if. .liznA .. >2... . 5. . 1 5L. .10,"le . 5!: .ll :5 :5. 14.1. L§.fiv : J. 5, . A... .. .3 .f. .x h$4psil .l ‘5' . $204!..L‘. . .11.... o 1.1!! .IL. _7..\ . . ...‘,.‘..z...»_w... Tm “'CH'GAN STATE umvens WIWIN/lWWII/Ill!NIH/Ill!!!KIWI/WW7" 3 1293 01019 7634 This is to certify that the dissertation entitled Analysis of pattern formation in the heterocyst—forming filamentous cyanobacterium Anabaena sp. strain PCC 7120 presented by Todd Alan Black has been accepted towards fulfillment of the requirements for Ph D degreein jimbenistry Major professor MS U i: an Affirmative Action/Equal Opportunity Institution 0- 12771 LIBRARY Michigan State University PLACE N RETURN BOX to remove thie checkout from your record. TO AVOID FINES return on or bdore date due. F————————___—————————= DATE DUE ' DATE DUE DATE DUE i i J LJ "—1 Jr JDL—J :i[__- ”if—Um MSU leAn Nflrnmive Action/Equal Opportunity inetituion ANALYSIS OF PATTERN FORMATION IN THE HETEROCYST-FORMING FILAMENTOUS CYANOBACTERIUM Anabaena SP. STRAIN PCC 7120 By Todd Alan Black A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry and MSU-DOE Plant Research Laboratory 1993 ABSTRACT ANALYSIS OF PATTERN FORMATION IN THE HETEROCYST-FORMING FILAMENTOUS CYANOBACTERIUM Anabaena SP. STRAIN PCC 7120 By Todd Alan Black Anabaena sp. strain PCC 7120 is a photoautotrophic organism that utilizes oxygen-evolving photosynthesis for growth. In Anabaena sp., oxygen-intolerant nitrogenase activity is segregated to specialized cells, called heterocysts, in which the concentration of 02 is kept very low. In nitrogen-deficient media, heterocysts are formed by the differentiation of semi-regularly spaced vegetative cells. Transposon insertions within the her}? gene prevented the evident differentiation of heterocysts (Het' phenotype). Transcriptional fusions with the reporter genes, luxAB, that encode Iuciferase, a protein complex that can produce light, showed that hetR is induced within two hours following the removal of nitrate from the medium. Photon-accumulating microscopy showed that by 3.5 hours, the induced expression of ham is limited to spaced groups of cells. The cells with enhanced luminescence are more narrowly delimited by 6 hours; and by 24 hours, cells midway between two mature heterocysts are most luminescent. The induction of hetR was found to require a functional copy of the her}? gene, indicating that her}? is autoregulatory. Transposon-generated mutant N10 is also Het'. The N10 mutant tends to change to a mutant that forms multiple contiguous heterocysts (Mch phenotype). Other insertions into the site of transposition typically directly generated the Mch phenotype. When borne on a plasmid, the wild-type gene, hetN, does not complement the Het' phenotype; rather, hetN suppresses heterocyst differentiation in wild-type cells. The suppression requires an intact copy of hetN. The sequences of the proteins encoded by hetN and by a large orf, the 5’ orf, found upstream from hetN, resemble the sequences of proteins that are involved in the syntheses of fatty acids and secondary metabolites. A third orf, hetl, is transcribed from the opposite strand of DNA and lies just 42 bp 3’ from hetN. The protein encoded by hetl shows sequence similarity to Bacillus spp. proteins that are required for the syntheses of secondary metabolites. The hetl gene may be required for vegetative growth. A physical and immunological approach was also undertaken in an attempt to study how cells within a filament communicate. lntercellular movement of fluorescent dyes was not observed. However, a protein associated with the plasma membranes of cells was identified that cross-reacts with antibodies raised against a gap junction protein. This observation is consistent with the possibility that aqueous channels provide cytosolic continuity between cells. To Laura, For Her Love and Support ACKNOWLEDGEMENTS I extend much gratitude to my advisor C. Peter Wolk for providing a laboratory environment that did not inhibit my ability to test new ideas, no matter how flawed they may have been sometimes, and the ability to make advances and grow freely as a scientist, learning through my mistakes and his advice. I thank Pam Green, Lee Kroos, Lee McIntosh, and Melvin Schindler for serving on my guidance committee and providing me with backup and advice during the course of my dissertation. Melvin Schindler and Sally Meiners also receive my great appreciation for providing me with advice, support, and materials for the work on the identification of the gap junction-like protein and the cell-cell studies. I also thank all the past and present members of the Wolk laboratory for their advice, camaraderie, and friendship over the years, and for providing me with many plasmids and strains that were used over the course of my studies. I thank Jeff Elhai, in particular, for writing the software to interface the photon- counting camera with our computer system and for providing the programs that allowed me to manipulate and analyze the photon images. Jeff was always a wealth of information on most matters and his unexpected outbursts of song would always bring a smile to my face. The support and love of my family and extended family was an essential ingredient for my success and for maintaining my perspective on life. Miles gave me the unconditional love that I could only hope to deserve, and his impish grin would always make me laugh no matter how bad the rest of my day had been. Chris Bauer at the University of Chicago pointed out some errors within the DNA sequence of the 5’ orf using data from Jim Golden’s laboratory at Texas A&M University. Accordingly, I have added a T at bp 941 , CA at bps 946 and 947, a C at bp 961 , and an A at hp 1476; removed a T from between bp 952 and 953 and a G from between hp 1490 and 1491; and converted a T to a G at hp 1045 and a C to a T at hp 1473. I thank my colleagues for sharing their data with me. This work was supported by a fellowship from the Nitrogen Availability Program at Michigan State University and the US. Department of Energy under grant DE-FGOZ-90ER20021. vi TABLE OF CONTENTS Page List of figures ................................................................ x List of abbreviations and symbols ..................................... xiii Chapter 1. General introduction Introduction .............................................................. 1 General description of development in some procaryotes ............................................................. 2 Processes involved in the initiation of differentiation Bacillus sub tilis .................................................... 7 Strep tomyces ...................................................... 1 6 Myxococcus xanthus ............................................ 18 Anabaena ........................................................... 21 Processes leading to localized gene expression during development 8. subti/is ............................................................ 28 Strep tomyces ...................................................... 36 M. xanthus .......................................................... 38 Anabaena ............................................................ 40 Chapter 2. Spatial expression and autoregulation of hetI-i, a gene involved in the control of heterocyst development in Anabaena Summary ................................................................. 48 Introduction .............................................................. 49 vii Materials and methods Plasmids and strains ............................................. Measurement of luminescence ............................... Resuhs Transposon insertions in MM ................................. Transcriptional control of hetR-luxAB fusions following the removal of fixed nitrogen ............................................................. Localization of expression of hetR-quAB during differentiation ........................................... Effect of hetR on expression of hepA ..................... Discussion ............................................................... Chapter 3. Analysis of N10, a Hot mutant in Anabaena sp., provides evidence that a secondary metabolite may regulate heterocyst spacing Summary ................................................................. Introduction .............................................................. Materials and methods Growth and manipulations of Anabaena sp. ............. Recovery of DNA contiguous with the transposon in mutant N10 .................................... Construction of a A-EMBLS library ......................... Subcloning the wild-type N10 DNA from the A—EMBL3 library, and subsequent plasmid constructions .......................................... Resuhs Isolation, recovery and reconstruction of the N10 mutation ............................................ Table 3.1. Construction, antibiotic resistances and phenotypes of the strains .............. Cloning and attempted complementation of the N10 Mutants ............................................. Sequence analysis of DNA near the transposon insertion in N10 (Genbank accession number) .............................................. Attempts to inactivate the 3’ orf ............................ Discussion ............................................................... Chapter 4. A study of intercellular communication in the filamentous cyanobacterium Anabaena sp. strain PCC 7120: immunological evidence for a gap junction- related peptide in a procaryote Summary ................................................................. viii 50 56 57 58 60 63 64 69 70 71 72 73 76 79 80 81 82 93 93 101 Introduction .............................................................. 102 Materials and methods Growth of strains ................................................. 105 Protein isolation ................................................... 105 Isolation of plasma membrane fractions ................... 106 Labeling the filaments with fluorescent dyes .................................................................. 107 Use of polar cytosolic dyes to measure cytoplasmic continuity ......................................... 108 Use of lipophilic membrane dyes to measure membrane continuity .............................. 109 Examination of intercellular movement of dye .......... 109 Resuhs Immunological analysis of Anabaena proteins ............................................................. 1 10 Examination of fluorescent labeling and FRAP ........................................................... 1 12 Discussion ............................................................... 1 20 Bibliography .................................................................. 1 30 LIST OF FIGURES Figure 1.1 1.2 1.3 1.4 1.5 2.1 2.2 2.3 Photomicrographs displaying filaments of Anabaena sp. strain PCC 7120 grown in the presence of nitrate (A) and 24 hours following the removal of nitrate from the medium (8) ........................................................... An idealized diagram of a vertical section through the center of a sporulating colony of Strep tomyces coelicalor .......................................... Diagram showing the proposed pathways of regulation for the activation of SpoOA and ComA ............................................................. Classes of developmental mutants of Myxacoccus xanthus ................................................................ Diagram showing the sequential expression and activation of sigma factors, and their transcriptional dependencies, required for sporulation in Bacillus subtilis .................................. Several DNA constructs used in this work ................. Relative light emission from suspensions of hetR-luxAB constructs SR884a, DR884a and DR937 .................. Images of luminescence from individual filaments of SR884a at (A) 0, (B) 3.5, (C) 6, and (D) 24 hours, and of DR884a at (E) O and (F) 24 hours, following nitrogen stepdown ................ X Page 20 31 53 59 62 3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 4.3 The region of the Anabaena genome that contains the 5’, hetN, and hetl ....... J ......................... Nucleotide sequence of the 4.7-kb fragment shown in Figure 3.1 ............................................... Alignment of the protein encoded by hetN with other proteins ................................................. Alignment of the protein encoded by hetl with Sfp and Oer from B. subtilis and B. brevis, respectively ....................................... Alignment of the entire protein encoded by the 5’ orf with the carboxy-terminal half of the LYSZ protein from S. cerevisiae ...................... Alignments of domains within the 5’ orf- encoded protein and (a) acyl carrier protein, (bl ketoreductase and (c) acyl-transferase domains from other proteins .................................... Proteins resolved on SDS-PAGE gels, transferred to nitrocellulose, and decorated with antibodies raised against the a—peptide of the rat liver gapjuncfion .......................................................... Proteins from Anabaena resolved on SDS-PAGE gels, transferred to nitrocellulose, and decorated with antibodies raised against the a-peptide of the rat liver gap junction ........................ Proteins from: 1) crude extract, 2) thylakoid membrane fractions, and 3) plasma membrane fractions of Anabaena resolved on SDS-PAGE gels, transferred to nitrocellulose, and decorated with antibodies raised against the a-peptide of the rat liver gap junction ......................................... xi 75 83 87 88 90 92 111 113 114 4.4 4.5 4.6 4.7 4.8 Absorption spectra of (A) thylakoid and (8) plasma membrane fractions from Anabaena sp. strain PCC 7120 ............................................... Vegetative filaments of Anabaena sp. PCC 7120 labeled with carboxyfluorescein diacetate .................. A single filament of Anabaena sp. PCC 7120, grown in nitrogen-free medium and containing a heterocyst, which was then labeled with carboxyfluorescein diacetate .................................... Vegetative filaments of Anabaena sp. PCC 7120 labeled with the lipophilic dye NBD-PC ...................... A single filament of Anabaena sp. PCC 7120, grown in nitrogen-free medium and containing a heterocyst, which was then labeled with the lipophilic dye NBD-PC ............................................. xii 115 116 118 121 122 AA ACP Ap ATP Bm bp BLOS Born BSA cAMP Chl LIST OF ABBREVIATIONS AND SYMBOLS Presence of a functional gene' or of a phenotype" Lack of a functional gene or of a phenotype Deletion Bacteriophage lamda Sigma factor Allen and Arnon medium Acyl carrier protein Ampicillin Adenosine 5'-triphosphate Bleomycin Basepair(s) A gene cassette that contains a born (oriT) region, the quA3 a transcriptional reporter genes, an oriV region, and the aad gene conferring Sm'lSp' Basis of mobilization during conjugation Bovine serum albumin Adenosine 3’:5’-cyclic monophosphate Chlorophyll xiii Cm cpm DNA DR EDTA Em FA FAS FRAP GTP HEPES IPTG kb kD Km LB MES NAD(P)H Nm orf PAGE Chloramphenicol Counts per minute Deoxyribonucleic acid Double recombinant Ethylenediaminetetraacetic acid Erythromycin Fatty acid Fatty acid synthase Fluorescence redistribution after photobleaching Gravity Guanosine 5’-triphosphate N-(hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) lsopropyl-thiogalactoside Kilobasepair(s) Kilodaltonls) Kanamycin Luria-Bertani medium 2-(N-morpholino)ethanesulfonic acid reduced form of nicotinamide adenine dinucleotide (3’- phosphate) Neomycin Open reading frame Polyacrylamide gel electrophoresis xiv PHB PK PKS PMSF ppGpp SDS Sm Sp SR Suc TBS TES Tn Tris Poly-B—hydroxybutyrate Polyketide Polyketide synthase Phenylmethanesulfonyl fluoride Guanosine tetraphosphate Resistant Sensitive Sodium dodecyl sulfate Streptomycin Spectinomycin Single recombinant Sucrose Tris buffered saline N-tris-(hydroxymethyl)methyl-2-aminoethanesulfonic acid Transposon Trislhydroxymethyl)aminomethane ' Names of genes and genetic loci are italicized " Names of phenotypes are capitalized XV Chapter 1 GENERAL INTRODUCTION INTRODUCTION Research on procaryotic development has focused on the mechanisms by which the cells sense environmental cues and thereupon initiate and coordinate the processes required for cellular differentiation. An in-depth understanding requires that the procaryotes whose development is studied can be analyzed both biochemically and genetically. The development of some procaryotes, whose development leads to a state that is distinct from that of vegetatively growing cells, has been subject to extensive investigation. Examination of the literature shows that even in these 'simple’ organisms, complex regulatory networks are utilized to achieve developmental transitions. In fact, numerous signals that are known to be used by eucaryotes to regulate differentiation and metabolism, including the production of extracellular peptidic factors, steroid hormone-like molecules, CAMP and ion gradients, are also utilized by procaryotes. In procaryotes, as in eucaryotes, specific receptors or sensors for the various signals regulate transcriptional responses through a complex of response regulator proteins that often have a variety of kinase and phosphatase activities. 2 In both procaryotic and eucaryotic organisms, temporally controlled processes are the most easily studied and best understood aspects of differentiation, whereas spatial control, which is required for pattern formation, is more difficult to approach experimentally and is poorly understood. All of the procaryotic developmental systems studied display spatial regulation that includes the controlled expression of genes within specific cells or compartments. In most cases it is unclear what generates the spatial differences, but the genetic manipulability of the procaryotic genome allows for a rapid dissection of the mechanisms involved in pattern formation. In this introduction, I will present a general description of the types of development that occur in some of the most extensively studied procaryotic developmental systems. Because processes that are known to control the initiation of, and spatially regulated gene expression during, development probably exert the greatest influence on pattern formation, I will give a more detailed account of these processes. GENERAL DESCRIPTION OF DEVELOPMENT IN SOME PROCARYOTES When filaments of the cyanobacterium Anabaena are placed in media that lack a source of fixed nitrogen, a small fraction of the vegetative cells differentiates into specialized cells called heterocysts (Figure 1.1). The primary known function of heterocysts is the fixation of dinitrogen (Walk, 1982). Differentiation into a heterocyst appears to be a terminal stage of development because i) heterocysts do not undergo cell division and ii) there are no Figure 1 .1 Photomicrographs displaying filaments of Anabaena sp. strain PCC 7120 grown in the presence of nitrate (a) and 24 hours following the removal of nitrate from the medium lb). In (b), the arrows indicate the positions of some of the heterocysts. Magnification = 1000x 4 confirmed accounts of the germination of mature heterocysts. lntercellular exchange of metabolites, nitrogenous compounds from the heterocysts and, it is thought, carbohydrates from the vegetative cells, is required for continued growth in media depleted of fixed nitrogen. Therefore, Anabaena has a truly multicellular nature in the sense that cells of different types interact to maintain growth. The only other free-living procaryote in which a similar type of metabolic exchange is known is the actinomycete, Frankie, in which nitrogen-fixing vesicles form at the tips of mycelia (Benson and Silvester, 1993). Very little information is available that concerns the regulation of development in Frankia; however, some of the morphological features of its development are similar to sporulation in another genus of the actinomycetes, Streptomyces, an organism that will be discussed below. The hallmark of heterocyst differentiation in Anabaena is that heterocysts arise non- randomly along the length of the filaments (Walk and Ouine, 1975). Therefore, Anabaena is a 'simple’ organism in which to study a process that forms a one- dimensional pattern. Some strains of Anabaena also form spores, called akinetes. Akinetes also appear to arise in a non-random pattern when heterocysts are present (Wolk, 1964). The pattern of akinete formation varies among strains. Some strains form akinetes adjacent to heterocysts whereas other strains form akinetes, initially, midway between two heterocysts. Because the patterns of both heterocyst formation and akinete formation are thought to depend on intercellular interactions, Anabaena may produce two distinct types of pattern-forming signals. 5 Sporulation of Bacillus subtilis is the most extensively studied and best characterized procaryotic developmental process. The great amount of knowledge of Bacillus sporulation that has been attained has made it the model system on which much research on other procaryotic systems is based. This fact will become apparent throughout this introduction. When Bacillus encounters nutrient limitation, cells undergo an asymmetric invagination of their membranes to form an internalized cell. The internal cell, the forespore, is transformed into a spore, while the surrounding cell, the mother cell, produces an environmentally resistant coat around the spore. During nutrient deprivation, B. subtilis also undergoes other types of development: the process of gaining genetic competence (the ability to take up exogenous DNA), and the production of antibiotics and extracellular proteases. Sporulation and competence development are clearly interrelated but also have unique and separable properties; competence development appears to be more closely associated with the production of antibiotics. The developmental biology of spore formation in the Streptomyces has not been studied as extensively; however, because of the commercial importance of the Streptomyces, a great deal has been learned about their biology. Sporulation of Streptomyces resembles, in many ways, the formation of conidia in filamentous fungi such as the well studied Aspergillus nidulans (Timberlake, 1991). In the Streptomyces, aseptate, filamentous, multinucleoid vegetative cells grow along a solid surface as substrate mycelia. When nutrients become limiting, the mycelia branch upward to produce aerial mycelia (Figure 1.2). Aerial mycelium Substrate mycelium Figure 1.2 An idealized diagram of a vertical section through the center of a sporulating colony of Streptomyces coelicolor. Black areas represent intact cells and the white areas represent cells that are lysing. Reproduced, with permission, from Figure 6-4 in M. Dworkin. 1985. Developmental Biology of the Bacteria. p.90. Benjamin/Cummings Publishing Company, Inc., Menlo Park, CA. 7 The aerial mycelia, deriving nutrients from the substrate mycelia (Mendez et al., 1985), undergo spatially regulated differentiation and septation into spores (Figure 1.2). The most extensively studied aspect of Streptomyces sporulation is the developmentally regulated induction of biosythesis of antibiotics. Interestingly, the antibiotics or metabolites produced are also, in some cases, required for morphological differentiation and sporulation. The developmental pattern of the myxobacteria mimics that of other lower eucaryotes, the slime molds, such as Dictyostelium (Loomis, 1982). Single Myxococcus cells can have an 'adventurous’ life style that is terminated in response to environmental cues, particularly nutrient limitation in conjunction with a previously established high cell density, that cause the cells to aggregate into macroscopic structures called fruiting bodies. Within the resulting aggregates of cells, the internal compartment appears to contain the spores that are formed, perhaps at the expense of the life of the outermost cells. During differentiation, Myxococcus, like Streptomyces, produces extracellular factors that are required for the completion of development. PROCESSES INVOLVED IN THE INITIATION OF DIFFERENTIATION B. subtilis Seven genetic loci have been found in B. subtilis that are required for the initiation of sporulation: spoOA, spoOB, spaOE, spoOF, spoOH, spoOJ, and spoOK (Hoch, 1976). SpoOA belongs to the response regulator family of procaryotic regulatory proteins that are activated by phosphorylation (Ferrari et 8 al., 1985b; Stock et al., 1989, 1990). SpoOA binds to DNA and can activate or repress transcription depending on the target (Strauch et al., 1990). SpoOA is also autoregulatory. repressing its own transcription during vegetative growth while inducing its own transcription when activated (Yamashita et al., 1986). A multicomponent phosphorelay system has been found to control the phosphorylation of SpoOA (Burbulys et al., 1991 ). The major components of this system include the proteins KinA, SpoOF and SpoOB (Figure 1.3). KinA is an autophosphorylating transmitter kinase; however, KinA does not appear to be an integral membrane protein that senses external stimuli (Perego et al., 1989). SpoOF is the principal substrate for phosphorylation by KinA (Burbulys et al., 1991; Perego et al., 1989). Although SpoOF is another response regulator-like protein, and contains a phosphate-accepting domain, it lacks a DNA-binding regulatory domain (Trach et al., 1985). SpoOB appears to be a phosphotransferase that mediates the relocation of the phosphate group from spoOF to SpoOA. The phosphotransferase reaction carried out by SpoOB is reversible, and SpoOB can therefore also specifically remove the SpoOA phosphate group (Burbulys et al., 1991). The eventual activation of SpoOA appears to be the decisive step in the initiation of sporulation. Whereas activated SpoOA can directly induce sporulation-specific genes, perhaps its most important effect is the inhibition of transcription of aer (Perego et al., 1988; Zuber and Losick, 1987). The Aer protein binds cooperatively to DNA and represses many operons that are induced at the onset of stationary growth, including some required for Peptide Signals 7 OtherSignals? ® ,WJ ><- *\ < 036 SpoOA SpoOA-F \ Cell Cycle Signals 7 7 V / onal activator repressor srfA transcription Figure 1.3 Diagram showing the proposed pathways of regulation for the activation of SpoOA and ComA. Adapted from Burbulys et al., 1991. 10 sporulation (Strauch et al., 1989). Binding by Aer reinforces other regulatory mechanisms to prevent the onset of sporulation during vegetative growth. By repressing the transcription of aer, SpoOA effectively amplifies its regulatory domain by derepressing operons that are controlled by Aer. The final step in the initiation of sporulation is the post-transcriptional activation of the earliest sporulation-specific sigma factor, a“ (Healy et al., 1991), encoded by spoOH (Carter and Moran, 1986; Dubnau et al., 1988; Zuber et al., 1989). a" directs the initiation of transcription of i) the sporulation-specific genes that are required to initiate septum formation as well as ii) a sigma factor, 0", that is needed for the following stages of development (Wu et al., 1989). The increase in activity of 0” appears to result from both increased stability of mRNA and decreased turnover of the a"-protein (Healy et al., 1991). It is not known how the activation of SpoOA accomplishes the increase in activity of a“. Most response regulators are directly phosphorylated by their cognate kinase (Stock et al., 1989, 1990); the numerous intermediates between KinA and SpoOA provide possible regulatory points at which multiple environmental signals may be integrated to control the initiation of sporulation. In some spoOA missense mutants, and in strains that overexpress KinA, the multicomponent phosphorelay is bypassed (Perego et al., 1989; Speigelman et al., 1990). It has been suggested that the presence of the multicomponent phosphorelay system in Bacillus reflects the "seriousness of the commitment to shut down cell division and initiate sporulation" (Burbulys et al., 1991). SpoOE appears to inhibit the activation of SpoOA by negatively regulating the phosphorelay 11 system (Perego et al., 1991a; Burbulys et al., 1991). Overexpression of spoOE inhibits sporulation while most mutations within spoOE cause hypersporulation (Perego et al., 1991a). However, some spoOE nonsense mutations inhibit sporulation, a phenomenon that is explicable if the carboxyl-terminus of the SpoOE protein can relieve an inhibition of sporulation mediated by the amino- terminus. Since all spoOA missense mutations that bypass the phosphorelay components, and some spoOA non-bypass missense mutations (Olmedo et al., 1990), can relieve the inhibition caused by the spoOE nonsense mutants, SpoOE appears to exert its control within the phosphorelay pathway, possibly by affecting the transfer of phosphate from SpoOB to SpoOA (Burbulys et al., 1991). Another protein that may regulate the phosphorelay pathway is Obg, which is encoded in the same operon as SpoOB (Ferrari et al., 1985; Trach and Hoch, 1989). Obg is a GTP-binding protein that is required for vegetative growth. Obg, like G-proteins, displays GTPase activity (Fohger and Hoch, unpublished; cited in Burbulys et al., 1991). It has been proposed that Obg may affect the phosphotransferase activity of SpoOB in response to the levels of GTP. Growth conditions that lower endogenous levels of GTP (reviewed by Freese et al., 1985) or direct lowering of levels of GTP by the addition of the drug decoyinine (Mitani et al., 1977), which inhibits GMP synthetase, can induce sporulation. Therefore, Obg-GTP may block SpoOB phosphotransferase activity while Obg-GDP stimulates SpoOB and allows sporulation to begin. Early experiments showed that sporulation is limited to a short phase of the cell cycle (Mandelstam and Higgs, 1974). Therefore, the similarity of Obg to the Ras 12 proteins, which regulate the cell cycle of eucaryotes, suggests that Obg may respond to cell cycle signals (Trach and Hoch, 1989). It is interesting that even when cells are grown in enriched media, lowering of the endogenous levels of GTP can initiate sporulation (Freese et al., 1985; Mitani et al., 1977), suggesting that GTP levels play the dominant role in signalling the onset of stationary growth. However, addition of decoyinine, which specifically inhibits GTP synthesis, does not induce the other late growth responses (i.e., competence development and production of antibiotics; Freese, 1981); therefore, divergent pathways may induce these responses. Additionally, the induction of sporulation by either nutrient stepdown or decoyinine is not very effective before cultures have achieved a high cell density, suggesting that an additional factor is involved in the induction of sporulation (Grossman and Losick, 1988). This matter will be discussed below. Induction of competence development also utilizes a signal transduction pathway. The srfA operon encodes the structural proteins required for the production of the cyclic lipopeptide antibiotic surfactin (Nakano et al., 1991 a). Induction of srfA occurs at the onset of stationary phase, as does the induction of sporulation, and usually requires comA, comP and spoOK (Figure 1.3; Hahn and Dubnau, 1991; Nakano et al., 1991b; Nakano and Zuber, 1991). ComA is a response regulator protein (Weinrauch et al., 1989, 1990) that, when activated by phosphorylation, binds to specific sites upstream from the srfA promoter to activate srfA transcription (Roggiani and Dubnau, unpublished; cited in Marahiel et al., 1993). ComP shows sequence similarity to the histidine 13 kinase signal transmitters; ComA is the primary substrate for phosphorylation (Weinrauch et al., 1990). ComP-independent induction of srfA can occur, but still requires ComA and spoOK as well as SpoOA, SpoOB, SpoOF and KinA. ComP-independent induction of srfA is independent of Aer function (Marahiel et al., 1993); it is unknown how ComP-independent induction of srfA occurs. The induced expression of srfA has been determined to be the primary function of ComA and ComP in competence development. Placing srfA under the control of an lPTG-inducible promoter allows for lPTG-induced competence development, even during exponential cell growth (Hahn and Dubnau, 1991; Nakano and Zuber, 1991). Another factor that is required for competence development and efficient sporulation is encoded by sfp (Nakano et al., 1992). Both inactivation and overexpression of sfp lead to reduced transcription of srfA and to a reduction in competence and sporulation. It follows that normal expression of srfA requires a specific level of expression of sfp. Some other genes that are induced late in growth, including a gene that encodes a protein similar to nitrite reductase, also require sfp for induced expression (Marahiel et al., 1993). That sfp strains can be complemented by growth in the presence of sfp+ strains suggests that Sfp is involved in the regulation or production of an extracellular factor. This observation is highly intriguing because evidence is accumulating that extracellular peptidic factors, produced by cells that have grown to a high density, are required for both efficient sporulation (Grossman and Losick, 1988) and induction of srfA (Magnusson and Grossman, unpublished; cited in Marahiel 14 et al., 1993). Furthermore, the spoOK operon, which is required for efficient initiation of sporulation as well as for the induction of srfA and competence development, has been found to contain five genes that encode proteins highly similar to oligopeptide permeases (Perego et al., 1991; Rudner et al., 1991). Surprisingly, disrupting the last gene of the spoOK operon does not affect sporulation but does prevent the development of competence (Rudner et al., 1991). Therefore, spoOK appears to play a role in sensing extracellular factors and may either sense different signal molecules or affect different signal transduction pathways that are used for efficient initiation of sporulation and competence development. The observation that overexpression of kinA alleviates the effect of spoOK mutations on sporulation lends support to the possibility that spoOK gene products may regulate the activity of KinA (Figure 1.3; Rudner et al., 1991). Although numerous factors appear to be required for the initiation of sporulation in Bacillus, nutrient limitation is evidently the major catalyzer (Freese et al., 1985), but how is nutrient limitation sensed? The molecule ppGpp has long been considered a signal for the onset of metabolic stress or stationary growth. In E. call, the cellular levels of ppGpp are inversely correlated with the growth rate (Lazzarini et al., 1971). It is generally considered that ppGpp synthesis is caused by uncharged tRNAs during amino acid starvation, a process that is referred to as the stringent response. The stringent response is mediated by the relA locus (Dennis, 1977; Dennis and Nomura, 1974); however, both E. coli (Atherly, 1979) and Bacillus (Fehr and Richter, 1981; 15 Price and Brown, 1981) also show reIA-independent ppGpp synthesis in response to carbon or energy limitation. The levels of ppGpp have been shown to increase in B. subtilis in response to nutrient limitation (Lapez et al., 1991). ppGpp inhibits the activity of IMP dehydrogenase, which converts IMP to GMP. Therefore, ppGpp may directly regulate the intercellular levels of GTP; alternatively, ppGpp may have a more direct role in the initiation of sporulation. Recent evidence suggests that KinA activity is repressed by free cis- unsaturated fatty acids (Strauch et al., 1992). In E. coli, the stringent response blocks an early step in the synthesis of fatty acids (Lueking and Goldfine, 1975; Nunn and Cronan, 1976; Spencer et al., 1977). If ppGpp also blocks fatty acid synthesis in B. subtilis, then inhibition of fatty acid production by ppGpp could derepress the activity of KinA. It is curious that lowering GTP levels with decoyinine does not induce competence development or production of antibiotics. Perhaps ppGpp is required to induce these late growth responses, and directly lowering GTP levels with decoyinine prevents both reIA-dependent and -independent ppGpp synthesis. It is apparent that extracellular peptide-like factors also have some control over both sporulation and competence development (Grossman and Losick, 1988). The exact nature of these factors and how they influence differentiation is, at present, unknown. 16 Streptomyces In the Streptomyces, increases in ppGpp levels also occur at the onset of the stationary or slow growth phase, and precede the induction of antibiotic production and sporulation. A decrease in the GTP pool also precedes and may also initiate differentiation in some species of Streptomyces. As in Bacillus, decoyinine can induce sporulation in the Streptomyces (reviewed by Ochi, 1988). Sporulation deficient mutants of some species of Streptomyces can be rescued by the addition of an exogenous factor. In S. griseus, the factor was found to be 2-isocapryloyl-3R-hydroxymethyl-r-butyrolactone (A-factor) (Khokhlov, 1985). Following the inception of nutrient limitation, A-factor is produced rapidly and diffuses from the cells, and thus appears to play an integral role in the initiation of sporulation. A receptor specific for A-factor has been identified as a soluble protein that represses antibiotic production and sporulation (Miyake et al., 1989, 1990). Binding of A-factor relieves the repression, as do mutations in the gene that encodes the A-factor—binding protein (Miyake et al., 1990). The regulation of biosynthesis of A-factor and antibiotic production share a feature with Bacillus sporulation in that the phosphorylation of DNA-binding proteins, such as Ast (Hong et al., 1991; Horinouchi and Beppu, 1992; Horinouchi et al., 1990), is required for the production of A-factor and antibiotics (Hara et al., 1983; Horinouchi et al., 1983; Stein and Cohen, 1989). Overexpression of Ast suppresses'the phenotype of a mutation in another 17 gene required for antibiotic production, afsB (Horinouchi and Beppu, 1984; Horinouchi et al., 1983; Stein and Cohen, 1989). Therefore, sensor/transmitter kinases and response regulator proteins may also be required for the initiation of sporulation in the Streptomyces. At least seven genetic loci have been identified, mutations in which block the formation of aerial mycelia in S. coelicolor. These mutants have been called bald (bld) because colonies lack the furry appearance produced by aerial mycelia. All of the bld mutations except bldC also adversely affect antibiotic production. Sporulation in most of the bld mutants is conditional, depending upon the carbon source that is supplied to the hyphae, whereas the defect in antibiotic production is not affected by the carbon source (Chater, 1989). Generally, the presence of glucose is required for expression of the Bid phenotype. Therefore, a catabolite repression mechanism may be involved in the conditional nature of the Bid mutants. The bldC mutants, like A-factor mutants, can be induced to sporulate in the presence of an unknown exogenous factor produced by sporulating colonies (Chater, 1989). Surprisingly, the entire pathway for the initiation of sporulation appears to be bypassed by the overexpression of whiG (Chater et al., 1989). The WhiG protein is a sigma factor whose overexpression causes formation of aerial mycelia and spores under typically vegetative conditions. Overexpression of WhiG also causes spores to form within the substrate mycelia, apparently circumventing the requirement for formation of aerial mycelia. If WhiG is inactivated, sporulation is completely blocked, although aerial mycelia still form 18 (Mendez and Chater, 1987). Myxococcus xanthus In M. xanthus, nutrient limitation is again correlated with an increase in the cellular levels of ppGpp (Manoil and Kaiser, 1980), although very little is known about any signal transduction pathway that may initiate the developmental response. It has been found that high concentrations of glycerol (0.5M) can effectively induce sporulation (Dworkin and Gibson, 1964). Similar levels of glycerol are found to inhibit the methylation of the methyl-accepting chemotaxis proteins that regulate chemotaxis in E. coli (Oosawa and lmae, 1983), which suggests that a signal transduction pathway may be required for the initiation of sporulation in Myxococcus. Accordingly, a gene encoding a serine/threonine kinase has been shown to be required for the initiation of sporulation (Munoz- Dorado et al., 1991; 1993). In fact, as many as 26 putative serine/threonine kinases, now identified in M. xanthus, may control a variety of environmental and developmental responses. Besides nutrient limitation, a high cell density (ca. 5 x 10° cells ml") is required for initiation of sporulation (reviewed by Shimkets, 1990). The cell density requirement is related to the observation that extracellular factors are required for all of the stages of development (Hagen et al., 1978; Kroos and Kaiser, 1987; Kuspa et al., 1986). The requirement for extracellular factors became apparent when four classes of conditional mutants were found. A mutant from one class was able to complement any of the other classes of 19 mutants in co-culture, and wild type cells could complement mutants from each class (Hagen et al., 1978). The complementation was shown to be due to the exchange of extracellular factors and not to genetic recombination (LaRossa et al., 1983). The genetic loci for the complementation groups have been designated asg, bsg, csg and dsg for A, B, C and D signalling. The gene bsgA has been found to affect the expression of genes that are normally induced at the onset of, or shortly after, the initiation of differentiation (Figure 1.4; Gill and Cull, 1986; Kroos and Kaiser, 1987). The sequence of the protein encoded by the bsgA gene is similar to the sequence of ATP-dependent proteases (Gill et al., 1993), such as La from E. coli (Amerik et al., 1988). In E. coli, the lan- encoded protein, La, activates transcription of genes by degrading the LexA repressor protein (reviewed by Goldberg, 1992). A second-site mutation has been isolated that alleviates the effect of the bsgA mutation and also causes the premature expression of genes that are normally expressed early in development (Gill et al., 1993), suggesting that BsgA may have a function similar to La. Since BsgA is an intracellular protein, it may control the synthesis of the extracellular factor involved in B-signalling. Because it had previously been suggested that the cell density signal required to initiate sporulation is an adenosine-containing factor (Shimkets and Dworkin. 1979), it is interesting that what appears to be an ATP-dependent protease is required early in differentiation. The other early-acting extracellular signalling factor, the A factor, is required for the induction of genes that are expressed approximately one hour after 20 I I l T 0 5 10 15 20 25 hours - — ——’ A A A ’ Starve Aggregate Sporulate Mutant Class s B l ‘5 .2 A I ' ' 0 U .L abnormal - g o - a - Figure 1.4 Classes of developmental mutants of Myxococcus xanthus. The thin lines below the time line indicate the periods during which distinguishable phases of development take place. The wide grey horizontal bars show the extent of development that occurs prior to the expression of the developmental defect. The columns to the right indicate the phenotypic defects with respect to aggregation and sporulation. Adapted from Kaiser and Losick, 1993. 21 development has begun (Kroos and Kaiser, 1987; Kroos et al., 1986; Kuspa et al., 1986). Half of the A factor activity has been attributed to extracellular proteases that can be functionally replaced by trypsin or other proteases (Plamann et al., 1992). The other half of the A factor activity has been attributed to a mixture of eight amino acids (Tyr, Pro, Phe, Trp, Leu, lle, Ala and Ser; [Kuspa et al., 1992a]). The amount of A factor released into the medium is proportional to the cell density, implicating A factor as a cell density signal (Kuspa et al., 1992b). Suppressor mutations have been isolated that allow induction of some A-dependent genes in an asg background. Because the suppressor mutations are recessive to the wild-type allele, they are thought to be the result of inactivation rather than enhancement of function (Kaplan et al., 1991). Therefore, the inactivated suppressor genes may encode negative regulators of the A-dependent genes examined. However, the suppressor mutations did not restore fruiting body formation or sporulation in the asg mutants, suggesting that not all A-dependent genes are regulated in a similar manner (Kaplan et al., 1991). Anabaena The initiation of heterocyst differentiation in Anabaena is also poorly understood. There are conflicting reports about whether ppGpp accumulates in filaments of Anabaena that are deprived of fixed nitrogen (Adams et al. 1977; R‘ogerson et al., 1978; Akinyanju and Smith, 1979). However, the level of cAMP, another molecule known to stimulate signal transduction pathways, 22 does increase following nitrogen stepdown (Hood et al., 1979) and other environmental perturbations (Ohmori, 1989). Furthermore, the addition of exogenous cAMP can induce formation of multiple contiguous heterocysts (Smith and Ownby, 1981). Because the activity of adenylate cyclase, which produces cAMP, is regulated by intracellular calcium ion concentrations (Bianchini et al., 1990), calcium ion concentrations may play a role in regulating the initiation of heterocyst differentiation. The NtcA protein, found in a unicellular cyanobacterium, is similar in sequence to CAMP-binding proteins and is required for the regulation of genes that are involved in nitrogen metabolism (Vega-Palas et al., 1990, 1992). NtcA apparently binds to specific DNA sequences in the promoters of regulated genes (Flores et al., 1993). A DNA-binding protein, BifA, found in Anabaena, is also similar to CAMP-binding bacterial regulatory proteins and is 77% identical to NtcA (Frias et al., 1993; Wei et al., 1993). The binding sites that have been characterized for BifA are near genes that are expressed during vegetative growth or late in heterocyst development, and inactivation of bifA prevents growth on nitrate as well as on N2 (Golden et al., personal communication). Transcriptional reporters within the nitrate reductase operon of Anabaena show an induction of transcription within 20 minutes following removal of nitrate from the medium (Wolk et al., 1991; Y. Cal, personal communication). The rapid induction of transcription suggests that sensory responses to the removal of nitrate from the medium may be mediated by signal transduction 23 pathways. Rapid changes in patterns of protein phosphorylation following nitrogen deprivation (Mann et al., 1991) are also consistent with the supposition that signal transduction pathways are activated by removal of fixed nitrogen from the medium. Characterization of the patA gene of Anabaena provided evidence that a signal transduction pathway also affects the positioning of heterocysts. Mutations within patA cause heterocysts to develop almost exclusively at terminal cells in the filament. The patA gene encodes a protein that resembles CheY (Liang et al., 1992), a phosphate-accepting response regulator protein that controls chemotaxis in E. coli. Although CheY does not bind to DNA but instead activates the flagellar motor apparatus (Matsumura et al., 1984), it has been proposed that the PatA protein is a transcriptional regulator (Liang et al., 1 992). The clearest example of early developmental control in Anabaena comes from studies on the hetl? gene. Inactivation of hetR blocks evident heterocyst differentiation, whereas multiple copies of hetR induce heterocyst formation in nitrogen-containing media and multiple contiguous heterocysts in Nz-grown filaments (Black et al., 1993; Buikema and Haselkorn, 1991a, 1991b). Although the behavior of hetl? is analogous to that described for WhiG in the Streptomyces (Chater et al., 1989), HetR bears no significant similarity to other characterized proteins and contains no previously identified DNA-binding motifs. A point mutation in hetl? that converts a serine codon to an asparagine codon causes a loss of HetR function (Buikema and Haselkorn, 1991a). Conceivably, 24 the serine residue could be a phosphate acceptor. The recent identification of a tyrosine/serine phosphatase in another filamentous, heterocyst-forming cyanobacterium, Nastoc commune (Potts et al., 1993), enhances the possibility that HetR is phosphorylated at a serine residue. Data presented in this dissertation (and in Black et al., 1993) show that hetl? is induced within 2 hours following the removal of nitrate from the medium, and that an intact copy of hetl? is required for the induction. hetR is the earliest induced gene known to be involved in heterocyst differentiation, and also regulates its own expression. Following 3.5 hours of nitrogen deprivation, only small groups of cells show enhanced levels of hetR expression. Therefore, the positions within the filaments where hetl? is induced may be preconstrained or, alternatively, may be the result of regulation that acts during the induction of hetR. The temporal and spatial regulation of hetl? may be directly related to the mechanism(s) of pattern formation because: i) the transcription of her}? is induced within two hours following the removal of nitrate from the medium, long before proheterocysts can be distinguished by microscopy, but at about the same time at which the development of proheterocysts can no longer be suppressed by the addition of ammonia and certain inhibitors (Bradley and Carr, 1977); ii) hetl? appears to become fully induced only within cells that are to differentiate into heterocysts; and iii) hetR is required in order for heterocysts to form. Hence, the induction of transcription of hetR may mark the initiation of differentiation. A possible role for hetl? in pattern formation will be discussed further in the following section of the introduction and in Chapter 2. 25 The production of, and requirement for, extracellular factors during the initiation of development of B. subtilis, Streptomyces and M. xanthus implies that a selective advantage is gained by the intrapopulation coordination of development. An extracellular factor has also been shown to stimulate akinete formation in the filamentous cyanobacterium, Cylindrospermum licheniforme (Fisher and Walk, 1976), whereas no such factor has been found to affect heterocyst differentiation. In contrast to the cell density requirement for development displayed by B. subtilis and M. xanthus, the only cell density effect on heterocyst differentiation is a decrease in heterocyst frequency in dense cultures where light limitation causes a slower rate of growth. However, filaments of Anabaena require a certain minimum number of vegetative cells before heterocysts can develop (Wilcox et al., 1973b; Wolk, 1967). Moreover, mutants that grow in nitrogen-containing media as short filaments, and other mutants that fragment following nitrogen stepdown, fail to form heterocysts (Buikema and Haselkorn, 1991; Ernst et al., 1992). In some cases, these mutants show no induction of hepA (my unpublished observations), a gene that is normally induced within differentiating cells ca. 7 hours after the initiation of nitrogen deprivation (Holland and Walk, 1990; Wolk et al., 1993). It is not known whether the requirement for a minimum filament length occurs prior to, or following, the initiation of differentiation. It is possible that there may be a relationship between filament length and the expression of hetR. Since a very short filament does not differentiate, if hetR is not induced in very short filaments, then a requirement for a minimum 26 filament length for differentiation may be expressed prior to the induction of hetR, possibly prior to the initiation of differentiation. On the other hand, if the timing and level of induction of hetl? are the same in short filaments as they are in long filaments, then i) a requirement for a minimum filament length for differentiation is expressed following the induction of hetR, or may be expressed prior to the induction of hetl? but does not affect the expression of hetR; and ii) the activation of hetR, alone, is not sufficient to induce heterocyst formation. It is also possible that within short filaments, hetR would be initially induced, but is then either re-repressed, or never achieves the level of induction that is obtained in long filaments. In this case, it is possible that hetl? has a direct role in preventing differentiation in short filaments. If hetl? is not induced in short filaments, the same mechanism(s) that spatially restricts the expression of hetR, perhaps the production of an inhibitor of differentiation (see below), may also prevent differentiation in short filaments. Therefore, unlike M. xanthus, S. griseus, and B. subtilis, which produce extracellular cell density signals that induce differentiation (e.g., A signal, A-factor, and possibly peptides, respectively), Anabaena may produce an intercellular signal that inhibits differentiation. Chapter 3 presents evidence that Anabaena may synthesize a metabolite that can suppress heterocyst differentiation and may regulate heterocyst spacing. Furthermore, a cell that has visibly begun to differentiate, forming a proheterocyst, may regress to a Vegetative state if a cell that is very close to the proheterocyst is destroyed (Wilcox et al., 1973b). The regression of the proheterocyst may be a 27 consequence of wounding of the filament (Wolk and Ouine, 1975) or, alternatively, sensing of filament length may control differentiation beyond the time of initiation of differentiation, a premise that is consistent with the notion that sensing itself occurs after initiation. The reversibility of the early stages of heterocyst differentiation may also be important for controlling the positions at which heterocysts differentiate, as will be discussed in the following section, and is in contrast to the irreversibility of sporulation in B. subtilis once sporulation is initiated by the activation of SpoOA. Another recently characterized mutation, P6, also blocks the evident differentiation of heterocysts (Ernst et al., 1992). The presence of the wild type gene, hetP, on a replicating plasmid causes multiple clustered heterocysts to form when filaments are grown on N, (F. Fernandez-Pines, personal communication). The function of the protein encoded by hetP is unknown. Whereas HetR has been found to be required for both heterocyst and akinete differentiation in Nastoc ellipsosporum, HetP is required only for heterocyst differentiation (Leganés et al., personal communication). Therefore, if HetR and HetP act within the same pathway of developmental regulation, HetP must act after HetR, and must be within a branch of the pathway that is specific to the regulation of heterocyst differentiation, whereas HetR would act prior to a branchpoint that separates the regulation of development of heterocysts and of akinetes. 28 PROCESSES LEADING TO LOCALIZED GENE EXPRESSION DURING DEVELOPMENT B. suba’lis The ordered expression of genes required for each stage of development in B. subtilis is controlled by the activities of a series of sigma factors (reviewed by Stragier and Losick, 1990). Sigma factors are involved in the recognition of specific promoter sequences in DNA and associate with the core RNA polymerase to initiate transcription (McClure, 1985). Sigma factors usually contain both a conserved region that interacts with RNA polymerase and a unique region within the DNA-binding domain. Because of their conserved region, it has been possible to identify many sigma factors by hybridization with heterologous sigma-factor DNA probes. The ordered appearance of activity of the various sigma factors in B. subtilis is controlled transcriptionally and by other regulatory mechanisms, including genomic rearrangement and post- translational modification. The activation of each sigma factor relies, at least in part, on the activity of the preceding sigma factor. As mentioned in the previous section,"Processes involved in the initiation of differentiation. B. subtilis", the increased activity of the initiating sigma factor, 0”, involves a stabilization of the corresponding mRNA and of the protein itself (Healy et al., 1991); the resulting increase in the level of 0“ leads to an induction of genes Whose promoters are recognized by a“. a“ directs the enhanced transcription of the ftsAZ operon (Beal et al., 1988), which is required for septum formation 29 (Beal and Lutkenhaus, 1991), and of the spa/IA operon (Figure 1.5), which contains the spa/[AC gene that encodes 0" (Figure 1.5; Wu et al., 1989). UP is required for the completion of septation. Recent results suggest that active 0‘ (shown as 0'" in Figure 1.5) is restricted to the forespore even though the spa/IA operon is transcribed in the predivisional cell (Gholamhoseinian and Piggot, 1989; Losick and Stragier, 1992) and the 0‘ protein is also present in the mother cell. The regulation of 0" is thought to be carried out by the promoter-proximal products of the spa/IA operon, SpollAA and SpollAB (Partridge et al., 1991; Schmidt et al., 1990). Genetic experiments indicate that SpollAB inhibits 0" activity and that SpollAA relieves the inhibition in a regulatable manner. Precisely how 0‘ is regulated is unknown. Transcription of the spa/l6 operon, in which spa/[GB encodes 0‘E (Figure 1.5; Stragier et al., 1984; Trempy et al., 1985), is initiated by the vegetative sigma factor, a‘, and also occurs in the predivisional cell (Gholamhoseinian and Piggot, 1989). The 0‘ protein is initially found as an inactive precursor called Pro-0*E (LaBell et al., 1987). Processing of Pro-0‘IE appears to require the product of the first gene in the spa/l6 operon, SpollGA, that encodes a protease (Jonas et al., 1988; Stragier et al., 1988). The activity of SpollGA, and consequently 0‘, has been proposed to require the completion of septation, and activation of (TE is restricted to the mother cell compartment (Stragier et al., 1988; Trempy et al., 1985). Such a regulatory network makes the temporal and spatial regulation of 0F- and (IE-dependent genes reliant upon the establishment of septation. 30 Figure 1.5 Diagram showing the sequential expression and activation of the various sigmas factors, and their transcriptional dependencies, required for sporulation in Bacillus subtilis. The arrows indicate either the action or the synthesis of sigma factors. Active 0‘F is denoted as a". Adapted from Stragier and Losick (1990) and Kaiser and Losick (1993). 32 The expression of spell/D is directed by 0‘; for that reason, SpolllD activity is limited to the mother cell compartment (Kroos et al., 1989; Kunkel et al., 1989; Stevens and Errington, 1990). SpolllD is a DNA-binding protein that regulates the expression of many late genes that are active in the mother cell, including the mother cell specific sigma factor, 0". 0" is encoded by spolVCB and spell/C, which are separated by a 42-kb DNA element called ’skin’. The 'skin’ element is excised exclusively from the mother cell chromosome and forms a circular, extrachromosomal element (Kunkel et al., 1990; Stragier et al., 1989). The excision is carried out by a site-specific recombinase encoded within ’skin’ (Kunkel et al., 1990; Sam et al., 1990) and requires SpolllD. However, the genomic rearrangement appears to be a redundant regulatory scheme: strains that lack the 'skin' insertion show proper regulation of UK expression (Kunkel et al., 1990). SpolllD is also required for the a‘E-directed transcription of the spliced spolVCB/spolllC gene (Kunkel et al., 1989). Like 0‘, a" is produced as an inactive precursor, Pro-0", which is activated by processing (Kroos et al., 1989; Lu et al., 1990). Processing of Pro-0'K requires the activity of SpolVFB, which is negatively affected by BofA and SpolVFA; all of these proteins are present within the mother cell or outer forespore membrane (Figure 1.5; Cutting et al., 1990, 1991b; Ricca et al., 1992). The activation of 0“ is stimulated by the protein encoded by the spa/VB gene, which is expressed only within the forespore. Transcription of spa/VB is limited to the forespore because its transcription requires the forespore-specific sigma factor, 0‘6 (Cutting et al., 1991a). SpolVB is made in the forespore and then acts, 33 directly or indirectly, across the forespore membrane to activate the product of the spa/VFB gene. The activated SpolVFB product then, directly or indirectly, controls the processing of Pro-0“ to an active form (Figure 1.5; Cutting et al., 1990, 1991a; Lu et al., 1990). Once activated, 0" can direct the initiation of its own transcription, thereby amplifying its expression within the mother cell. 0" also directs the transcription of gerE, which encodes a DNA-binding protein that acts in conjunction with 0“ to activate transcription of late mother cell- specific genes (Kroos et al., 1989; Zheng and Losick, 1990). The mechanism that leads to the initiation of a“ transcription in the forespore is still not clear, but once a6 is activated it is capable of directing its own transcription to amplify its presence within the forespore (Karmazyn- Campelli et al., 1989). The promoter for the a“ gene is recognized by active 0", so regulation of 0‘ activity by the SpollAA/SpollAB system may control the time at which 0G is expressed (Partridge et al., 1991; Schmidt et al., 1990; Sun et al., 1991), while the activity of 0° may also be held in check by the SpollAA/SpollAB system that also regulates 0‘ activity (Figure 1.5; Rather et al., 1990; Schmidt et al., 1990; Stragier, 1991). The mechanisms that regulate the activity of 0G and 0" coordinate forespore and mother cell development. The development of competence in Bacillus is not well understood. As mentioned in the previous section, the expression of the surfactin synthetase operon appears to be the sole or major requirement for competence development (Hahn and Dubnau, 1991; Nakano and Zuber, 1991). Surfactin is composed of seven amino acids and a B-hydroxy fatty acid moiety (Vater, 34 1989). The constituent amino acids are activated by adenylylation to form acyl adenylate intermediates. Each aminoacyl adenylate is covalently bound to the enzyme complex via a carboxyl thioester linkage to a cofactor, apparently 4’- phosphopantetheine (Kleinkauf et al., 1970; Kleinkauf and von Ddhren, 1990; Lipmann, 1980). Each functional domain that adds a single constituent amino acid contains within a highly conserved region a serine residue to which the cofactor is covalently attached (D’Sousa et al., 1993). The attached cofactors act as arms that shuttle the growing peptide chain to the various enzymatic domains that carry out the sequential condensations of the constituent amino acids. The synthesis is similar to syntheses carried out by multifunctional (Type 1) fatty acid and polyketide synthases that shuttle growing acyl chains through multiple domains that contain 4'-phosphopantetheine attachment sites and carry out sequential specific chain elongations and modifications (Hopwood and Sherman, 1990). The surfactin synthetase operon, srfA, is composed of four genes. The first three genes in the srfA operon encode multifunctional proteins, each of which contains separate, linearly arranged domains; each domain incorporates a single constituent amino acid. The fourth gene encodes a thioesterase-like protein that has been proposed to release the completed lipopeptide product (Marahiel et al., 1993). The srfAA gene encodes the protein that initiates surfactin biosynthesis with a glutamate residue, to which the fatty acyl moiety is added Water, 1989). The other domains of SrfAA catalyse the incorporation of L- leucine and b-Ieucine. SrfAB catalyses the addition of valine, asparagine and D- 35 leucine, and SrfAC completes surfactin synthesis by adding a final L-leucine residue. Converting the cofactor-binding serine residue to alanine in any one of the domains blocks the further synthesis of surfactin. However, none of the serine to alanine conversions blocks competence development (D’Sousa et al., 1993). Thus, despite the requirement for expression of the surfactin synthetase operon, surfactin itself is not required for cells to gain competence. However, other evidence suggests that some activity associated with the valine-activating domain of srfAB is required for competence development (van Sinderen et al., 1993). What the activity is and how it induces competence development is still unknown. Streptomyces Spatial control of gene expression and pattern formation in the Streptomyces has not received a great deal of attention. All of the studies mentioned in this section were performed on S. coelicolor which, because it is the species of Streptomyces most amenable to genetic manipulation, has been the favored organism for developmental studies. Promoters that are turned on in a temporally and spatially controlled manner were found by cloning random fragments of chromosomal DNA upstream from the luciferase genes, luxAB (Schauer et al., 1988; Schauer, 1988; Schauer and lm, 1988). However, these promoter fragments have not been further characterized. The sapA gene was identified using reverse genetics on the spore associated proteins (Saps) that are abundantly produced in aerial mycelia and are associated with the spore 36 coat. sapA-luxAB fusions show that the sapA gene is expressed only in the aerial mycelia, demonstrating that spatially regulated gene expression does occur during sporulation in Streptomyces (Guijarro et al., 1988). An 18-residue peptide, called SapB, has also been identified (Willey et al., 1991). SapB, however, appears to be produced by a non-ribosomal mechanism (Willey et al., 1993), probably using a thiotemplate mechanism similar to that used by surfactin synthetase and other peptide antibiotic synthetases. Extracellular addition of SapB to many bld mutants rescued their ability to produce aerial mycelia. In fact, it was found that coculture of pairs of different bld mutants, which individually are incapable of producing SapB, rescued SapB production and the formation of aerial mycelia in one member of the pair (Willey et al., 1993). Only one member of the pair was rescued, and the aerial mycelia formed by the rescued cells extended toward the donor-cell colony. In addition, cells that acted as a donor in one pairing could become a responder in a different pairing. The complementation analysis led to the proposal that at least four separate signal molecules are involved in the extracellular complementation and regulation of SapB expression. It has been proposed that SapB acts as a morphogenic component that reduces surface tension, thereby allowing aerial growth (Willey et al., 1991, 1993). A unique regulatory phenomenon seen in the Streptomyces was discovered by the analysis of bldA mutants. The gene affected in these mutants was found to encode a tRNA for the rare UUA leucine codon (Lawlor et al., 1987). A majority of the genes found to contain a UUA codon are directly involved in 37 regulating antibiotic production and sporulation (Chater, 1989; Wright and Bibb, 1992). It was recently shown that the low levels of the UUA-tRNA in vegetative cells are a result of inefficient processing of the 5' end of the primary transcript and not of transcriptional regulation (Leskiw et al., 1993). However, other recent work suggests that the bldA gene does not play a central role in regulating antibiotic production (Gramajo et al., 1993), and the fact that the bldA mutant phenotype is only expressed in certain media suggests that bldA has only a limited role in regulating sporulation. The whi mutations block spore formation and the production of a grey pigment encoded by the whiE locus (Davis and Chater, 1990). The whi designation was given to describe what were therefore white colonies. The strong effects associated with the overproduction of the WhiG protein (see above under ”Processes involved in the initiation of differentiation. Streptomyces") suggest that the localization of sigma-WhiG expression may be the primary contribution of earlier regulatory factors, and is required for spatially regulated expression of sporulation-associated genes (Chater et al., 1989). The role of sigma-WhiG in controlling differentiation may be different than that expressed by the cascade of sigma factors associated with sporulation in B. subtilis. Early expression of late sigma factors in B. subtilis does not usually cause sporulation to occur in vegetative cells, nor does it lead to defective spatial regulation. Other sigma factors have been identified in the Streptomyces both biochemically and by heterologous hybridization (Buttner, 1989; Tanaka et al., 1988; Westpheling et al., 1985), and promoter switching has been found 38 for a number of genes during development (Buttner et al., 1990). Nonetheless, no developmental role has been found for these other sigma factors. M. xanthus The production and exchange of extracellular factors is required for each stage of morphological differentiation of Myxococcus xanthus (Hagen et al., 1978). However, the cells must also be motile in order to aggregate, form fruiting bodies, and sporulate (Kim and Kaiser, 1990b,c; Kroos et al., 1988; Sager and Kaiser, 1993). There are two types of motility systems, called A (for adventurous) and S (for social); each has unique genetic loci. Additional loci, denoted mgl (for mutual gliding), are required for all motility (Hodgkin and Kaiser, 1977, 1979). Most mutations within the S system adversely affect development, while mutations in the A system were initially shown to have little effect on development. However, strains that carry mutations within two different A system loci -- loci that are induced by dissimilar signals -- fail to sporulate (Kalos and Zissler, 1990). Therefore, the A motility system appears to be essential for differentiation, and may have multiple pathways for stimulation. The best characterized signal in M. xanthus is the C-signal, which shows the clearest relationship between cell motility and development. The C factor is a 17-kD protein, encoded by the csgA gene, that is present on the cell surface (Hagen and Shimkets, 1990; Kim and Kaiser, 1990a; Shimkets and Rafiee, 1990). lmmunogold labeling showed that between 1000 and 2000 molecules 39 of C factor are present on the surface of a single cell (Shimkets and Rafiee, 1990). csgA mutants fail to form fruiting bodies or sporulate, a phenotype that is similar to the phenotype of non-motile (mgl) mutants (Figure 1.4; Kroos et al., 1988). Furthermore, csgA mutants and non-motile strains both fail to induce developmental genes that are normally expressed after six hours into development. These data suggested that motility is required for C factor to function properly. Subsequent experiments showed that motility is required for both donation and response to C factor; that is, non-motile cells (which produce C factor) cannot induce sporulation in csgA mutants (which are motile), and wild-type cells cannot induce sporulation in non-motile cells. Surprisingly, purified C factor is able to induce sporulation in non-motile cells even though non-motile cells already produce C factor (Kim and Kaiser, 1990b). Non-motile mutants that are artificially aligned also regain the ability to sporulate (Kim and Kaiser, 1990c), which shows that motility is normally required to achieve a crucial spatial alignment between cells that allows for C factor signalling. It is not known whether and to end or side by side is the crucial cellular arrangement. C factor has been shown to influence two distinct morphological and transcriptional responses in csgA cells in a concentration-dependent manner (Kim and Kaiser, 1991). Addition of low levels of purified C factor to csgA mutants restores transcription of early C-dependent genes (those induced at six hours of development) and aggregation of cells. Adding higher levels of C factor restores the induction of late C-dependent genes (those induced at ca. 12 hours 40 of development) and complete sporulation. Additionally, it was found that csgA itself was induced by low levels of C factor. C factor is therefore required for aggregation and then, through auto-induction, increases its cellular concentration to induce sporulation genes. The relationship between cell motility, alignment, and C-dependent gene expression suggests that induction of sporulation genes requires tight packing of cells lie, a fruiting body). A dense population of germinating spores may be ecologically advantageous. Anabaena In Anabaena sp. strain PCC 7120 the transcription of genes that are involved in various aspects of growth and metabolism has been shown to be restricted to specific cells. For instance, the transcription of genes encoding ribulose bisphosphate carboxylase, required for CO2 fixation, is primarily limited to vegetative cells while transcription of the nif genes, encoding nitrogenase, and hepA, required for heterocyst envelope formation (Holland and Walk, 1990), is limited to the heterocysts (Elhai and Wolk, 1990; Wolk et al., 1993). In many species of Anabaena and Nastoc, the induction of nitrogenase within heterocysts has been found to require the excision of intervening DNA from the chromosome of developing heterocysts (Golden et al., 1985, 1988; Matveyev et al., 1993). However, loss of the intervening DNA from vegetative cells does not affect heterocyst differentiation, just as loss of the 'skin' element from vegetative cells of B. subtilis does not affect sporulation. Therefore, genomic rearrangements appear not to be required as regulatory mechanisms except, 41 perhaps, for preventing expression in vegetative cells. In Anabaena sp. strain PCC 7120, unlike in most other diazotrophs, the induction of the nif genes has been found to be under developmental control and not simply to be inducible in an environment devoid of fixed nitrogen with low oxygen tension (Elhai and Wolk, 1990). The developmental program is required in order to produce nitrogen-fixing cells. The questions raised by heterocyst differentiation are i) what form of regulation causes genes to be differentially transcribed in the different cells and ii) what aspects of transcriptional regulation are involved in the patterned formation of heterocysts? Two models have been proposed that could account for the non-random distribution of heterocysts (Wolk, 1989). One model postulates that the first cells to sense nitrogen deprivation start to differentiate into heterocysts. Differentiating cells produce an inhibitor that prevents nearby cells from differentiating and can even reverse early stages of differentiation of neighboring cells. Because localization of differentiation precedes nitrogen fixation, the inhibitor cannot be a direct product of nitrogen fixation (Neilson et al., 1971 ). According to a second model, cells that sense nitrogen deprivation induce uptake systems by which they scavenge nitrogenous compounds from neighboring cells. The last cells to sense nitrogen deprivation would be subjected to extreme deprivation that would induce heterocyst differentiation. The first model provided an interpretation for early experiments that showed that differentiating cells are able to prevent or reverse differentiation in neighboring cells (Wilcox et al., 1973a; Wolk, 1967). It was suggested that 42 'zones of inhibition' (Mitchison and Wilcox, 1976) are generated by the most advanced differentiated cells. The second model is primarily based on the fact that other procaryotes induce uptake systems upon nutrient deprivation, and that high-affinity amino acid transport systems in cyanobacteria have been described (Flores and Muro—Pastor, 1988; Herrero and Flores, 1990). A distinguishing aspect of the two models is whether the first or last cell to sense nitrogen deprivation will be the first to differentiate. The second model would appear to be supported by the fact that hetR, which clearly plays a central role in development, is not induced until 2 hours following nitrogen deprivation, whereas cells are capable of sensing and responding to nitrogen deprivation within 20 minutes (e.g., by activation of the nitrate reductase operon; Wolk et al., 1991; Y. Cai, personal communication). However, the cellular localization of hetl? expression, using luxAB fusions, shows that small foci of cells emit enhanced levels of light 3.5 hours after the removal of nitrate, and that by ca. six hours the light emission is mostly localized to single cells. Following 24 hours of nitrogen deprivation, light is emitted most intensely from within cells that lie approximately midway between the positions of two existing heterocysts, i.e., at the position at which a heterocyst would be expected to form. These data indicate that mechanisms that act following the induction of hetl-“i can limit the positions at which the induced expression of hetR is maintained. The production of a diffusible inhibitor within the most advanced differentiated cells could limit the positions of hetR expression. However, the induction of uptake systems could also control hetl? expression, with the last 43 cell to induce an uptake response continuing to transcribe hetR. How hetl? is regulated may be of central importance in understanding how heterocyst differentiation is initiated and perhaps how pattern formation occurs. Characterization of the N10 mutant and of the genes that surround the site of insertion of the transposon in N10, presented in chapter 3 of this dissertation, leads me to suggest that a secondary metabolite, perhaps a polyketide, may be involved in regulating the spacing of heterocyst differentiation. As discussed in chapter 3, the gene that is interrupted in N10, hetN, and the genes that surround hetN may be required for the production of a diffusible compound that inhibits heterocyst differentiation. Because the presence of hetN on a replicating plasmid suffices to suppress heterocyst differentiation, hetN appears to play a central role in the process of suppression. A gene, called hetl, that lies adjacent to hetN may regulate the production of the metabolite because the presence of hetl on the same plasmid as hetN relieves the inhibition of heterocyst differentiation. hetl also appears to be essential for vegetative growth because, thus far, I have been unable to inactivate hetl. The information obtained from my work on the N10 locus lends support to the inhibitor model of pattern formation. On the basis of what was known, and what has been shown in this dissertation, about the genes involved in the patterned differentiation of heterocysts, I propose to specify the functions within a mathematical model that has been developed to interpret one-dimensional pattern formation (Meinhardt, 1982). The basis of the model is that two components are essential 44 for the de novo production of pattern. The first component is an inhibitor that can rapidly diffuse or be transported along the filament. The second component is an activator that cannot readily diffuse from the site of activation. The activator must induce its own expression, thereby ensuring that activation by the activator exceeds inhibition by the inhibitor. In order to produce a stable gradient of inhibitor that can prevent the differentiation of neighboring cells, production of the inhibitor should also be induced within differentiating cells. During vegetative growth, the system is maintained in a homeostasis that prevents the differentiation of heterocysts. Following nitrogen stepdown, the system may be perturbed by fluctuations that either decrease the amount of the inhibitor or increase the amount of the activator. Even if the sites of initial perturbations are relatively random, the gradient of the inhibitor produced by the differentiating cells would lead to a patterned spacing of the final sites of differentiation (Meinhardt, 1982). The molecule proposed to be produced by hetN and the surrounding genes could fulfill the role of the inhibitor. My preliminary experiments have suggested that hetN is induced within 2 to 3 hours following nitrogen stepdown, which (if hetN is involved in production of an inhibitor) is consistent with the possibility that enhanced levels of inhibitor are produced following nitrogen stepdown. The auto-activation of hetl? and the dependence of differentiation upon hetR, as well as the ability of hetR to induce differentiation when present in multiple copies, shows that HetR can fit the role of an auto-inducible activator. 45 What perturbation of the system would lead to the initiation of differentiation? A sensory transduction pathway that responds to environmental or internal cues (or perhaps an integration of signals from both environmental and internal stimuli) could act to disrupt the balance of inhibitor and activator. Perhaps the sensory transduction pathway acts through PatA, an apparent response regulator protein. Although PatA has been proposed to be a transcription factor (Liang et al., 1992), it is possible that PatA is required to activate the protein(s) that can inactivate or act counter to the inhibitory molecule produced by differentiating cells. In fact, in a patA mutation that was constructed in our laboratory, many filaments failed to form heterocysts (Cal and Wolk, unpublished observations). lf patA strains cannot inactivate or counteract the inhibitor, they may be self-inhibited. If the putative inhibitor were able to diffuse slowly out of cells, then the additional surface area of the terminal cells might permit enough diffusion of the inhibitor to allow differentiation sometimes to proceed within terminal cells. Therefore, the formation of terminal heterocysts, which is the phenotype reported for patA mutants (Liang et al., 1992), may result from physical differences in terminal cells rather than from a specific regulatory role for PatA. Hence, the sensory response to nitrogen stepdown may act through PatA to decrease the levels (or perhaps the activity) of the inhibitor, thereby allowing initiation of differentiation to take place. Because the perturbation could occur randomly Within filaments, the reversal of heterocyst differentiation, perhaps due to the production of an inhibitor by nearby differentiating cells, may be required for 46 the semiregular spacing of the final sites of differentiation. If the production and movement of an inhibitory molecule plays a direct role in pattern formation, then an important question remains: how does the inhibitor move between cells? Because no extracellular control over the spacing of heterocysts is known, the presumptive inhibitory molecule may be polar and therefore remain within the filaments rather than being secreted into the medium. To move from cell to cell, the inhibitor would have to be transported or there would have to be channels for intercellular diffusion. Electron microscopy of thin sectioned (Lang and Fay, 1971; Wilden and Mercer, 1963) and freeze fractured material (Giddings and Staehlin, 1978, 1981) suggests that microplasmodesmata may provide aqueous channels between cells. However, proof of the existence of such channels has not been obtained. In other filamentous cyanobacteria, in which presumed microplasmodesmata have also been visualized, there is evidence that low-resistance electrical coupling exists between cells (Hader, 1987; Severina et al., 1988). Although microplasmodesmata would provide low resistance coupling, the exact nature of the electrical continuity is not known. It is also possible that specific transport complexes could facilitate movement of the inhibitor. It has been shown that, in E. coli, flagellar proteins are specifically targeted to cell poles (Maddock and Shapiro, 1993); perhaps a distinct feature of the membrane that is made during septation can direct proteins to the poles of the cells. If a similar feature exists in Anabaena, then specific transport systems could be localized at the cell poles to provide intercellular transport. It is also possible that 47 receptor proteins can be specifically positioned at the cell poles. This may be the case in M. xanthus, where cells may require an end-to-end alignment in order to transmit and receive the C factor signal. Because Anabaena cells are already aligned end to end within the filament, a signal peptide or some other factor produced and exported through the cell pole, or incorporated into the outer membrane, could transmit signals between cells. Whether promoter switching or use of alternative, development-specific sigma factors is required for heterocyst differentiation is still unclear. There are examples of developmentally regulated genes, such as hetl? (Buikema and Haselkorn, 1992), glnA (Turner et al., 1983) and sigA (the major vegetative sigma factor; Brahamsha and Haselkorn, 1991), that use alternative transcriptional start sites. Two genes encoding alternative sigma factors, sigB and sigC, have been identified by heterologous hybridization. The levels of mRNA for both genes increase following the inception of nitrogen deprivation; however, inactivation of either or both of the genes does not affect heterocyst differentiation (Brahamsha and Haselkorn, 1992). Therefore, temporal and spatial control of the various stages of differentiation by a succession of sigma factors, which appears to be the primary means of developmental regulation in B. subtilis, may not be of central importance for the differentiation of heterocysts. Chapter 2 SPATIAL EXPRESSION AND AUTOREGULATION OF hetR. A GENE INVOLVED IN THE CONTROL OF HETEROCYST DEVELOPMENT IN Anabaena’ SUMMARY The spatially patterned differentiation of heterocysts in the filamentous cyanobacterium Anabaena requires a functional hetR gene. Transcriptional fusions to luxAB show that net}? is transcribed at a low level throughout the filament when Anabaena is grown with combined nitrogen, and that induction of the gene begins within 2 hours following nitrogen deprivation. By 3.5 hours, induction is localized to spaced foci. By 6 hours, there is an overall induction of at least three fold in whole cultures, reflecting at least a 20-fold increase within spatially separated cells. The induction requires the presence of a functional hetl? locus, indicating that hetl? is autoregulatory. Full induction of a heterocyst structural gene, hepA, also requires a functional hetR locus. ' The essential content of this chapter has been published in Molecular Microbiology (Black. Cal, and Walk, 1993). 48 49 INTRODUCTION Study of nifH-luxAB fusions and of mutants that fail to differentiate have shown that in Anabaena sp. strain PCC 7120, nitrogenase is expressed only in heterocysts and that its transcriptional induction is developmentally regulated (Elhai and Wolk, 1990; Ernst et al., 1992). The induction of nifH at about 16 hours after the removal of nitrate (Elhai and Wolk, 1990) is a late event in heterocyst development. hepA (see Ernst et al., 1992), a gene that is required for normal elaboration of the heterocyst envelope (Wolk et al., 1988), is induced within 7 hours of nitrogen deprivation (Holland and Wolk, 1990), before morphological changes in the vegetative cells can be visualized by bright-field light microsc0py. Luminescence from hepA-lax fusions is confined to semi-regularly spaced cells that (once morphological differentiation is discernible) are seen to be differentiating into heterocysts (Wolk et al., 1993). Buikema and Haselkorn (1991b) identified a gene, hetR, that is essential for heterocyst differentiation. Its presence on a replicating plasmid leads to the formation of heterocysts in the presence of fixed nitrogen and to the formation of multiple, contiguous heterocysts following nitrogen deprivation. The level of hetl? mRNA increases within 6 hours of nitrogen stepdown (idem). Our laboratory has reported (Ernst et al., 1992) that many but not all of the transposon-generated mutations that arrest heterocyst development prior to morphological differentiation lie within hetR. Although it is apparent that hetR plays a regulatory role in heterocyst differentiation, HetR has shown no sequence similarities to regulatory (or other) proteins. Because of the 50 importance of hetl? in differentiation, l have examined its temporal and spatial expression in differentiating filaments and its effect on the spatially and developmentally regulated gene, hepA. MATERIALS AND METHODS Plasmids and strains. Strains of Anabaena were grown at 30° with continuous illumination (ca. 3500 erg cm'2 s") on a rotary shaker in 125-ml Erlenmeyer flasks containing 50 ml of AA/8 nitrate (Hu et al., 1982). Transposons and plasmids containing hetl? constructions were introduced by conjugation (Wolk et al., 1984, 1991), and double recombinants were selected as described by Cal and Walk (1990). Suspension cultures of DR884a and of strains containing transposon insertions were supplemented with 50 pg neomycin sulfate (Nm) ml", while suspension cultures of SR884a and of DR908-DR1069 were supplemented with 50 pg Nm ml" + 5 pg Spectinomycin (Sp) ml". DR937 and DR1069 were grown in the presence of 5 pg Sp ml". Several plasmids that were designed to facilitate the sacB-mediated positive selection for double recombinants (Cai and Wolk, 1990) were used in this study. Plasmid pRL271 was derived from pRL270 (Maldener et al., 1991) by (1) cutting dam-methylated plasmid with Xbal, treating with Klenow fragment, and religating; and (2) inserting between the Xhal and Sstl sites of the plC20R polylinker (Marsh et al., 1984; erroneously given as the plC20H polylinker by Maldener et al., 1991) a 527-bp Xhol-Sstl polylinker derived from hp 1943 to 1572 and bp 1443 to 1289 of plasmid pJRD184 (Heusterspreute et al., 1985; 51 bp 1571 to 1444 were deleted by treatment with T4 DNA polymerase of the Bglll and Pstl ends, and subsequent ligation). The polylinker region of pRL271 contains, from Cm' to oriV, unique sites for EcoRV, Xbal (methylated in dam“ strains), Bglll, Xhol, Nael, BssHll, BstEll, EcoNl, Spel, Spll, Xbal (not methylated in dam+ strains), Avril, Aflll, Asull, Sstl, Nrul, Sphl, Pstl, Asel, Bgll, Eam1105l, and Bsle. The anl and Seal sites (both blunt) in the 3’ portion of the Em' and Cm' genes, respectively, permit convenient replacement of antibiotic resistance cassettes to generate derivatives of pRL271. A Sm'Sp' derivative, pRL277, and a Km'Nm' derivative, pRL278, were made by ligation of the larger anl-Scal fragment of pRL271 with the aadA-containing Dral fragment from the D interposon (Prentki et al., 1991), and with the nptll- containing cassette C.K3 ,(Elhai and Walk, 1988a; Xbal ends made blunt by filling-in reactions), respectively. Plasmid pRL1075 (Figure 2.1E), based on pRL271 but bearing an additional oriT, was constructed by inserting most of the polylinker region, from EcoRV to Oral (between Eam1105| and BspHI), in inverted orientation into the (blunted) Mlul site of pRL271, and then adding the oriT region of RK2 from pAT187 (Trieu-Cuot et al., 1987) to the unique Sall site of the resulting plasmid. The Pstl ends of the RK2 oriT fragment were converted to Sail ends by passage through pRL500 (Elhai and Wolk, 1988a). Plasmid pRL1075 is useful for reconstructing mutations generated by transposons Tn5-1058 and Tn5-1063 (Wolk et al., 1991). These transposons are conveniently excised from the chromosome of Anabaena with Clal, compatible with Asull; 52 Figure 2.1 Several DNA constructs used in this work. See text for details. (A) The hetl? region of the genome of mutant 041, interrupted by Tn5-1058 which generates 9-bp direct repeats (B), was excised with EcoRV, and religated, forming pRL880. To introduce luxAB as a reporter, the transposon was replaced with pRL579 (Elhai and Walk, 1990), yielding plasmid pRL881a (C); sacB was then added by ligation to a part of pRL277, producing plasmid pRL884a (D). Plasmid pRL271, derived from pRL270 (Maldener et al., 1991) by addition of a multiple cloning site (MCS; see text), served as a precursor of pRL1075 (E) and pRL277. Plasmid pRL759 (F), cassette BLOSZ with its ends joined by an extended polylinker (see Materials and methods), was fused by Scal-Nrul junctions with pW8216N9.5 (Buikema and Haselkorn, 1991a) to produce plasmid pRL936 (G), of which only a portion is shown and in which luxAB serves as reporter. pRL880 and the EcoRV fragment of pRL936 that is shown were inserted separately between EcoRV sites of pRL1075 to form, respectively, pRL908 (H) and pRL937 (I). O, oriTof pMBI; open rectangle, oriV of pMBl; thick hatched arc, oriT region of RK2; solid bar, Tn5-1058 (Wolk et al., 1991) including oriV of p15A; thin cross-hatched line or curve, Anabaena DNA; Tn’ase, Tn5 transposase gene; Cm', Em', Nm', Sp’, genes conferring resistance to Chloramphenicol, erythromycin, neomycin (and kanamycin), and Spectinomycin (and streptomycin), respectively; *, sites useful for adding sacB, Cm'Em' and oriT (RK2) to transposon-bearing fragments of the Anabaena chromosome (see Materials and methods). The bar at the upper right shows the length of 1 kb of DNA, valid for horizontal (linear) and circular portions of the plasmids illustrated. 53 PRLOBIa (C) ucs pRL1075 (a. to“, (E) Cmr I I IIIIIB 1' ‘—SL l-lllee! pRL271 M ”lull! ‘ ///////////. PRI;277 :9 o" 0 neo|V pRL937 (I) 7/////////// pRL759 (F) ‘f (6) mm I RL938 J I (Onlyagortlonleehownfh ‘ m" pW8216N9.5 54 Spel, compatible with Avril and Spel; and Oral, EcoRV (see above), Scal and SnaBl, compatible with Fspl, Ecl136ll (a neoisoschizomer of Sstl), and EcoRV (however, ligation to that site somewhat impairs the activity of sacB). A set of qu fusion cassettes, collectively called BLOS, was designed to generate qu fusions in recovered Anabaena DNA fragments. Cassette BLOS1 was derived from a portion of pRL1050 (Wolk et al., 1993): bp 1 to 2567 are Sstl (5’ from luxAB) to Oral (destroyed; 3' from the T7 terminator); bp 2568 to 5391 are Stul (within cassette C.S4, destroyed; a description of C.S4 was given by Bancroft and Walk, 1989, and amended by Wolk et al., 1993) to Alel. To this fragment was ligated Alel, bp 2884, to bp 3230 (Oral, destroyed) from pBR322, and an 8-bp linker Smal (destroyed)-BamHl, completing 5666-bp cassette BLOSl. An EcoRV site that had been generated 5’ from C.S4 in pRL1050 by an SI nuclease digestion and subsequent ligation was eliminated by cutting with EcoRV and Eco47lll, and religating, thus deleting 54 bp and forming BLOSZ. The ends of cassette BLOS1 are connected by different polylinkers to form plasmids pRL739, pRL7398, and pRL739S, and from BLOSZ to form pRL759 (Figure 2.1F) and pRL759B. The polylinker in pRL739 and pRL759 is (from the luxAB-distal end) BamHl-XbaI-Sall-Pstl-Sphl- Hindlll-Nrul-Sstl-Xhol-Bg/lll-[Xball-G—Kpnl-Sstl (BamHl to [Xbal] from plCZOR [Marsh et al., 1984], the rest from pUC [Vieira and Messing, 1982]); in pRL7398 and pRL7598 is BamHl-Smal-Kpnl-Sstl; and in pRL7398 is an Sstl deletion of the polylinker in pRL739. The Xbal site in brackets is methylated by dam+ strains of E. coli. The BamHI site 3’ from luxAB in BLOSZ in pRL759 55 was eliminated by partial digestion with BamHI followed by a filling-in reaction using T4 DNA polymerase, producing plasmid pRL759D and cassette BLOSB. Plasmid pRL58 consists of the large Sa/l fragment of pRL139 (Elhai and Wolk, 19888) to which was ligated the following 4379-bp Sall fragment that can be excised with Sell or Smal, and that lacks EcoRV sites: Vibrio harveyi luxAB (as reporter) from the Sall site 137 bp 5' from luxA (Cohn et al., 1985) through to the Pvull (destroyed) 195 bp 3’ from luxB (Johnston et al., 1986), followed by bases GG, a BamHI site, andthen polylinker Xbal-EcoRl-Xbal from L.HEH2 (Elhai and Wolk, 19888) with C.S4 (Bancroft and Wolk, 1989; corrected in Wolk et al., 1993; antiparallel to luxAB) inserted in the EcoRI, site followed by Sell. That lax-bearing cassette, excised with Smal, was inserted into the Nrul site of hepA (Holland and Wolk, 1990; renamed by Ernst et al., 1992) in the 2.1-kb EcoRV fragment shown in Figure 5 of Holland and Wolk (1990), and was oriented to have luxAB report on transcription from PM“. The entire resulting 6.5-kb EcoRV fragment was then transferred to the unique Nrul site of pRL271, yielding plasmid pRL1069. The genomic EcoRV fragment bearing the transposon in 041 was cloned as plasmid pRL880 as described (Wolk et al., 1991), and was maintained in E. coli H8101. DNA contiguous with the transposon was subjected to automated sequencing (Applied Biosystems Inc., Foster City, CA) outwards from primers (CTCAAGAAGATCATC and CACATGGAATATCAG) corresponding to unique sequences near the left and right ends of transposon Tn5-1058, respectively. Plasmids pRL881a and pRL881b were constructed from pRL880 by excising, 56 with Xholl, all of Tn5-1058 but the terminal 60 bp at the left end through to bp 607 3’ from the translational start of hetR (Figure 2.1A) and by inserting, in its place, the lax-fusion vector pRL579 (Elhai and Walk, 1990) cut with Bglll. Insertion of qu fusion cassette BLOS2 (Nrul-cut pRL759) into the Seal site 21 bp downstream from the stop codon of hetl? in pW8216N9.5 (Buikema and Haselkorn, 1991b) produced plasmid pRL936 (Figure 2.16). To permit selection for double recombinants (Cal and Walk, 1990) of pRL881a and pRL881 b in Anabaena, these plasmids were partially digested with EcoRV and ligated with the larger, sacB-containing EcoRV fragment of pRL277 yielding pRL884a (Figure 2.1D) and pRL884b, respectively. For the same reason as well as to permit mobilization, EcoRV-cut pRL880 and pRL936 were each ligated with the larger EcoRV fragment of pRL1075 (Figure 2.15) that contains the Cm'Em'-oriT(RK2)-sacB cassette, yielding plasmids pRL908 (Figure 2.1H) and pRL937 (Figure 2.11), respectively. The latter two plasmids were recombined with hepA::luxAB strain DR1069 and with wild type Anabaena sp. strain PCC 7120, respectively, to produce double recombinants DR908- DR1069 and DR937. Measurement of luminescence. Three-microliter portions of lightly greened liquid cultures were spotted onto filters atop solidified AA medium containing 10 mM nitrate and lacking antibiotics (Allen and Arnon, 1955). Following four days of growth on the solid medium, individual spots were moved to solidified AA medium at various times. Following the final time point, whole plates were exposed to vapor of n-decanal for 2 min and then immediately assayed for 57 luminescence using a Hamamatsu Photonics System, model C1966-20 (Wolk etaL,1991L Microscopy and assays of luminescence of suspension cultures were performed as described by Elhai and Walk (1990). For induction experiments, cultures grown with AA/8 nitrate plus appropriate antibiotics were washed three times with AA/8, and were resuspended in AA/8 without antibiotics. The cultures were maintained under growth conditions throughout the time course. Images of luminescence are 20-min integrals. Concentrations of chlorophyll were measured in methanolic extracts (Mackinney, 1941). Results Transposon insertions in hetR. Transposon-generated mutants 041, 063, T34, T47, and YC49 (Ernst et al., 1992) showed little sign of differentiation after 24 to 48 hours of nitrogen-deprivation, although some cells within the filaments displayed a less granular, more homogeneous appearance suggestive of a very early stage of differentiation. Southern analysis (Maniatis et al., 1982) of genomic DNA (data not shown) disclosed that, as expected for insertions within the hetl? open reading frame, the Clal, EcaRl, EcoRV, and Spel fragments bearing the transposons were the same size for each of the mutants. Plasmid pRL880, containing the transposon-bearing EcoRV fragment from 041 (see Materials and methods), hybridized to the appropriate 5-kb EcoRV fragment of hetR clone pW8216N9.5 (Buikema and Haselkorn, 1991 a; data not shown). The transposon in 041 was found to have inserted 101 bp 3' from the 58 translational start site of hetR, generating a 9-bp direct repeat as is characteristic of Tn5 (Figure 2.1A,B). Transcriptional control of hetR-luxAB fusions following removal of fixed nitrogen. To measure transcription of hetR, Tn5-1058 (Wolk et al., 1991) in pRL880 was replaced (see Materials and methods) with linearized plasmid pRL579, which bears the Vibrio fischeri bioluminescence genes, luxAB, as a transcriptional reporter (Elhai and Wolk, 1990). Plasmids were obtained with the reporter inserted in two orientations: pRL881a (Figure 2.1C) reports transcription initiated 5’ from the site of insertion of the transposon in hetl? and pRL881 b reports possible transcription initiated 3' from that site (Figure 2.1C). The larger EcoRV fragment of the positive selection vector pRL277 containing the Bacillus subtilis levansucrase gene (sacB) and conferring resistance to streptomycin (Sm) and Spectinomycin (Sp) was placed into the EcoRV site used for the genomic cloning, yielding pRL884a,b (Figure 2.1D). These plasmids were transferred by conjugation to Anabaena sp. strain PCC 7120. Single recombinant exconjugants (SR884a,b; shown by our unpublished data to have recombined at the hetl? locus) were identified as Sm'Sp'Nm'Suc' and had a Fox“ Het“ phenotype (terminology of Ernst et al., 1992; Fox refers to the ability to grow in an oxygenated environment on media lacking a source of fixed nitrogen). Plating the single recombinants on sucrose-containing medium (Cal and Walk, 1990) gave rise to Sm'Sp'Nm'Suc’ double recombinants (DR884a,b) at the hetl? locus that were Fox' Het‘. 59 5 I I r I I i a 4— ~ '03 m ‘i g /A E 3 — A - m *r ‘"' ._ g l. l-J m 2 F a a -_ :5 i- S -- J- a: 1 - " . 0— ' I l I I 0 2 4 6 8 10 HOURS OF NITROGEN DEPRIVAT‘ION Figure 2.2 Relative light emission from suspensions of hetR—luxAB constructs SR884a (I), DR884a (e) and DR937 (A) at various intervals after removal of nitrate from the medium. The results graphed are averages derived from three, three and two independent experiments, respectively. Values of luminescence during successive measurements on a given suspension were, for the sake of clarity of the graph, normalized to the corresponding mean values at 2 hours, and results for different suspensions then averaged. The mean values of light emission (x10 '7) at 2 heurs for SR884a, DR884a, and DR937 were 278.4 :1: 121.3, 126 :I: 13.8 and. 16.1 a: 0.54 quanta (pg Chl)" min", respectively. 60 Luminescence from spots of filaments (see Materials and methods) of SR884b and DR884b, in which luxAB was oriented opposite to the direction of transcription of hetR, was very low, and constitutive (data not shown). Luminescence from spots of filaments of SR884a was initially low, but increased within 2 hours following removal of nitrate, and within 8 hours reached a level 3-4 fold greater than that seen on nitrate (data not shown). Nitrogen-deprived suspension cultures of SR884a showed a rapid increase in luminescence (Figure 2.2). In contrast, the luminescence from spots of filaments (data not shown) and suspension cultures of DR884a, whose level of expression at 0 hours was similar to that of SR884a, showed no significant increase (Figure 2.2). A luxAB fusion generated 21 bp downstream from the hetR termination codon (pRL937; see Figure 2.1i) was introduced as a double recombinant (DR937) into the wild type strain. Heterocyst differentiation proceeded normally, and the luminescence of this recombinant, like that of SR884a (Figure 2.2), increased rapidly after removal of nitrate, although absolute values of luminescence were much lower. Localization of expression of hetR-luxAB during differentiation. At various times after nitrogen stepdown of suspension cultures of SR884a and DR884a, small aliquots were removed from the cultures, a suspension of n-decanal was added, and luminescence was collected from filaments through an oil immersion lens (Elhai and Wolk, 1990). SR884a initially displayed a uniform distribution of luminescence throughout the filament (Figure 2.3A). After 3.5 hours without nitrate, single cells or in some cases clusters of 2 or 3 cells luminesced 61 Figure 2.3 Images of luminescence (20-min integrals corrected for charge spreading: see Elhai and Walk, 1990) from individual filaments of SR884a at (A) 0, (B) 3.5, (C) 6, and (D) 24 hours, and of DR884a at (E) 0 and (F) 24 hours, following nitrogen stepdown. Adjacent analog images of the filaments have not been manipulated to eliminate an instrumentally generated honeycomb pattern (compare Elhai and Wolk, 1990; Wolk et al., 1993). In (D), the arrows point to mature heterocysts that are non-luminescent due either to a lack of 02, a substrate of luciferase, within the heterocyst, or possibly to diminished transcription of hetR in mature heterocysts. Values of light flux density are presented as means :1: standard deviations of the mean for areas the size of cells, after correcting for charge spreading. In (A), (E) and (F), the light flux density averaged over the filament was 0.38 :l: 0.03, 0.32 :t 0.06, and 0.29 :l: 0.03, and outside of the filaments was 0.06 :l: 0.01, 0.04 :1: 0.006, and 0.05 :l: 0.01 cpm pm", respectively. In (B), the light flux densities of the two bright cells toward the left, the single bright cell near the center, the remainder of the filament, and the background were 1.69, 2.08, 2.93, 0.72 :l: 0.05, and 0.18 :l: 0.02 cpm pm", respectively. In (C), the single bright cell, the remainder of the filament and the background had respective values of light flux density of 2.50, 0.09 :l: 0.02, and 0.012 :I: 0.003 cpm pm". In (D), the central bright cell, the left-hand and right-hand heterocysts (arrows), the other cells, and the background had values of light flux density of 3.26, 0.13, 0.19, 0.75 :l: 0.07 and 0.12 :t 0.01 cpm pm", respectively. The bar represents 10 pm. 62 63 much more brightly than did the remainder of the cells of the filament. Noncontiguous, especially luminescent cells, when observed, were separated by about 89 less luminescent cells (Figure 2.38). As of 6 hours, expression was particularly clearly localized (Figure 2.3C). After 24 or more hours of incubation of the filaments in the absence of fixed nitrogen, the positions of especially luminescent cells approximately bisected the intervals between observed heterocysts (Figure 2.3D). Luminescence of DR884a was weak and constitutive in all cells in the presence or absence of nitrate (Figure 2.3E,F). Effect of hetl? on expression of hepA. The expression of luciferase from fusions to hepA has also been found to occur in a patterned array prior to morphological differentiation of heterocysts (Wolk et al., 1993). The induction of hepA fusions, however, does not occur until 57 hours after the removal of fixed nitrogen, much later than the induction of hetR. To determine whether a mutation in hetl? would affect the expression of hepA, we made use of Sm'Sp' strain DR1069, which bears a hepA-luxAB fusion. Plasmid pRL908 (Figure 2.1 H), bearing an insertion within hetR, was transferred by conjugation to strain DR1069. There, it recombined into the genome at the hetR locus (data not shown). Sucrose-resistant double recombinants (DR908-DR1069) displayed a typical Het ' phenotype, characteristic of a hetR mutant. Twenty-four hours after nitrogen stepdown (and averaged over three pairs of independent experiments), DR1069 showed a 334-fold increase in luminescence to 1635 ' :1: so x lo7 quanta (pg Chl)" min" while DR908-DR1069 showed only a 19- fold increase to 6.5 :l: 1.0 x 107 quanta (pg Chl)" min". 64 DISCUSSION Buikema and Haselkorn (1991b) showed both that hetR is required for differentiation of heterocysts and that it leads to formation of supernumerary heterocysts when present in multiple copies. Those authors therefore suggested that her}? is a positive regulator of heterocyst differentiation. Our results confirm their suggestion and show, further, that he”? is also involved in regulation of its own expression. Transcriptional fusions to luxAB within the coding region of hetl? block induction of the reporter following nitrogen deprivation unless a wild type copy of hetR is maintained. Transcripts that range in size from 1.4 to 2.4 kilobases have been reported for hetR, whose open reading frame spans 0.9 kb (Buikema and Haselkorn, 1991b). It was therefore possible that the regulatory effect that we saw in hetl? strains was due to a polar effect of the insertion on a cotranscribed factor lying downstream from hetR. To examine this possibility, a lux fusion was made downstream from the hetl? coding region. Although the absolute values of luminescence obtained from this fusion were much lower than from fusions within hetR, induction of hetl-1 was found to proceed normally as reported by (i) increased luminescence measured following removal of fixed nitrogen, and (ii) unimpaired differentiation. The lower levels of luminescence seen in DR937 can potentially be accounted for in many ways but is probably based, in part, on the use of a different reporter construct. The very low level of luminescence from DR937 at zero hours apparently cannot have resulted from the presence of only one hetl? promoter region compared with two promoter regions in the 65 single recombinant strain because the levels of luminescence of SR884a and DR884a are very similar at zero hours. However, the greater relative increase in luminescence seen in DR937 compared with SR884a could result from the extra hetR promoter in SR884a, although there are also numerous other possible explanations for this difference. Nonetheless, it is apparent that HetR, the product of hetR, is required for enhanced expression of hetR. The positive feedback seen with hetR is similar to that seen with the B. subtilis sporulation factors 0“ (Kroos et al., 1989) and 0G (Sun et al., 1989; Karmazyn-Campelli et al., 1989). Once these sigma factors are synthesized, they activate their own transcription from specific promoters to assure their enhanced expression in the appropriate compartment and at the appropriate stage in development (Stragier and Losick, 1990). The deduced sequence of the HetR protein, however, shows no similarities to the sequences of known sigma factors. The expression of 0" requires the presence of an auxiliary DNA- binding protein (Kroos et al., 1989), while 0G may require an effector encoded by the spell/A locus (Stragier and Losick, 1990) to overcome repression by SpollAB (Rather et al., 1990). Although HetR contains no known DNA-binding motifs (Buikema and Haselkorn, 1991b), it could interact with a development- specific sigma factor or another DNA-binding protein and thereby enhance the expression of hetl? as well as of other heterocyst-specific genes. Full induction of hepA depends upon the presence of a functional hetR gene, "but may depend only indirectly on the hetl? product because hepA is first activated several hours after hetl? is induced. The time course of induction of 66 a hepA-lax fusion has been published (Wolk et al., 1993). The DR1069 hepA- qu fusion has a similar increase in light emission between 4 and 20 hours after nitrogen stepdown. In the presence of the DR908 mutation, an increase in light emission is also seen, but the much lower values of light emission preclude accurate quantitation at early times following nitrogen stepdown. The presence of DR908 reduces the measured expression of DR1069 (quanta min" [pg chll") by typically >99%. The spatially constitutive luminescence seen in nitrate-grown filaments, the results of Northern blot analysis (Buikema and Haselkorn, 1991b), and the higher levels of luminescence seen in cultures of DR884a than in cultures of DR884b (both, in the presence of nitrate) indicate that hetl? is transcribed at a low level in all vegetative cells. Upon nitrogen stepdown, induction of hetl? begins within 2 hours; and within 3.5 hours, enhanced transcription of hetl? can be visualized within specific cells. Because expression of hetR in differentiating filaments is restricted to no more than 10% of the cells, and because, later, luminescence from hetR-luxAB fusions is seen in the cells that lie midway between two pre-existing heterocysts, i.e., at the sites at which new heterocysts would be anticipated to form, it appears that transcription of hetl? is enhanced only in cells that have initiated differentiation into heterocysts. If the overall induction of at least 3 fold by 6 hours is attributed to the increase in at most 10% of the cells, then within those cells, hetl? must have been induced at least 20-fold. Induction of hetR thus appears to occur within a pattern of cells and may, through positive autoregulation, contribute 67 to the stabilization of the pattern of heterocyst spacing; that it contributes to the initial establishment of that pattern is not excluded. Wilcox et al. (1973a,b) observed that incompletely differentiated cells, called proheterocysts, often arise in clusters with all but one cell in a cluster eventually reverting to a state of vegetative growth. Those authors considered the theory that a differentiation-inhibitory substance generated by a proheterocyst increases in concentration until it represses differentiation of contiguous cells, even if those cells have also initiated differentiation. According to this theory, the differentiating cell must be immune to the effects of the inhibitor that it generates. If the hetl? product were a factor that confers immunity to (or, more generally, acts counter to) such an inhibitor, and if the level of inhibitor were low in nitrate-grown cells, differentiation of nitrate-grown cells in the presence of multiple copies of hetl? (Buikema and Haselkorn, 1991b, and my confirmatory observations) might be explained. Likewise, the appearance of multiple heterocysts in NZ-grown filaments can be interpreted as implying that there is no change in the specification of the foci of differentiation, but that the balance between inhibitory and counteracting factors is shifted in favor of differentiation when hetl? is present in higher copy number than in the wild type strain. Strains lacking HetR would be unable to overcome self—repression, and therefore unable to differentiate. According to this interpretation, hetR would be involved not in the generation of the primary pattern, but in the resolution of clusters and stabilization of the differentiation process. Such a conclusion would be consistent with our observations and 68 that of Buikema and Haselkorn (1991a) that in a hetR mutant, ”non-granular cells are visible and may represent cells that have degraded their internal stores of nitrogen more fully than other cells." Nonetheless, as has been pointed out (Wolk, 1982), several substances that are known to elicit the formation of clusters of heterocysts also lead to decreased spacing of the clusters, suggesting that mechanisms underlying the resolution of clusters, and spacing, may well be related. I presume that cellular factors help to determine the non-uniform expression of her}? along a filament: the identification of these factors is crucial for elucidation of the involvement of hetR in heterocyst differentiation. Other genes are also required for all, or all but the earliest, stages of heterocyst differentiation (Ernst et al., 1992); it will be of great interest to determine whether they or their products interact with HetR or affect transcription of hetR. Chapter 3 ANALYSIS OF N10, A Het' MUTANT IN Anabaena SP., PROVIDES EVIDENCE THAT A SECONDARY METABOLITE MAY REGULATE HETEROCYST SPACING SUMMARY Transposon-generated mutant N10 has a Het' phenotype (Ernst et al., 1992). Reconstruction of the mutation has reproduced a Het' phenotype, but other insertions at the position of the transposon have produced strains that form multiple contiguous heterocysts. Sequence analysis around the site of insertion of the transposon showed that the insertion lies within the 5’ end of an 861 -bp orf (hetN). The product of translation of hetN shows extensive similarity to NAD(P)H-dependent oxido-reductases that are involved in biosyntheses of fatty acids, poly-B-hydroxybutyrate, nod factor and polyketides. A second, 1518-bp orf (5' orf) that ends 557 bp 5’ from the start of hetN appears to encode a protein that has at least two functional domains: its amino-terminus is similar to an acyl-carrier protein (ACP), while its central portion is similar to domains of proteins that perform reductive reactions. A third 711-bp orf (hetl) encoded on the opposite strand ends 42 bp away from the 3’ end of hetN. The 69 70 translated product of hetl shows similarity to the B. subtilis protein Sfp that regulates biosynthesis of the cyclic peptide, surfactin. Clones from a A-EMBL3 library that contain the wild-type hetN gene do not complement the N10 mutant. The presence of hetN, as the only orf, on a replicating plasmid suppresses heterocyst formation in wild-type cells, while the additional presence of hetl alleviates this effect. INTRODUCTION Study of heterocyst differentiation and pattern formation has been greatly stimulated by the development of methods for genetic analysis ofAnabaena sp. (Wolk, 1991; Wolk et al., 1984). Despite facile transposon mutagenesis of Anabaena sp. with derivatives of transposon Tn5 (Ernst et al., 1992; Wolk et al., 1991), only one mutation has been reported that clearly alters pattern formation: the PAT-1 mutant causes heterocysts to form almost exclusively at the ends of filaments (Liang et al., 1992). Several mutations have been found that prevent heterocyst differentiation. Inactivation of hetl? by a point mutation that converts a serine residue to an asparagine (Buikema and Haselkorn, 1991a,b) or by insertion of a transposon (Black et al., 1993) prevents evident differentiation of heterocysts. Supernumerary copies of hetl? lead to the formation of heterocysts when filaments are grown on media containing combined nitrogen, and to the formation of multiple contiguous heterocysts in the absence of fixed nitrogen (Buikema and Haselkorn, 1991 b; and unpublished I observations). Like mutations of hetl-i, an insertion in hetP prevents heterocyst differentiation (Ernst et al., 1992) while extra copies of hetP lead to the 71 production of multiple contiguous heterocysts during growth on N2 (F. Fernandez Pifias and C. P. Wolk, unpublished observations). Another mutant, N10, was reported to be similar to hetl? in that heterocyst differentiation was blocked at an early stage; however, this mutant has a tendency to revert to a form that generates multiple contiguous heterocysts (strain N16) when inoculated on media lacking fixed nitrogen (Ernst et al., 1992). The site of transposition in the N10 mutant has been mapped near the arbitrary origin of the physical map (Kuritz et al., 1993). I report a characterization of the sequence of the DNA surrounding the site of transposition in mutant N10 and a genetic analysis of the function of that DNA in heterocyst differentiation. My data lead me to suggest that the gene inactivated by the transposon in N10 may be involved in the production of a secondary metabolite that can inhibit heterocyst differentiation. MATERIALS AND METHODS Growth and manipulations of Anabaena sp. Wild-type Anabaena sp. strain PCC 7120 and its various mutant derivatives were grown at 30°C in the light (ca. 3,500 ergs cm'zs") on a rotary shaker in 50 ml of AA/8 medium (Hu et al., 1982) supplemented with 5 mM nitrate in 125-ml Erlenmeyer flasks. Plasmids containing the various insertions were introduced by conjugation (Elhai and Wolk, 1988b; Wolk et al., 1984), and double recombinants were selected as described by Cal and Wolk (1990). Mutant strains were grown in the presence of appropriate antibiotics (Table 3.1). To induce heterocyst formation, portions 72 of rapidly growing cultures (ca. 1 to 2 pg of chlorophyll a ml") were washed twice with AA/8, suspended in an equal volume of AA/8 without antibiotics, and incubated under growth conditions. Filaments were examined by microscopy 48 to 72 hours following nitrogen stepdown. Recovery of DNA contiguous with the transposon in mutant N10. Ca. 1 pg of DNA isolated from mutant N10 was cut with 10 u of EcoRV. The reaction mixture was then heated to 65°C for 15 min, diluted 20-fold, subjected to self- ligation for four hours at 25°C, and the DNA sedimented. Half of the DNA was used for electroporation into E. coli H8101. Colonies that grew on plates of L- agar containing 50 pg kanamycin-sulfate (Km) ml'1 were analyzed for the presence of the transposon, Tn5-10638 (Wolk et al., 1991). Plasmid pRL916, containing 2.1 kb of Anabaena DNA, was isolated from a positive colony. Larger regions of Anabaena DNA, ca. 3.5, 2.1 and 13 kb, respectively, flanking the left end of the transposon were obtained by recovering transposon- containing fragments after digestion of total N10 DNA with Eco47lll, Mlul and Munl. pRL921 was made by ligating EcoRV-cut pRL916 to an EcoRV fragment from plasmid pRL1075 (Black et al., 1993) that contains the genes conferring resistance to Chloramphenicol (Cm) and erythromycin (Em), and sensitivity to sucrose (sacB), and the oriT region of RK2. To replace the transposon, pRL921 was cut with Alel, the ends blunted with the Klenow fragment, and then cut with BamHI. This fragment containing Anabaena DNA was ligated with the BLOSZ vectors (Black et al., 1993) 73 pRL759 (cut with Bglll + Nrul), generating pRL944a, and pRL759D (cut with BamHI + Nrul), generating pRL944b. The BLOSZ insert in pRL944b was removed by cutting with BamHI + Bglll and was replaced with the S.K3 vector, pRL498 (Elhai and Walk, 1988a). that had been cut with BamHI, producing pRL945a (Figure 3.1) and pRL945b. To generate insertions within hetl, plasmid pRL94Sa was cut with Ncol + Nhel and the fragment containing hetl was isolated and ligated with the positive selection vector pRL271 (Black et al., 1993) producing pRL975. pRL975 was then cut with SgrAl, and the C.K3 cassette, excised from pRL648 [pUC/L.EHEl/(XbaI)/C.K3: nomenclature of Elhai and Wolk, 1988a] with Aval, was inserted producing pRL978a and pRL978b. The C.K3 insertion in pRL978a was then removed by cutting with BamHI and replaced with the luxAB-C.S4 cassette (Black et al., 1993) present in pRL65 with Bglll-ends, generating pRL979a and pRL979b. Construction of a A-EMBL3 library. Chromosomal DNA from wild-type Anabaena sp. strain PCC 7120 was partially digested with Sau3Al. The digestion mixture was separated on a 10-40% continuous sucrose gradient. Fractions were collected and examined on 0.2% agarose gels. Fractions that contained fragments between ca. 15 and 25 kb in length were pooled, concentrated and ligated with A-EMBLS arms that had been cut with BamHI and treated with calf intestinal alkaline phosphatase (purchased from Promega, Madison, WI). The ligated DNA was packaged in vitro and plated on E. coli KW251 (mcrA'mch) and LE392 (mcrA‘ mch”). The packaged mixture formed ca. 1.5 x 10° pfu ml" on KW251 but gave ca. 1 x 104 pfu ml" on LE392, while 74 Figure 3.1 The region of the Anabaena genome that contains the 5' orf, hetN, and hetl. Insertions shown above the genomic map were made using plasmids derived from mutant N10. Reconstructions shown below the genetic map were made using a wild-type clone of the region obtained from a A-EMBL3 library. With the exception of pRL989, each insert was obtained in both orientations, although only the "a" orientation is shown. eoauma 153. IIVW flléaanfi _ e: emmeuma oEm mml I. x: I I+moow 25m :20 _m_o>m00m Eoom Agli :obIrFAIzioia _ Ti tab Tm 5 7 mmnmumg :3qu .e_eomav , (Eta sell I emsElVV MAIImto Eon MW Mil Ecl “glam naval—ma 76 a control insertion gave equal plating efficiency on the two strains. The library was amplified on KW251; the eluted lysate contained ca. 109 pfu ml". Subcloning the wild-type N10 DNA from the A-EMBL3 library, and subsequent plasmid constructions. Plasmid pRL916 was labeled with [0- 32PldATP by using a random primer labeling kit from GIBCO BRL (Gaithersburg, MD). The pRL916 probe was used to screen two sets of duplicate filters containing a total of ca. 1200 plaques. A single positive plaque was amplified on E. coli LE392 and the phage were eluted with 3 ml of 10 mM Tris, pH 7.5, 10mM MgSO,. The plate-amplified lysate was used to produce 10 ml of liquid lysate that was used for isolation of phage DNA (Maniatis et al., 1982). The ca. 15-kb insert, cut from the phage DNA with Sell, was ligated with Sell-cut pRL500 (Elhai and Wolk, 1988a). yielding plasmid pRL946. EcoRl fragments ca. 4.8 and 4.5 kb in length were subcloned into EcoRI-cut pRL500 giving pRL947 and pRL948, respectively. Restriction analysis showed that pRL948 contained the 2.1-kb EcoRV genomic fragment that had been disrupted in N10. To place interposons within this fragment, it was first sub-cloned into the Nrul site of pRL271, producing pRL981. pRL981 was cut at the unique Nhel site within hetN and then ligated with the C.K3 cassette, excised from pRL648 with Xbal, generating pRL984a and pRL984b. Alternatively, the ends of Nhel-cut pRL981 were filled in with the Klenow fragment and then ligated with the luxAB-C.S4 cassette, removed from pRL58 (Black et al., 1993) as a Smal fragment, to generate pRL985a and pRL985b. Transfer of the same cassette to SgrAl-cut pRL981, filled in, generated pRL989. 77 Plasmids pRL959 and pRL982, containing wild-type N10 DNA in a (pDU1- based) plasmid that can replicate in Anabaena sp. were made by inserting the 4.5-kb EcoRI fragment and the 1.3-kb Clal fragment, respectively, into the unique EcoRI and Clal sites of pRL1049, the lux—less immediate precursor of pRL1050 (Wolk et al., 1993). The same single orientation was obtained for both insertions with hetN oriented parallel to the aedA gene. Deletions pRL9828N and pRL982RN were generated from pRL982 by cutting pRL982 with BemHl and Nhel or EcoRI and Nhel, respectively, blunting with the Klenow fragment, and self-ligating. Cassette C.A1 (Elhai and Walk, 1988a). excised from pRL494 (S.K4/L.XSX1(EcoRl)/C.A1; nomenclature of Elhai and Wolk, 1988a) with Xbal, was inserted into the Nhel site of pRL982 to generate pRL982C.A1. Plasmids pRL967a and pRL967b that bear the hetl? gene and can replicate in Anabaena sp. were made as follows. The 5.3—kb, hetR-containing, EcoRV fragment from pW8216N9.5 (Buikema and Haselkorn, 19913), provided with BemHl ends by transfer to HincIl-cut pRL500, was ligated into the BamHI site of pRL1049. Two orientations of the insert were obtained; pRL967b has the hetl? gene oriented parallel to the aadA gene. Plasmids pRL1450a and pRL1450b (F. Fernandez-Pines and C. P. Wolk, unpublished results), that contain the wild-type DNA surrounding the site of the transposon insertion in mutant P6, were made as follows. First, pRL1404 was made by adding a portion of the polylinker L.HEHI and the C.C1 cassette, excised from pRL171 (Elhai and Walk, 1988a) with a partial BamHl digestion, 78 to the BamHI site of pRL1049. Next, pRL1404ACN was made by cutting pRL1404 with Clal and Natl, blunting the ends with the Klenow fragment, and then self-ligating to remove an additional EcoRI site and a portion of pDUI that is not required for replication in Anabaena. Finally, a 4.3-kb EcoRV fragment from a A-EMBL3 clone was ligated between the Ecl136ll sites of pRL1404CN. Southern analysis. Double-recombinant insertions were confirmed by Southern analysis of chromosomal DNA that was isolated from the mutants and digested with EcoRV. Probes were made with the 2.1-kb EcoRV fragment from pRL948 that was labeled with [0-32PldATP or digoxigenin-I 1-dUTP using kits purchased from GIBCO BRL or Boehringer Mannheim (Indianapolis, IN). Sequencing. Automated sequencing (Applied Biosystems Inc., Foster City, CA) was performed using universal primers on a variety of fragments that had been subcloned from pRL916 and pRL948 into vectors pUC18, pUC19, pUC118, and pUC119 (Vieira and Messing, 1982; 1987). Database comparisons and alignments of the translated sequences were performed, using the default settings of the algorithm developed by Altschul et el. (1990), at the National Center for Biotechnology Information (NCBI) using the BLAST network service. Single-sequence alignments obtained from NCBI were combined to give the multiple-sequence alignments. In some cases, sequence in addition to that returned from BLAST was added. 79 RESULTS Isolation, recovery and reconstruction of the N10 mutation. Whether the Het' phenotype of mutant N10 (Ernst et al., 1992) was a result of the transposon insertion was tested by regenerating the mutation in the wild-type strain as described by Black et al. (1993). The transposon and flanking Anabaena DNA, cloned from the N10 chromosome as pRL916, were combined with the sacB- containing portion of pRL1075, and the resulting plasmid (pRL921) was transferred by conjugation into wild-type Anabaena. The locus of insertion was confirmed by Southern analysis (data not shown). The resulting Nm'Suc' double recombinants (DR921) displayed the same Het’ phenotype as N10. However, when the transposon in pRL921 was replaced with a variety of other insertions, the most frequently derived phenotype of the double recombinants was the production of multiple contiguous heterocysts (Mch; Table 3.1). The double recombinants (DRs) from at least three independent single recombinant strains were examined in each case, and all isolated DRs had the same phenotype (for DR984a, the same combination of phenotypes) for each particular construction. Two to three rounds of growth, fragmentation and selection on plates containing 5% sucrose were used in order to obtain fully segregated double recombinants with plasmids pRL944a, pRL944b, pRL945a, pRL945b, and pRL984a, whereas double recombinants were more readily obtained with pRL944a and pRL944b in the hetR strain, DR908 (Black et al., 1993; data not shown). 8C) TABLE 3.1 Construction, antibiotic resistances and phenotypes of the strains. Strain Source of Antibiotic Phenotype* insert resistance‘ Insertions in hetN N10 Tn5-1063a BmNmSm Het' DR92 1 pRL92 1 BmNmSm Het‘ DR944(a,b) pRL944Ia,b) Sm/Sp Mch DR94S(a,b) pRL945(a,b) Nm Mch DR984a pRL984a Nm wild type and Bet' DR984b pRL984b Nm Mch DR985(a,b) pRL985(a,b) Sm/Sp Mch Insertions in hetl. DR978a pRL978a Nm wild type and Mch DR978b pRL978b Nm wild type and Het' DR979(a,b) pRL979(a,b) Sm/Sp wild type DR989 pRL989 Sm/Sp wild type but fragmented 13m = bleomycinfi Nm = neomycinfi Sm = streptomycinfi and Sp = spectinomycinfi Concentrations used in liquid media were 50 pg Nm ml" or a combination of 1 pg Sm ml" + 5 pg Sp ml". Concentrations used in solid media were 200 pg Nm ml" or a combination of 1 pg Sm ml" + 10 pg Sp ml". ° In no instance were doubly recombinant chromosomes completely segregated. * Observed 48-72 h after the removal of nitrate. Mch = multiple contiguous heterocysts. In cases where two phenotypes are listed, clonal cultures had filaments that displayed different phenotypes. 81 Cloning and attempted complementation of the N10 mutants. Plasmid pRL916 was used to probe a A-EMBL3 library. The wild-type DNA for the region interrupted by the transposon was subcloned from the library as a 4.5-kb EcoRI fragment into pRL500 (Elhai and Wolk, 1988a), producing pRL948. The wild- type EcoRl clone contained 400 bp upstream and ca. 3.3 kb downstream from the orf that had been interrupted by the transposon in N10 (hetM. This clone, in replicating vector pRL959, was unable to complement the N10 or DR921 mutants. This result was surprising because sequence analysis confirmed that the wild-type clone was identical in sequence to the DNA surrounding the transposon (see below). Plasmid pRL959 also produced no effect in wild-type Anabaena sp. strain PCC 7120. However, a 1.3-kb Clal subclone, containing 48 bp upstream and 371 bp downstream from the hetN and no other complete orf, inhibited heterocyst formation by wild-type cells when in replicating vector pRL982. Like pRL959, pRL982 was unable to complement N10 or DR921. Heterocyst differentiation in wild-type cells was not suppressed by plasmid pRL9828N, which has the region 3’ from the Nhel site in hetN deleted; pRL982RN, which has the region 5’ from the Nhel site in hetN deleted; or pRL982C.A1, which contains an insertion within the Nhel site in hetN. To test whether the Het' phenotype of N10 and DR921 was due to a second-site mutation in a gene known to inhibit heterocyst formation, her}? and the cloned wild-type DNA containing hetP were introduced into the N10 and I DR921 strains on replicating vectors (pRL967 for hetR, and pRL1450a and pRL1450b for hetP). None of these plasmids complemented the Het‘ 82 phenotype of N10 or DR921. It was possible that the phenotypic differences seen in the various reconstructions were due to a second mutation contained within the DNA near the transposon. To examine this possibility, plasmids (pRL984a, pRL984b, pRL985a, and pRL985b) constructed from the ll-EMBL3 clone were used to generate insertion mutations. The phenotypes of the resulting double- recombinant strains DR984a, DR984b, DR9853 and DR985b, are shown in Table 3.1. Except for DR984a, these disruptions of hetN consistently yielded a Mch phenotype. Sequence analysis of DNA near the transposon insertion in N10. Sequence analysis (sequence is available from Genbank under accession number L22883; Figure 3.2) showed that the transposon in N10 had inserted 179 bp from the 5' end of an 861-bp orf, called hetN, and that other orfs were nearby (Figures 3.1 and 3.2). A 2.3-kb EcoRV fragment from the A-EMBL3 clone containing hetN was also sequenced, and was found to be identical in sequence to the clone recovered with the transposon. The predicted amino acid sequence from each reading frame was compared to those contained in the NCBI database using the blastp program. A large number of highly similar proteins was found for the protein sequence encoded by hetN. Alignments with some of the most similar proteins are shown in Figure 3.3. This class of proteins can be grouped as a superfamily of NAD(P)H-dependent oxido-reductases. The HetN protein aligns with known oxido-reductases for approximately 190 amino acid residues from the amino terminus, before diverging. The protein sequences that show 83 1 mum. OOOOOGA. WWW. MWAMWO {MTOOGOAOGMONI’OHOOYOATO QIIOOPAIOOPIDIVROOYLABAIIRBVVDVAB 101 103mmmemntcecamammmmmémrmcmimcummnmrcceonce-tn IIIQGAAPLYRIQBLIIQOIQIPDBYOQYBPVR! 201 mmmmfimmmammeccaaéncmrcaécmanarieecmmicflacrménmeema O 301 mentions-ennui:mmrmmiammmnémrmmmmeemmmemammm lenses!eaanteawureanasnnevoebsI 401 anmfimummnmmeemimmmémmcnmmficcamcm barrenaetenoeaoauvnveansonnepopep 501 Wanamamaécemrmmmmurémmncmncmmmacm Lanna!petsenssanassnasosssoarbsara 601 Wmmmdmmmnnumcmmmmmmm Ineleepvnonoaavvnbpvrspeearevolvev 101 mom. met-0mm. . Mmeeémnmmmmémecmamm sopaxvrneeeeornearnrannneeeQADvrc 001 mmmrmtcmcmrhamécmicmmefimxummmiummmnamc LvaaabaoaexaxIoeanaetaxllosntssnz19 901 WIRBWOWWMWWMMWWMAMTWM V V O D In A I P In L G In B B 9 Q P Q I In I A I I D '1' I Y B 8 G A I; 1001 man-same. mmmmecmmemmmm' ASNOOTTOOOAMHMMTOMGCOTONO In I I V P P Y D A I; I I A I V I. O f 0 I V In I! In A O Q I I V K P V 1101 A O O O O O I O O O c I V V 9 B V A V P I B P V Y A G I V V l I B D D P 8 B I B G I P Io O 1301 memmoemmmmmmmmmacrnmrmacuccmnémmr VDQYEIVAIILVKIIGDROLPV!IHRPPLIOOD 1301 mnemonicmccmeamufiufimmaieeemmmfiecmcccoimaenui-arecuean eeeeroaeabrxlnuvaecnouocrpbvstnnb 1401 mmmmCANONOATO‘PATOTAWW‘POOOTWMAGO‘I'TTOCAOOTAOMOAOOOAOMOOMYATO IDPVDYVBKAIVBLDIQIIBLOKAPILQBPBPVB 1501 mmmmmmmimmmmrfim nevnvovvaererlzxenetasveasnxllaee 1601 mmmrmmmmmmmmmfimmfimmmiccrcnmccmcccc VDIPLI‘PLRPDLLIRIBDIQLTIPDLYLKAKRP 1701 mmmmccmmméamammcmimamfiammfimmmmr a:ecobsnxanaereIecsprbexnrlrtrarnx 1001 mmmmmhmmmcflmsficamemhccmmmmmc .seolu . 1901 eeoeHmccmocaeeanmaameemecaamnécceccmmmmcnimmemmnceeenmccrcar 2001 mficmnfimmhcmmmmammnammmm 2101 mmmmmammaflWWamimmamfinm 2301 mmmmmmficcmmmnmme Figure 3.2 Nucleotide sequence of the 4.7-kb fragment shown in Figure 3.1. The underlined sequences indicate putative ribosome-binding sites. The double underline shows the position of a possible factor-independent transcriptional terminator following the 5' orf. The overlined sequence shows the 9 bp that were duplicated during transposition. Restriction sites mentioned in the text are shown in bold type. 84 Figure 3.2 (cont'd) 3301 3001 3501 3001 3701 3001 3901 3001 3101 3301 3301 3001 3501 3601 3101 3001 3901 0001 0101 0301 0301 0001 0501 0501 0701 . . .BcolW/ClaI mammmmmmmmmmmknmnmmmmmr 111-rare WWmeocmecmmmunocmmcmcmemcmémmcmm xevnnreaenenevrxnnnnnxon-rvvcvsno necmmocccnnc’ emu-11mm. naeeca-ec'reeeeeeamc' «Memen'rerommcmrcncunncoecrcr ceennorcnnvxnneexaxazproves-2901.911. mm. earimrceecccr' amuse-remix. men. rec-roecA-rconn' nuc'eem'cmflAc-rcrc' raccoon voqnovaepxnvnxlunexnzuerrnnrenne ammmmmrfixmunennmcmmiccmmimmeeumnu IOIIPITILLAAI3L93LLLP8IIBROBORIVI . . . . .nd . O . . mmrmmcmmmnnmmrmmcammu Inennexxovnrlevzenexnenxuwrnauaosn mmmimmmeénmmmmnhcmmccmnmm veeevnrevvcPerveqreurvprnvsnpxnne nmémmémumcéemanfimmuimmémrfimnmnnflmcnmcm I029xsvaunvvxnzxnxrravzvnqnp1239:. mmmmfimmmmcfimmmmmeémmmcnccmcmwm rsnunnvoo:sprevnnzrnurovvprloxnaeu mm. cocci-1m. m. rceércnmcéecmecfimeeuremme' Am. rcnrnéoccmm nvxpervnvnennet -reurx mmficcmcccmmmmimmmnmmmcmmomma-09mm POIOPOAVAVAAVCIDDPVLBLLSIAPATQLKA Wemcnmfimmufimmmcmfinmmmcfiemm 239921.313:llnxnzeneraxnrnnxcrwrnrr . I .u. C . . m. 130303091.ennnnrnnpnrraxaneznoerpn .claI GIOISISTOTIBASOOIIOOOGATTTOGOOOOTOTAARTOIOTOOAOACAAOOOOAAOTTOTOGGAOTOTOAOAAOTTAAASAAEAAACOAOTOTCGOOA L Y I L D I O I Q R T Y I V' I O L A L I O 8 I 8 L I P L L O 8 I A mmmeccrnnmmccmeeonmamimamncfieccceeeficcncmccinmccccccmnncmcemnm PRDOLIPKORBBYDFIVQGPBVGLYOGLISRLIO u 901'“ e e e a e o e OIOOLOOAOOAOEIAAAOGOOOOOONIOTTOOOGAAAASAIAAIO301110OAOOGOOTAASTOGTOAOTRGASAhflcfhflcoocfifihfiflcOTOTAGC90 I O I f P R R R I I P P Y P I I A I A L I D 8 8 L T A A L D Q L Q fGISTOTOOIIOOTOAAOGOGABIOOOOOIGAGASOIACéQOASOEOAOLARAAOGTTIIASTTGGOGGéTTTGGTAGCOBAGTAIGOTOOAAOAATIRA I I P O D L P I R I L I V I D 0 L L f L I P P K P L I f B Q L L mmmreucemmfifiurmaflimmcfinmmhmnmreoecmflaicccremcc'rcrc'rr WWmmmémmmccficmcmmmiccccccocfiecmmm mammemcucneecm' WWW. donuts-011mm. mom-1mm. nucce u n q a I I n Wammmmflimmfimminmnmcfimmmccflmm noxxpnorrrnsnelpxenpuunIpeorrnues . BcoRV mmmfinmnmmccxmmxmmmummumummr rpennnnnnzneponlvrvexrruen!011-0110 ' mommrnomnmnfimccmaecnmucinumeinnmeooimnenmfieemnemc nvvnllnxzplannppnelrxzproezpnwrnp mflmonfimnmmnfimmnémmefinmomflcwumfiumrm vsevrunnnxprcennenxreaereneonsnare nmecoeeammmcmmfim mm. nan-re. on IncnnorerprnreexIrppnauro 85 Figure 3.3 Alignment of the protein encoded by hetN with other proteins: 0rf4 from S. cinnamonensis (S.cin [Arrowsmith et al., 1992]), Actlll from S. coelicolor (S.coe [Hallam et al., 1988]), NodG from R. meliloti (R.mel [Debelle and Sharma, 1986]). FabG from E. coli (E.col [Rawlings and Cronan, 1992]), 3—oxoacyI-ACP-reductase from Arabidopsis thaliana (A.tha [Slabas et al., 1 992]), acetoacetyl-CoA-reductase from Alcaligenes eutrophus (A.eut [Peoples et al., 1989]) and the estradiol 17B—dehydrogenase from Homo sapiens (H.sap [Peltoketo et al., 1988]). The residues showing the greatest conservation are shaded. 86 "new "use "map "H00 ”Hoe "woo "cflo " zuo "new "use "may "H00 "H03 "moo "Can " zuu "new "use. "ecu. "Hon. "Hoe. "woo. “can. " zuu new. "use. "can. "H00. “H03. "moo. "GHQ. " zau hm 0mmU¢A>Mmmammn mflHthUQ¢MOOmA3MmmmmmszUU€>>MmOD Q>QMH€O0>b940m 87 the greatest similarity, as much as 40% identity through the first 190 amino acid residues, are 3-ketoacyI-(acyl carrier protein [ACP]) reductases involved in the biosyntheses of polyketides (PKs), poly-B-hydroxybutyrate (PHB) and fatty acids (FAs). The 100 carboxy-terminal amino acid residues of HetN do not show significant sequence similarity to other protein sequences contained in the databases. A 711-bp orf, hetl, lies 3’ from hetN and is encoded on the opposite strand of DNA. This orf ends 42 bp from the 3' end of hetN. hetl lacks an apparent ribosome binding site and an initiator AUG codon but has two adjacent UUG codons that could initiate translation. The sequence of the protein encoded by hetl shows similarity (28% identity) to portions of the proteins encoded by the sfp (Nakano et al., 1992) and oer (Kréitzschmar et al., 1989) genes from Bacillus subtilis and B. brevis, respectively (Figure 3.4). During the late-growth phase of B. subtilis, Sfp regulates the production of surfactin, a cyclic lipopeptide; affects genetic competence and sporulation; and may regulate various genes that are induced in the presence of alternative nitrogen sources (Marahiel et al., 1993). oer, only partially sequenced, lies adjacent to the 5’ end of the operon that encodes the proteins involved in the biosynthesis of the cyclic lipopeptide, gramicidin. A third complete reading frame, the 5’ orf, lies 5’ from hetN and is encoded on the same strand of DNA as hetN. The 5’ orf is 1518 bp in length, ends 557 'bp from the start of hetN, and appears to be followed by a transcriptional terminator (Figure 3.2; Carafa et al., 1990). The protein sequence encoded by 88 0:11: o ........................... IDBHVIN'LBSIVBIIIOQAIVIIVNVKDAIRSLLGILLIRKYLIQVL leeeeleee e It as I I eel. I I 3.01 3 LLOBIILPIPPILTLLEDIVHLIIIPLDOPIDQLODLILTLSSDILARANIFYFPIHRRRPEAGRQILRBILOGYL II‘I '0. e e 0'... at e||| | ’I I college Sfp: ............... IIIYOIIIDRPLBOIIII o83"!I8PIIIIIORRFIIIIDAEIILLGDVLVRBVI880! 0r£2r IIPIIIILIIIII!GIP.... .PVDPDII'IIBISDIIVVOIIB. IEPVOIDII. .RIBID .0 ti. | it II. It. It ||.'||t. .t'ltt tt00"'il I t. tfi|ttll t'. not: 1 mnmzmneemasommvmmmmmnuemrum “ ““l' “Ill'l “ Il‘lll‘“l'“*l|ll*l *I'I ll*lllll II' 8298 QLDIBDIIIBTOIIOKPOIPD...LPDAEPIIBBBGIIVIGIID.BQPIOIDIIIIIPIS.LI.IAIRPI 03:33 ILOBIIONDQVBBP'IL'TIIISIIIAIGIGI!. IPIIB'IIDIIOEQTVIYKONKKIPVTIYIPILFIOIIOSOO 1" see e 0 || II l|e|e|'e'e|e e e e Iotlz LLRBLPDIQKOKIPPRIITOIIAILKATGDGII. KLIIIIIALTPTIPAKLOIAPAISLLILVPDDNOVIAVIVIQ I O O... O 'O. I. 'l...|ifil.l.. '..000 I .00..... I O I * t 9.. 82p: DLLAIDIDIOTDYFYHLIEIIIBIIIDIGIDLBLPLDBISVILBODOOVBIILPDBHBPOYIKTYIVDPGYIIINO 02111 SLIBSVTILBITILQVOILOILILDBTIBIIII e e IOtI: PGIOPIPIBY Sfp: IAIPDPPIDIEIVBYIILL Figure 3.4 Alignment of the protein encoded by the hetl with Sfp (Nakano et al , 1992) and 0er (Kre’itzschmar et al., 1989) from B. subtilis and B. brevis, ‘ respectively. Identical residues are marked with an H) and similar residues with an 1*). 89 the 5’ orf shows great similarity (26% identity) over its entire length to the carboxyl-terminal half of the yeast LYSZ protein (Morris and Jinks-Robinson, 1991; Figures 3.5 and 3.6), and a small region of both proteins, beginning at ca. residue 27 in the protein encoded by the 5' orf, is similar to the 4’- phosphopantetheine-binding domains of putative and confirmed ACPs and ACP domains (Figure 3.6A). The ACP domains from the multifunctional eryA gene products of Saccharopo/yspora erythraea (Cortes et al., 1990; Donadio et al., 1991; Donadio and Katz, 1992) which produce the polyketide-derived macrolide antibiotic, erythromycin, show particularly strong similarities to the V ACP domain in the 5’ orf. Beginning at amino acid residue 136 in the protein encoded by the 5' orf, and in LYSZ, lies a consensus NAD(P)H binding site that is also found in EryA and other polyketide synthases (PKSs) and fatty acid synthases ([FASs] Figures 3.5 and 3.63). The remainder of the protein encoded by the 5’ orf has a region of similarity to the acyl transferase domains that are weakly conserved in PKSs (Figure 366). Another long orf, only partially sequenced (Figures 3.1 and 3.2), lies just upstream from, and is transcribed in the same direction as, the 5' orf. Preliminary alignments with the translated protein from this orf also show similarity to PKSs (data not shown). Another apparent orf, also only partially sequenced, is found to extend beyond the 3’ end of the fragment shown in Figures 3.1 and 3.2. 90 30 00 60 80 IBLNOIIBIADIOKIIIBNLLILLGVDGDIIDLTVNLIBYOLDBIOAIVLVBKLIOLLGPQP8PLLLIEYPTIISLBIRL cl see one. Q el to accent on e | 'I | to 0'0. I e e eee|'|'eeeee e VIIIDDBOPIIVIRIVIDLILSILPTIPABVBPDDBPIDLOGHBILAEIIIPTLKIKLQVDLPLOTIPKIPTIKAPAAII 100 130 100 160 PVIEPOGIIPVD'PVEOPIIVPLTGGEOILGIFLIII e to ace... nee. e at e 'l .9. e e e I e e e e||e|l IIIII"'* ’1' one DIIIBSOOBBOGIVVIIVTIIIIIDAIILVITLPBBYP8IIIPVIPIELIGIITIIV'VTGVTGFLOBIILADLLGRBPI 100 300 220 no.ooVICLNIAInIOIGKIKIQEILICIII'OIIIISIIIPVVUDLIIPLLOLBBTQIQLLRIBIDT e I. ||| It... Ieelee o'eeIQIeeelel jell'eee eIIIeeee elleeeIel lejjleeeIe LBKBQ'GLBDIKIIDLRNTVDIIIENGILVE'VY 300 360 380 300 QIKVIPVI!VBSVIVIISPVY..o.o....ICK.VVKIBDDFBHW.ICIPLGYSQTK'V II. I.“II"I 'l’*I| I I "l||"'*" I ‘II ' I'll. ‘ *I’ ll‘l‘ll‘ LAKVUIPIIID'VBBESTLDTIYYIILBDILVEIGKPOILIBDDLIIBABGLIGGIGQSKIA 330 300 360 380 LPVTIHRPPLIGGDBITOI .. II II "‘ ||*||l* I ll “|‘l"l *ll‘ll‘ “l|"l‘l ‘II'I *‘l lll‘|‘***| I* ‘ll DPLLIPLIOCVDLGIIPDIINBVNIVPVDHVIRVVVITSLNPPKI 000 030 000 060 SLGIAIILOBPIPVBLSVLVDIVRBPGYIIKELPIIIIOIILIII..AEBVDIPLYTLRPLLLBRIBDIQLTIPDLYLKA e e to e to an e eeee||eeee o'eele 0| 0. e e e|e||el eee'ee see a '0' ea IILNVIOVTGEPRILFIDYLITLEDIUIDVIIIS!BIIIISLIABVIDRIIIIILYPLLEIVLDNLPIST.KAPILDDRN 080 500 .OTBIGOPPIDGILIITYEKILIQTOILELA O O. 9.. Q... .*t'l II 0 .9. I.‘ .."| INIBLIIDEIITGVDIBNGIGVTPIIVOIYI...APLNKVUPLPPPTHIDKLPLP8IILTOAOISLVASGAGARQBSAAA Figure 3.5 Alignment of the entire protein encoded by the 5’ orf (shown on top) with the carboxy-terminal half of the LYSZ protein from S. cerevisiae (Morris and Jinks-Robinson, 1991), starting at amino acid residue 837. Identical residues are marked with an (H and similar residues with an (*l. The numbers shown above the sequence refer to the position within the translated 5’ orf. 91 Figure 3.6 Alignments of domains within the 5' orf-encoded protein and (A) acyl carrier protein, (B) ketoreductase and (C) acyl-transferase domains from other proteins. eryA( S,1-5) represent the subunits within EryAI,-ll and -l|l from S. erythraea (Cortes et al., 1990; Donadio et al., 1991; Donadio and Katz, 1992). MSAS represents domains from the 6-methylsalicylic acid synthase of PeniciI/ium patulum (Beck et al., 1990), ACV is the a-aminoadipyI-cysteinyl- valine synthetase from Acremonium chrysogenum (a.k.a. Cephalosporium acremonium [Gutierrez et al., 1991)), SrfA is a domain from the surfactin synthetase from B. subtilis (Fuma et al., 1993), LYSZ is from S. cerevisiae (Morris and Jinks-Robinson, 1991), Act is the ACP from the actinorhodin PKS of S.coelicolor (Hallam et al., 1988), and FAS is the FAS-ACP domain from A.thaliana (Lampa and Jacks, 1991) in (A) and the 2-oxoacyl reductase from the chicken FAS (Chirala et al., 1989) in (A). The serine residue shown in bold type in (A) represents the putative 4’-phosphopantetheine attachment site. Residues showing the greatest conservation are shaded. Z—I >IMI—J—l>- ounumanzz <<<>rxc¥ CGGCD<_I_IDI.|J :nu>>._.<r_m.:..r._.:_ =o>44mm >:3._._.:_mn_o —l><< UW<_l> IDI> LLLDO.)—>CDU- 'zz<<' wm<< rum—JI—d (SCSI-UL“ I-ZCDCJ mam—r— :r:>>r—c I—H—rww PUD-z ‘4—NC‘OV’" 00— com .- >-LW LN~< L<<s>a>a- m 0.1 cu 0’sz 93 Preliminary alignments with the translated protein from this orf also show similarity to PKSs (data not shown). Another apparent orf, also only partially sequenced, is found to extend beyond the 3’ end of the fragment shown in Figures 3.1 and 3.2. Attempts to inactivate the hetl. The hetN gene alone, but not together with hetl, on a replicating plasmid, inhibited heterocyst differentiation. To examine the possibility that hetl, or its product, regulates hetN or its product, I attempted to inactivate hetl. None of the insertion mutations attempted yielded fully segregated double recombinants despite multiple rounds of growth and selection for sucrose-resistant colonies. DISCUSSION Testing that a transposon insertion is the cause of an observed phenotype is normally essential when working with Anabaena sp. strain PCC 7120 (see Liang et al., 1992; 1993; and personal observations). Confirmation that the original Het' phenotype of mutant N1 0 was due to the Tn5—1063a insertion was apparently achieved when the reconstructed mutant (DR921) also proved to be Het'. However, when the transposon was replaced with a variety of other insertions, the predominant phenotype was the production of multiple (i.e., two or more) contiguous heterocysts, a result that contrasts with the original phenotype. Moreover, the presence of the wild-type hetN gene on a replicating vector failed to complement the (original) Het‘ N10 mutation or its reconstructed version, DR921. In fact, the wild-type hetN gene inhibited the 94 formation of heterocysts in wild~type cells when it was the only Anabaena DNA present in pRL982. Partial deletions of, or an insertion within, hetN alleviated the suppression of heterocyst formation in wild-type cells, indicating that suppression required the product encoded by hetN. Several explanations could account for these results. (i) Perhaps a second- site mutation that occurred during the original transposon mutagenesis led to the Het' phenotype of N10. In that case, however, the same (or a phenotypically identical) mutation must have arisen during the reconstruction of the mutation in DR921, a result that seems improbable. However, the inhibition of heterocyst formation seen when hetN is carried on a plasmid suggests that hetN may be involved in repressing heterocyst differentiation and may be essential for preventing all vegetative cells from differentiating. There may, therefore, be very strong selective pressure for second-site Het' mutations in N10 and DR921 if the transposon-mutated cells are to survive and grow. Indeed, double recombinant insertions within hetN were more readily obtained in a hetR strain, which was already Het'. Why the other insertion mutations that were generated did not also lead to a Het' phenotype is unclear. Perhaps hetN retains partial function when differently interrupted. Although hat]? and the gene that complements the P6 mutation were unable to rescue the Het‘ phenotype of N10, the occurrence of a differentiation-blocking mutation at an as yet unidentified locus remains possible. (ii) The recovered clone of the N10 region may have contained an additional mutation. As mentioned above, the N10 mutant had a tendency to revert to 95 a strain that produced multiple contiguous heterocysts. Perhaps the DNA recovered from N10 contained such an additional mutation. The phenotypes of the subsequent double recombinant strains would have then been dependent upon the sites of recombination. This interpretation appears unlikely because the sequence of the DNA recovered from N10 matches, base for base, that of a wild-type clone that includes hetN and hetl. (iii) Perhaps the transposon and the various other insertions had unique, cis- dominant effects that gave rise to the variety of phenotypes, e.g., by altering the expression of hetl. This explanation is tempting in view of the arrangement and proximity of hetN and hetl. My inability to inactivate hetl also suggests that it may play an essential role in maintaining vegetative growth: if cis-dominant effects of the interposons generated the Het‘ and Mch phenotypes, this could also explain why the wild-type clone was unable to complement the Het‘ mutations. I am tempted to attribute to cis interactions between hetN and hetl, the observation that plasmid pRL959, which contains those two orfs, did not inhibit heterocyst formation in wild-type cells whereas plasmid pRL982, which lacks hetl, did. However, the differing effects of pRL959 and pRL982 might also be attributed to the additional absence from pRL982 of a 352-bp sequence, upstream from hetN, that may contain regulatory sites. A different pDU1-based vector that contained both hetl and hetN, but stops 48 bp upstream from hetN at the Clal site that terminates the sequence in pRL982, I also failed to inhibit heterocyst differentiation (J. Golden, personal communication). This observation appears to strengthen the interpretation that 96 it is the presence of hetl on pRL959, and not the presence of bp -48 to -352 upstream from hetN, that is responsible for the loss of heterocyst suppression. The protein encoded by hetN is similar to a large number of oxido- reductases. Most similar are synthases of PKs, PHB and FAs, including the FA moiety found on some NOD factors that are required for host-specific recognition of symbiotic bacteria by legumes. These proteins also contain regions that are highly conserved in the steroid hormone-modifying dehydrogenases (Baker, 1991). However, other proteins involved in dehydrogenase-like reactions also contain these highly conserved domains, including the proposed NAD(P)H binding site. Thus, sequence comparisons do not narrowly delimit the possible activities of the protein encoded by hetN. The protein encoded by the nearby 5’ orf shows similarities to multifunctional PKS and FAS proteins, and particularly to domains within the EryA proteins from S. erythraea. It is of potential evolutionary interest that the putative transposase of insertion sequence l81136‘ that is found between the first and second subunits of the EryA PKS gene cluster (Donadio and Staver, 1993) is similar in sequence to that encoded by insertion sequence IS891 of Anabaena (Bancroft and Wolk, 1989). The similarity of the 5' orf to the LYSZ protein of Saccharomyces cerevisiae may reflect the similarity of the latter to PKSs and FASs, such as the presence of a possible 4’-phosphopantetheine attachment site and a domain that carries out an NADPH-dependent reduction of the bound substrate. The strong similarity between the 5’ orf and LYSZ may reflect similarities in the substrates that are bound to the enzymes. It is known 97 that PKSs can initiate with, or add, a variety of molecules to the growing acyl chain, which increases the diversity of the side chains that provide unique properties to the metabolites that are generated (Hopwood and Sherman, 1990). The consortium of LYSZ and LYSS, referred to as a-aminoadipate reductase, carries out the NADPH-dependent reduction of a-aminoadipate to a- aminoadipate-6-semialdehyde, a biosynthetic precursor of lysine in S. cerevisiae (Bhattacharjee, 1985). The similarity between LYSZ and the protein encoded by the 5' orf begins at a putative 4'-phosphopantetheine attachment site (Figures 3.5 and 3.6A). The amino-terminal half of LYSZ is homologous to the synthetases that produce cyclic peptides, such as surfactin and gramicidin, and to 6—(L-a—aminoadipyl)-L-cysteine-D-valine (ACV)-synthetases (ACV is a precursor in the biosynthesis of penicillin). These enzymes appear to use a thiotemplate mechanism in which acyladenylated substrates are covalently attached to a 4’-phosphopantetheine cofactor; the cofactor supports the condensation and translocation of the growing peptide chain (Lipmann, 1980). The thiotemplate mechanism is analogous to the mechanism used by FASs and PKSs in which the 4'-phosphopantetheine of the ACP or ACP domain supports the condensation and translocation of growing acyl chains (Hopwood and Sherman, 1990). Consequently, the putative cofactor attachment sites are conserved among the cyclic peptide synthetases, FASs and PKSs, particularly around the highly conserved serine residue that provides the ester linkage to the cofactor (D’Sousa et al., 1993; Hopwood and Sherman, 1990; Lipmann, 98 1980). In fact, PKS ACPs can be functionally replaced by FAS ACPs (Khosla et al., 1992).The possible attachment site in the protein encoded by the 5’ orf shows greater similarity to ACPs and ACP domains found in PKSs and FASs than it does to LYSZ and cyclic peptide synthetases (Figure 3.6A). Because the protein encoded by the 5’ orf may contain at least two functional domains, an ACP domain and a ketoreductase domain, it may be a type-l PKS. Furthermore, if the LYSZ protein maintains a covalently attached cofactor that has a thioester linkage to the substrate, then it may also possess thioesterase or acyltransferase functions that release the reduced substrate, since free a—aminoadipate is produced in vitro by a-aminoadipate reductase. Although the LYSS protein could carry out this function, a domain within LYSZ, and the 5’ orf protein, shows similarity to the proposed acyltransferase domains from other PKSs. Thus, the protein encoded by the 5’ orf may contain a third functional domain. The protein encoded by hetl shows less similarity to other entries in the databases, but it is intriguing that it is most similar to a gene (sfp) that is proximal to, and involved in the regulation of, genes that encode structural proteins for the biosynthesis of a secondary metabolite, surfactin (Nakano et al., 1992). The late growth phase expression of genes whose regulation is dependent on the source of nitrogen may also be influenced by sfp (Marahiel et al., 1993). Interestingly, it appears that sfp strains can be complemented by an extracellular factor produced by sfp” strains (Marahiel et al., 1993). Another similar protein is encoded by oer from (B. brevis [Krétzschmar et al., 99 1989]), a presumptive gene that lies adjacent to, and may possibly regulate, the operon that encodes the proteins required for the biosynthesis of gramicidin (Krétzschmar et al., 1989; Turgay et al. 1992). The sequence of hetN and of the DNA surrounding it raises new possibilities concerning the control of pattern formation in Anabaena. l conjecture that hetN and the 5' orf encode, and hetl regulates, functions that are required for the synthesis of a PK, PA, or modified amino acid. Because most insertions within hetN result in the production of multiple contiguous heterocysts, while supernumerary copies of hetN suppress heterocyst formation, a substance produced or modified by the product of hetN may inhibit heterocyst differentiation. Such a substance may even mediate intercellular interactions that regulate development. Therefore, an endeavor should be undertaken to identify a substance whose synthesis is affected by the integrity, and presence on plasmids, of the orfs that l have studied. Reduced synthesis of the substance might lead to massive initiation of differentiation, which could in turn prevent vegetative growth, and thus be effectively lethal. My inability, to date, to inactivate hetl suggests that it is essential for maintaining vegetative growth. Perhaps hetl acts, via hetN, to inhibit differentiation. There is no indication, in Anabaena, that cell density affects the differentiation of heterocysts, other than in terms of growth-rate effects on spacing (Antarikanonda and Lorenzen, 1982; Fogg, 1949). A putative inhibitory metabolite may, therefore, remain within the filament. Procaryotic secondary metabolites, some that have highly specific receptors, 100 have been found to control and coordinate cellular differentiation within populations of cells (Beppu, 1992). Such compounds are considered to be the evolutionary precursors to the hormones that play a vital role in coordinating eucaryotic metabolism, growth and differentiation. It will be of great interest to determine whether (and if so, how) a secondary metabolite acts intercellularly to control pattern formation within ancient, heterocyst-forming procaryotes. Chapter 4 A STUDY OF INTERCELLULAR COMMUNICATION IN THE FILAMENTOUS CYANOBACTERIUM Anabaena SP. STRAIN PCC 7120: IMMUNOLOGICAL EVIDENCE FOR A GAP JUNCTION-RELATED PEPTIDE IN A PROCARYOTE SUMMARY The existence of intercellular channels between cells of Anabaena sp. has been examined by physical and immunological techniques. Fluorescence redistribution assays on filaments labeled with a variety of dyes were employed to assess intercellular movement of the dyes; however, no dye movement was visualized. Antibodies raised against the a-peptide of the 26-kD rat liver gap junction protein were used to probe protein extracts from Anabaena. Primarily a 52-kD protein was labeled by the antibodies. The 52-kD protein was localized to sedimented fractions of crude extracts, and appeared to be enriched within isolated plasma membrane fractions. 101 102 INTRODUCTION lntercellular interactions are clearly required for pattern formation in Anabaena (Adams, 1992; Wilcox et al., 1973a,b; Walk, 1967; WoIk and Ouine, 1975). Nonetheless, the means by which cells communicate is essentially unknown. Speculation on the existence of cytoplasmic continuity between cells in filamentous cyanobacteria began years ago when the dyes methylene blue (Brand, 1901; Schmid, 1923) and ruthenium red (Metzner, 1955) were reported to move through filaments via the cell septa. Drews (1959) reported that filaments of Phormidium uncinatum, a motile, non-heterocyst-forming organism, reversed their direction of movement when the front 10% of the cells in the filament were shaded. Shading the rear cells had no effect. The generation of an electrochemical potential, induced by light and mediated by fluxes of protons and of calcium ions, and propagated along the filaments through low resistance channels (Hader, 1978,1981; Hader and Poff, 1982), appears to be responsible for controlling the direction of movement. Similarly, illumination of a portion of the filament was found to generate a membrane potential that was rapidly propagated along the entire filament (Severina et al., 1988). Low resistance electrical coupling, shown thus to occur in cyanobacteria, is mediated by gap junctions in animal cells and by plasmodesmata in plants (Goodwin, 1976). More evidence for the existence of structures mediating intercellular communication in Anabaena was obtained from thin section electron micrographs (Lang and Fay, 1971; Wildon and Mercer, 1963). These 103 micrographs showed intercellular connections that were 5 - 20 nm in diameter and 2 40 nm long. These connecting elements, called microplasmodesmata after their presumed plant counterparts (Clowes and Juniper, 1968), may provide aqueous channels that mediate cytoplasmic continuity between cells. Freeze fracture electron micrographs display pits on the protoplasmic fracture face and protrusions on the exoplasmic fracture face of plasma membranes (Giddings and Staehlin, 1978), characteristics that are consistent with the premise that protein complexes traverse the cytoplasmic membrane and cell wall. During a survey of a number of filamentous species (Giddings and Staehlin, 1981), it was noted that heterocyst-forming species tend to form a greater number of microplasmodesmata between vegetative cells (100 to 250 per cell) than do non-heterocyst-forming species (30 to 40 per cell). This finding was interpreted as meaning that heterocyst-forming species require a greater degree of intercellular communication, and that the increased presence of microplasmodesmata may fulfill this requirement. The same study showed that there are 80% fewer microplasmodesmata between vegetative cells and heterocysts than between two vegetative cells, an observation that is consistent with the fact that the area of contact at the septa between heterocysts and vegetative cells is approximately one-tenth of that between neighboring vegetative cells. The decreased contact area and fewer microplasmodesmata may prevent excessive entry, into the heterocyst, of oxygen that could decrease the nitrogenase activity within the heterocyst. 104 Although the observations presented are consistent with the idea that microplasmodesmata exist and confer cytoplasmic continuity, the actual existence of microplasmodesmata remains. in doubt. Furthermore, little additional work has been performed to characterize presumptive microplasmodesmata despite their potential importance in regulating cell-cell ' interactions in Anabaena. Gap junctions and plasmodesmata are structures that provide cyt0plasmic continuity between animal and plant cells, respectively (Robards, 1976; Unwin and Zampighi, 1980). The passage of regulatory molecules and metabolites through these channels allows cells to interact and function in a synchronized manner. My study of microplasmodesmata was inspired by investigations that demonstrated antigenic relatedness of the proteins that constitute gap junctions and the proteins that are thought to constitute plasmodesmata (Meiners and Schindler, 1987, 1989; Yahalom et al., 1991). Transport via plasmodesmata and gap junctions is also regulated by similar factors, such as calcium ions and calmodulin-dependent protein kinases, pH, and phorbol esters (Baron-Epel et al., 1988; Rink et al., 1980; Wade et al., 1986), providing further evidence of their similarity. Therefore, the evolution of intercellular channels may have an early phylogenetic history. Consequently, I inspected Anabaena cells for a protein immunologically related to a gap junction protein, and thereby to the putative plasmodesmatal protein, and also looked for intercellular movement of fluorescent dyes. 105 MATERIALS AND METHODS Growth of strains. The laboratory strain Anabaena sp. strain PCC 7120 was maintained in liquid culture with Allen and Arnon medium diluted eight-fold (AA/8 [Hu et al., 1981]). 5 mM nitrate (2.5 mM NaN03 + 2.5 mM KNOa) was added to repress heterocyst formation and maintain vegetative growth of filaments. Cells were grown at 30°C with gentle shaking under continuous illumination. To obtain heterocyst-containing filaments, nitrate-grown cells were washed once and then resuspended in AA/8 to give a final concentration of 0.3 pg of chlorophyll a (Chl a) ml“. The cells were allowed to grow under normal conditions for 24-28 hours. Mature heterocysts were present after 20 hours. Protein isolation. Cells grown in AA/8 + NO3‘ were washed and then concentrated 10-fold in lysis buffer (50 mM Tris [pH 7.5], 0.1 M NaCl, 0.5 mM CaCIZ). Cells were lysed by cavitation, and cell wall material and unlysed cells were removed by sedimentation at 3000 X g for 10 min. The supernatant solution was used as the crude extract. For further fractionation, the crude extract was centrifuged at 100,000 X g for one hour; the supernatant solution was retained and the pellet was resuspended in the lysis buffer. Proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS PAGE; 12% acrylamide) under reducing conditions (Laemmli, 1970). For western blot analysis, proteins were transferred to nitrocellulose using a Gelman Bio Trans semidry electrophoretic transfer unit (Gelman, Ann Arbor, MI) as described by the manufacturer. After transfer, the nitrocellulose was washed in T83 (20 mM Tris [pH 7.5], 0.5M NaCl) and then blocked for 2 hours in TBS 106 + 5% BSA. The filter was washed in TBS + 0.05% TWEEN 20 and then rinsed in distilled water. Antibody raised against a 27-amino acid peptide (a- peptide) from the rat liver gap junction 26-kD polypeptide (connexin32; Meiners et al., 1991; provided by Dr. Melvin Schindler, Dept. of Biochemistry, Michigan State University, East Lansing) was diluted 1:50 (final concentration: 10 pg/ml) in TBS + 1% BSA and incubated with the filter overnight at 4°C. The filter was washed in TBS + 0.05% TWEEN 20, incubated with alkaline phosphatase- linked goat anti-rabbit antibody (purchased from Sigma Chemical Company, St. Louis, MO), and developed according to the manufacturer’s instructions. Isolation of plasma membrane fractions. Plasma membranes were isolated as described by Murata and Omata (1988). Approximately 3 grams of cells (wet weight) were washed with 60 ml of 10 mM TES (pH 7.0) and resuspended in 30 ml of spheroplast buffer (10 mM TES [pH 7.0] + 600 mM sucrose + 2 mM EDTA), and 0.05% (w/v) of lysozyme then added. Digestion with lysozyme was carried out under growth conditions for two hours; the cells were collected by centrifugation at 5000 X g for 10 min, and then washed twice with extraction buffer (20 mM TES [pH 7.0] + 600 mM sucrose). The final pellet was resuspended in 30 mls of ice cold extraction buffer and passed through a French press at 5800 psi. The cell extract obtained from the French press was incubated 15 min with DNase) by adding 0.1 ml of 3% PMSF and 0.05 ml of DNase buffer (10 mM NaOAc [pH 5.6] + 1 mM MgCl2 + 0.1% [w/v] DNasel). Cell debris was removed by centrifugation for 10 min at 5000 X g. The supernatant solution was adjusted to 50% (w/v) sucrose by adding 107 0.74 volumes of a 90% (w/v) sucrose solution containing 24 mM TES [pH 7.0] + 12 mM EDTA + 24 mM NaCl. Fifteen-ml portions of the cell extract, adjusted to 50% sucrose, were placed into 35-ml centrifuge tubes and discontinuous sucrose gradients were made by overlaying with 8 ml of 39%, 3 ml of 30%, and 7 ml of 10% sucrose solutions (w/v), each containing 10 mM TES (pH 7.0) + 5 mM EDTA + 10 mM NaCI + 1 mM PMSF. The sucrose gradients were centrifuged in a swinging bucket rotor at 130,000 X g for 16 hours at 4°C. A yellow-orange band (plasma membrane fraction) that formed within the 30% sucrose layer, and a green band (thylakoid membrane fraction) at the interface of the 39% and 50% sucrose layers were collected using a syringe, and pooled with the equivalent fractions collected from other tubes. The pooled material was diluted with 3 volumes of 10 mM TES (pH 7.0) + 10 mM NaCl, placed into a centrifuge tube, and sedimented for 1 hour at 300,000 X g. The pellets were resuspended in 10 mM TES (pH 7.0) + 10 mM NaCl. The purity of the fractions was established by measuring the absorption spectra of diluted portions of the resuspended membranes. Labeling the filaments with fluorescent dyes. One-ml portions of rapidly growing filaments were washed and concentrated 10-fold in either AA/8 or AA/8 + N03' media buffered at pH 7.3 with 5 mM HEPES (final cell concentration: 3 pg Chl a ml"). Fluorescent dyes were then added as described below. All the dyes, unless otherwise indicated, were provided by Dr. Melvin Schindler. The cells were incubated with the dyes under growth conditions. After incubation with the dyes, cells were washed 3 times with fresh media as 108 follows: sedimentation for one minute in a microcentrifuge was followed by rapid decanting and resuspension in one ml of fresh medium. The cells were kept at ca. 25°C both during and after the washings. Approximately 10 pl of each cell suspension was placed onto a microscope slide and covered with a coverslip. The edges of the coverslip were sealed with paraffin wax. Each slide was examined within five minutes. Use of polar cytosolic dyes to measure cytoplasmic continuity. Carboxyfluorescein or fluorescein, each in the non-fluorescent, membrane- permeable diacetate form, was added to the cells at a final concentration of 10 pg ml". Once inside the cells, the acetoxymethyl esters were cleaved by internal esterases to give the fluorescent products, carboxyfluorescein or fluorescein. When carboxyfluorescein was used to label developing heterocysts, vegetatively growing filaments were washed once in AA/8 and concentrated 10-fold in AA/8. Ten ug ml" of carboxyfluorescein was added, and filaments were placed under normal conditions of growth for 24 hours. 4-Methylumbelliferyl B—D-galactopyranoside (Sigma) and 3-carboxyumbelliferyl ls-D-galactopyranoside (Molecular Probes, Eugene, OR) were added to filaments at a final concentration of 0.1 mg ml". The carboxyl derivative was also added in media with pH's of 4.5, 5, and 6 (adjusted with 5 mM MES). The two umbelliferyl galactosides are non-fluorescent until cleaved by Is-galactosidase. To ensure the presence of high levels of B—galactosidase, cells were used that contained the plasmid pRL544. pRL544 has the [602 gene fused to the tee promoter which, although constitutively very active, can be induced 34 fold 109 by the addition of 10 pM IPTG (J. Elhai, personal communication, and my unpublished observations). Filaments were labeled with the dyes ca. five hours after IPTG had been added and fluorescence was semi-quantitatively analyzed in photographs by comparing ultraviolet light-induced fluorescence from spots of labeled cells and control spots of solution containing known concentrations of free umbelliferone. Use of lipophilic membrane dyes to measure membrane continuity. N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)dipalmitoyl-L-phosphatidylcholine (NBD-PC),4-(N,N—dihexadecyl)amino-7-nitrobenz-2-oxa-1,3-diazole (NBD-diCl6amine) and 5IN-dodecanoyl)aminofluorescein were all added to filaments at a final concentration of 50 pg ml". The mixtures of dye and filaments were shaken vigorously and examined. Examination of intercellular movement of dye. A Meridian Instruments ACAS 470 workstation was used to measure fluorescence and the cell to cell movement of the dyes. This instrument contains a 5-watt argon laser that generates narrow band excitation beams to induce fluorescence of the dyes. To examine a field, the stage moved at a speed of 2 cm s‘1 with a step size of 0.25 pm. At each step, the emitted light is detected by a photomultiplier tube that generates quantitative electronic signals. The signals are accumulated by a microprocessor that generates a false color image of the labeled cells indicating relative fluorescence. A 530/540 nm narrow band-pass filter was placed in front of the photomultiplier tube to block autofluorescence. Control experiments (data not shown) showed only much lower levels of fluorescence 110 from unlabeled filaments. To examine the movement of the dyes, the fluorescence redistribution after photobleaching (FRAP [Baron-Epel et al., 1988]) technique was utilized. In this technique, a single cell within a filament is bleached by a high intensity laser beam. The recovery of fluorescence within the bleached cell is then followed by a series of scans with a lower intensity beam. The timed scans, in my case, 1 to 2 minutes between scans, reveal any changes in intensity of fluorescence of the cells as a function of time. Recovery of fluorescence in a bleached cell indicates that dye is diffusing into the cell from adjacent cells. Therefore, a decrease in fluorescence from within the adjacent cells is expected. At least 50 individual filaments labeled with carboxyfluorescein, 20 filaments labeled with fluorescein, and 20 filaments labeled with NBD-PC were examined. RESULTS Immunological analysis of Anabaena proteins. To determine whether Anabaena filaments contain a polypeptide similar to a gap junction polypeptide, antibodies raised against, and purified with, the a-peptide from rat liver gap junction connexin32 were used to probe crude extracts from vegetative filaments of Anabaena sp. strain PCC 7120. Figure 4.1 depicts Anabaena proteins and rat liver gap junction proteins separated by electrophoresis on a 12% SDS PAGE gel and probed with antibodies against the a—peptide. A number of protein bands are seen: rat liver gap junction has bands at 27, 50, 70, and 100 kD, and Anabaena has bands at 25, 48, 68, 92, and 100 kD. 111 92- . 66- R 45- a3: 29- Figure 4.1 Proteins resolved on an SDS-PAGE gel, transferred to nitrocellulose, and decorated with antibodies raised against the a-peptide of the rat liver gap junction (10 pg ml") in lanes A and B or preimmune serum (25 pg ml") in lanes C and D. Lanes A and C contain fractions with rat liver gap junction (15 pg each) and lanes 8 and D contain crude extracts of Anabaena (60 pg each). The positions of molecular weight standards are shown to the left. The rat liver gap junction fraction was provided, and the western blot was performed, by Sally Meiners, Dept. of Biochemistry, Michigan State University. 112 The similarity in the pattern of the aggregates formed is striking. Immunological cross-reactivity with the anti-rat liver gap junction antibody and aggregate formation has also been seen for proteins from a variety of higher plant extracts (Meiners and Schindler, 1987, 1989). However, I was never able to obtain a similar result showing the pattern of aggregate formation. In my experiments, a 52-kD protein band was consistently found in crude extracts of Anabaena, often along with a 45-kD band and occasionally another band at 50 kD. The 52- and 45-kD bands were found to be present in the pellet of crude extracts that had been subjected to high speed centrifugation and virtually absent from the soluble fraction of the same extract (Figure 4.2). The same protein bands along with a 27-kD protein were enriched in plasma membrane fractions, and absent from the thylakoid membrane fractions (Figure 4.3). The purity of the membrane fractions was ascertained by measuring the absorption spectra of the fractions. The spectra (Figure 4.4) are nearly identical to those reported by Murata and Omata (1988). Examination of fluorescent labeling and FRAP. Vegetative filaments labeled with carboxyfluorescein are shown in Figure 4.5A. The false color image displays two filaments that cross each other, as well as a single cell. The filaments demonstrate a high degree of labeling, which shows that carboxyfluorescein is able to enter the cells and is maintained within them for up to 3 hours. Figure 4.58 shows the same filaments after photobleaching. The two cells that were bleached show a significant decrease in fluorescence, while the remainder of the cells appear unchanged. Figure 4.5C shows the filaments 113 21* Figure 4.2 Proteins from Anabaena resolved on an SDS—PAGE gel, transferred to nitrocellulose, and decorated with antibodies raised against the a—peptide of the rat liver gap junction (10 pg ml"). Each lane was loaded with 30 pg of protein. Lane 1 contains the soluble fraction of proteins from crude extract that was subjected to high speed centrifugation while lane 2 contains the sedimented proteins from the same sample. Lane 3 contains crude extract that was not fractionated further. The arrows on the left indicate the positions of molecular weight standards. The arrows on the right indicate the positions of the protein bands typically labeled by the antibody. 114 1.2 3 66*» ; 42." ' <52 31* ...—_.——.— -r. ———— Figure 4.3 Proteins from: 1) crude extract, 2) thylakoid membrane fraction, and 3) plasma membrane fraction of Anabaena resolved on an SDS-PAGE gel, transferred to nitrocellulose, and decorated with antibodies raised against the a~peptide of the rat liver gap junction (10 pg ml"). Each lane was loaded with 30 pg of protein. The positions of molecular weight standards are shown on the left. Arrows indicating the positions of labeled protein bands found in the plasma membrane fraction are shown on the right. ABSORBANCE 115 Q 1.0 J ‘3 0.8 - A Q g; ‘3 v 2 0.6 '1 co 3 8 (’3 V 0.4 - g 3 (‘3 0.2 "' 2 A CO L 0.0 I I I I l 300 400 500 600 7°° WAVELENGTH lnm) Figure 4.4 Absorption spectra of (A) thylakoid and (8) plasma membrane fractions from Anabaena sp. strain PCC 7120. The spectra were measured at room temperature (ca. 28°C). 116 Prebleach @ minutes 18 minutes Figure 4.5 Vegetative filaments of Anabaena sp. strain PCC 7120 labeled with carboxyfluorescein diacetate as presented by the ACAS 470 workstation. Two separate, crossing filaments as well as a single cell (box 1) are shown (A). The fluorescence from the same filaments is shown immediately after photobleaching the cells in boxes 1 and 2 (B). The photobleached cells show a significant decrease in fluorescence. Fluorescence of the same filaments 18 minutes later (C) shows no recovery of fluorescence within the photobleached cells. Note that the filaments have moved slightly and the photobleached cells are no longer contained within the boxes. (Magnification = ca. 1500x) 117 after 18 minutes of recovery. It is apparent that no recovery of fluorescence has occurred within the bleached cells.lt should also be noted here that neither of the bleached cells loses any additional fluorescence after bleaching. Therefore, the bleaching process does not appear to disrupt the membranes, a consequence that would allow dye to leak out of the cells. The intensity of fluorescence of the unbleached cells also remains unchanged, indicating that the cells are not bleached by the scanning process. Therefore, the overall procedure does not appear to be detrimental to the cells, and the dyes are sufficiently photostable for use in the assay. Carboxyfluorescein labeling of a filament containing a heterocyst is shown in Figure 4.6. The vegetative cells are highly labeled, whereas the heterocyst displays a low level of fluorescence. Even filaments that are incubated with the dye for up to five hours have only weakly labeled heterocysts. Therefore, carboxyfluorescein does not appear to enter heterocysts either from the medium or from the adjacent vegetative cells. The assay of photobleaching of heterocysts or vegetative cells in these filaments gave the same result as with vegetative filaments: no movement of the dye was observed. In an attempt to obtain labeled heterocysts, vegetative filaments were induced to form heterocysts in the presence of carboxyfluorescein diacetate. Although normal heterocysts were formed, no label was seen within them. Hence, the metabolism of the cells and their ability to differentiate is not affected by the dye. 118 Color Values — 499s ” 3939 3593 3327 E; - 397i *’ 28lS 2559 2393 2947 - I79) — 1535 - 1279 - 1923 - 757 — 511 — 255 Figure 4.6 A single filament of Anabaena sp. strain PCC 7120, grown in nitrogen-free medium and containing a heterocyst (box 1), which was then labeled with carboxyfluorescein diacetate. The heterocyst is very weakly labeled in comparison to the vegetative cells within the filament. The color bar at the right displays relative values of fluorescence. The increasing size of the vegetative cells at the ends of the filament is due to over- and under-focusing, the result of the narrow depth of focus at high magnifications. (Magnification = ca.3000x) 119 Because the negative charge on the carboxyl group of the carboxyfluorescein molecule may have prevented movement of the dye between cells, the FRAP assay was attempted using fluorescein diacetate. The experiment was more difficult to perform because fluorescein diffuses out of the cells with a half-time of 10-15 min. Nonetheless, the results of experiments with fluorescein (data not shown) did not differ from those obtained with carboxyfluorescein. The other cytosolic dyes utilized, methylumbelliferone and carboxyumbelliferone were not helpful in measuring cell-cell communication in Anabaena. The methylumbelliferone diffused out from the cells into the medium much too rapidly (t1 ,2 s 5 min) to allow measurements to be made, whereas the carboxyumbelliferyl B-D-galactopyranoside was never hydrolyzed to its fluorescent form and may not have entered the cells. The pKa of the carboxyl group is very low, so that the dye may have been too highly charged to penetrate the cell membrane. Figure 4.7A shows vegetative filaments labeled with the lipophilic probe NBD-PC. The borders of the cells are intensely labeled, especially at the cell septa, suggesting that the membranes are labeled. However, it is also possible that the outer membrane or cell wall may have absorbed the dye. The lower level of intensity of fluorescence in the interior of the cells suggests that NBD- PC does not enter the cells and label the thylakoid membranes. After photobleaching and scanning, the dye is rapidly lost into the medium (Figures 4.78 and 4.7C). Dispersion of the dye, perhaps a result of hyperpolarization of the membranes due to the intensity of the scanning light, probably makes NBD- 120 PC and similar dyes useless for further studies of Anabaena. Although the NBD- PC should be isolated from the labeled cells to ensure that the dye is intact in these cells and not metabolized to another form. Likewise, labeled cells should be washed with salt solutions or other solvents to ensure that the dye is, in fact, labeling the cytoplasmic membrane of the cells. Figure 4.8 shows a heterocyst-containing filament that was labeled with NBD-PC. The heterocyst does not show any preferential degree of labeling despite the presence of a thickened glycolipid layer. DISCUSSION lntercellular communication in plant and animal cells is mediated by protein channels called plasmodesmata and gap junctions, respectively (Robards, 1976; Unwin and Zampighi, 1980). Structural and regulatory similarities between the two types of channels have been demonstrated (Baron-Epel et al., 1988; Rink et al., 1980; Robards, 1976; Unwin and Zampighi, 1908; Wade et al., 1986). Anti-gap junction antibodies have also shown that various plants contain an immunologically related peptide (Meiners and Schindler, 1987,1989; Yahalom et al., 1991). Although it has not been proven that the cross-reactive protein is associated with the 121 8 minutes ‘. . . .. . ffif: m. 19 minutes Figure 4.7 Vegetative filaments of Anabaena sp. strain PCC 7120 labeled with the lipophilic dye NBD-PC (A). Two filaments that cross as well as a three- celled filament that lies below and adjacent to the horizontal filament are shown. The fluorescence from the same filaments is shown immediately after photobleaching the cells in boxes 1 and 2 (B). The photobleached cells show a significant decrease in fluorescence. Fluorescence of the same filaments 10 minutes later (C) shows no recovery of fluorescence within the photobleached cells and shows that NBD-PC is lost to the medium. Note that the filaments have moved slightly and the photobleached cells are no longer contained within the boxes. (Magnification = ca. 1500x) 122 Color Values - 4095 - 3839 g - 3583 '7 — 3327 . 8:- 3671 .- ‘w - 28lS ' - 2559 _’ ‘ - 2383 ’2. - 2847 -- - 1791 - 153s - 1279 - 1023 - 767 - 511 - 255 - 0 Figure 4.8 A single filament of Anabaena sp. strain PCC 7120, grown in nitrogen-free medium and containing a heterocyst (box 1), which was then labeled with the lipophilic dye NBD-PC. The slightly more intense areas of labeling correspond to the cell septa. (Magnification = ca. 3000x) 123 plasmodesmata, immunolocalization and cell fractionation strongly suggest that such is the case. Results from the current study on microplasmodesmata show that Anabaena also contains a protein with a peptide that is immunologically related to gap junction and plasmodesmatal polypeptides. The western blot performed by Sally Meiners (Figure 4.1) shows a similar behavior of aggregate formation by rat liver gap junction and the cross-reactive protein in Anabaena. The formation of aggregates by rat liver gap junction connexin32 has been well documented. The way in which samples are handled prior to loading on gels greatly influences the pattern of aggregation. Typically, extensive boiling of connexin32 in gel loading buffer leads to the formation of aggregates. Storage of frozen protein extracts has also been found to increase the presence of aggregates. It might also be that differences in the age of the reducing component of the sample buffer could result in different degrees of protein denaturation and of subsequent formation of aggregates. My results indicate that a 52-kD protein is consistently labeled and is present in cytoplasmic membrane fractions. It is possible that the 52-kD band represents a dimer of the 25-kD band seen in Figure 4.1 or of the 27-kD band seen in Figure 4.3. Whether the protein that gives rise to the 52-kD band is localized at the cell septa remains to be determined. The presence of a protein that is antigenically related to a gap junction protein poses the possibility that channel-forming proteins may provide conduits between cells in filaments of Anabaena. The a- peptide forms an alpha helix that is positioned within the cytoplasmic portion of the rat liver gap junction protein (Meiners et al., 1991). Therefore, the a- 124 peptide portion of connexin32 may be involved in interactions between the homohexamers that comprise a gap junction (Hertzberg and Johnson, 1988), or could be involved in regulating the channel. Either of these roles, and therefore the peptide sequence and structure, may have been conserved during evolution or obtained through convergent evolution. In more recent studies, Yahalom et al. (1991) found that antibodies raised against connexin32 decorated a 27-kD protein band (PAP27) that was present in both cell wall and membrane fractions from maize. The western blots for PAP27, like my own blots with extracts from Anabaena, did not show the formation of aggregates that is characteristic of connexin32, but there was an apparent 54—kD dimer band present in their blots. lmmunolocalization of PAP27 showed that the neck region (cytoplasmic portion) of plasmodesmata was most highly labeled, suggesting that the antibodies recognize a cytoplasmic loop of PAP27 just as anti-a antibodies recognize a cytoplasmic domain of connexin32. That plasmodesmata are aqueous channels that provide intercellular cytoplasmic continuity has been confirmed by a number of studies in which the intercellular movement of fluorescent probes was examined (Baron-Epel et al., 1988; Terry and Robards, 1987). Carboxyfluorescein as well as larger fluorescein derivatives (up to MW = 1000) have been shown to permeate plasmodesmata freely. The analogous experiment, using the FRAP assay, in filaments of Anabaena has failed to demonstrate any intercellular movement of these dyes. FRAP assays on plant cells showed that there were two rate components associated with recovery of fluorescence (Baron-Epel et al., 1988). 125 A short phase of recovery that occurred immediately following photobleaching which lasted for ca. 30 seconds was attributed to intracellular redistribution of unbleached dye. The second component was slower than the first, but still gave a measureable rate of recovery of fluorescence within one minute after photobleaching. By comparison, the length of time in which the filaments of Anabaena were examined (18 minutes in Figure 4.5) should have been more that sufficient for visualizing movement of the dye after photobleaching. The external dimensions of plasmodesmata (diameter = 60 nm; Robards, 1976) are much larger than the dimensions of microplasmodesmata (diameter = 5 to 20 nm). However, the morphology of the plasmodesmata is quite complex, and a single plasmodesma may contain six to nine small pores (pore diameter = 1 to 2 nm; Baron-Epel et al., 1988; Terry and Robards, 1989) rather than one large single pore. Therefore, the actual pore size may not vary much between a plasmodesma and a microplasmodesma, especially if each microplasmodesma contains a single pore. Consequently, it is somewhat surprising that carboxyfluorescein does not move between the cyanobacterial cells if microplasmodesmata are in fact aqueous channels between cells. The largest molecule that has been proposed to move between cells is a disaccharide from vegetative cells (Wolk, 1982), possibly sucrose (Schilling and Ehrnsperger, 1985), which may provide the heterocyst with a source of reductant required for nitrogen fixation. The molecular weight of sucrose (MW =342; atomic radius =0.44 nm) is just slightly less than that of carboxyfluorescein (MW = 378; atomic radius =0.63 nm), although fluorescein 126 is a rigid molecule and has different charge characteristics. Work by Terry and Robards (1987) has shown that the hydrodynamic radius of a molecule is the sole factor affecting its ability to permeate plasmodesmata. It may be that sucrose is near the size exclusion limit of the microplasmodesmata and that fluorescein is too large to permeate these channels. It is also possible that sequestration of the fluorescein compounds within cells, possibly within the intra-thylakoid compartment or other compartments of Anabaena cells (if that is where the dyes had been deesterified) or by binding of the charged carboxyl group to cellular components, prevents movement of the dyes. Heterocyst differentiation is known to be accompanied by extensive changes in cellular physiology. Lack of esterase activity in the heterocyst, which could have prevented labeling of the heterocyst, coupled with the slow loss of the dye out of all cells may explain why heterocysts are not labeled even when vegetative cells are labeled prior to the formation of heterocysts. However, failure to visualize labeling of the heterocyst might also be due to other differences associated with the heterocyst, such as the possible presence of compounds that can quench the fluorescence of fluorescein. Some of these possible explanations may be verified or refuted by isolating some of the cellular components and determining whether they show any preferential labeling by the dyes. Isolated heterocysts could be used to check for the presence of any compounds that may quench fluorescence of fluorescein, or dyes with different spectral properties could be used. The only way to prove definitively that cytoplasmic continuity exists will be 127 to find a probe that cannot be specifically transported by cells but can move between cells. A probe with a radius that is smaller than or equal to that of sucrose would determine whether sucrose or other disaccharides can diffuse between cells. Unfortunately, bacterial cells are notoriously difficult to infuse with cytosol-specific dyes (i.e., polar dyes) because i) the small size of bacterial cells precludes the use of microinjection techniques, ii) the complex bacterial cell wall prevents the use of Iiposome fusions, and iii) very few dyes exist that are membrane-permeable and modified to a polar form within cells. To my knowledge, fluoresceins and umbelliferones are the only dyes available in membrane-permeable, hydrolyzable forms that can label the cytoplasm. I have shown that fluorescein and carboxyfluorescein do not move between cells and that umbelliferones diffuse out of cells too rapidly to allow analysis of their intercellular movement. In the case of probes of membrane potential, such as the rhodamine dye used by Severina et al. (1988), the electrical coupling that is observed lends strong support to the presence of aqueous channels, but is correlative and not definitive as ion currents could also propagate through contiguous membranes. However, if probes of membrane potential are used in conjunction with genetic and ultrastructural analysis, it may be possible to prove, or disprove, that electrical coupling occurs via microplasmodesmata. Use of 1“C-labeled substrates has shown that substances do move between cells (Walk, 1968); however, the method does not provide information as to how the movement occurs. It is possible that the movement is directed by specific transport processes. The gene that is disrupted in a heterocyst 128 developmental mutant, M7 (Ernst et al., 1992), in which the protoplasts of differentiating cells do not mature, has been found to encode a protein that is similar to the ATP-binding component of binding-protein-dependenttransporters (l. Maldener, personal communication). Therefore, maturation of heterocysts may require the uptake of compounds from adjacent vegetative cells, with the uptake mediated by specific pumps rather than by diffusion through aqueous channels. Recently, Maddock and Shapiro (1993) showed that chemoreceptor complexes are specifically targeted to the cell poles in E. coli. They suggested that a structure at the cell pole, such as the periseptal annulus made during septation (Cook et al., 1987), may provide the target for localization. The septa of cells of Anabaena lie at the cell junctions. Therefore, a similar mechanism of protein targeting could place specific transporters at the junctions, and the transporters could mediate intercellular communication. The localization of porins at cell poles, where the cell wall is relatively thin, might also provide for electrical coupling between cells. A porin protein from Anabaena variabilis has been partially characterized (Benz and Bt'ihme, 1985). It would be of interest to determine whether this porin, or other porins, is present primarily at the cell septa, and whether porins are required for heterocyst development. A mutant of Anabaena sp. has been found that does not adsorb a particular bacteriophage, and is incapable of aerobic nitrogen fixation (I. Khudyakov, personal communication). Perhaps the binding site for the phage is a porin protein that is required for heterocyst maturation. In conclusion, the presence of a protein that contains a peptide 129 immunologically related to a gap junction protein, and the structural similarities found between microplasmodesmata, plasmodesmata, and gap junctions, make further studies on intercellular communication in Anabaena very attractive. However, the existence of microplasmodesmata has still not been proven. Nonetheless, the potential exists to utilize the genetic manipulability of Anabaena to examine the involvement of the putative microplasmodesmata) protein in cell-cell communication and perhaps to provide a model system for examining how plasmodesmata and gap junctions function. This study has shown that a cytosolic dye different from the ones used in this study, or a membrane potential dye, will have to be used in order to measure intercellular connectivity. Membrane potential dyes may also be useful, along with genetic analysis, in examining whether porins, which may provide interperiplasmic connections, rather than microplasmodesmata, which would provide intercytoplasmic connections, are required for intercellular electrochemical coupfing. BIBLIOGRAPHY BIBLIOGRAPHY Adams, D. G. 1992. Multicellularity in cyanobacteria, p. 341-384. In S. Mohan, C. Dow, and J. A. Cole (ed.), Procaryotic Structure and Function: A New Perspective. Symp. Soc. Gen. Microbiol. 47, Cambridge Adams, D. (3., D. 0. Phillips, J. M. Nichols, and N. G. Carr. 1977. The presence and absence of magic spot nucleotide modulation in cyanobacteria undergoing nutritional shift-down. FEBS Lett. 81:48-52. Akinyanju, J., and R. J. Smith. 1979. Accumulation of ppGpp and pppGpp during nitrogen deprivation of the cyanophyte Anabaena cylindrica. FEBS Lett. 107:173-176. Allen, M.B., and D. l. Arnon. 1955. Studies on nitrogen-fixing blue-green algae. l. Growth and nitrogen fixation by Anabaena cylindrica Lemm. Plant Physiol 30:336-372. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D.J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403-410. Amerik, A. Y.. L. G. Chistyakova, N. I. Ostroumova. A. I. Gurevich, and V. K. Antonov. 1988. Cloning, expression and structure of the shortened, functionally active Ion gene of Escherichia coli. Bioorg. Khim. 14:408-411 (cited in Gill et al., 1993). Arrowsmith, T. J., F. Malpartida, D. H. Sherman, A. Birch, D. A. Hopwood, and J. A. Robinson. 1992. Characterization of actI-homologous DNA encoding polyketide synthase genes from the monensin producer Streptomyces cinnamonensis. Mol. Gen. Genet. 234:254-264. Atherly, A. 1979. Deletion of re/A and relX has no effect on basal or carbon- downshift ppGpp synthesis, p. 53-66. In G. Koch and D. Richter (ed.), Regulation of macromolecular synthesis by low molecular weight mediators. Academic Press, Inc., New York. 130 131 Baker, M. E.. 1991. Geneology of regulation of human sex and adrenal function, prostaglandin action, snapdragon and petunia flower colors, antibiotics, and nitrogen fixation: functional diversity from two ancestral dehydrogenases. Steroids 56:354-360. Bancroft. l., and C. P. Wolk. 1989. Characterization of an insertion sequence (l889 7) of novel structure from the cyanobacterium Anabaena sp. strain M- 131. J. Bacteriol. 171:5949-5954. Baron-Epel, 0., D. Hernandez, L.-W. Jiang, S. Meiners, and M. Schindler. 1988. Dynamic continuity of cytoplasmic and membrane compartments between plant cells. J. Cell Biol. 106:715-721. Baal, B., and J. Lutkenhaus. 1991. FtsZ in Bacillus subtilis is required for vegetative septation and for asymmetric septation during sporulation. Genes Dev. 5:447-455. Baal, B., M. Lowe, and J. Lutkenhaus. 1988. Cloning and characterization of Bacillus subtilis homologs of Escherichia coli cell division genes ftsZ and ftsA. J. Bacteriol. 170:4855-4864. Beck, J., S. Ripka, A. Siegner, E. Schiltz, and E. Schweizer. 1990. The multifunctional 6-methylsalicylic acid synthase gene of Penicil/ium paw/um; its gene structure relative to that of other polyketide synthases. Eur. J. Biochem. 192:487-498. Benson, D. R., and W. B. Silvester. 1993. Biology of Frankia strains, Actinomycete symbionts of actinorhizal plants. Microbiol. Rev. 57:293-319. Benz, R., and H. Btihme. 1985. Pore formation by an outer membrane protein of the cyanobacterium Anabaena variabilis. Biochim. Biophys. Acta. 81 2:286- 292. Beppu, T.. 1992. Secondary metabolites as chemical signals for differentiation. Gene 115:159-165. Bhattacharjea, J. K.. 1985. a-Aminoadipate pathway for the biosynthesis of lysine in lower eukaryotes. Crit. Rev. Microbiol. 12:131-151. Bianchini, G. M., A. C. Pastini, J. P. Muschietti, M. T. Téllez-lfldn, H. E. Martinetto, H. N. Torres, and M. M. Flawia. 1990. Adenylate cyclase activity in cyanobacteria: activation by Ca“-calmodulin and calmodulin-Iike activity. Biochim. Biophys. Acta. 1055:75-81. 132 Black, T. A., Y. Cai, and C. P. Wolk. 1993. Spatial expression and autoregulation of hetR, a gene involved in the control of heterocyst development in Anabaena. Mol. Microbiol. 9:77-84. Bradley, S., and N. G. Carr. 1977. Heterocyst development in Anabaena cylindrica: the necessity for light as an initial trigger and sequential stages of commitment. J. Gen. Microbiol. 101:291-297. Brahamsha, B., and R. Haselkorn. 1991. Isolation and characterization of the gene encoding the principal sigma factor of the vegetative cell RNA polymerase from the cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol. 173: 2442-2450. Brahamsha, 8., and R. Haselkorn. 1992. Identification of multiple RNA polymerase sigma factor homologs in the cyanobacterium Anabaena sp. strain PCC 7120: cloning, expression, and inactivation of the sigB and sigC genes. J. Bacteriol. 174:7273-7282. Brand, F. 1901. Bemerkungen (iber Grenzzellen und iiber spontan rothe lnhaltskdrper der Cyanophyceen. Ber. Deut. Bot. Ges. 19:152-159. Buikema, W. J., and R. Haselkorn. 1991a. Isolation and complementation of nitrogen fixation mutants of the cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol. 173:1879-1885. Buikema, W. J., and R. Haselkorn. 1991b. Characterization of a gene controlling heterocyst differentiation in the cyanobacterium Anabaena 7120. Genes Dev, 5:321-330. Buikema, W. J., and R. Haselkorn. 1992. The hetR gene controls heterocyst development in Anabaena 7120. Abstr. no. 601, 9th lntl. Cong. Nitrogen Fixation. Cancun, Mexico. Burbulys, D., K. A. Trach, and J. A. Hoch. 1991. Initiation of sporulation in B. subtilis is controlled by a multicomponent phosphorelay. Cell 64:545-552. Buttner, M. J. 1989. RNA polymerase heterogeneity in Streptomyces coelicolor A3(2). Mol. Microbiol. 3:1653-1659. Buttner. M. J., K. F. Chater, and M. J. Bibb. 1990. Cloning, disruption and transcriptional analysis of three RNA polymerase sigma factor genes in Streptomyces coelicolor A3(2). J. Bacteriol. 172:3367-3378. 133 Cal. Y., and C. P. Walk. 1990. Use of a conditionally lethal gene in Anabaena sp. strain PCC 7120 to select for double recombinants and to entrap insertion sequences. J Bacteriol. 172:3138-3145. Carafa, Y. d'A., E. Brody and C. Thermes. 1990. Prediction of Rho-independent Escherichia coli transcriptional terminators: statistical analysis of their RNA stem-loop structures. J. Mol. Biol. 216:835-858. Carter, I", H. L., and C. P. Moran, Jr. 1986. New RNA polymerase sigma factor under spoO control in Bacillus subtilis. Proc. Natl. Acad. Sci. USA. 83:9438- 9442. Chater, K. 1989. Multilevel regulation of Streptomyces differentiation. Trends Genet. 5:372-383. Chater, K. F., C. J. Bruton, K. A. Plaskitt, M. J. Buttner, C. Mendez and J. Helmann. 1989. The developmental fate of Streptomyces coelicolor hyphae depends crucially on a gene product homologous with the motility sigma factor of Bacillus subtilis. Cell 59:133-143. Chirala, S. 8., R. Kasturi, M. Pazirandeh, D. T. Szowlow, W.-Y. Huang, and S. J. Wakil. 1989. A novel cDNA extension procedure: isolation of chicken fatty acid synthase cDNA clones. J. Biol. Chem. 264:3750-3757. Clowes, F. A. L., and B. E. Juniper. 1968. Plant Cells. Blackwell Scientific Publications. Oxford. Cohn, D.H., Mileham, A.J., Simon, M.l., Nealson, K.H., Rausch, S.K., Bonam, D., and Baldwin, T.O. 1985. Nucleotide sequence of the luxA gene of Vibrio harveyi and the complete amino acid sequence of the a subunit of bacterial luciferase. J. Biol. Chem. 260:6139-6146. Cook, W. R., F. Kepes, D. Joseleau-Petit. T. J. MacAlister. and L. l. Rothfield. 1987. Proposed mechanism for generation and localization of new cell division sites during the division cycle of Escherichia coli. Proc. Natl. Acad. Sci. USA 84:7144-7148. Cortes, J., S.H. Haydock, G. A. Roberts, D. J. Brevitt, and P. F. Leadlay. 1990. An unusually large multifunctional polypeptide in the erythromycin-producing polyketide synthase of Saccharopolyspora erythraea. Nature 348:176-178. Cutting, S., V. Oke, A. Driks, R. Losick, 8. Lu, and L. Kroos. 1990. A forespore checkpoint for mother-cell gene expression during development in Bacillus subtilis. Cell 62:239-250. 134 Cutting, S., A. Driks, R. Schmidt., B. Kunkel, and R. Losick. 1991a. Forespore- specific transcription of a gene in the signal transduction pathway that governs Pro-0" processing in Bacillus subtilis. Genes Dev. 5:456-466. Cutting, S., S. Roels, and R. Losick. 1991 b. Sporulation operon spa/VF and the characterization of mutations that uncouple mother-cell from forespore gene expression in Bacillus subtilis. J. Mol. Biol. 221:1237-1256. D'Souza, C., M. M. Nakano, N. Corbell, and P. Zuber. 1993. Amino-acylation site mutations in amino acid-activating domains of surfactin synthetase: effects on surfactin production and competence development in Bacillus subtilis. J. Bacteriol. 175:3502-3510. Davis, N. K.. and K. F. Chater. 1990. Spore colour in Streptomyces coelicolor A(3)2 involves the developmentally regulated synthesis of a compound biosynthetically related to polyketide antibiotics. Mol. Microbiol. 4:1679-1692. Debelle, F. and S. B. Sharma. 1986. Nucleotide sequence of Rhizobium meliloti RCR2011 genes involved in host specificity of nodulation. Nucleic Acids Res. 14:7453-7472 Dennis, P. P. 1977. Influence of the stringent control system on the transcription of ribosomal ribonucleic acid and ribosomal protein genes in Escherichia coli. J. Bacteriol. 129:580-588. Dennis, P., and M. Nomura. 1974. Stringent control of ribosomal protein gene expression in Escherichia coli. Proc. Natl. Acad. Sci. USA. 71:3819-3823. Donadio, S., and L. Katz. 1992. Organization of the enzymatic domains in the multifunctional polyketide synthase involved in erythromycin formation in Saccharopolyspora erythraea. Gene 111:51-60. Donadio, S., and M. J. Staver. 1993. ISII36, an insertion element in the erythromycin gene cluster of Saccharopolyspbra erythraea. Gene 1 26: 1 47-1 51 . Donadio, S., M. J. Staver, J. B. McAlpine, S. J. Swanson, and L. Katz. 1991. Modular organization of genes required for complex polyketide biosynthesis. Science 252:675-679. Drews, G. 1959. Beitr'a‘ge zur Kenntnis der phototaktischen Reaktionen der Cyanophyceen. Arch. Protistenk. 104:389-430. Dubnau, E., J. Weir, G. Nair, H. L. Carter III, C. P. Moran, Jr., and I. Smith. 1988. Bacillus sporulation gene spoOH codes for 03° (0“). J. Bacteriol. 170:1054-1062. 135 Dworkin M., and S. M. Gibson. 1964. A system for studying microbial morphogenesis: rapid formation of microcysts in Myxococcus xanthus. Science 146:243-44. Elhai, J., and C. P. Wolk. 1988a. A versatile class of positive-selection vectors based on the non-viability of palindrome-containing plasmids that allows cloning into long polylinkers. Gene 68:119-138. Elhai, J., and C. P. Wolk. 1988b. Conjugal transfer of DNA to cyanobacteria. Methods Enzymol. 167:747-754. Elhai, J., and C. P. Walk. 1990. Developmental regulation and spatial pattern of expression of the structural genes for nitrogenase in the cyanobacterium Anabaena. EMBO J. 9:3379-3388. Ernst, A., T. Black, Y. Cai, J.-M. Panoff, D. N. Tiwari, and C. P. Wolk. 1992. Synthesis of nitrogenase in mutants of the cyanobacteriumAnabaena sp. strain PCC 7120 affected in heterocyst development or metabolism. J. Bacteriol. 174:6025-6032. Fehr, S., and D. Richter. 1981. Stringent response of Bacillus stearothermophilus: evidence for the existence of two distinct guanosine 3’,5'- polyphosphate synthetases. J. Bacteriol. 145:68-73. Ferrari, F. A., K. Trach, and J. A. Hoch. 1985a. Sequence analysis of the spoOB locus reveals a polycistronic transcription unit. J. Bacteriol. 161:556- 562. Ferrarl, F. A., K. Trach, D. LeCoq, J. Spence, E. Ferrari, and J. A. Hoch. 1985b. Characterization of the spoOA locus and its deduced product. Proc. Natl. Acad. Sci. USA 82:2647-2651. Fisher, R. W., and C. P. Walk. 1976. Substance stimulating the differentiation of spores of the blue-green alga Cylindrospermum licheniforme. Nature 259:394-395. Flores, E., and A. M. Mum-Pastor. 1988. Uptake of glutamine and glutamate by the dinitrogen-fixing cyanobacterium Anabaena sp. PCC7120. FEMS Microbiol. Lett. 56:127-130. Flores, E., I. Luque, J. E. Fries, A. Mérida, and A. Herrero. 1993. Control of cyanobacterial nitrogen metabolism by the NtcA regulatory protein, p. 45. Abstract of The Cyanobacterial Workshop. Pacific Grove, CA. 136 Freese, E. 1981. Initiation of bacterial sporulation, p. 1-12. In H.S. Levinson, A. L. Sonenshein, and D. J. Tipper (ed.), Sporulation and Germination. Am. Soc. Microbiol. Washington, D.C. Freese, E., E. B. Freese, E. R. Allen, 2. Olempska-Beer, C. Orrego, A. Varma, and H. Wabiko. 1985. Metabolic initiation of spore development, p. 194-202. In J. A. Hoch and P. Setlow (ed.), Molecular Biology of Microbial Differentiation. American Society for Microbiology. Washington, D.C. Frlas, J. E., A. Mérida, A. Herrero, J. Martin-Nieto, and E. Flores. 1993. General distribution of the nitrogen control gene ntcA in cyanobacteria. J. Bacteriol. 175:5710-5713. Fuma, 8., Y. Fujishima, N. Corbell, C. D'Souza, M. M. Nakano, P. Zuber, and K. Yamane. 1993. Nucleotide sequence of 5’ portion of srfA that contains the region required for competence establishment in Bacillus subtilis. Nucleic Acids Res. 21:93-97. Gholamhoseinian, A., and P. J. Piggot. 1989. Timing of spa/l gene expression relative to septum formation during sporulation of Bacillus subtilis. J. Bacteriol. 171 :5747-5749. Giddings, T. H. and L. A. Staehelin. 1981. Observation of microplasmodesmata in both heterocyst-forming and non-heterocyst forming filamentous cyanobacteria by freeze-fracture electron microscopy. Arch. Microbiol. 129:295-298. Giddings, T. H. and L. A. Staehelin. 1978. Plasma membrane architecture of Anabaena cylindrica: occurrence of microplasmodesmata and changes associated with heterocyst development and the cell cycle. Cytobiologie 16:235-249. Gill, R. E., and M. G. Cull. 1986. Control of developmental gene expression by cell-to-cell interactions in Myxococcus xanthus. J. Bacteriol. 168:341-347. Gill, R. E., M. Karlok, and D. Benton. 1993. Myxococcus xanthus encodes an ATP-dependent protease which is required for developmental gene transcription and intercellular signaling. J. Bacteriol. 175:4538-4544. Goldberg, A. L. 1992. The mechanism and functions of ATP-dependent proteases in bacterial and animal cells. Eur. J. Biochem. 203:9-23. Golden, J. W., S. J. Robinson, and R. Haselkorn. 1985. Rearrangement of nitrogen fixation genes during heterocyst differentiation in the cyanobacterium Anabaena. Nature 314:419-423. 137 Golden, J. W., C. D. Carrasco, M. E. Mulligan, G. J. Schneider, and R. Haselkorn. 1988. Deletion of a 55-kilobase-pair DNA element from the chromosome during heterocyst differentiation of Anabaena sp. strain PCC 7120. J. Bacteriol. 170:5034-5041. Goodwin, P. B. 1976. Physiological and electrophysiological evidence for intercellular communication in plant symplasts, p. 121-129. In B. E. S. Gunning and A. W. Robards (ed.), lntercellular Communication in Plants: Studies on Plasmodesmata. Springer, New York. Gramajo, H. C., E. Takano, and J. B. Mervyn. 1993. Stationary-phase production of the antibiotic actinorhodin in Streptomyces coelicolor A3(2) is transcriptionally regulated. Mol. Microbiol. 7:837-845. Grossman, A. D., and R. Losick. 1988. Extracellular control of spore formation in Bacillus subtilis. Proc. Natl. Acad. Sci. USA 85:4369-4373. 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. Bacteriol. 170: 1 895-1 901 . Gutierrez, 8., B. Diez, E. Montenegro, and J. F. Martin. 1991. Characterization of the Cephalosporium acremonium pchB gene encoding a-aminoadipyl- cysteinyl-valine synthetase, a large multidomain peptide synthetase: linkage to the pch gene as a cluster of early cephalosporin biosynthetic genes and evidence of multiple functional domains. J. Bacteriol. 173:2354-2365. Hfider, D.-P. 1978. Evidence of electrical potential changes in photophobically reacting blue-green algae. Arch. Microbiol. 118:115-119. Hider, D.-P. 1981. Electrical and proton gradients in the sensory transduction of photophobic responses in the blue-green alga, Phormidium uncinatum. Arch. Microbiol. 130:83-86. Hider, D.- P. 1987. Photosensory behavior in procaryotes. Microbiol. Rev. 51:1-21. Hider, D.-P., and K. L. Poff. 1982. Dependence of the photophobic response of the blue-green alga, Phormidium uncinatum, on cations. Arch. Microbiol. 132:345-348. Hagen, D. C., A. P. Bretscher, and D. Kaiser. 1978. Synergism between morphogenetic mutants of Myxococcus xanthus. Dev. Biol. 64:284-296. 138 Hagen, T. J., and L. J. Shimkets. 1990. Nucleotide sequence and transcriptional products of the csg locus of Myxococcus xanthus. J. Bacteriol. 172:15-23. Hahn, J., and D. Dubnau. 1991. Growth stage signal transduction and the requirements for srfA induction in development of competence. J. Bacteriol. 173:7275-7282. Hallam, 8.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. Hara, 0., S. Horinouchi, T. Uozumi, and T. Beppu. 1983. Genetic analysis of A-factor synthesis in Strep tomyces coelicolor A3(2) and Streptomyces griseus. J. Gen. Microbiol. 129:2939-2944. Healy. J., J. Weir, l. Smith. and R. Losick. 1991. Post-transcriptional control of a sporulation regulatory gene encoding transcription factor a“ in Bacillus subtilis. Mol. Microbiol. 5:477-487. Hertzberg, E. L., and R. G. Johnson (eds.). 1988. Gap Junctions. Alan R. Liss, Inc., New York. Heusterspreute, M., Thi, V.H., Emery, S. Tournis-Gamble, 8., Kennedy, N., and Davison, J. 1985. Vectors with restriction site banks. IV. pJRD184, a 3793-bp plasmid vector having 43 unique cloning sites. Gene 37:299-304. Hoch, J. A. 1976. Genetics of bacterial sporulation. Adv. Genet. 18:68-99. Hodgkin, J., and D. Kaiser. 1977. Cell to cell stimulation of movement in nonmotile mutants of Myxococcus. Proc. Natl. Acad. Sci. USA 74:2938-2942. Hodgkin, J., and D. Kaiser. 1979. Genetics of gliding motility in Myxococcus xanthus (Myxobacterales): two gene systems control movement. Mol. Gen. Genet. 172:177-191. Holland, D., and C. P. Wolk. 1990. Identification and characterization of hetA, a gene that acts early in the process of morphological differentiation of heterocysts. J. Bacteriol. 172:3131-3137. Hong, S.-K., K. Morikazu, 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. 139 Hood, E. E., S. Armour, J. D. Ownby, A. K. Hands, and R. A. Bressan. 1979. Effect of nitrogen starvation on the level of adenosine 3’,5’-monophosphate in Anabaena variabilis. Biochim. Biophys. Acta. 588:193-200. Hopwood, D. A., and D. H. Sherman. 1990. Molecular genetics of polyketides and its comparison to fatty acid biosynthesis. Ann. Rev. Genet. 24:37-66 Horinouchi, 8., and T. Beppu. 1984. Production in large quantities of actinorhodin and undecylprodigiosin induced by afsB in Streptomyces lividans. Agric. Biol. Chem. 48:2131-2133. Horinouchi, 8., 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, 8., 0. Here, 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. 1 55: 1 238-1 248. Horinouchi, 8., M. Kito, M. Nishiyama, K. Furuya, 8.-K. Hong, 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. Hu, N.-T., T. Thiel, T. H. Giddings, and C. P. Walk. 1982. Anabaena and Nastoc cyanophages from sewage settling ponds. Virology 114:236-246. Johnston, T.C., R. B. Thompson, and T. 0. Baldwin. 1986. Nucleotide sequence of the quB gene of Vibrio harveyi and the complete amino acid sequence of the (3 subunit of bacterial luciferase. J. Biol. Chem. 261:4805- 4811. Jonas, R. M.,.E. A. Weaver, T. J. Kenney, C. P. Moran, Jr., and W. G. Haldenwang. 1988. The Bacillus subtilis spa/[G operon encodes both 0E and a gene necessary for a"E activation. J. Bacteriol. 170:507-511. Kaiser, D., and R. Losick. 1993. How and why bacteria talk to each other. Cell 73:873-885. Kalos, M., and J. F. lesler. 1990. Defects in contact-stimulated gliding during aggregation by Myxococcus xanthus. J. Bacteriol. 172:6474-6493. 140 Kaplan, H. B., A. Kuspa, and D. Kaiser. 1991. Suppressors that permit A-signal- independent developmental gene expression in Myxococcus xanthus. J. Bacteriol. 1 73: 1 460-1 470. Karmazyn-Campalli, C., C. Bonamy, B. Savelli, and P. Stragier. 1989. Tandem genes encoding a-factors for consecutive steps of development in Bacillus subtilis. Genes Dev. 3:150-157. Khokhlov, A. 8. 1985. Actinomycete autoregulators, p. 791-798. In G. Szabo, 8. Bird, and M. Goodfellow (ed.), Proceedings of the Sixth International Symposium on Actinomycete Biology. Akadémiai Kiado, Budapest. Kim, 8. K., and D. Kaiser. 1990a. C-factor: a cell-cell signaling protein required for fruiting body morphogenesis of M. xanthus. Cell 61:19—26. Kim. S. K., and D. Kaiser. 1990b. Cell motility is required for the transmission of C-factor, an intercellular signal that coordinates fruiting body morphogenesis of Myxococcus xanthus. Genes Dev. 4:896-905. Kim, 8. K., and D. Kaiser. 1990c. Cell alignment required in differentiation of Myxococcus xanthus. Science 249:926-928. Kim, 8. K., and 0. Kaiser. 1991. C-factor has distinct aggregation and sporulation thresholds during Myxococcus development. J. Bacteriol. 173:1722-1728. Khosla, C., R. McDaniel, 8. Ebert-Khosla, R. Torres, D. H. Sherman, M. J. Bibb, and D. A. Hopwood. 1993. Genetic construction and functional analysis of hybrid polyketide synthases containing heterologous acyl carrier proteins. J. Bacteriol. 175:2197-2204. Kleinkauf, H., and H. van Dfihren. 1990. Nonribosomal biosynthesis of peptide antibiotics. Eur. J. Biochem. 192:1-15. Kleinkauf, H., W. Gevers, R. Roskoski, and F. Lipman. 1970. Enzyme-bound phosphopantetheine in tyrocidine biosynthesis. Biochem. Biophys. Res. Commun. 41:1218-1222. Krfitzschmar, J., M. Krause and M. A. Marahiel. 1989. Gramicidin S biosynthesis operon containing the structural genes grsA and grsB has an open reading frame encoding a protein homologous to fatty acid thioesterases. J. Bacteriol. 1 71 :5422-5429. 141 Kroos, L., and 0. Kaiser. 1987. Expression of many developmentally regulated genes in Myxococcus xanthus depends on a sequence of cell interactions. Genes Dev. 1:840-854. Kroos, L., A. Kuspa, and 0. Kaiser. 1986. A global analysis of developmentally regulated genes in Myxococcus xanthus. Dev. Biol. 1 17:252-266. Kroos, L., P. Hartzell, K. Stephens, and 0. Kaiser. 1988. A link between cell movement and gene expression argues that motility is required for cell-cell signaling during fruiting body development. Genes Dev. 2:1677-1685. Kroos, L., B. Kunkel, and R. Losick. 1989. Switch protein alters specificity of RNA polymerase containing a compartment-specific sigma factor. Science 243:526-529. Kunkel, B., L. Kroos, H. Poth, P. Youngman, and R. Losick. 1989. Temporal and spatial control of the mother-cell regulatory gene spell/D of Bacillus subtilis. Genes Dev. 3:1735-1744. Kunkel, B., R. Losick, and P. Stragier. 1990. The Bacillus subtilis gene for the developmental transcription factor 0“ is generated by excision of a dispensable DNA element containing a sporulation recombinase gene. Genes Dev. 4:525- 535. Kuritz, T.. A. Ernst, T. A. Black, C. P. Walk. 1993. High-resolution mapping of genetic loci of Anabaena sp. PCC 7120 required for photosynthesis and nitrogen fixation. Mol. Microbiol. 8:101-110. Kuspa, A., L. Kroos, and 0. Kaiser. 1986. lntercellular signaling is required for developmental gene expression in Myxococcus xanthus. Dev. Biol. 1 17:267- 276. Kuspa, A., L. Plamann, and 0. Kaiser. 1992a. Identification of heat-stable A- factor from Myxococcus xanthus. J. Bacteriol. 174:3319-3326. Kuspa, A., L. Plamann, and 0. Kaiser. 1992b. A-signalling and the cell density requirement for Myxococcus xanthus development. J. Bacteriol. 174:7360- 7369. LaBell, T. L., J. E. Trempy, and W. G. Haldenwang. 1987. Sporulation-specific sigma factor 0'”, of Bacillus subtilis, is synthesized from a precursor protein, P31. Proc. Natl. Acad. Sci. USA 84:1784-1788. 142 Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 277:680-685. Lampa, G., and C. Jacks. 1991. Analysis of two linked genes coding for the acyl carrier protein (ACP) from Arabidopsis thaliana (Columbia). Plant Mol. Biol. 16:469-474. Lang, N. J., and P. Fay. 1971. The heterocysts of blue-green algae. Il. Details of ultrastructure. Proc. Roy. Soc. Lond. B 178:193-203. LaRossa, R., J. Kuner, D. Hagen, C. Manoil, and D. Kaiser. 1983. Developmental cell interactions of Myxococcus xanthus, analysis of mutants. J. Bacteriol. 153:1394-1404. 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. Lazzarini, R. A., M. Cashel, and J. Gallant. 1971. On the regulation of guanosine tetraphosphate levels in stringent and relaxed strains of Escherichia coli. J. Biol. Chem. 246:4381-4385. Leskiw, B. K., R. Mah, E. J. Lawlor, and K. F. Chater. 1993. Accumulation of bldA-specified tRNA is temporally regulated in Streptomyces coelicolor A3(2). J. Bacteriol. 175:1995-2005. Liang, J., L. Scappino, and R. Haselkorn. 1992. The patA gene product, which contains a region similar to CheY of Escherichia coli, controls heterocyst pattern formation in the cyanobacterium Anabaena 7120. Proc. Natl. Acad. Sci. USA 89:5655-5659. Liang, J., L. Scappino, and R. Haselkorn. 1993. The patB gene product, required for growth of the cyanobacterium Anabaena sp. strain PCC 7120 under nitrogen-limiting conditions, contains ferredoxin and helix-turn-helix domains. J. Bacteriol. 175:1697-1704. Lipmann, F. 1980. Bacterial production of antibiotic polypeptides by thiol-linked synthesis on protein template. Adv. Microbiol. Physiol. 21:227-260. Loomis, W. F. (ed.). 1982. The Development of Dictyostelium discoideum. Academic Press, New York. Lapez, J. M., A. Dromerick, and E. Freese. 1981. Response of guanosine-5'- triphosphate concentration to nutritional changes and its significance for Bacillus subtilis sporulation. J. Bacteriol. 146:605-613. 143 Losick, R., and P. Stragier. 1992. Crisscross regulation of cell-type-specific gene expresssion during development in B. subtilis. Nature 355:601-604. Lu, 8., R. Halberg, and L. Kroos. 1990. Processing of the mother-cell a factor, a“, may depend on events occurring in the forespore during Bacillus subtilis development. Proc. Natl. Acad. Sci. USA 87:9722-9726. Lueking, D. R., and H. Goldfine. 1975. The involvement of guanosine 5’- diphosphate-3'-diphosphate in the regulation of phospholipid biosynthesis in Escherichia coli: lack of ppGpp inhibition of acyltransfer from acyl-ACP to sn- glycerol 3-phosphate. J. Biol. Chem. 250:4911-4917. Mackinney, G. 1941. Absorption of light by chlorophyll solutions. J. Biol. Chem. 140:315-322. Maddock, J. R., and L. Shapiro. 1993. Polar location of the chemoreceptor complex in the Escherichia coli cell. Science 259:1717-1723. Maldener, I., W. Lockau, Y. Cai, and C. P. Wolk. 1991. The calcium-dependent protease of the cyanobacterium Anabaena: molecular cloning and expression of the gene in Escherichia coli, sequencing and site-directed mutagenesis. Mol. Gen. Genet. 225:113-120. Mandelstam, J. and S. A. Higgs. 1974. Induction of sporulation during synchronized chromosome replication in Bacillus subtilis. J. Bacteriol. 120:38— 42. Maniatis. T.. E. F. Fritsch. and J. Sambrook. 1982. In Molecular cloning. A laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Mann, N. H., R. Rippka, and M. Herdman. 1991. Regulation of protein phosphorylation in the cyanobacterium Anabaena strain PCC 7120. J. Gen. Microbiol. 137:331-339. Manoil, C., and 0. Kaiser. 1980. Guanosine pentaphosphate and guanosine tetraphosphate accumulation and induction of Myxococcus xanthus fruiting body development. J. Bacteriol. 141:305-15. Marahiel, M. A., M. M. Nakano, and P. Zuber. 1993. Regulation of peptide antibiotic production in Bacillus. Mol. Microbiol. 7:631-636. Marsh. J. L., M. Erfle, and E. J. Wykes. 1984. The plC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional activation. Gene 32:481-485. 144 Matsumura, P., J. J. Rydel, R. Linzmeier, and D. Vacante. 1984. Overexpression and sequence of the Escherichia coli cheY gene and biochemical activities of the CheY protein. J. Bacteriol. 160:36-41. Matveyev, A. V., E. Rutgers, E. deerbfick, and B. Bergman. 1993. A third genome rearrangement involved in Anabaena 7120 heterocysts differentiation revealed by pulsed-field gel electrophoresis, p.65. Abstract of The Cyanobacterial Workshop. Pacific Grove, CA. McClure, W. R.. 1985. Mechanism and control of transcription initiation in procaryotes. Annu. Rev. Biochem. 54:171-204. Meiners, 8., and M. Schindler. 1987. Immunological evidence for gap junction polypeptide in plant cells. J. Biol. Chem. 262:951-953. Meiners, 8., and M. Schindler. 1989. Characterization of a connexin homologue in cultured soybean cells and diverse plant organs. Planta 179:148-155. Meiners, 8., A. Xu, and M. Schindler. 1991. Gap junction protein homologue from Arabidopsis thaliana: evidence for connexins in the plant family. Proc. Natl. Acad. Sci. USA 88:4119-4122. Meinhardt, H. 1982. Models of Biological Pattern Formation. Academic Press. New York. Mendez, C., A. F. Brafia, M. B. Manzanal, and C. Hardisson. 1985. Role of substrate mycelium in colony development in Streptomyces. Can J. Microbiol. 31:446-450. Mendez, C., and K. F. Chater. 1987. Cloning of WhiG, a gene critical for sporulation of Streptomyces coelicolor A3(2). J. Bacteriol. 169:5715-5720. Metzner, l. 1955. Zur Chemie und zum submikroskopischen Aufbau der Zellwande, Scheiden und Gallerten von Cyanophyceen. Arch. Mikrobiol. 22:45- 77. - Mitani, T., J. E. Heinze, and E. Freese. 1977. Induction of sporulation in Bacillus subtilis by decoyinine or hadacidin. Biochem. Biophys. Res. Commun. 77:1118-1125. Mitchison, G. J., and M. Wilcox. 1976. Measurement of an inhibitory zone. Science 191:866-868. 145 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 griseus. J. Bacteriol. 171:4298-4302. Miyake. K., T. Kuzuyama, S. Horinouchi, and T. Beppu. 1990. The A-factor- binding protein of Streptomyces griseus negatively controls streptomycin production and sporulation. J. Bacteriol. 172:3003-3008. Morris, M. E., and 8. Jinks-Robinson. 1991. Nucleotide sequence of the LY82 gene of Saccharomyces cerevisiae: homology to Bacillus brevis tyrocidine synthetase 1. Gene 98:141-145. Munoz-Dorado, J., S. lnouye, and M. lnouye. 1991. A gene encoding protein serine/threonine kinase is required for normal development of M. xanthus, a Gram-negative bacterium. Cell 67:995-1006. Munoz-Dorado, J., S. lnouye, and M. lnouye. 1993. Eukaryotic-like protein serine/threonine kinases in Myxococcus xanthus, a developmental bacterium exhibiting social behavior. J. Cell. Biochem. 51:29-33. Murata N. and T. Omata. 1988. Isolation of cyanobacterial plasma membranes. Methods Enzymol. 167:245-251. Nakano, M. M., and P. Zuber. 1991. The primary role of comA in the establishment of the competent state in Bacillus subtilis is to activate the expression of srfA. J. Bacteriol. 173:7269-7274. Nakano, M. M., R. Magnuson, A. Myers, J. Curry, A. D. Grossman, and P. Zuber. 1991 a. srfA is an operon required for surfactin production, competence development, and efficient sporulation in Bacillus subtilis. J. Bacteriol. 173:1770-1778. Nakano, M. M., L. Xia, and P. Zuber. 1991 b. The transcription initiation region of the srfA operon which is controlled by the comP-comA signal transduction system in Bacillus subtilis. J. Bacteriol. 173:5487-5493. Nakano, M. M., N. Corbell, J. Benson, and P. Zuber. 1992. Isolation and characterization of sfp: a gene that functions in the production of the lipopeptide biosurfactant, surfactin, in Bacillus subtilis. Mol. Gen. Genet. 232:313-321. Neilson, A., R. Rippka, and R. Kunisawa. 1971. Heterocyst formation and nitrogenase synthesis. Arch. Mikrobiol. 76:139-150. 146 Nunn, W. D., and J. E. Cronan, Jr. 1976. Evidence for a direct effect on fatty acid synthesis in relA gene control of membrane phospholipid synthesis. J. Mol. Biol. 102:167-172. Ochi, K. 1988. Nucleotide pools and stringent response in regulation of Streptomyces differentiation, p. 330-337. In Y. Okami, T. Beppu, and H. Ogawara (ed.), Biology of Actinomycetes '88. Japan Scientific Societies Press, Tokyo. Ohmori, M. 1989. cAMP in Anabaena cylindrica: rapid changes in cellular levels in response to changes in extracellular environments. Plant Cell Physiol. 30:91 1-914. Olmedo, G., E. Ninfa, J. Stock, and P. Youngman. 1990. Novel mutations that enhance sporulation in Bacillus subtilis: evidence that phosphorylation of regulatory protein SpoOA controls the initiation of sporulation. J. Mol. Biol. 215:359-372. Oosawa, K., and Y. lmae. 1983. Glycerol and ethylene glycol: members of a new class of repellants of Escherichia coli chemotaxis. J. Bacteriol. 154:104- 1 12. Partridge, S. R., D. Foulger, and J. Errington. 1991. The role of 0’ in prespore- specific transcription in Bacillus subtilis. Mol. Microbiol. 5:757-767. Peltoketo, H., V. lsomaa, 0. Maentausta, and R. V. Vihko. 1988. Complete amino acid sequence of human placental 17B-hydroxysteroid dehydrogenase. FEBS Lett. 239:73-77. Peoples, O. P. and A. J. Sinskey. 1989. Poly-B-hydroxybutyrate biosynthesis in Alcaligenes eutrophus H16. J. Biol. Chem. 264:15293-15297. Perego, M., G. B. Spiegelman, and J. A. Hoch. 1988. Structure of the gene for the transition state regulator, aer: regulator synthesis is controlled by the SpoOA sporulation gene in Bacillus subtilis. Mol. Microbiol. 2:689-699. Perego, M., S. P. Cole, 0. Burbulys, K. Trach, and J. A. Hoch. 1989. Characterization of the gene for a protein kinase which phosphorylates the sporulation-regulatory proteins SpoOA and SpoOF of Bacillus subtilis. J. Bacteriol. 1 71 :6187-61 96. Perego, M., J.-J. Wu, G. B. Spiegelman, and J. A. Hoch. 1991a. Mutational dissociation of the positive and negative regulatory properties of the SpoOA sporulation transcription factor of Bacillus subtilis. Gene 100:207-212. 147 Perego, M., C. F. Higgins, S. R. Pearce, M. P. Gallagher, and J. A. Hoch. 1991 b. The oligopeptide transport system of Bacillus subtilis plays a role in the initiation of sporulation. Mol. Microbiol. 5:173-185. Plamann, L., A. Kuspa, and D. Kaiser. 1992. Proteins that rescue A-signal- defective mutants of Myxococcus xanthus. J. Bacteriol. 174:3311-3318. Potts, M., H. Sun, K. Mockaitis, P. J. Kennelly, 0. Reed, and N. K. Tonks. 1993. A protein-tyrosinelserine phosphatase encoded by the genome of the cyanobacterium Nastoc commune UTEX 584. J. Biol. Chem. 268:7632-7635. Prentki, P., A. Binda, and A. Epstein. 1991. Plasmid vectors for IS I-promoted deletions in cloned DNA: sequence analysis of the omega interposon. Gene 103: 1 7-23. Price, V. L., and L. R. Brown. 1981. Transcriptional inhibition and production of guanosine polyphosphate in Bacillus subtilis. J. Bacteriol. 147:752-756. Rather, P.N., R. Coppolecchia, H. De Grazia, and C. P. Moran, Jr. 1990. Negative regulator of aG-controlled gene expression in stationary-phase Bacillus subtilis. J. Bacteriol. 172:709-715. Rawlings, M. and J. E. Cronan, Jr. 1992. The gene encoding Escherichia coli acyl carrier protein lies within a cluster of fatty acid biosynthetic genes. J. Biol. Chem. 267:5751-5754. Ricca, E., 8. Cutting, and R. Losick. 1992. Characterization of bofA, a gene involved in intercompartmental regulation of Pro-0" processing during sporulation in Bacillus subtilis. J. Bacteriol. 174:3177-3184. Rink, T. J., L. W. Tslen, and A. E. Warner. 1980. Free calcium in Xenopus embryos measured with ion-selective electrodes. Nature 283:658-660. Robards, A. W. 1976. Plasmodesmata in higher plants, p. 15-37. In B. E. S. Gunning and A. W. Robards (ed.), lntercellular Communication in Plants: Studies on Plasmodesmata. Springer, New York. Rogerson, A. C., P. Rowell, and W. D. P. Stewart. 1978. Guanosine 5’- diphosphate-3’-diphosphateintheheterocystouscyanobacteriumAnabaenopsis circularis: possible lack of involvement in cellular metabolism. FEMS Microbiol. Lett. 3:299-303. Rudner, D. 2., J. R. LeDeaux, K. Ireton, and A. D. Grossman. 1991. The spoOK locus of Bacillus subtilis is homologous to the oligopeptide permease locus and is required for sporulation and competence. J. Bacteriol. 173:1388-1398. 148 Sager, B. and 0. Kaiser. 1993. Two cell-density domains within the Myxococcus xanthus fruiting body. Proc. Natl. Acad. Sci. USA 90:3690-3694. Sato, T., Y. Samori, and Y. Kobayashi. 1990. DNA rearrangement during sporulation in Bacillus subtilis. J. Bacteriol. 172:1092-1098. Schauer, A., M. Ranes, R. Santamaria, J. Guijarro, E. Lawlor, C. Mendez, K. Chater, and R. Losick. 1988. Visualizing gene expression in time and space in the morphologically complex, filamentous bacterium Streptomyces coelicolor. Science 240:768-772. Schauer, A. T. 1988. Visualizing gene expression with Iuciferase fusions. Trends Biotechnol. 6:23-27. Schauer, A. T., and H. Im. 1988. Visualizing gene expression in Streptomyces coelicolor. p. 346-350. In Y. Okami, T. Beppu, and H. Ogawara (ed.), Biology of Actinomycetes '88. Japan Scientific Societies Press, Tokyo. Schilling, N., and K. Ehrnsperger. 1985. Cellular differentiation of sucrose metabolism in Anabaena variabilis. Z. Naturforsch. 40c:776-779. Schmid, G. 1923. Das Reizverhalten kiinstlicher Teilstiicke, die Kontraktilitét und das osmotische Verhalten der Oscillatoria jenensis. Jahrb. Wiss. Bot. 62:328-419. Schmidt, R., P. Margolis, L. Duncan, R. Coppolecchia, C. P. Moran, Jr., and R. Losick. 1990. Control of developmental transcription factor 0‘ by sporulation regulatory proteins SpollAA and SpollAB in Bacillus subtilis. Proc. Natl. Acad. Sci. USA 87:9221-9225. Severina, l. I. and V. P. Skulachev. 1984. Ethylrhodamine as a fluorescent penetrating cation and a membrane potential-sensitive probe in cyanobacterial cells. FEBS Lett. 165:67-71. Shimkets, L. J. 1990. Social and developmental biology of the myxobacteria. Microbiol. Rev. 54:473-501. Shimkets, L. J., and M. Dworkin. 1979. Excreted adenosine is a cell density signal for the initiation of fruiting body formation in Myxococcus xanthus. Develop. Biol. 84:51-60. Shimkets, L. J., and H. Rafiee. 1990. ngA, an extracellular protein essential for Myxococcus xanthus development. J. Bacteriol. 172:5299-5306. 149 Simon, R. D. 1971. Cyanophycin granules from the blue-green alga Anabaena cylindrica: a reserve material consisting of copolymers of aspartic acid and arginine. Proc. Natl. Acad. Sci. USA 68:265-267. Slabas, A. R., 0. Chase, l. Nishida, N. Murata, C. Sidebottom, R. Safford, P. S. Sheldon, R. G. 0. Kekwick, D. G. Hardie. and R. W. MacKintosh. 1992. Molecular cloning of higher-plant 3-oxoacyl-(acyl carrier protein) reductase. Sequence identities with the nodG-gene product of the nitrogen-fixing soil bacterium Rhizobium meli/oti. Biochem. J. 283:321-326. Smith, G., and J. D. Ownby. 1981. Cyclic AMP interferes with pattern formation in the cyanobacterium Anabaena variabilis. FEMS Microbiol. Lett. 11:175-180. Spencer, A., E. Muller, J. E. Cronan, Jr., and T. A. Gross. 1977. relA gene control of the synthesis of lipid A fatty acid moieties. J. Bacteriol. 130:114- 1 17. Spiegelman, G. B., B. Van Hoy, M. Perego, J. Day, K. Trach, and J. A. Hoch. 1990. Structural alterations in the SpoOA regulatory protein of Bacillus subtilis which suppress mutations at several spoO loci. J. Bacteriol. 172:5011-5019. Stein, 0., and S. N. Cohen. 1989. A cloned regulatory gene of Streptomyces lividans can suppress the pigment deficiency phenotype of different developmental mutants. J. Bacteriol. 171:2258-2261. Stevens, 0. M., and J. Errington. 1990. Differential gene expression during sporulation in Bacillus subtilis: structure and regulation of the spoil/D gene. Mol. Microbiol. 4:543-552. Stock, J. B., A. J. Ninfa, and A. M. Stock. 1989. Protein phosphorylation and regulation of adaptive response in bacteria. Microbiol. Rev. 53:450-490. Stock, J. B., A. M. Stock, and J. M. Mottonen. 1990. Signal transduction in bacteria. Nature 344:395-400. Stragier, P. 1991. Establishment of forespore-specific gene expression during sporulation of Bacillus subtilis, p. 385-405. In J. A. Cole, F. Mohan, and C. Cow (ed.), Procaryotic Structure and Function. Symp. Soc. Gen. Microbiol. Cambridge University Press, Cambridge. Stragier, P., and R. Losick. 1990. Cascades of sigma factors revisited. Mol. Microbiol. 4:1801-1806. 150 Stragier, P., J. Bouvier, C. Bonamy, and J. Szulmajster. 1984. A developmental gene product of Bacillus subtilis homologous to the sigma factor of Escherichia coli. Nature 312:376-378. Stragier, P., C. Bonamy, and C. Karmazyn-Campelli. 1988. Processing of a sporulation sigma factor in Bacillus subtilis: how morphological structure could control gene expression. Cell 52:697-704. Stragier, P., B. Kunkel, L. Kroos, and R. Losick. 1989. Chromosomal rearrangement generating a composite gene for a developmental transcription factor. Science 243:507-512. Strauch, M. A., G. B. Spiegelman, M. Perego, W. C. Johnson, D. Burbulys, and J. A. Hoch. 1989. The transition state transcription regulator aer of Bacillus subtilis is a DNA binding protein. EMBO J. 8:1615-1621. Strauch, M. A., V. Webb, G. Spiegelman, and J. A. Hoch. 1990. The SpoOA protein of Bacillus subtilis is a repressor of the aer gene. Proc. Natl. Acad. Sci. USA 87:1801-1805. Strauch, M. A., D. de Mendoza, and J. A. Hoch. 1992. cis-Unsaturated fatty acids specifically inhibit a signal-transducing protein kinase required for initiation of sporulation in Bacillus subtilis. Mol. Microbiol. 6:2909-2917. Sun, 0., P. Stragier, and P. Setlow. 1989. Identification of a new a-factor involved in compartmentalized gene expression during sporulation of Bacillus subtilis. Genes Dev. 3:141-149. Sun, 0., R. M. Cabrera-Martinez, and P. Setlow. 1991. Control of transcription of the Bacillus subtilis spolllG gene, which codes for the forespore specific transcription factor O'G. J. Bacteriol. 173:2977-2984. Tanaka, K., T. Shiina, and H. Takahashi. 1988. Multiple principal sigma factor homologs in eubacteria: identification of the "rpoD box." Science 242:1040- 1042. Terry, 8. R.. and A. W. Robards. 1987. Hydrodynamic radius alone governs the transport of molecules through plasmodesmata. Planta 171:145-157. Thomas, J., J. C. 'Meeks, C. P. Wolk, P. W. Shaffer, S. M. Austin. and W.-S. Chien. 1977. Formation of glutamine from [‘3Nlammonia, [‘ledinitrogen, and [“Clglutamate by heterocysts isolated from Anabaena cylindrica. J. Bacteriol. 1 29:1545-1555. 151 Timberlake, W. E. 1991. Temporal and spatial controls of Aspergillus development. Curr. Opin. Gen. Dev. 1:351-357. Trach, K., and J. A. Hoch. 1989. The Bacillus subtilis spoOB stage 0 sporulation operon encodes an essential GTP-binding protein. J. Bacteriol. 171:1362-1371. Trach, K., J. W. Chapman, P. J. Piggot, and J. A. Hoch. 1985. Deduced product of the stage 0 sporulation gene spoOF shares homology with the SpoOA, OmpR and SfrA proteins. Proc. Natl. Acad. Sci. USA. 82:7260-7264. Trempy, J. E., C. Bonamy, J. Szulmajster, and W. G. Haldenwang. 1985. Bacillus subtilis sigma factor, a”, is the product of the sporulation-essential gene spollG. Proc. Natl. Acad. Sci. USA 82:4189-4192. Trieu-Cuot, P., C. Carlier, P. Martin, and P. Courvalin. 1987. Plasmid transfer by conjugation from Escherichia coli to Gram-positive bacteria. FEMS Microbiol. Lett. 48:289-294. Turner, N. E., S. J. Robinson, and R. Haselkorn. 1983. Different promoters for the Anabaena glutamine synthetase gene during growth using molecular or fixed nitrogen. Nature 306:337-342. Turgay, K., M. Krause, and M. A. Marahiel. 1992. Four homologous domains in the primary structure of GrsB are related to domains in a superfamily of adenylate-forming enzymes. Mol. Microbiol. 6:529-546. Vater, J. 1989. Lipopeptides, an interesting class of microbial secondary metabolites, p. 27-37. In U. P. Schlunegger (ed.), Biologically active molecules. Springer-Verlag, Berlin. Unwin, P. N. T., and G. Zampighi. 1980. Structure of the junction between communicating cells . Nature 283:545-549. Vega-Palas, M. A., F. Madueno, A. Herrero, and E. Flores. 1990. Identification and cloning of a regulatory gene for nitrogen assimilation in the cyanobacterium Synechococcus sp. strain PCC 7942. J. Bacteriol. 172:643-647. Vega-Palas, M. A., E. Flores, and A. Herrero. 1992. NtcA, a global nitrogen regulator from the cyanobacterium Synechococcus that belongs to the Crp family of bacterial regulators. Mol. Microbiol. 6:1853-1859. Vieira, J. and J. Messing. 1982. The pUC plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19:259-268. 152 Vieira, J., and J. Messing. 1987. Production of single-stranded DNA. Methods Enzymol. 1 53:3-1 1. Wade, M. H., J. E. Trosko, and M. Schindler. 1986. A fluorescence photobleaching assay of gap junction-mediated communication between human cells. Science 232:525-528. Wei, T.-F., T. S. Ramasubramanian, F. Pu, and J. W. Golden. 1993. Anabaena sp. strain PCC 7120 bifA gene encoding a sequence-specific DNA-binding protein cloned by in vivo transcriptional interference selection. J. Bacteriol. 175:4025-4035. Weinrauch. Y., N. Guillen, and D. A. Dubnau. 1989. Sequence and transcription mapping of Bacillus subtilis competence genes comB and comA, one of which is related to a family of bacterial regulatory determinants. J. Bacteriol. 171:5362-5375. Weinrauch, Y., R. Penchev, E. Dubnau, I. Smith, and D. Dubnau. 1990. A Bacillus subtilis regulatory gene product for genetic competence and sporulation resembles sensor protein members of the bacterial two-component signal- transduction systems. Genes Dev. 4:860-872. Westpheling, J., M. Ranes, and R. Losick. 1985. RNA polymerase heterogeneity in Streptomyces coelicolor. Nature 313:22-27. Wilcox, M., G. J. Mitchison, and R. J. Smith. 1973a. Pattern formation in the blue-green alga, Anabaena. I. Basic mechanisms. J. Cell Sci. 12:707-723. Wilcox, M., G. J. Mitchison, and R. J. Smith. 1973b. Pattern formation in the blue-green alga, Anabaena. ll. Controlled proheterocyst regression. J. Cell Sci. 13:637-649. Wildon. D. C., and F. V. Mercer. 1963. The ultrastructure of the heterocyst and akinete of the blue-green algae. Arch. Mikrobiol. 47:19-31. Wildermuth, H. 1970. Development and organization of the aerial mycelium in Streptomyces coelicolor. J. Gen. Microbiol. 60:43-50. Willey, J., R. Santamaria, J. Guijarro, M. Geistlich, 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. 153 Willey, J., J. Schwedock, and R. Losick. 1993. Multiple extracellular signals govern the production of a morphogenetic protein involved in aerial mycelium formation by Streptomyces coelicolor. Genes Dev. 7:895-903. Wolk, C. P. 1964. Experimental studies on the development of a blue-green alga. Dissertation. Rockefeller Institute. New York City. Wolk, C. P. 1967. Physiological basis of the pattern of vegetative growth of a blue-green alga. Proc. Natl. Acad. Sci. USA 57:1246-1251. Wolk, C. P. 1968. Movement of carbon from vegetative cells to heterocysts in Anabaena cylindrica. J. Bacteriol. 96:2138-2143. Wolk, C. P. 1982. Heterocysts. In The Biology of Cyanobacteria. Carr, N.G., and Whitton, B.A. (eds.) London: Blackwell Scientific Publ., pp. 359-386. Wolk, C. P. 1989. Alternative models for the development of the pattern of spaced heterocysts in Anabaena (Cyanophyta). Pl. Syst. Evol. 164:27-31. Walk, 0. P. 1991. Genetic analysis of cyanobacterial development. Curr. Opin. Gen. Dev. 1:336-341. Walk, 0. P. and M. P. Ouine. 1975. Formation of one-dimensional patterns by stochastic processes and by filamentous blue-green algae. Develop. Biol. 46:370-382. Wolk, C. P., A. Vonshak, P. Kehoe, and J. Elhai. 1984. Construction of shuttle vectors capable of conjugative transfer from Escherichia coli to nitrogen-fixing filamentous cyanobacteria. Proc. Natl. Acad. Sci. USA 81:1561-1565. Wolk, C. P., Y. Cai, L. Cardemil, E. Flores, B. Hohn, M. Murry, G. Schmetterer, B. Schrautemeier, and R. Wilson. 1988. Isolation and complementation of mutants of Anabaena sp. strain PCC 7120 unable to grow aerobically on dinitrogen. J. Bacteriol. 170:1239-1244. Wolk, C. P., Y. Cai, and J.-M. Panoff. 1991. Use of a transposon with Iuciferase as a reporter to identify environmentally responsive genes in a cyanobacterium. Proc. Natl. Acad. Sci. USA 88:5355-5359. Wolk, C. P., J. Elhai, T. Kuritz, and D. Holland. 1993. Amplified expression of a transcriptional pattern formed during development of Anabaena. Mol. Microbiol. 7:441-445. Wright, F., and M. J. Bibb. 1992. Codon usage in the G +C-rich Streptomyces genome. Gene 113:55-65. 154 Wu, J. J., M. 0. Howard, and P. J. Piggot. 1989. Regulation of transcription of the Bacillus subtilis spa/IA locus. J. Bacteriol. 171:692-698. Yahalom, A., R. D. Warmbrodt, D. W. Laird, 0. Traub, J.-P. Revel, K. Willecke, and B. L. Epel. 1991. Maize mesocotyl plasmodesmata proteins cross-react with connexin gap junction protein antibodies. Plant Cell 3:407-417. Yamashita, 8., H. Yoshikawa, F. Kawamura, H. Takahashi. T. Yamamoto, Y. Kobayashi. and H. Saito. 1986. The effect of spoO mutations on the expression of spoOA- and spoOF-lacZ fusions. Mol. Gen. Genet. 205:28-33. Zheng, L., and R. Losick. 1990. Cascade regulation of spore coat gene expression in Bacillus subtilis. J. Mol. Biol. 212:645-660. Zuber, P., J. Healy, H. L. Carter III, 8. Cutting, C. P. Moran, Jr., and R. Losick. 1989. Mutation changing the specificity of an RNA polymerase sigma factor. J. Mol. Biol. 206:605-614. Zuber, P. and R. Losick. 1987. Role of Aer in SpoOA- and SpoOB-dependent utilization of a sporulation promoter in Bacillus subtilis. J. Bacteriol. 169:2223- 2230.