160 the chromosome of S. coelicolor. Both loci are distinct from the loci of the four antibiotic gene clusters and from the locations of genes, which when overexpressed, stimulate antibiotic production such as, abaA, ast, afsQ, and asaA. Crosses of the absA mutants against each other produced an extremely small number of recombinant progeny with the wild-type phenotype, suggesting that all absA mutations are very close to each other and may belong to the same gene. The same results were also obtained with the absB mutants. Both absA and absB genes were found to regulate transcriptionally the antibiotic production since a promoterless reporter gene fused to the promoters of actinorhodin and undecylprodigiosin biosynthetic genes was unable to be expressed in the strains carrying either absA or absB mutations. There are two main differences in the phenotype between absA and absB mutants. absB mutants appear to be somewhat leakier in antibiotic production on same complex media and do not accumulate spontaneous suppressor mutations as absA mutants do. Another difference between absA and absB mutants is the frequency at which they were isolated. absA mutants appear to be extremely rare with a frequency of 5 X 10°, which is similar to the frequency with which the double mutants were found. absB mutants were found at a frequency similar to the frequency for loss-of-function mutants found in S. coelicolor. The low frequency with which absA mutants were found is curious. One possibility for the rarity of absA mutants might be inadequate screening for these mutants. However, the detection of the double mutants, act red, at the expected frequency argues against this possibility. A need for multiple mutations to produce the AbsA phenotype is another 161 possibility. However, the mapping data requires that if such multiple mutations occurred, they must be clustered together in the absA locus. Furthermore, they would all be suppressed by a single mutation, since absA mutants acquire spontaneous suppressors (P. Riggle, unpublished results). A third possibility is that the absA gene is essential for S. coelicolor and can withstand only certain mutations. This work on absA and absB has lead to three conclusions: 1) Str. coelicolor’s four antibiotic gene clusters are subject to a common control from which control of sporulation can be genetically uncoupled; II) the absA and absB loci are likely to encode regulatory genes in a global regulatory pathway for antibiotic synthesis; and III) the block to antibiotic synthesis in the abs mutants occurs at the level of transcription of the biosynthetic genes. Three other classes of mutants having the Abs phenotype were isolated. Three different loci have been defined by these mutants, abs-95, abs-8752, and abs- 155, which map at 9 o’clock, 10 o’clock and 10 o’clock, respectively. All three of these mutant strains were found to be leakier than absA and absB mutants, with respect to antibiotic production, in all media tested. Perhaps due to their leaky phenotype, each class is represented by only one mutant i.e., being leaky, such mutants would have been less likely to have been picked in the mutant hunt for the strong Abs' phenotype. These mutants produced reduced amounts of all four antibiotics with the exception of mutant C155 (abs-155) which was able to produce methylenomycin. Further characterization of these mutants is underway. We also looked for mutants which produce antibiotics but do not sporulate. N a such mutant was recovered suggesting that a mutation in an essential gene is 162 required or such genes are redundant. However this mutant hunt was not as extensive as the Abs mutant hunt. Three libraries were constructed for shot-gun cloning of the absB" allele. Two of the libraries used a high-copy number vector pIJ702 while the third library utilized the conjugative functions of the low copy number plasmid pIJ922. Two putative absB " clones, pTA108 and pTA128, were isolated from the low copy number library. Hybridization experiments showed the presence of common DNA sequences between the two clones. Both clones were able to restore antibiotic production in absB mutants to wild type levels. Three lines of evidence suggest that pTA108 and pTA128 carry the absB" gene and not a bypassing gene: 1) Both were able to restore antibiotic production in absB mutants to wild type levels; 2) both clones were incapable of restoring any of the three antibiotics tested (Act, Red, and CDA) when transformed to different developmental mutants other than absB mutants; 3) all genes cloned up to now which have been shown to regulate antibiotic production in Str. coelicolor were able to activate expression of actinorhodin in Str. lividans while both of the putative absB" clones fail to do so. The possibility that pTA108 and pTA128 carry a "bypassing" clone still exists, thus we have designated the clones as absB’. Experiments to prove that absB’ gene corresponds to absB" allele are underway. Two types of experiments will prove that these clones carry the absB" allele. The absB' clone will be mapped on the physical map of Str. coelicolor. Following separation of AseI fragments of Str. coelicolor chromosome by CHEF electrophoresis, hybridization experiments using the absB' smallest complementing 163 region as a probe will reveal if absB' maps to the same fragment that absB mutations map. In the second experiment the absB mutation will be transferred from strain to strain using the vector ¢KC516. ¢KC516 can lysogenize into the Str. coelicolor chromosome only through recombination between the cloned insert and the homologous region of the chromosome. Phage released from the lysogens will be a mixture of absB" and absB‘ only if the cloned insert was absB". Phage that carry the absB“ allele will be used to infect an absB mutant. Some of the lysogens should have an Abs' phenotype indistinguishable from the AbsB phenotype (i.e, those infected with absB‘ phage). Phage will be analyzed to verify that the phage have not undergone a rearrangement. These experiments should determine whether the absB’ clones carry the absB" allele. Even if the absB' clones do not correspond to absB", further analysis will be performed because these clones represent novel clones capable of globally regulating the antibiotic production in absB mutants. During the course of cloning the absB" gene, several clones were found that were able to "bypass" the absB mutation. The actII-orf4 and redD genes which are the positive regulators of actinorhodin and undecylprodigiosin biosynthesis, respectively, were cloned. Both clones were found to be able to restore the relevant antibiotics in all developmental mutants tested. This result suggests that the regulation to antibiotic production in abs mutants may be due to a failure to express sufficient ActII-orf4 and Red. Alleles of another antibiotic regulatory gene, ast alleles were isolated from both high copy and low copy number libraries. The ast gene is not clonable in high copy number, but the ast allele isolated from the high copy number library was 164 shown to carry only part of the ast gene that encodes the carboxyl-terminus of the Ast protein. Interestingly, this clone behaves as though it carries the entire ast gene since it is able to restore actinorhodin and undecylprodigiosin production in most of the developmental mutants tested. Hybridization experiments showed that pWC5 isolated from the low copy number library carries the entire ast gene. Upon transformation of the pWC5 clone to different developmental mutants, the ast gene was able to stimulate the production of actinorhodin and undecylprodigiosin but not the production of calcium dependent antibiotic and methylenomycin. Clones of ast can also stimulate Act and Red in most bld mutants and stimulate overproduction of Act and Red in vegetative phase. The ast" gene cannot stimulate actinorhodin production in bldA mutants (bldA gene encodes a tRNA that recognizes the UUA codon for leucine). The actII-orf4 gene carries a TTA codon and thus is translationally dependent on the bldA gene. This result suggests that ast exerts its function through the actII-orf4 gene and cannot substitute for the actII-orf4 gene. 81 nuclease experiments will determine whether ast exerts its antibiotic stimulatory function during vegetative growth by turning on actII-orf4 and redD transcription. Clone pTA227 was isolated from the high copy number library and restriction analysis showed that it carries the whiE gene cluster which is responsible for the spore pigment. Although whiE is temporally and spatially regulated, our clone expresses the whiE genes constitutively even in the substrate mycelium of absB and absA mutants which may be due to a plasmid or cloned promoter. Several other clones have been isolated from the high and low copy number library which are able to "bypass" the absB mutation. Some of the clones isolated 165 from the high copy number library were unable to exert their stimulatory function when subcloned in a low copy number vector. The creation of null phenotypes of these genes will determine whether they are required for antibiotic production. In summary, through the discovery of the absA and absB loci these studies have revealed the existence of a global regulatory mechanism for Streptomycetes antibiotic biosynthesis. Further studies will determine if the genetic elements described here function in regulation of antibiotics in other members of the genus Streptomyces. ADDENDUM In order to more clearly define the roles played by the different authors who contributed work to this dissertation, I have listed the parties who were responsible for the individual experiments. In Chapter 2, Perry Riggle, a graduate student of Dr. Wendy Champness’ laboratory, subcloned the ast gene to phage ¢KC5 16 and he mapped the ast gene in Str. colicolor chromosome. Chapter 2, is reproduced with the permission of the American Society for Microbiology, from the Journal of Bacteriology (1990) 172: 2962-2969. Chapter 3, is reproduced with the permission of American Society for Microbiology, from the Journal of Bacteriology (1992) 174: 4622-4628. In Chapter 6, Perry Riggle constructed the P11922 low copy number library. In Chapter 7, Perry Riggle transformed the pIJ702-Ap22 plasmid to different developmental mutants and he assayed them for calcium dependent antibiotic production. 166