Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies on the transcription of three overlapping operons encoding photosynthesis genes from the phototrophic… Wellington, Cheryl Lea 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A1 W44.pdf [ 8.91MB ]
Metadata
JSON: 831-1.0101048.json
JSON-LD: 831-1.0101048-ld.json
RDF/XML (Pretty): 831-1.0101048-rdf.xml
RDF/JSON: 831-1.0101048-rdf.json
Turtle: 831-1.0101048-turtle.txt
N-Triples: 831-1.0101048-rdf-ntriples.txt
Original Record: 831-1.0101048-source.json
Full Text
831-1.0101048-fulltext.txt
Citation
831-1.0101048.ris

Full Text

STUDIES O N THE TRANSCRIPTION OF THREE OVERLAPPING OPERONS ENCODING PHOTOSYNTHESIS GENES F R O M T H E PROTOTROPHIC BACTERIUM RHODOBACTER CAPSULATUS by Cheryl Lea Wellington B.Sc. (Honors), University of Alberta, 1985 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF G R A D U A T E STUDIES (DEPARTMENT OF MICROBIOLOGY) We accept this thesis as conforming to the required standards THE UNIVERSITY OF BRITISH COLUMBIA OCTOBER 1990 © Cheryl Lea Wellington, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of jfy ifsisPk-iiyCo, The University of British Columbia Vancouver, Canada DE-6 (2/88) ii ABSTRACT A Rhodobacter capsulatus photosynthesis gene was isolated by creating in-frame fusions in a lacZ transcriptional/translational vector, and selecting for those that directed oxygen-regulated levels of /?-galactosidase in R. capsulatus. One lacZ fusion isolate was used to identify an open reading frame (ORF) of unknown function and flanking sequences that promoted initiation of transcription. Interposon mutagenesis experiments identified the ORF as the bchC gene, which encodes an enzyme that catalyses the penultimate step in the biosynthesis of bacteriochlorophyll a, and also showed that the bchC gene formed an operon with the bchA gene. The nucleotide sequence of this bchC gene and its 5' regulatory region were determined. The deduced amino acid sequence showed that the bchC gene encodes a 33 kDA protein that has hydrophobic segments that could interact with a lipid membrane, and that this putative BchC protein contains a potential bacteriochlorophyll a binding site. Deletion analysis, Sl-nuclease protection, and primer extension experiments showed that promoter activity was associated with sequences to which a 5' end mapped, and that these sequences had significant similarity to the proposed promoter regions of several other R. capsulatus photosynthesis genes. RNA blotting and Sl-nuclease protection end-mapping experiments using bipartite probes provided direct evidence that the mRNA transcripts of the bchCA operon overlap those of the two flanking operons, the crtEF and the puf operons, such that the crtEF, bchCA, and puf operons may be cotranscribable, and that RNA polymerase may initiate transcription at one of several promoters. The significance of these overlapping mRNAs was evaluated using two interposon mutant strains, one that prevented crtEF transcripts from overlapping those the bchCA and puf operons, and the other that prevented both crtEF and bchCA transcripts from overlapping those of the puf operon. The results suggested that transcriptional readthrough stimulates promoter activity. Moreover, a pufB::lac'Z fusion could be expressed from the bchCA promoter equally as well as from the puf promoter, suggesting that these overlapping transcripts are functionally significant in the chromosomal context. iii TABLE OF CONTENTS A B S T R A C T i i L I S T O F T A B L E S v L I S T O F F I G U R E S vi A B B R E V I A T I O N S A N D S Y M B O L S v i i i A C K N O W L E D G E M E N T S x D E D I C A T I O N xi INTRODUCTION 1 MATERIALS AND METHODS 18 1. G r o w t h and maintainance of bacterial strains 18 2. D N A purification and manipulation 20 3. P lasmids 20 4. Measurement of /?-galactosidase specific activity 27 5. Absorption and fluorescence spectroscopy 28 6. D N A sequence analysis 28 7. Isolation of R. capsulatus R N A 29 8. D N A and R N A blotting and preparation of probes 29 9. Primer extension analysis 30 10. Sl-nuclease protection m R N A end site mapping and preparation of probes 31 RESULTS 32 1. Isolation and identification of the bchC gene from R. capsulatus 32 la . Isolation of an oxygen-regulated promoter and the 5' coding region of a gene from the photosynthesis gene cluster of R. capsulatus 32 l b . Identification of the gene fused to lac'Z 34 2. Sequence of the bchC gene and characterization of its product 38 2a. Nucleotide sequence analysis of the bchC gene 38 2b. Analysis of the putative amino acid sequence of the BchC protein 42 iv 3. Promoter mapping of the bchCA operon 46 3a. Genetic analysis 46 3b. Biochemical analysis 47 3c. Comparison with other photosynthesis gene promoter sequences 51 4. Characterization of the bchCA mRNA transcripts 53 4a. R N A blot analysis of bchCA mRNA transcripts 53 4b. Sl-nuclease protection 5' end site mapping of bchCA mRNA transcripts 57 4c. Sl-nuclease protection 3' end site mapping of crtEF mRNA transcripts 60 4d. Sl-nuclease protection end site mapping of the 3' end within the bchC gene 63 4e. Sl-nuclease protection 3' end site mapping of bchCA mRNA transcripts 68 4f. Sl-nuclease protection 5' end site mapping of puf mRNA transcripts 71 5. Functional significance of the overlapping crtEF, bchCA, and puf mRNA . transcripts 76 5a. Consequences of interrupting crtEF transcriptional readthrough on expression of the bchCA operon 76 5b. Consequences of interrupting crtEF transcriptional readthrough on initiation of puf operon transcription 85 5c. Consequences of interrupting all transcriptional readthrough on initiation of puf operon transcription 91 5d. Cis/Trans test of puf promoter stimulation by bchCA readthrough transcription 94 5e. Contribution of bchCA readthrough transcription to the expression of a pufB::lac'Z fusion 95 D I S C U S S I O N 100 1. Overlapping mRNA transcripts of the crtEF, bchCA, and puf operons 100 2. Functional significance of overlapping crtEF, bchCA, and puf mRNAs 106 3. Analysis of the bchCA promoter 112 4. Comparison of Bch protein predicted sequences 115 5. Concluding remarks 118 R E F E R E N C E S 120 V LIST OF TABLES Table I. Bacterial strains 19 Table II. Plasmids 21 Table III. Codon usage in the bchC gene 44 Table IV. j3-galactosidase specific activities of a pufB::lac'Z fusion in Bchl" and Bchl + strains 96 Table V. Alignment of Ala/Gly-X-X-X His sequences in predicted protein sequences of bch genes 117 vi LIST O F F I G U R E S Figure 1. Diagrammatic representation of the photosynthetic apparatus of R. sphaeroides 2 Figure 2. Genetic map of the R. capsulatus photosynthesis gene cluster in pRPS404 6 Figure 3. Tetrapyrrole biosynthesis: Early steps 12 Figure 4. Bchl a biosynthesis: Late steps 14 Figure 5. Construction of plasmids pJPl, pJPlOO, and pJPlOl 23 Figure 6. Construction of plasmids pXCA6::HQ and pXCA6::HQA44 24 Figure 7. Genetic map of the 5.5 kb EcoRI-H fragment of pRPS404 33 Figure 8. Southern blot of bchC interposon mutant strains 36 Figure 9. Absorption spectrum of pigments accumulated in cultures of R. capsulatus CW100 and R. capsulatus CW100 (pCW2) 37 Figure 10. Sequencing strategy for the bchC gene and flanking sequences 39 Figure 11. Nucleotide and deduced amino acid sequence of the bchC gene and flanking regions 40 Figure 12. Codon preference plot of the nucleotide sequence in Figure 12 43 Figure 13. Secondary structure analysis of the putative BchC protein 45 Figure 14. /J-galactosidase specific activities of cells containing bchCv.lac'Z 5' deletion constructs : 48 Figure 15. Low resolution Sl-nuclease protection 5' end site mapping of bchCA mRNA transcripts 49 Figure 16. Primer extension analysis of the predominant 5' end of bchCA mRNA 50 Figure 17. Comparison of the bchCA promoter sequence with other R. capsulatus photosynthesis gene promoter sequences 52 Figure 18. Blot of R. capsulatus BIO and R. capsulatus BIO (pCW2) R N A 54 Figure 19. Sl-nuclease protection 5' end site mapping of bchCA mRNA transcripts 58 Figure 20. Sl-nuclease protection 3' end site mapping of crtEF mRNA transcripts 61 Figure 21. Sl-nuclease protection mapping of the 3' ends within the bchC coding region 64 Figure 22. Titration of bc/zC-internal Sl-nuclease protection 3' end site mapping probe 67 Figure 23. Comparison of bcftC-ihternal 3' ends in R. capsulatus SB1003 and vii R. capsulatus DE324 69 Figure 24. Sl-nuclease protection 3' end site mapping of bchCA mRNA transcripts 72 Figure 25. Sl-nuclease protection 5' end site mapping of puf mRNA transcripts 74 Figure 26. Absorbency scans of R. capsulatus CW100, R. capsulatus DE324, R. capsulatus SB1003, and R. capsulatus BP503 77 Figure 27. Titration of the bchCA Sl-nuclease protection 5' end site mapping probe 82 Figure 28. Comparison of bchCA 5' mRNA ends in R. capsulatus SB1003 and R. capsulatus DE324 83 Figure 29. Titration of the puf Sl-nuclease protection 5' end site mapping probe 86 Figure 30. Comparison of puf 5' mRNA ends in R. capsulatus SB1003 and R. capsulatus DE324 88 Figure 31. Comparison of puf 5' mRNA ends in R. capsulatus BIO and R. capsulatus CW100 92 Figure 32. Contribution of bchCA readthrough transcription to expression of a pufBr.lac'Z fusion 98 Figure 33. Model of the overlapping transcripts of the R. capsulatus crtEF, bchCA, and puf operons 101 viii ABBREVIATIONS A N D SYMBOLS Ap ampicillin ATP adenosine triphosphate bch R. capsulatus bacteriochlorophyll biosynthesis gene Bchl bacteriochlorophyll bp base pair(s) BSA bovine serum albumin B800-850 R. capsulatus light-harvesting II (LH II) complex B870 R. capsulatus light-harvesting I (LH I) complex ca. approximately cfu colony forming units C i Curie cpm counts (Cerenkov) per minute crt R. capsulatus carotenoid biosynthesis gene cyt cytochrome c^-dGTP 7-deaza-2'-deoxyguanosine-5'-triphosphate 5-ALA 5-aminolevulinic acid dATP deoxyadenosine-5'-triphosphate ddNTP dideoxynucleoside-5'^triphosphate dGTP deoxyguanosine-5'-triphosphate DMSO dimethylsulfoxide D N A deoxyribonucleic acid dNTP deoxynucleoside-5'-triphosphate EDTA ethylenediamihetetraacetic acid EtBr ethidium bromide GC guanosine/cytosine ICM intracytoplasmic membrane kb kilobase(s) Km kanamycin lacZ E. coli bfeffl-galactosidase gene ix LH light harvesting antenna MOPS 4-morpholinepropanesulfonic acid mRNA messenger ribonucleic acid NADPH nicotinamide adenine dinucleotide phosphate (reduced) nt nucleotide (s) ntr E. coli nitrogen regulation genes oNPG orr/zo-nitrophenyl-befa-D-galactoside ORF open reading frame PEP phosphoenolpyruvate puc R. capsulatus puc operon puf R. capsulatus puf operon puh R. capsulatus puh operon R resistance/resistant RC reaction center RNA ribonucleic acid rpm revolutions per minute rRNA ribosomal ribonucleic acid S sensitive/sensitivity SDS sodium dodecyl sulfate T thymine Tc tetracycline Tn transposon Tris Tris (hydroxyrhethyl) aminomethane U units XGal 5-bromo-4-chloro-3-indoyl-/?-D-galactopyranoside denotes novel joint ' (prime) denotes a truncated gene at the indicated 5' or 3' side () or [] denotes plasmid-carrier state X ACKNOWLEDGEMENTS I would like to thank the many friends and colleagues who have made the last five years much more than just work. I wish to thank Tom Beatty, my research supervisor, for his incredible enthusiasm, for his critical and creative thinking, and for his unlimited patience in teaching an inexperienced graduate student the many joys of doing science. The members of Tom's lab also merit recognition for their generosity and friendship. These exceptional people are Mary Forrest, Anthony Zucconi, Tim Lilburn, Joanna Zilsel, Farahad Dastoor, and Heidi LeBlanc. I cannot adequately express how much Tom and the members of the "Beatty bunch" have enriched my stay at UBC, both from scientific and personal perspectives. I would also like to acknowledge that this work was supported by Natural Sciences and Engineering Council of Canada grant #A-82796. I am grateful for the support of an NSERC postgraduate fellowship, a Killam predoctoral fellowship, and a University of British Columbia graduate fellowship. I would like to express my appreciation to those who have contributed directly to this work. Many thanks to Barry Marrs and Debra Young for providing numerous different R. capsulatus strains and for many helpful discussions, to John Priatel for constructing three of the plasmids used in this thesis, to Carl Bauer and Marie Alberti for sharing unpublished results, and to the members of my thesis advisory committee, Beverley Green, Tony Warren, and George Spiegelman, for their suggestions and support. I would very much like to acknowledge Nancy and Fereydoun Sajjadi, John McClean, and Patrick Blaney for the friendships that have meant so much along the way. Also, many thanks to the members of the Keen and Wellington families, for the love that has sustained me throughout all the ups and downs of the last several years. Most significantly, I wish to acknowledge my husband, Stephen, who has been the greatest friend and most valued colleague of all, and who has brought so much meaning into my life. XI DEDICATION This thesis is dedicated to the memory of my mother, Elsie Doreen Keen, whose strength and dignity will always be an inspiration to me. 1 INTRODUCTION A distinguishing feature of the purple nonsulfur bacteria is their ability to use anaerobic photosynthesis as a means of generating the energy required for growth. In addition to their photosynthetic capability, these Gram-negative organisms are equipped with metabolisms of extraordinary diversity. Every known type of energy conversion have been observed in these bacteria (113). One particular species, Rhodobacter capsulatus, is unique among living cells in that it can grow photoautotrophically, photoheterotrophically, chemoautotrophically, chemoheterotrophically, and fermentatively (67, 68, 69). This degree of metabolic flexibility among some members of the Rhodospirillaceae immediately raises the question of how these cells sense their surroundings and adjust their metabolism in order to exploit the potential of the environment. A striking metabolic change these bacteria undergo can be seen as a dramatic increase in pigmentation as the cells switch from aerobic chemoheterotrophic growth to anaerobic photosynthetic growth. Aerobic chemoheterotrophic growth of R. capsulatus is supported by respiration using an electron transport pathway similar to that found in other Gram-negative respiring cells (47, 73, 74, 78). When the oxygen tension is reduced below a threshold level of approximately 2% p02 (21), the cell induces the differentiation of the cytoplasmic membrane into the physically continuous but functionally and structurally distinct intracytoplasmic membrane system (the ICM) that contains all of the components necessary for photosynthetic electron transport (see 32, 53, and 99 for reviews). The ICM of R. capsulatus contains three pigment-protein complexes unique to the photosynthetic electron transport chain (Figure 1). These include two light-harvesting (LH) antenna complexes that have been designated B870 and B800-850 based on their absorption maxima (22). These antenna complexes absorb the majority of photons striking the cell, and funnel this energy by a process known as exciton transfer to a third pigmented complex known as the reaction center (RC) complex (77,112). Each of these three complexes have bound to them the photosynthetic pigments bacteriochlorophyll (Bchl) a and several aliphatic carotenoids derived from neurosporene; it is these pigments that give oxygen-limited cells their visible coloration (112). Absorption of light energy by the RC complex initiates cyclic electron transport, which results in the production of a proton-motive force across the photosynthetic membrane that is used to drive the synthesis of 2 Figure 1. Diagrammatic representation of the photosynthetic apparatus of R. sphaeroid.es, a facultatively phototrophic bacterium that is closely related to R. capsulatus. The periplasmic side of the membrane is at the top, and the cytoplasmic side is on the bottom. Shown are the B800-850 and B875 (comparable to the B870 complex of R. capsulatus) complexes as well as the reaction center (RC) complex. The cyt bc\ complex (with the mobile cyt ci protein) and the ATP phosphohydrolase complex are also shown. Arrows depict the movement of electrons and protons during cyclic electron transport. The succinate-DH and N A D H - D H complexes act to maintain the correct redox state of the quinone pool. Abbreviations are: Q and QH2, oxidized and reduced (respectively) quinones; ADP, adenosine 5'-diphosphate; ATP adenosine 5'-triphosphate; D H , dehydrogenase; N A D H , reduced nicotinamide adenine dinucleotide. Reproduced from Kiley and Kaplan (53) with permission of the American Society for Microbiology. 3 ATP (4, 54, 78, 87). Synthesis of the ICM thus equips oxygen-deprived cells with an alternative electron transport pathway to that used in aerobic respiration, and photosynthetic growth of pigmented cells can begin if sufficient light is available. Conversion of light energy into a form usable by the cell normally begins with the absorption of a photon by one of the antenna complexes. In R. capsulatus, each B800-850 complex is thought to be an oligomer, the minimal unit of which is composed of three molecules of Bchl a and one molecule of carotenoid that are bound to two integral membrane polypeptides known as B800-850« and B800-850/J, as well as a third polypeptide, named B800-850y, that is thought to be associated with the B800-850 complex, but whose precise function remains unknown (32,34, 35,101,102). Once induced by a decrease in oxygen tension, the number of B800-850 complexes relative to RC complexes varies inversely with incident light intensity, in order to maximize the surface area of antenna Bchl molecules available for efficient photon capture (33, 96, 97). It is for this reason that the B800-850 complexes are also known as the variable antenna complexes, and they are not essential for photosynthetic growth if the light intensity is adequate (100). Transfer of exciton energy from the B800-850 complex to the RC complex is thought to involve the B870 complex as an intermediate (77). The R. capsulatus B870 complex contains two molecules of Bchl a and two molecules of carotenoids per two hydrophobic polypeptides (B870a and B870j3) in a 1:1 ratio (32, 88). In contrast to the B800-850 complexes, the ratio of B870 complexes to RC complexes remains fixed at approximately 15-20:1 regardless of light intensity (96, 97). The RC complex contains three polypeptides designated RC-H (heavy), RC-M (medium), and RC-L (light), based on their apparent mobilities through SDS polyacrylamide gels (32, 84). They are present in a 1:1:1 stoichiometric ratio along with four molecules of Bchl a (two of which form the "special pair"), two molecules of bacteriopheophytin, two molecules of ubiquinone, a non-heme iron, and one carotenoid (32, 86). All of these cofactors are bound in a precise symmetrical arrangement by the integral membrane RC-M and RC-L subunits (17). It is the particular orientation with which these cofactors are positioned by the RC-M and RC-L subunits that allows vectorial electron flow to result in the formation of an electrochemical gradient across the photosynthetic membrane. Analysis of the crystal structure of the RC from the related species Rhodopseudomonas viridis by Deisenhofer et al. (30) was a tremendous accomplishment because it revealed the exact position of the chromophores within the RC complex, and gave direct support to the path of electron flow through the RC as was proposed by 4 many previous spectroscopic studies. Since that time, the RC from the closely related species Rhodobacter sphaeroides has also been crystallized (2). Because the diffraction data show that the major structural features of both the R. viridis and R. sphaeroides RC complexes have been conserved (2), and studies of the R. capsulatus RC indicate that its structure is very similar to that of R. sphaeroides (37, 120), the overall molecular architecture of the purple bacterial RC can most likely be extended to include the R. capsulatus RC. Cyclic electron transfer (see 4, 54, 87 for reviews, and Figure 1) begins with the photooxidation of a "special pair" of Bchl a molecules in the RC. The cofactors bound by the RC allow movement of electrons across the membrane, where a loosely bound quinone molecule is reduced. After acquisition of protons obtained from the cytoplasm, the reduced quinone diffuses through the membrane, and delivers the electrons and protons to the membrane-bound ubiquinol cytochrome C£. oxidoreductase (cyt bc\) complex. The cyt bc\ complex is responsible for transfer of electrons back across the membrane to the mobile periplasmic cyt c^, and also for extrusion of protons to the periplasmic side of the membrane. Cyclic electron flow is completed by the transfer of electrons from the cyt ci carrier to photooxidized RCs. Although recent studies have shown that cyclic electron flow can be completed in R. capsulatus by reduction of the RC in the absence of cyt C2, this direct transfer is not as efficient and results in a slower growth rate, especially at low light intensities (27). The net result of photosynthetic electron flow is the formation of a proton gradient across the photosynthetic membrane, which enables the cell to synthesize ATP by movement of protons through the membrane-bound ATP phosphohydrolase complex (4). The goals of much current research are to understand, at the molecular level, the mechanisms by which oxygen tension regulates the induction of the ICM, and how light intensity regulates the final composition of the photosynthetic apparatus. The complexity of this problem can be appreciated by considering that not only must the cell coordinate the expression of at least thirty known genes (1, 3,106,110, 111, 121,125, 126) in order to satisfy the pigment-protein stoichiometry required by each complex, but also that their expression must be regulated in a sufficiently flexible manner to allow for adaptation to fluctuations in incident light intensity (55, 85, 96, 97, 130). Because the facultative metabolism of R. capsulatus makes the maintainance and study of photosynthetically incompetent mutant strains possible, R. capsulatus has emerged as an attractive model system with which to investigate photosynthetic differentiation. Moreover, the development of suitable genetic systems in R. 5 capsulatus over the past ten years has made possible the analysis of photosynthesis gene expression by using the techniques of molecular biology (99). As a result, our understanding of the regulation of photosynthesis gene expression has recently advanced very rapidly. The photosynthesis genes in question can be divided into three groups: those coding for the enzymes of the Bchl biosynthetic pathway (the bch genes), those coding for the enzymes of the carotenoid biosynthetic pathway (the crt genes), and those coding for the structural polypeptides of the L H and RC complexes. The B800-850 a, B, and y polypeptides are encoded by the pucB, pucA, and pucE genes, respectively, and comprise three genes of the proposed pucBACDE operon (111). The first gene in the puf operon, pufQ, encodes a product that is essential for Bchl a biosynthesis (9), and is thought to be important for assembly of B800-850 complexes (see below and ref. 38). The four central genes of the puf operon, pufB, puf A, pufL, and pufM, encode the j3 and a polypeptides of the B870 complex and the L and M subunits of the RC complex, respectively (125). The final gene in the puf operon, pufX, encodes a product that is required for photosynthetic growth in minimal media but whose specific function remains obscure (65). The RC-H subunit is encoded by the puhA gene as part of an independent transcriptional unit (125). All known photosynthesis genes from R. capsulatus (except the puc genes which encode the high-light dispensable B800-850 polypeptides) have been mapped to a 50 kb region of the R. capsulatus chromosome, and have been isolated as an intact group on the plasmid RP1 derivative pRPS404 (Figure 2, and ref. 110). These essential photosynthesis genes are arranged such that the bch and crt genes are flanked by the puh and puf operons. Because this pattern of photosynthesis gene clustering has also been found in the related species R. sphaeroides (26), it is suggestive that such an arrangement may have significance for the evolution of photosynthetic organisms, perhaps to facilitate a single-step lateral transfer of essential photosynthesis genes between species. Alternatively, tight clustering of functionally related genes may be important for coordinated regulation of their expression. Regulation of the expression of these photosynthesis genes occurs at several levels, including transcription initiation, mRNA turnover, activity of the Bch and Crt biosynthetic enzymes, and assembly of mature photosynthetic complexes. Shifting the oxygen tension from 20% to 2% pC>2 induces the transcription of these photosynthesis genes, but the magnitude of the response seems to vary with the specific gene assayed. Although transcription initiation of puf, 6 Figure 2. Genetic map of the R. capsulatus photosynthesis gene cluster in pRPS404, showing the 50 kb region of the R. capsulatus chromosome that contains all essential photosynthesis genes. The shaded boxes represent the coding regions of genes for which sequence data is available. Arrows represent the direction of transcription. Genetic designations are: puhA, gene encoding the RC-H subunit; F1696; gene of unknown function (see 121); art, carotenoid biosynthesis gene; bch, Bchl biosynthesis gene; puf, operon encoding the B870a, B870/J, RC-L, RC-M, and two other genes (PufQ and PufX) with less well defined functions (see text). The relevant restriction sites that separate this cluster into specific fragments are: EcoRl (E) and BamHI (B). pufX pufM pufL puf A pufB pufQ bchA.3 bchA.2 bchA.1 bchC c r t F c r t E crtD c r t C crtK crtB c r t I crtA bchD bchG bchE bchB bchF bchK bchH bchB bchE F1696 puhA m o o w a o u w o u w o o w o o w o o w w o o w s o o pa o o o w o o w w o O (X, W O o w o o W Cd w PQ u E fl CQ CQ m CQ e fO PQ CQ PQ CQ fl CQ CQ PQ ffl CQ o 6 fl CQ Q e CQ CQ CQ fl PQ fl CQ 43 8 puc, and bch genes is oxygen-regulated, the increase in transcription, or at least mRNA accumulation, of the puf and puc genes appears to be approximately 5-10 fold greater than that observed for the bch genes (21,127,128). The observed modest increase in mRNA levels observed for the bch genes relative to the puf and puc genes can be rationalized in terms of the relative amounts of protein products required for formation of the photosynthetic apparatus. The Bch enzymes produce Bchl a catalytically, whereas this Bchl a is bound stoichiometrically by the puf- and puc- encoded polypeptides. Zhu and Hearst (127, 128) have shown with dot blots that the level of crt mRNA seems to be either unresponsive to this stimulus, or to be somewhat greater in cells grown under high oxygen conditions (127,128). These results were surprising, since carotenoids accumulate significantly only in cells grown under low oxygen or photosynthetic conditions (24). However, Giuliano et al. (40) have shown that the 5' mRNA ends of the crt A, crtl, crtC, and crtE genes are present at much higher levels in oxygen-limited cells than in highly aerated cells, which suggests that the earlier results of Zhu and Hearst (127,128) may have been due to the lack of gene-specific probes. Also, Biel and Marrs (15) have demonstrated that oxygen-regulated accumulation of carotenoids is dependent upon Bchl a production. The link between Bchl and carotenoid regulation may be advantageous for the cell, because a major function of carotenoids is to protect the organism from photooxidative damage that may result when singlet oxygen is produced by the interaction of 0"2 with Bchl in the triplet state (23, 58). Because the promoters of the R. capsulatus puf, and presumably other, photosynthesis genes are not recognized by Escherichia coli DNA-dependent RNA polymerase (50), one of the aims of current research is to define the sequence of a photosynthesis gene promoter. Most studies of photosynthesis gene promoter structure to date have focussed primarily on the puf operon promoter. Adams et al. (1) have provided the most detailed analysis of the puf operon promoter structure, and have defined by site-directed mutagenesis two nucleotides that are essential for puf promoter activity. A great deal of D N A sequence data from many crt genes (3) as well as the puc (111) and bchCA (117; M . Alberti, pers. comm.) operons is now available, which has allowed an extensive comparative analysis between the puf promoter sequences and the putative promoter sequences of these other photosynthesis genes. However, because high-resolution structural and functional studies have not yet been performed on any other photosynthesis gene promoter, it is not yet possible to define unambiguously a consensus promoter 9 sequence for an R. capsulatus photosynthesis gene. Regulation of photosynthesis gene expression also occurs at many posttranscriptional levels. Belasco et al. (11) have proposed that the fixed molar ratio of B870 to RC complexes in the ICM is regulated primarily at the level of selective segmental degradation of the polycistronic puf mRNA, and have subsequently shown that a relative short intercistronic mRNA sequence, thought to be acting as a barrier to mRNA degradation, is one of the factors required for longevity of the more stable segment (pufBA) of the puf message (20). As well, recent studies have indicated that the light-regulated variability of the proportion of B800-850 complexes in the ICM is due to some unknown translational or posttranslational mechanism (130). Assembly of the various pigments and polypeptides to form functional complexes appears to be very tightly regulated. No pools of free Bchl, or RC or L H pigment-binding polypeptides can be detected in wild-type cells (24, 57), suggesting that mechanisms exist to coordinate the synthesis of the individual elements that make up the photosynthetic apparatus. Supporting this idea are the results of several investigations that report tantalizing interactions between the products of various photosynthesis genes. Bauer et al. (9) have demonstrated that the formation of Bchl a requires expression of pufQ, and that the level of Bchl a synthesized correlates with the level of pufQ gene expression. They conclude that the pufQ gene product may act as a "carrier" protein whose function is to shuttle intermediates of Bchl biosynthesis between the various enzymes of the Bchl biosynthetic pathway, as was originally proposed by Lascelles (61). Supporting their hypothesis is the observation that the PufQ protein is of the same size and hydrophobic character as a protein found associated with intermediates of Bchl biosynthesis (92). Forrest et al. (38) have shown that PufQ is also required for assembly of B800-850 complexes, although this dependency may be due to the role of PufQ in Bchl a biosynthesis. Also intriguing is the interaction between carotenoid biosynthesis and formation of B800-850 complexes. It has been observed that some mutant strains unable to synthesize coloured carotenoids do not form B800-850 complexes. Bartley and Scolnik (6) have shown that the crtl gene product is required for assembly of B800-850 complexes, and the amount of B800-850 complexes formed is correlated with the amount of expression of the crtl gene. One possible reason for the connection between carotenoid and B800-850 biosynthesis is that carotenoid-10 deficient cells are susceptible to photooxidative killing (23, 58). Therefore, by decreasing the size of the variable antenna complex in the ICM, the cell may reduce the probability of receiving photooxidative damage. Recently, Tichy et al. (Ill) have discovered that there are three more genes downstream of the R. capsulatus pucBA operon, called pucC, pucD, and pucE, respectively. Because insertion of Tn5 into the pucC gene eliminated assembly of B800-850 complexes, the authors suggest that the pucC gene product may be a regulatory protein that is important for B800-850 complex assembly. As noted above, the pucE gene product encodes the B800-850/ subunit. The product of the pucD gene has no defined function at present. A recent report by Lee et al. (63) indicates the R. sphaeroid.es puc operon may be similarly extended. Finally, Sockett et al. (106) have reported that sequences adjacent to the puh A gene are required for formation of B875 complexes in R. sphaeroxd.es, suggesting that the puhA gene may also be part of larger operon, and that the other genes play regulatory roles in assembly of ICM photosynthetic complexes. Taken together, the results of these studies suggest that there are many regulatory interactions that control the coordinated biosynthesis and assembly of the components of the ICM, and that we have but a very preliminary understanding of how the assembly of this specialized structure is orchestrated. It is becoming increasingly apparent that the molecular mechanisms by which R. capsulatus produces and maintains a functional photosynthetic membrane in response to the environmental signals of oxygen and light are very complex and multilayered. The tetrapyrrole pigment Bchl a in R. capsulatus is necessary for charge separation in the RC and for transfer of light energy between the three pigment-protein complexes in the photosynthetic apparatus. Synthesis of mature Bchl a is also required for stabilization of the peptide components of these complexes. When mutants blocked in Bchl biosynthesis are grown under conditions that would normally induce the formation of the photosynthetic apparatus, the structural polypeptides of the RC and LH complexes are present at very low steady-state amounts, apparently because they are synthesized and then rapidly degraded (31, 56, 57). It has also been demonstrated that mutants blocked in Bchl biosynthesis are impaired in carotenoid accumulation under low oxygen conditions (15). These observations suggest that the production of Bchl is required for formation of many of the components of the photosynthetic apparatus. Therefore, an improved understanding of the mechanism and regulation of Bchl biosynthesis 11 would greatly enhance our understanding of the biogenesis of the photosynthetic apparatus. Bchl a, like other tetrapyrroles, is synthesized by a series of reactions that can be conveniently divided into two halves, the early steps of which are common to the biosynthetic pathways of other tetrapyrroles (such as hemes and corrinoids), and begin with the formation of ^aminolevulinic acid (5-ALA). Eight molecules of 5-ALA are condensed and modified in a series of reactions to form the tetrapyrrole intermediate protoporphyrin IX (Figure 3, and refs. 51, 52). The pathways of Bchl and heme biosynthesis diverge at this point, by the insertion of either Mg2+ or Fe^+ into the tetrapyrrole macrocycle of protoporphyrin IX, whereas the vitamin Bi 2 biosynthetic branch splits off from the common pathway before protoporphyrin IX is synthesized. The work of Lascelles and colleagues has been instrumental in elucidating many of these early steps in tetrapyrrole biosynthesis, and also in developing purification strategies and assays for most of these early enzymes (51, 52, 62, and references therein). In contrast, the enzymes that convert protoporphyrin IX to Bchl a have been much less amenable to experimental analysis. These enzymes are encoded by the bch genes, of which thirteen are currently known in R. capsulatus (14,110,121,123, M . Alberti, pers. comm.). A mutation in any of the bch genes blocks a biosynthetic reaction and results in the accumulation of a colored intermediate that is usually easily identifiable on the basis of its absorption or fluorescence emission spectra. Through the analysis of many bch mutant strains, a "metabolic grid" has been proposed for ordering the various reactions that comprise the late steps in Bchl biosynthesis (Figure 4, and ref. 89), and the locations of many of the bch genes have been mapped on the R. capsulatus chromosome (14, 110,121,123). However, because the activities encoded by all the cloned bch genes have not been demonstrated, we cannot be certain that the proposed path of the late steps of Bchl a biosynthesis is complete. As noted above, transcription of bch genes is only moderately responsive to changes in oxygen tension (differences of approximately 2-4 fold, refs. 14,21,127,128) and is unresponsive to variations in light intensity (13). These findings were surprising because the specific Bchl content of low oxygen or low light-grown cells is on the order of at least 50 times greater than the content of high oxygen or high light-grown cells (96, 97). In addition, introduction of oxygen into or increasing the light intensity incident upon an anaerobic culture of wild-type cells results in an almost instantaneous cessation of Bchl biosynthesis, without accumulation of Bchl intermediates or diversion of these intermediates into heme or corrinoid biosynthesis (24, 62). This 12 Figure 3. Tetrapyrrole biosynthesis: Early steps. Shown is the biosynthetic sequence leading to the formation of protoporphyrin IX and the branch points at which the routes to Bchl a and heme biosynthesis diverge. Reproduced from Jones (51) with the permission of Plenum Press. 13 COOH CH, CH, S-CoA Succinyl CoA CH, NH, I COOH Glycine ALA hv»iheta«e ALA rlchyflf *K COOH I CH, i CH, I C 0 I CH, I NH, 5-Aminolevulinic acid (ALA) COOH I CH, I CH, \ c— II H C V COOH CH, CH,NH, Porphobilinogen (PBGI COOH COOH COOH Protoporphyrinogen IX CH, I CH, 1 1 CH, 1 1 CH, 1 1 COOH 1 COOH Piotopofphyrin IX Mg PtotopOfphyrin Pfotoheme • Bacieriochlorophyll a 14 Figure 4. Bchl a biosynthesis: Late steps. Shown is the proposed biosynthetic route from protoporphyrin IX leading to the formation of Bchl a. The red-most absorption peak is shown in brackets beside the name of characterized intermediates. The Bch enzymes that catalyse each step, where known, are also shown. 15 BchE C O O C H , ' Magnesium protoporphyrin Magnesium 2,4-divinylpheoporphyrin as monomethyl ester (P590) monomethyl ester (P631) C H - C H , C , H , BchF C H , + H2O 2-Devinyl-2-hydroxyethyl chlorophyllide a (P668) + 2H H,C C , H , BchF C H , + H2O H XH HC C = 0 C H , , C H ] C O O C H , C O O H 2-Desacetyl-2-hydroxyethyl bacteriochlorophyllide a (P720) •2H BchC + 2H Chlorophyllide a (P665) + 2 H BchA c,H, C H , 2-Desacetyl-2-vinyl bacteriochlorophyllide a (P730) C O O C H , Protochlorophyllide H,C H C — c , I I ^ o H,C C O O C H , C O O H Bacteriochlorophyllide a (P770) C O O - C H ^ ; H - C - ( C H » ) « - C H - ( C H , > , _ C H - < C H , > , - C H - C H , > I I I Bacteriochlorophyll a C H ' CH> CH> CH> 16 interruption of Bchl biosynthesis is just as quickly reversible, provided that cell growth during the period of inhibition has not markedly diluted the existing Bchl biosynthetic enzymes (24). These observations suggest that although transcription of bch genes is subject to regulation by oxygen to a moderate degree, the regulation of the activities of most or all of the enzymes of the Bchl biosynthetic pathway by environmental factors is likely to be quantitatively more important for control of Bchl biosynthesis. Despite this information, detailed knowledge of the structure and regulation of bch genes and the enzymes encoded by them remains fragmentary. At the time I began my thesis work, no bch gene had been sequenced, although several had been cloned. As well, the lack of specific assays for most of the Bchl biosynthesis enzymes had significantly impeded their purification. However, progress has been made by the partial purification of certain enzymes of the magnesium branch of tetrapyrrole biosynthesis: the S-adenosyl-L-methionine:magnesium protoporphyrin methyltransferase from chromatophores of R. sphaeroides (45) and from the cytoplasmic fraction of etiolated wheat seedlings and Euglena gracilis (44), the 4-vinylreductase from the cytoplasmic fraction of etiolated wheat seedlings (90), and the NADPH:protochlorophyllide reductase from the wheat etioplast membrane (59). Because of the difficulties noted above in the direct detection of bch and other photosynthesis gene products, I chose to identify potential R. capsulatus photosynthesis genes by fusing their 5' and regulatory regions to a truncated E. coli lac'Z gene, and screening for host cells that contain a recombinant plasmid that displays the desired regulation of fi-galactosidase activity. In the case for a search for oxygen-regulated photosynthesis genes from R. capsulatus, the screen simply consists of plating R. capsulatus cells containing a fusion bank on a medium containing Xgal, and choosing for study those colonies whose anaerobic centers have turned blue. After isolation of putative photosynthesis genes by this method, their function in photosynthesis can be evaluated by in vitro mutation and replacement of the chromosomal allele with the mutated copy. By use of this approach, I cloned and sequenced an open reading frame (ORF) that I subsequently identified as the bchC gene, which encodes an enzyme that catalyzes the penultimate step in the biosynthesis of Bchl a (Figure 4, and refs. 14, 72, 91). The nucleotide and predicted amino acid sequence of the BchC protein was the first primary structure of a chlorophyll biosynthesis gene and its product to be published. I demonstrated that the bchC 17 and bchA genes are cotranscribed as the bchCA operon, which, in agreement with the results of Young et al. (124), helped to settle a controversy in the literature over the direction of transcription of this operon (14,129). The results of 5' end site mapping experiments showed that a 5' mRNA end site is located within the cloned regulatory region, and that it is positioned very near sequences with significant similarity to the puf operon promoter (1, 8). Analysis of the specific /?-galactosidase activities of bchCdac 'Z fusion constructs containing various 5' deletions showed that the presence of this 5' end is correlated with bchCA promoter activity. RNA blotting experiments and extensive end site mapping analyses showed that the mRNA transcripts of the bchCA operon overlap those of its two flanking operons, the crtEF and the pw/operons (110), such that is is possible that these three operons are cotranscribable, from crtEF, through bchCA, to puf. Finally, I demonstrated that this arrangement of overlapping mRNA transcripts has considerable functional significance to the expression of these related photosynthesis genes. 18 MATERIALS A N D METHODS 1. Growth and maintainance of bacterial strains. The R. capsulatus bacterial strains used in this study have each been described. Their relevant genotypes and phenotypes are listed in Table I. E. coli strains C600 (r"m +, ref. 12), RB404 (16), JM83 (122), and S-17 (103) were used as hosts for the various plasmids used in these studies. Recombinant M13 phage was maintained in E. coli JM101 (122). E. coli HB101 containing the conjugative plasmid pRK2013 was used to supply mobilization functions in triparental matings between a donor E. coli strain and a recipient R. capsulatus strain, as described previously (95). R. capsulatus was routinely grown in rich YPS media (0.3% yeast extract, 0.3% peptone, 2 m M MgS04, 2 m M CaCl2r p H 6.8), or in a malate-minimal salts medium (RCV, ref. 10). High oxygen cultures were grown in Erlenmeyer flasks filled to 8% of their nominal capacity and shaken at 300 rpm in a rotary shaking water bath. Low oxygen cultures were grown in Erlenmeyer flasks filled to 80% of their nominal volume and shaken at 150 rpm. Photosynthetic liquid cultures were grown in completely filled screw-cap tubes that were held in a glass-sided water bath in front of a bank of Lumiline 60W lamps (General Electric Co.) at an incident light intensity of 55 W/m^, and photosynthetic plate cultures were grown in an anaerobic jars (BBL) that were similarly incubated. All liquid cultures were inoculated to an initial optical density between 10 and 20 Klett units (between 4 x lCr7 and 8 x 10^ cfu/ml) and growth was monitored by measuring the optical density of the culture (at 660 nm) using a Klett-Summerson colorimeter. High and low oxygen cultures were inoculated from highly aerated precultures in log phase. Photosynthetic cultures were inoculated from oxygen-limited precultures in log phase. Cells grown for /J-galactosidase specific activity measurements or spectrophotometric scans were cultured in RCV media containing various supplements as detailed below. The antibiotic concentrations used to select resistant R. capsulatus strains were: tetracycline (0.5 ug/ml); kanamycin (10 pg/ml); and spectinomycin (10 pg/ml). All E. coli strains were maintained in rich LB media (70), except for E. coli JM101 which was grown in minimal M9 media (70). The antibiotic concentrations used to select resistant E. coli strains were: ampicillin (250 pg/ml); tetracycline (10 pg/ml); and kanamycin (50 pg/ml). Plate cultures for both R. capsulatus and E. coli strains were grown on the appropriate 19 Table I. R. capsulatus s t r a i n s . S t r a i n Genotype Phenotype Reference BIO SB1003 CW100 wild-type rif-10 bchC: : K r p R PS+ a PS+ PS -, P658 accumulated, derived from BIO 116 107a 117 DE324 crtF: : Q , rif-10, aad+ PS +, demethylsphaeroidene accumulated, derived from SB1003 124 BP503 crtF129, hsd-1, str-2 PS +, demethylsphaeroidene accumulated 110 BRP15 bchH615, crtF129, hsd-1, str-2 PS -, no Bchl or precursor, demethylsphaeroidene accumulated, derived from BP503 110 BRP4 bchE604, crtF129, hsd-1, str-2 PS -, P590 accumulated, demethylsphaeroidene accumulated, derived from BP503 110 Photosynthetic compet'ence 20 medium containing 15 g agar/1. Bacteriological supplies used for media preparation were obtained from Difco. Reagents used to prepare stock salt solutions for minimal media and other buffers were supplied by BDH Inc. and Fisher Scientific. Antibiotics were supplied by Sigma Chemical Co. 2. DNA purification and manipulation. Plasmid DNAs were isolated by Triton lysis followed by isopycnic centrifugation in a CsCl (Bethesda Research Laboratories) gradient (70). Chromosomal DNA was purified by a modified Triton lysis procedure as follows. Cells were grown to a density of 7.0 x 10^  cfu/ml, harvested by centrifugation, and resuspended in 0.01 volumes of 25% sucrose, 50 mM Tris HC1 (pH 8.0). Lysozyme (Sigma Chemical Co.) and EDTA were added to final concentrations of 1.0 mg/ml and 50 mM, respectively, and the mixture was incubated on ice for 10 min. Cells were gently but completely lysed by the addition of 0.8 volumes of 2% Triton X-100, 50 mM Tris HC1 (pH 8.0), and 10 mM EDTA (pH 8.0), followed by heating at 65°C for 10 min. CsCl was added to the lysate to a final concentration of 1.0 mg/ml, and the mixture was centrifuged to establish a density gradient from which chromosomal DNA was purified. DNA restriction and modification enzymes obtained from various sources (Bethesda Research Laboratories, Boehringer Mannheim GmbH, and New England Biolabs) were used according to the manufacturers' specifications. DNA fragments were purified from agarose (Bio-Rad) gels in 0.5X TBE buffer (70) by electroelution followed by NACS prepac 52 column chromatography (Bethesda Research Laboratories), or by adsorption to glass beads according to protocols supplied in the GeneClean kit (BIO 101 Inc.). The various recombinant DNA techniques used were performed as described by Maniatis et al. (70). 3. Plasmids. Relevant features of the plasmids used in this thesis are summarized in Table II. The source of cloned R. capsulatus DNA used in this study was plasmid pRPS404, which contains a 50 kb R. capsulatus DNA fragment encoding all known essential photosynthesis genes (110). Segments of R. capsulatus DNA containing putative promoter sequences were inserted into the mobilizable promoter fusion vectors pTB931 (117) and pXCA601 (1), both of which contain a truncated E. coli lac'Z gene to monitor promoter activity. R. capsulatus DNA fragments 21 Table I I . Plasmids. Plasmid Descri p t i o n Reference pRPS404 pTB9 31 pXCA601 pJAJ9 pCWl pCW2 pJPl pJPlOO pJPlOl pXCA6::935 pXCA6::935A4 4 pXCA6::EHQ pXCA6::EHQA44 contains a l l e s s e n t i a l R. capsulatus photosynthesis genes lac'Z promoter fusion vector lac'Z promoter fusion vector expression vector containing the R. capsulatus puf promoter bchCA promoter and 5' terminus of the bchC fused i n frame to the lac'Z gene of pTB931 BamHI-C fragment of pRPS404 i n s e r t e d i n t o pJAJ9 1.5 kb EcoRI-BamHI segment of the EcoRI-H fragment of pRPS404 fused i n frame to the Jac'Z gene of pXCA601 0.46 kb Smal-BamHI segment of the EcoRI-H fragment of pRPS404 fused i n frame to the lac'Z gene of pXCA601 0.37 kb BcII-BamHI segment of the EcoRI-H fragment of pRPS404 fused i n frame to the lac 'Z gene of pXCA601 puf operon extending from the wild-type puf promoter to an in-frame fusion of pufB to the lac'Z gene of pXCA601 puf operon extending from the mutant (A44) puf" promoter to an in-frame fusion of pufB to the lac'Z gene of pXCA601 wild-type bchCA and puf operons extending to an in-frame fusion of pufB to the lac'Z gene of pXCA601 wild-type bchCA operon and the mutant (A44) puf operon extending to an i n -frame fusion of pufB to the lac'Z gene of pXCA601 110 117 1 50 117 117 t h i s work t h i s work t h i s work t h i s work t h i s work 22 lacking promoter sequences were inserted into the mobilizable expression vector pJAJ9 (50), which uses the oxygen-regulated puf operon promoter to drive expression of appropriately positioned genes. Commercially supplied pUC or pBR322 plasmids were used for routine manipulation of cloned R. capsulatus D N A segments. The construction of plasmids pCWl, and pCW2 have been described previously (117). Briefly, plasmid p C W l contains a 450 bp fragment of R. capsulatus D N A encoding the regulatory region and first 84 codons of the bchC gene. Plasmid pCW2 contains a 13 kb R. capsulatus fragment encoding the carboxy terminus of the bchC gene, the bchA gene cluster, and the entire puf operon. Plasmids pJPl, pJPlOO, and pJPlOl were constructed in our laboratory by John Priatel. Each plasmid contains an in-frame fusion of the bchC gene to the eighth codon of the E. coli lac 'Z gene, but each contains a different amount of R. capsulatus D N A 5' to the bchC coding region. These plasmids were constructed as follows (Figure 5). The EcoRI-H fragment of pRPS404 contains the 3' terminus of the crtF gene, the entire bchC , bchA.l, and bchA.2 genes, and all but the last 821 nt of the bchA.l gene (124). The 1.5 kb EcoRl to BamHI segment of the EcoRI-H fragment, containing the 3' terminus of the crtF gene and approximately 50% of the bchC gene, was purified and inserted into both pUC13 and pBR322, each of which had been digested with EcoRl and BamHI, creating plasmids pUC13::EB and pBR322::EB respectively. To construct pJPl, the entire R. capsulatus segment was excised from pBR322::EB as a PstI to BamHI fragment, purified by gel electrophoresis, and inserted into pXCA601 that had been cut with PstI and BamHI. To construct pJPlOO, the 0.46 kb Smal to BamHI portion of the R. capsulatus insert in pUC13::EB was removed, purified, and inserted into pUC13 that had been digested with Hindll and BamHI to create plasmid pUC13::SB. This 0.46 kb R. capsulatus segment was then excised from pUC13::SB as a Psfl to BamHI fragment, purified, and inserted into pXCA601 that had been cut with Psfl and BamHI. Similarly, the 0.37 kb Bell to BamHI segment of the R. capsulatus insert in pUC13::EB was removed, reinserted into pUC13 cut with BamHI to create pUC13::BcB, excised from pUC13::BcB as a PstI to BamHI fragment, and inserted into pXCA601 that had been cut with Psfl and BamHI to construct plasmid pJPlOl. A flow chart showing the construction of plasmids pXCA6::HQ and pXCA6::HQA44 is shown in Figure 6. Both plasmids contain an R. capsulatus insert that extends from the EcoRl site in the crtF gene to an in-frame fusion between the pufB and lac'Z genes, and are isogenic 23 Figure 5. Construction of plasmids pJPl, pJPlOO, and pTPlOl. See MATERIALS A N D METHODS section 3 for details. Relevant restriction sites are shown as follows: BamHI (B); Bell (Be); EcoRl (E); PstI (P); and Smal (Sm). A "0" symbolizes the loss of a restriction site. 24 A c r t ' F b c h C b c h A . l bchA. 2 b c h A . 3 pufB pufL pufM pufX pufQ pufA EcoH EcoQ L_J BamC Figure 6. Construction of plasmids pXCA6::HQ and pXCA6::HQA44. Panel A: Key to the shading of the various R. capsulatus photosynthesis genes present on the EcoRI-H (EcoH), EcoRI-Q (EcoQ), and the BamHI-C (BamC) fragments of pRPS404. Relevant restriction sites are shown as follows: BamHl (B); CZal (C); EcoRl (E); Nael (Na); Ncol (Nc); Sail (S); and XfcoIKXh). Panel B: Method of plasmid construction. See MATERIALS A N D M E T H O D S section 3 for details. The additional restriction site not present in panel A is Xbal (Xb). Restriction endonuclease digestions were complete unless otherwise indicated. The two dots (• •) represent the two nt puf promoter mutation (A44), which inactivates the activity of the puf promoter. 25 26 except that pXCA6::HQA44 contains a two nucleotide change in the puf promoter that is known to eliminate approximately 95% of puf promoter activity (1). The starting plasmids were pUC13::EcoH and pUC13::BamC, which contained either the EcoRI-H, or BamHl-C fragments, respectively, of pRPS404 inserted into pUC13. Also, pJAJ21A44 (1) was used as the source of the A44 puf promoter mutation. The first two steps (a and b of Figure 6) involved insertion of the EcoRI-Q fragment of pRPS404 with either the wild-type or the mutant puf promoter sequence in a pUC13 derivative. In step a, the 1.0 kb EcoRI-Q (wild-type sequence) was excised from pUC13::BawC by complete digestion with EcoRl followed by purification of the 1.0 kb fragment. This fragment was inserted into pUC13 that was cut with EcoRl to give plasmid pUC::EcoQ. In step b is shown the construction of plasmid pUC13::EcoQA44 (mutant sequence). The 0.385 kb Nael to Ncol fragment from plasmid pJAJ21A44 was purified and used to replace the same 0.385 kb Nael to Ncol fragment that was removed from pUC13::EcoQ. To verify replacement of the wild-type sequence with the mutant sequence, recombinant plasmids were screened by sequencing across the substituted region. A 15-mer oligonucleotide of the sequence 5'-GCGTCGAACAGTCCG-3' was synthesized (courtesy of Tom Atkinson, Department of Biochemistry, University of British Columbia) on an ABI Applied BioSystems Model 380B D N A synthesizer, purified by polyacrylamide gel electrophoresis, and used as a primer in the sequencing reaction. This oligonucleotide is complementary to nucleotides 262 to 278 in Figure 4 of Adams et al. (1), and hybridizes to a region of the bchA.3 gene approximately 30 nt downstream of the puf operon promoter. The double-stranded template used for this reaction was pUC13::£coHQA44 that had been made linear by digestion with Xbal. This template was prepared as described (43), and was sequenced with the reagents from a Sequenase version 2.0 kit (United States Biochemical Corporation) according to the Sequenase double-stranded sequencing protocol, with 5' [a-^P] dATP (800 Ci/mmol) obtained from New England Nuclear Research Products. The next two steps involved joining the EcoRI-Q and EcoRI-QA44 fragments to the EcoRI-H fragment. In step c, plasmid pUC13:: BamC was digested completely with Clal, partially with EcoRl, and the 2.5 kb Clal to EcoRl fragment was purified. This fragment was inserted into pUC13::EcoH that had been prepared by complete digestion with CZal, partial digestion with EcoRl, and purification of the 6.7 kb fragment. Step d illustrates the joining of the EcoRI-QA44 fragment to the EcoRI-H fragment. The 1.0 kb EcoQA44 fragment was 27 released from pUC13::EcoQA44 by complete digestion with EcoRl. This fragment was purified and inserted into pUC13: :EcoH that had been partially digested with EcoRl. Recombinant plasmids were screened by digestion with Sail for those that received the insert at the correct EcoRl site and in the correct orientation. This step generated plasmid pUC13: :EcoHQA44. The final two steps (e and f) were designed to place the wild-type and mutant EcoRI-HQ sequences into the recombinant pufB::lac'Z fusion vector pXCA6::935 (1). Both pUC13: :EcoHQ and pUC13: :EcoHQA44 were digested completely with Xbal and Sail, and the 6.1 kb fragment released from each plasmid was purified by gel electrophoresis. The vector, pXCA6::935, was prepared by complete digestion with Xbal, partial digestion with Sail, and purification of the 16 kb fragment. Following ligation of the 6.1 kb fragment purified above into the the prepared vector, the correct final constructions were verified by analytical digestions with BamHl. The final wild-type construct was named p X C A 6 : : H Q , and the final mutant construct was named p X C A 6 : : H Q A 4 4 . 4. Measurement of /?-galactosidase specific activity. Vectors containing R. capsulatus in-frame fusions to the lac'Z gene were mobilized by conjugation into the desired R. capsulatus host strain as described (95). Exconjugant cells were grown at 34° C in R C V medium supplemented with 0.1% yeast extract, 8.0 m M phosphate buffer ( p H 6.8), and 0.5 ug T c / m l . Cultures (40 ml) were grown with high or low aeration as described above, and were harvested by centrifugation at a density of 80 - 100 Klett units. Care was taken to ensure that highly aerated cells and oxygen-limited cells were harvested at the same Klett density. Under these growth conditions, a cell density of 80-100 Klett units was at the mid-log phase of growth for highly aerated cells that had a doubling time between 2.0 and 2.2 h. The growth rate of oxygen-limited cultures gradually decreased from approximately 2 h to approximately 12 h as oxygen was consumed. This change in growth rate was reproducibly observed when oxygen-limited cultures reached a density between 50 - 60 Klett units. Cell pellets were resuspended in 1 ml Z buffer (80), and disrupted by sonication (two 15 sec. bursts, microtip intensity 2.5, on a Branson Model 350 sonifier). Extracts were cleared of cell debris by centrifugation at 12000 X g for 1 min. The specific activities of /J-galactosidase in extracts of R. capsulatus cells containing recombinant lac'Z fusion plasmids were measured as described (80) except that the rate of o N P G cleavage was determined directly by a continuous 28 spectrophotometric assay. Values were normalized to the protein content of the extracts as determined by a Lowry protein assay (66), using BSA (Sigma Chemical Co.) as standard. The values presented are the average of at least four independent assays of each fusion construct. 5. Absorption and fluorescence spectroscopy. Cultures of R. capsulatus cells were grown with low aeration or photosynthetically in R C V medium supplemented with 0.6% glucose, 0.5% pyruvate, and 50 m M dimethylsulfoxide (DMSO) . These additives enabled the cultures to reach a significantly higher cell density than in R C V medium alone (14). Cells were harvested by centrifugation either at mid-log or stationary phase of growth as specified in the figure legends. For absorbency scans of intact cells, equal numbers of cells were resuspended in 22.5% BSA in R C V (107) and scanned from 900 n m to 350 n m in a Hitachi U2000 U V / V i s i b l e spectrophotometer. Culture fluids were scanned after removal of cells by two 4 min centrifugations at 12000 X g. Identification of Bchl a biosynthetic intermediates accumulated by bch mutant strains was performed by absorption and fluorescence spectroscopy of extracted pigments. Total Bchl and carotenoids were extracted from cell pellets containing equal numbers of cells with a mixture of acetone:methanol (7:2) as described (24). Absorbency scans of the extracts were performed in a Hitachi U2000 U V / V i s i b l e spectrophotometer as described above, and fluorescence scans were performed in a Perkin-Elmer 650-10S fluorescence spectrophotometer, scanning emitted light from 450 n m to 800 n m with the excitation wavelength set at 400 nm. 6. D N A sequence analysis. The bchC gene and its flanking regions were sequenced by subcloning specific fragments of D N A into M13 m p l 8 and m p l 9 vectors (122). Dideoxy chain-termination reactions were carried out essentially as described (105) except that c^-dGTP (Boehringer Mannheim G m b H ) was used instead of d G T P (5,81). Unlabelled d N T P s and ddNTPs , and M13 sequencing primers were obtained from P.L. Biochemicals and the Regional D N A Synthesis Laboratory, University of C a l g a r y . Computer-assisted analyses of the nucleotide and proposed amino acid sequences were performed using algorithms as cited in the text and figure legends. 29 7. Isolation of R. capsulatus R N A . T w o standard conditions were used for growth of cultures of R. capsulatus cells for R N A isolation, depending upon whether the entire culture was harvested at a single time point or whether samples of the culture were harvested at various time points. For single time-point batch R N A isolations, cultures of R. capsulatus cells were grown at 3 4 ° C in R C V medium with high aeration (40 ml of culture in a 500 m l flask shaken at 300 rpm) to a density of 4 x 10^ c f u / m l , then shifted to low aeration by transfer of the entire culture to a 50 m l flask that was shaken at 150 r p m to stimulate synthesis of m R N A from photosynthesis genes. Induced cells were collected at 30 m i n after the shift to low aeration by centrifugation through an ice slurry. For multiple time-point R N A extractions, care was taken to ensure that the aeration of the culture d id not change appreciably during removal of the samples. This was achieved by growing 400 m l of culture with high aeration (equally divided between two 2 1 flasks that were shaken at 300 rpm) to a density of 4 x 10^ c f u / m l , and then combining these cultures into a single 500 m l flask that was shaken at 150 rpm. At the time of the shift, and at 10 or 15 min intervals for 60 m i n past the shift, 10 or 12 ml samples, respectively, were removed from the culture and cells were pelleted through an ice slurry. Because the total volume of culture removed from the 500 m l flask during the time course was 60 ml (15% of the total volume) the samples were assumed to be under relatively continuous oxygen-limitation. Total R N A was extracted from cell pellets with water-saturated unbuffered phenol as described previously (115). RNase-free DNase I was obtained from Boehringer Mannheim G m b H . Prior to use, the integrity of the R N A was checked by visualization of EtBr-stained R N A following electrophoresis through an agarose minigel. If the r R N A bands were clearly defined, the m R N A was assumed to be of good quality. 8. D N A and R N A blotting and preparation of probes. Each lane of the 1% agarose gels used for D N A blots contained either 5 ug of chromosomal D N A from the appropriate R. capsulatus strains digested to completion with EcoRl, or 2 ng of purified EcoRI-H or E c o R I - H : : K m ^ D N A fragments that were mixed with 5 ug of sheared salmon sperm D N A (Sigma Chemical Co.). Following agarose gel electrophoresis in 0.5X T B E buffer, the D N A fragments were transferred by capillary action to a sheet of nitrocellulose paper as described (70). A D N A ladder consisting of lambda phage D N A cut with HmdIII was 30 used as size markers. Each lane of the 1% agarose, 2.2 M formaldehyde gels used for R N A blots contained 10 ug of R. capsulatus R N A . Following electrophoresis in M O P S buffer (40 m M M O P S [pH 7.01,10 m M sodium acetate, 1 m M E D T A [ p H 8.01), gels were equilibrated in 0.5X T B E buffer, and the R N A was transferred electrophoretically to a BioTrans (ICN Biomedicals, Inc.) nylon membrane at 30 V for 16 h in 0.5X T B E buffer. R N A size markers (0.3 to 9.5 kb) were purchased from Bethesda Research Laboratories. Radioactive probes for D N A and R N A blots were prepared by the method of primer extension using random hexadeoxynucleotide primers as described (36). Probes were labeled with 5' [cx-32p] d A T P (800 C i / m m o l ; N e w England Nuclear Research Products) to an average specific activity of 1 X 10? cpm/u.g. Blots were prehybridized with denatured sheared salmon sperm D N A (500 pg salmon sperm D N A / m l in 5X SSC [20X SSC: 3 M N a C l , 0.3 M sodium citrate, p H 7.0], 1% SDS, 10 m M E D T A [pH 8.01, and 50% formamide) for a minimum of two hours at 42° C . Radioactive probe was then added to the prehybridization mix, and hybridization was performed for 16 h at 42° C . Blots were washed twice for 10 min in 2X SSC, 0.1% SDS at room temperature, twice for 10 min in 2X SSC, 0.1% SDS at 50° C , and as necessary (determined by measuring radioactivity bound to the blot with a Geiger counter) in several changes of 0.2X SSC, 0.1% SDS at 5 5 ° C . Washed membranes were wrapped in plastic fi lm while damp and used for autoradiography. 9. Primer extension analysis. A 24-mer deoxyoligonucleotide of the sequence 5 ' - G C C T T G G G C C C G G A C A T T A T G A C G - 3 ' (complementary to nucleotides 199 to 223 in Figure 11) was synthesized (by T o m Atkinson on a A B I A p p l i e d Biosystems M o d e l 380B D N A synthesizer, Department of Biochemistry, U B C ) and purified by polyacrylamide gel electrophoresis. The purified oligonucleotide was radioactively labeled at the 5' end with T4 polynucleotide kinase (Bethesda Research Laboratories) and 5' [y-32p] A T P (4500 C i / m m o l ; I C N Biomedicals, Inc.) as described (70) to a specific activity of 2.4 X 10^ c p m / p g . Three pmol of labeled oligonucleotide were hybridized to 15 ug of R. capsulatus R N A and extended with Moloney murine leukemia virus reverse transcriptase as described (1). The same oligonucleotide used for the primer extension reaction was also used as a primer in a dideoxy chain-terminating sequencing reaction, using as template a purified 4.1 kb Fspl D N A 31 fragment from p C W l that contains the entire cloned R. capsulatus region. Double stranded D N A sequence analysis using 5' [a-32p] d A T P (800 C i / m m o l ; N e w England Nuclear Research Products) was performed as described (43). 10. Sl-nuclease protection m R N A end site mapping and preparation of probes. The bipartite probes used in most of the Sl-nuclease protection experiments were prepared by digestion of a recombinant plasmid containing an R. capsulatus fragment of interest at a restriction endonuclease site located within the R. capsulatus segment of the plasmid. The ends thus generated were radioactively labeled at either the 5' or 3' ends using standard protocols (70). The 5' termini were labeled with T4 polynucleotide kinase (Bethesda Research Laboratories) and 5' [y-32p] A T P (4500 C i / m m o l ; I C N Biomedicals, Inc.) as described (70), and the 3' termini were labeled with the Klenow fragment of E. coli D N A polymerase I (Bethesda Research Laboratories) and 5' [a-^^P] d A T P (800 C i / m m o l ; N e w England Nuclear Research Products) as described (70). Probes were labeled to a minimal specific activity of 1 X 10^ c p m / u g . Following digestion of the labeled plasmid with a restriction endonuclease that cut in the vector rather than at the vector/R. capsulatus junction, the probe was purified by gel electrophoresis. Between 50 and 400 ng of labeled probe were ethanol precipitated with 10 ug of R. capsulatus R N A for each sample used in the Sl-nuclease protection experiments. Samples were resuspended, denatured, hybridized, and digested as described (130) except that hybridizations were performed at 53° C for 3 h. Sl-nuclease was supplied by Bethesda Research Laboratories, and was used at the concentrations specified in the figure legends. Sl-resistant hybrids were ethanol precipitated and resuspended in formamide loading dye prior to electrophoresis in 5% polyacrylamide gels containing 8 M urea. Single stranded Haelll-digested M13 m p l l phage D N A fragments that were radioactively labeled at the 5' ends as described above served as size markers. 32 R E S U L T S 1. Isolation and identification of the bchC gene from R. capsulatus. l a . Isolation of an oxygen-regulated promoter and the 5' coding region of a gene from the photosynthesis gene cluster of R. capsulatus. The known photosynthesis genes that map to the 5.5 kb EcoRI-H fragment of pRPS404 are crtF, bchC, and bchA (Figure 7, and refs. 110, 123, 124), but at the time this work was done, the precise boundaries of these genes had not been established. The products of a Saw3A-l partial digest of this fragment were randomly inserted into the unique BamHI site of the broad host-range promoter fusion vector, pTB931 (117), in order to generate a fusion bank. Because the source of R. capsulatus D N A was restricted to the EcoRI-H fragment, the fusion bank thus created would be biased toward the isolation of a pigment biosynthesis gene promoter. The resultant collection of fusion plasmids was conjugated into R. capsulatus BIO and T c ^ exconjugants were screened for blue colonies, indicating that the cloned insert in pTB931 resulted in the synthesis of /}-galactosidase. One recombinant plasmid, p C W l , was isolated and found to contain an insert of approximately 450 bp. Extracts of R. capsulatus cells harboring p C W l grown with low aeration showed a level of /?-galactosidase specific activity approximately five times that observed in extracts of cells grown with high aeration (122 nmoles O N P G / m i n / m g in low O2 extracts compared with 22.2 nmoles O N P G / m i n / m g in high O2 extracts). Because it was later demonstrated that the pTB931 vector used in these initial experiments contained at least one promoter that could result in transcription of the plasmid-borne truncated lac 'Z gene, I could not be certain that all p C W l transcripts resulting in the synthesis of /J-galactosidase had originated within the R. capsulatus D N A insert. To overcome the complicating effects of potential readthrough transcription from pTB931 sequences, these experiments were repeated with the R. capsulatus insert subcloned into plasmid pXCA601 (1), which contains a transcriptional terminator to prevent transcription that initiates on the vector from extending into sequences fused to the lac 'Z gene. The level of oxygen-regulated expression of /3-galactosidase observed in extracts of cells containing the recombinant pXCA601 plasmids was very similar to that observed in extracts of cells harboring p C W l , thus confirming that I had isolated an oxygen-regulated promoter, as well as the 5' coding region of an R. capsulatus gene ( R E S U L T S section 3a, and Figure 14). 33 Figure 7. Genetic map of the 5.5 kb EcoRI-H fragment of pRPS404 showing the approximate locations of the genes that map partially (crtF and bchA.3) or completely (bchC, bchA.l, and bchA.2) to this fragment. Relevant restriction sites are shown as follows: EcoRl (E), Mlul (M), Smal (Sm), Bell (Be), Apal (Ap), H i n d l l l (H), BamUl (B), Accl (Ac), and Sau3A-l (S3). The R. capsulatus insert in p C W l was found to correspond to the area between the two Sau3A-l sites, and the direction of transcription initiated within this cloned insert is designated by the arrow. The dotted genetic boundaries indicate that portions of D N A from the crtF and bchA.3 genes extend beyond the EcoRl sites shown. 34 Because the D N A insert in p C W l originated from the EcoRI-H fragment of pRPS404, it was likely to contain the promoter, regulatory region, and 5' terminus of one of the crtF, bchC, or bchA genes, or possibly some other gene that had not yet been identified. Southern blotting (data not shown) was used to demonstrate that the insert in p C W l was obtained from an area near the left-most genetic boundary of the bchC gene, as indicated in Figure 7. Alignment of restriction maps generated from this insert and from the EcoRI-H fragment indicated that the direction of transcription originating from the cloned promoter would be from left to right in Figure 7. l b . Identification of the gene fused to lac 'Z. A 1.6 kb K m R cartridge encoding the aph (aminoglycoside 3' phosphotransferase) gene from Tn903 was excised from plasmid p U C 4 K (114) and inserted into the BamHI site of the EcoRI-H fragment, in order to create a mutation within the open reading frame derived from the nucleotide sequence (RESULTS section 2a). This BamHI site contains the codon for amino acid 120 in the O R F , and was believed to lie within the bchC gene because mutations resulting in a BchC" phenotype had been found to map to either side of this BamHI site (14). If the gene fused to lacZ was bchC, recombination of this mutation into the chromosome of an otherwise wild-type cell should have resulted in the accumulation of an intermediate of Bchl a biosynthesis and the loss of the ability to grow photosynthetically. T w o previous studies had proposed that the bchC and bchA genes form an operon, but there was uncertainty about the direction of transcription. Biel and Marrs (14) had suggested that transcription progressed from the bchC gene toward the bchA genes, whereas Zsebo and Hearst (129) had concluded that transcription occurred in the opposite direction, from the bchA genes toward the bchC gene. Therefore, even if the O R F was the bchC gene, there was uncertainty of what the phenotype of the O R F " strain would be. The 7.1 kb E c o R I - H : : K n R fragment was inserted into the "suicide" vector, pSUP202 (103), mobilized by conjugation into R. capsulatus BIO, and K m R , T c s exconjugants were screened for replacement of the wild-type gene with the mutated copy by Southern blotting. Four candidates were tested using radioactively labeled EcoRI-H D N A as a probe, and the resultant autoradiogram revealed only one band in the EcoRI-digested D N A from each mutant strain that comigrated with a purified E c o R I - H : : K m R D N A fragment, indicating that the mutated 35 copy of this gene had replaced the wild-type copy (Figure 8, panel A ) . To test if the K m R cartridge D N A had integrated into only one site in the chromosome, I used radioactively labeled K m R cartridge D N A as a probe that was hybridized to D N A immobilized on a second, identical filter. The resultant autoradiogram showed hybridization of the probe to only one fragment in the EcoRI-digested D N A s from each of the four strains, and that this fragment comigrated with the authentic E c o R I - H : : K m R segment (Figure 8, panel B). The combined results of these two blots showed that in each of the four strains, the K m cartridge seemed to have integrated into only one site in the R. capsulatus chromosome, which was within the EcoRI-H fragment. These interposon mutants were incapable of photosynthetic growth and were found to accumulate pigments with a red-most absorption peak of 668 n m (Figure 9, scan a), and a red-most fluorescence emission peak of 660 n m in acetone:methanol (7:2) extracts (data not shown). These absorbency and emission spectra identified the major accumulated pigment as 2-devinyl-2-hydroxyethyl chlorophyllide a, and indicated that the mutants were blocked in BchA activity (123). Because the bchA gene products act before that of the bchC gene product in the Bchl biosynthetic pathway (Figure 4, refs. 14, 89, 124), the phenotype observed in the mutant strains was consistent with the hypothesis that the O R F encodes the bchC gene, and that the bchC and bchA genes form an operon with the direction of transcription proceeding from bchC toward bchA (from left to right in Figure 7). Because the mutation created in the O R F appeared to have a polar effect upon bchA, thereby masking the consequences of the mutation of the O R F sequence, conclusive identification of the O R F depended upon overcoming the polar effect of this mutation, so that the phenotype of the O R F " strains could be observed directly and compared with the phenotype of known bchC mutant strains. O u r laboratory has constructed a plasmid RK2-derived expression vector, pJAJ9, which employs the oxygen-regulated puf promoter from R. capsulatus to obtain transcription of appropriately positioned genes (50). The BamHl-C fragment of pRPS404, which contains a portion of the bchC gene and the entire bchA region (110), was inserted into pJAJ9. One of the resultant plasmids, p C W 2 , contained the BamHl-C fragment inserted into pJAJ9 in an orientation that resulted in transcription of the bchC gene segment and the bchA genes from left to right as shown in Figure 7. Plasmid p C W 2 was conjugated into one of the K m R O R P strains of R. capsulatus, designated CW100, to test the phenotype of cells bearing the mutation in the O R F when complemented in trans with the 36 Figure 8. Southern blot of bchC interposon mutant strains showing replacement of the wild-type copy of the bchC gene with a copy containing the cartridge. Lanes 1 and 9 contain purified 5.5 kb EcoRI-H fragment (the faint band is due to a small amount of the 2.7 kb pUC13 plasmid vector that co-purified with the EcoRI-H fragment), lanes 2 and 10 contain purified 7.1 kb EcoRI-H::KmR fragment, lanes 3 and 8 contain R. capsulatus BIO chromosomal DNA, and lanes 4 to 7 contain chromosomal DNA from four different Km R , Tc^ isolates. The DNA in lane 4 was obtained from a strain that was designated CW100. Panel A: Purified EcoRI-H DNA was radioactively labeled and used as a probe. Panel B: Purified Km* DNA was radioactively labeled and used as a probe. 37 0.550 I Figure 9. Absorption spectrum of pigments accumulated in cultures of R. capsulatus CW100 (scan a) and R. capsulatus CW100 (pCW2) (scan b). Cellular pigments were extracted from equal numbers of oxygen-limited cells with acetone:methanol (7:2) in the dark and characterized by scanning the absorbency from 800 to 500 nm, as shown on the X axis. The absorbency values obtained are shown on the Y axis. 38 bchA genes. Exconjugant cells remained photosynthetically incompetent, but now accumulated an intermediate of Bchl a biosynthesis with a red-most absorption peak centered at 710 n m (Figure 9, scan b) when grown with low aeration. The presence of this product, 2-desacetyl-2-hydroxyethyl bacteriochlorophyllide a, which also accumulates in cells with known mutations in the bchC gene (Figure 4, refs. 14, 91, 110), confirmed the identity of the O R F as the bchC gene. These results argue that the bchC and bchA genes do form an operon, and that the direction of transcription proposed by Biel and Marrs (14) is correct. 2. Sequence of the bchC gene and characterization of its product. 2 a . Nucleotide sequence analysis of the bchC gene. Both strands of the R. capsulatus insert on p C W l were sequenced entirely. Because the 5'-terminus fused to the lac'Z gene in p C W l resulted in expression of /J-galactosidase, the reading frame of the bchC gene could be deduced by alignment of the lacZ reading frame across the junction with the R. capsulatus insert. The reading frame of the cloned 5'-terminus is proposed to begin at an A T G near a potential R. capsulatus ribosome-binding site ( G G A G in Figure 11; ref. 39a). Segments of the EcoRI-H fragment that overlapped this sequenced region were subcloned into phage M13 vectors and used to extend the sequence to two tandem termination codons 942 nt downstream from the proposed A T G initiation codon. The sequencing strategy used in this study is shown in Figure 10, and the nucleotide sequence of the cloned promoter/regulatory region, along with the proposed amino acid sequence of the bchC gene product are presented in Figure 11. It is noteworthy that another open reading frame was found to begin at base position 1129 in Figure 11, such that it overlaps that tandem termination codons T A A T G A of the bchC gene. Because the A T G start codon of this second open reading frame was preceded by a possible ribosome-binding site ( G G A A A G ) , and contains typical R. capsulatus codons (see below), it may encode the 5' terminus of the bchA.l gene. In order to check for possible frameshifts in the sequence shown in Figure 11, a codon preference plot (40) was generated for all three reading frames of the sequenced region using a codon preference statistic compiled from twenty-four R. capsulatus genes (3, 27,28, 75, 98,125, 126), for a total of 7017 codons. This program is able to locate potential errors in the sequence by identifying deviations from the observed R. capsulatus codon bias. The codon preference plot 39 0 142 S 3 Be 452 382 541 657 H N B N 899 834 939 1059 A A F S 1296 B l crt 'F bchC h— bchA.l ' 1 Figure 10. Sequencing strategy for the bchC gene and flanking sequences. Relevant restriction sites are shown as follows: SHM3A-1 (S3), Bell (Be), HmdIII (H), N a r l (N), BamHI (B), Alul (A), Fspl (F), Sful (S), and Ball (Bl). Numbers above the restriction sites correspond to the nucleotide position of the first nucleotide of the restriction sequence. Arrows represent the direction and extent of sequence data obtained from various M13 subclones. 40 Figure 11. Nucleotide and deduced amino acid sequence of the bchC gene and flanking regions. Proposed coding sequences (in single letter code) extending from nucleotides 1 to 19 for the 3' portion of the crtF gene, nucleotides 185 to 1129 for the bchC gene, and nucleotides 1129 to 1194 for the 5' portion of the bchA.l gene, are shown above the nt sequence. Putative ribosome-binding sites (SD) are indicated. The site at which transcription of the bchCA operon begins is indicated by the dot (•) above the nt sequence. Restriction sites of interest are underlined, and include the two Sau3A-l sites that flank the R. capsulatus insert in p C W l , the H m d l H site used as a reference point in the Sl-nuclease protection experiments, and the BamHI site used for interposon mutagenesis, as the fusion point to lac'Z in the pJPl deletion series, and as a reference point in the Sl-nuclease protection experiments. The nucleotide sequence is numbered above the sequence (with the last digits aligned with the corresponding nt), and the amino acid sequence of the bchC gene is numbered on the right margin. 41 10 20 30 40 50 60 70 80 i e a e r g * GJ^GAGGCCGAACGCGGCTGACACGGCTGCGTTCGGACCCGGCTTTGACCCGGGGGTCAGAAAGTCGCACATCCGTCTG S a u 3 A l 90 100 110 120 130 140 150 160 TCGCAAAAGTGTCTAATCAAATTGACAGTCGGGCGTGTAAGTTCAATGATACACACAGGCGTGATCAGCCCGACTCTCCG 170 180 190 200 210 220 230 240 SD m e t q v v i m s g p k a i s t g i a 19 GCCCGATCATACCGGGAGCAAGAAATGGAAACGCAAGTCGTCATAATGTCCGGGCCCAAGGCCATCTCGACGGGCATCGC 250 260 270 280 290 300 310 320 g l t d p g p g d l v v d i a y s g i s t g t e k l f 46 CGGTCTGACCGACCCCGGGCCGGGGGACCTCGTCGTGGATATCGCCTATTCCGGCATTTCGACTGGCACCGAGAAATTGT 330 340 350 360 370 380 390 400 w l g t m p p f p g m g y p l v p g y e s f g e v v 72 TCTGGCTCGGCACCATGCCACCCTTCCCGGGCATGGGATATCCGCTTGTTCCCGGCTACGAAAGCTTCGGAGAGGTCGTT H i n d l l l 410 420 430 440 450 460 470 480 q a a p d t g f r p g d h v f i p g a n c f t g g l r 99 CAAGCCGCCCCCGACACCGGCTTCCGACCGGGC£AlCJiCGTCTTCATTCCCGGCGCCAACTGCTTCACCGGCGGGTTGCG S a u 3 A I 490 500 510 520 530 540 550 560 g l f g g a s k r l v t a a s r v c r l d p a i g p e 126 CGGGCTGTTCGGCGGGGCGTCGAAGCGCCTTGTCACGGCCGCCTCGCGCGTTTGTCGGCTGGATCCCGCCATCGGCCCCG BaaU I 570 580 590 600 610 620 630 640 g a l l a l a a t a r h a l a g f d n a l p d l i v 152 AGGGCGCGCTTCTGGCCCTTGCCGCCACCGCGCGGCATGCGCTGGCCGGGTTTGACAATGCTCTGCCGGATCTGATCGTC 650 660 670 680 690 700 710 720 g h g t l g r l l a r l t l a a g g k p p m v w e t n 179 GGCCACGGCACCCTCGGGCGCCTTCTGGCCCGTCTGACCCTGGCTGCCGGTGGCAAGCCGCCGATGGTCTGGGAAACCAA 730 740 750 760 770 780 790 800 p a r r t g a v g y e v l d p e a d p r r d y k a i y 206 TCCTGCCCGTCGCACGGGCGCGGTCGGCTACGAGGTTCTGGACCCCGAAGCCGATCCCCGGCGCGACTACAAGGCCATCT 810 820 830 840 850 860 870 880 d a s g a p g l i d q l v g r l g k g g e l v l c g 232 ATGACGCCTCGGGCGCGCCCGGTCTGATCGACCAGCTCGTCGGGCGTCTGGGCAAGGGCGGGGAACTGGTGCTGTGCGGC 890 900 910 920 930 940 950 960 f y t v p v s f a f v p a f m k e m r l r i a a e w q 259 TTCTATACGGTGCCGGTCAGCTTCGCCTTTGTTCCCGCCTTCATGAAGGAAATGCGCCTGCGCATCGCCGCCGAATGGCA 970 980 990 1000 1010 1020 1030 1040 p a d l s a t r a l i e s g a l s l d g l i t h r r p 286 GCCGGCCGACCTTTCGGCCACGCGCGCGCTGATCGAAAGCGGGGCGCTCTCGCTGGATGGTCTCATCACGCATCGTCGCC 1050 1060 1070 1080 1090 1100 1110 1120 a a e a a e a y q t a f e d p d c l k m i l d w k d 312 CCGCGGCGGAGGCGnCCGAGGCCTATCAGACCGCTTTCGAAGACCCTGACTGCCTGAAGATGATCCTTGACTGGAAAGAT SD 1130 1140 1150 1160 1170 1180 1190 a k * m t d a p n l k g f d a r l r e e a a e e p 314 GCAAAATAATGACTGACGCACCCAACCTGAAGGGATTTGACGCCCGTCTGCGGGAAGAAGCCGCCGAAGAGCCC 42 showed that codon usage in the proposed reading frame corresponded well with the codon bias observed in other sequenced R. capsulatus proteins (Figure 12, plot 2), whereas there was very poor correspondence in each of the other two reading frames (Figure 12, plots 1 and 3). Therefore, although confirmation of the proposed amino acid sequence of the bchC gene requires receipt of the true amino acid sequence of the BchC protein itself, it seems likely that the reading frame that I propose is correct. If this is true, the molecular mass of the BchC protein is 33006. Examination of codon usage for the bchC gene (Table III) revealed that, like all other R. capsulatus genes sequenced to date, the bchC gene showed a strong bias against codons ending in A or T, and an extreme bias against the codons T T A , C T A , G T A , and A T A . O f the 7331 codons now sequenced in R. capsulatus, only one G T A codon has been found in the crtK gene (3), and the codon A T A has been found for the first time within the bchC gene. 2b. Analysis of the putative amino acid sequence of the BchC protein. Examination of the hydrophobiaty plot (Figure 13A) generated by the method of Kyte and Doolittle (60) showed that the BchC protein is predicted to be slightly hydrophobic overall and that it has four more pronounced hydrophobic regions that may interact with the cell membrane: from Ala-122 to Ala-139; from Gly-142 to Gly-169; from Val-219 to Met-250; and from Ala-265 to Leu-275. These regions, however, are significantly less hydrophobic than the membrane spanning regions of the puf- and puc-encoded integral membrane proteins (125,126). Secondary structure analysis of the proposed bchC sequence predicted several regions that may form a-helices (Figure 13B): Arg-108 to Pro-121; Ala-128 to Ala-146; Arg-159 to Gly-169; Val-187 to Asp-193; Ala-241 to Pro-260; Ala-265 to Leu-275; His-283 to Pro-301. Because the BchC enzyme has a chlorin tetrapyrrole as a substrate, I examined the proposed amino acid sequence for regions that resemble chlorin binding sites in other pigment-binding proteins. Although ligands to Bchls in both R C and L H polypeptides are conserved histidine residues found within a hydrophobic region, the sequence of amino acids flanking these conserved histidine residues have different features that presumably reflect the different functions of the R C and L H Bchl molecules. Bylina et al. (19) have noted that a common motif of purple bacterial L H Bchl-binding sites is the amino acid sequence A l a / G l y - X - X - X - H i s found within a hydrophobic region, where "X" may be, as far is known, any amino acid. This motif is conserved in the R. capsulatus, R. sphaeroid.es, and R. viridis L H polypeptides (119). 43 B - 1 — 11 I i . . i . i . i . 3 • — • — " — - — • • — i — • • • — i • • •—" • • - 1 200 400 600 800 1000 1 200 crtF bchC bchA? Figure 12. Codon preference plot generated for the nucleotide sequence in Figure 11 using a program written by Gribskov et al. (41). Codon preference values calculated for this sequence using a window of 24 nucleotides were plotted using the computer program TS Graph II. Each of the three plots in panel A correspond to translation of the nucleotide sequence (Figure 11) in one of three reading frames: plot 1 represents reading frame 1 (translation begins with base position 1); plot 2 represents reading frame 2 (translation begins with base position 2); and plot 3 represents reading frame 3 (translation begins with base position 3). Nucleotide positions are numbered along the horizontal axis, and relative codon preferences, measured as distance from a baseline (indicated by the dashed lines across each of the plots) are given along the vertical axis. Peaks above the dashed line indicate more frequently used codons, according to a codon frequency table compiled from nucleotide sequences of 24 independently sequenced R. capsulatus genes (see text). The open reading frames of the crtF, bchC, and bchA.l genes are shown in panel B and are proposed to be translated in reading frame 2 for the crtF 3' segment and the bchC gene, and reading frame 1 for the bchA.l 5' segment. 44 T a b l e I I I . Codon usage i n t h e R. capsulatus bchC g e n e , s h o w i n g t h e number o f t i m e s e a c h codon was f o u n d i n t h e 316 codons ( i n c l u d i n g t h e two tandem t e r m i n a t i o n codons) o f t h e bchC g e n e . TTT phe F 2 TTC phe F 11 TTA l e u L 0 TTG l e u L 2 CTT l e u L 7 CTC l e u L 6 CTA l e u L 0 CTG l e u L 19 ATT i l e I 2 ATC i l e I 11 ATA i l e I 1 ATG met M 8 GTT v a l ' V 5 GTC v a l V 11 GTA v a l V 0 GTG v a l V 3 TCT s e r S 0 TCC s e r S 2 TCA s e r S 0 TCG s e r S 7 CCT p r o P 2 CCC p r o P 13 CCA p r o P 1 CCG p r o P 9 ACT t h r T 1 ACC t h r T 10 ACA t h r T 0 ACG t h r T 7 GCT a l a A 3 GCC a l a A 27 GCA a l a A 1 GCG a l a A 11 TAT t y r Y 5 TAC t y r Y 3 TAA OCH Z 1 TAG AMB Z 0 CAT h i s H 2 CAC h i s H 2 CAA g i n Q 2 CAG g i n Q 3 AAT a s n N 2 AAC a s n N 1 AAA l y s K 3 AAG l y s K 7 GAT asp D 7 GAC asp D 12 GAA g l u E 9 GAG g l y E 6 TGT c y s C 1 TGC c y s C 3 TGA OPA Z 1 TGG t r p W 4 CGT a r g R 4 CAC a r g R 10 CGA a r g R 1 CGG a r g R 3 AGT s e r S 0 AGC s e r S 3 AGA a r g R 0 AGG a r g R 0 GGT g l y G 4 GGC g l y G 22 GGA g l y G 2 GGG g l y G 11 45 100 200 300 1—1 1—1 L 100 2 0 0 3 0 0 Figure 13. Secondary structure analysis of the putative BchC protein. A m i n o acids are numbered along the horizontal axis. Panel A : Hydropathy plot of BchC generated by the computer program D N A Strider (71) using the algorithm of Kyte and Doolittle (60) with a window of 15 amino acids. Positive values indicate hydrophobic regions, and negative values indicate hydrophilic regions. Panel B: Prediction of a-helix formation in BchC using the algorithm of Gamier et al. (39) with a window of 15 amino acids. Positive values indicate regions likely to form an a-helix, and negative values indicate a low probability of a-helix formation. 46 Although no such conserved alanine or glycine residue is found four amino acids prior to the histidine ligand of the R C special pair, Brunisholz and Zuber (18) have observed that there are several clusters of aromatic acids that flank the R C histidine ligand. Analysis of the three dimensional structure of crystallized RCs from R. viridis indicates that several of these aromatic amino acids are in close proximity to the histidine ligand of the special pair (79). The proposed BchC sequence contains four histidine residues at amino acid positions 85,138,154, and 283 (Figure 11). Of the four histidine residues found within the putative BchC sequence, His-138 resembles a Bchl-binding site most strongly, being the only histidine residue that is both within a hydrophobic region and that is preceded by an alanine residue four amino acids prior to the histidine. Another possible ligand is His-85, which is located near four aromatic (phenylalanine) residues that are found in approximately the positions proposed by Brunisholz and Zuber (18). However, because it is currently impossible to predict a chlorin tetrapyrrole binding site from a primary sequence, it will be necessary to test the BchC protein and mutant derivatives experimentally to determine which, if any, of these structural features have functional significance for enzymatic activity. 3. Promoter M a p p i n g of the bchCA operon. 3a. Genetic analysis. Sequences necessary for the activity of the bchCA promoter were determined by analysis of a series of bchCv.lac'Z fusions, each of which contained the 5' part of the bchC gene joined in frame at the BamHI site to the eighth codon of the E. coli lac'Z gene in pXCA601, but which differed in the amount of R. capsulatus D N A 5' to the bchC gene. A flow chart for the construction of these plasmids is shown in Figure 5 in M A T E R I A L S A N D M E T H O D S . The 1.5 kb R. capsulatus fragment in plasmid pJPl extended from the EcoRl site in the crtF gene to the BamHI site in the bchC gene, and therefore contained all the R. capsulatus D N A present in the original oxygen-regulated clone, p C W l . Plasmid pJPlOO contained a 0.49 kb R. capsulatus fragment that extended from the Smal site between the crtF and bchC coding regions to the BamHI site in the bchC gene. Finally, the 0.40 kb R. capsulatus segment in plasmid pJPlOl extended from the Bell site between the crtF and bchC coding regions to the BamHI site in the bchC gene. The /?-galactosidase specific activities in 47 extracts of R. capsulatus cells harboring each of these plasmids were determined as outlined in M A T E R I A L S A N D M E T H O D S . The results (Figure 14) showed that the j3-galactosidase specific activities of pJPl and pJPlOO were similar, which suggested that the bchCA promoter is located downstream of the Smal site between the crfF and bchC genes. Because deletion of the R. capsulatus segment between the Smal and Bell sites on plasmid pJPlOl resulted in a drastic reduction of the /?-galactosidase specific activity, the sequences important for promoter activity were determined to lie between the Smal and Bell sites in the crtF/bchC intergenic region. 3b. Biochemical analysis. I performed a 5' end site mapping Sl-nuclease protection experiment with m R N A isolated from R. capsulatus BIO. The D N A probe was prepared by radioactive labeling of p C W l (117) at the 5' ends generated by digestion with Hzndll l , which cut within the bchC gene (at nucleotide position 382 in Figure 11). Following digestion of the labeled plasmid with Fspl, which cut in the vector 15 nt from the junction to the R. capsulatus insert, the 415 bp HmdlH-FspI fragment was purified by gel electrophoresis. Because this probe was designed to contain 15 bp of heterologous vector D N A as a tail extending from its unlabeled end, it allowed distinction on the basis of size between protection from Sl-nuclease digestion of the 400 bp R. capsulatus-derived segment of the probe by an m R N A molecule that would initiate upstream of the 5'-most nt of the R. capsulatus segment, and protection of the complete probe by D N A - D N A reannealing. Probes that consist of an R. capsulatus-derived segment as well as a vector-derived segment shall be referred to as "bipartite" probes throughout this thesis. Two bands that each resulted from protection of the probe by m R N A were observed on an autoradiogram of a polyacrylamide gel of Sl-protected hybrids (Figure 15). The predominant band approximately 250 nt in length indicated a 5' end site around nt position 150 in Figure 11, and a minor band approximately 400 nt in length indicated readthrough transcription into the region delimited by the probe from an upstream promoter. The exact position of the predominant 5' end site observed in Figure 15 was determined by the method of primer extension. A 24-mer antisense oligonucleotide complementary to nucleotides 199 to 223 (Figure 11) was used as a primer to extend by reverse transcriptase to the 5' end Of the bchCA m R N A . Figure 16 shows an autoradiogram of a polyacrylamide gel on which the primer extension products were run alongside a dideoxy 48 CONSTRUCT REPRESENTATION LOW 02 HIGH 02 L/H pJPl S Be \ / B 1 4 8 . 5 ( 1 7 . 6 ) 2 6 . 6 ( 5 . 6 ) 5.6 c r t 'F bchC S Be pJPlOO 1 9 1 . 3 ( 2 6 . 9 ) 2 3 . 8 ( 4 . 5 ) 8.0 Be B pJPlOl 3.6 ( 0 . 7 ) 0.7 ( 0 . 0 1 ) 5.0 Figure 14. /3-galactosidase specific activities of cells containing bchCr.lac'Z 5' deletion constructs. 1 Representation of R. capsulatus D N A fused to the lac'Z gene. Restriction sites are represented as follows in the top construct: EcoRl (E); Smal(S); Bell (Be); and BamHI (B). The structural genes encoded by the R. capsulatus inserts in each plasmid are shown, and are labeled in the top construct. 2 j3-galactosidase specific activities are expressed as nmoles o N P G c l e a v e d / m i n / m g protein. The values in brackets are the standard deviations of four independent assays that were performed for each construct. 3 The ratio of j3-galactosidase specific activities obtained from cells grown with low aeration versus with high aeration. 49 Figure 15. Low resolution Sl-nuclease protection 5' end site mapping of bchCA mRNA transcripts. A 415 bp HmdIII-FspI bipartite probe was prepared from pCWl as outlined in the text. Fifty ng of this probe were hybridized with either 10 ug of E. coli tRNA (lane 2), or 10 ug of RNA extracted from R. capsulatus BIO (lanes 3,4, and 5). Hybrids were trimmed with 500 U (lanes 2 and 3), 1000 U (lane 4), or 1500 U (lane 5) of Sl-nuclease, and separated through a 5% polyacrylamide gel containing 8M urea. The sizes, in nucleotides, of Haelll fragments of single-stranded M13 mpll DNA (lane 1) are given on the left margin. 50 G A T C Figure 16. Primer extension analysis of the predominant 5' end of bchCA mRNA. Three pmoles of a 24-mer oligonucleotide primer complementary to nucleotides 199-223 (Figure 12), radioactively labeled at the 5' end, were hybridized to 10 |J.g of total RNA isolated from R. capsulatus BIO. After extension with Moloney murine leukemia virus reverse transcriptase, the extension produces (lane 1) were compared with a sequencing ladder generated by using the same primer and a 4.1 kb Fspl double-stranded D N A fragment from pCWl containing the R. capsulatus insert as a template in chain-terminating sequencing reactions (lanes G, A, T, and C). To allow direct comparison with the sequences in Figures 11 and 17, this autoradiogram is intentionally inverted and lanes labeled with the complementary base. 51 sequencing ladder that was generated by using the same oligonucleotide primer and a D N A fragment that contained the cloned promoter region as template. The band in lane 1 of Figure 16 corresponds to the major band detected in the 5' Sl-nuclease protection experiment (Figure 15), and its position showed that this 5' end site maps to nt 137 of the sequenced region, 48 bp upstream of the proposed A T G start codon of the bchC gene (Figure 11). 3 c . Comparison with other photosynthesis gene promoter sequences. The nucleotide sequences flanking the bchCA major 5' end site are shown aligned with several other proposed photosynthesis gene promoter sequences in Figure 17. It should be stressed that of these aligned sequences, only three have been shown to contain nucleotides necessary for promoter activity as well as nucleotides to which a 5' m R N A end maps. The puf operon promoter sequence has been extensively characterized by functional activity assays, 5' m R N A end site mapping and capping experiments, and site-directed mutagenesis studies (1, 8). The major 5' end of the bchCA m R N A transcript has been localized by functional activity assays and 5' m R N A end site mapping studies (this work and refs. 117,124), but a fine structure analysis of the specific nucleotides that make up the bchCA promoter sequence has not yet been done. Similarly, the puc operon promoter has been mapped by functional activity assays and low resolution end site mapping of two 5' triphosphate m R N A ends to the regions shown in Figure 17 (130). Therefore, the promoters of the puf, puc, and bchCA operons are the best understood. Although the positions of 5' m R N A ends for the crt A, crtlBK, crtC, and crtEF operons have been mapped by low resolution Sl-nuclease protection experiments (40), correlation of the presence of these 5' ends with promoter activity has not yet been demonstrated. Finally, the proposed promoter sequences for the puh operon (125) and the crtD gene (3) have been found by examination of nucleotide sequences alone. With these uncertainties in mind, examination of the aligned sequences showed that the bchCA promoter sequence contained several features found in several of these other promoters. These features are the " A C A " motif, the " T G T A A -N g - T T A C A " palindromic sequence (which overlaps the " A C A " motif), an E. coli a ^ - l i k e recognition sequence that overlaps the palindromic sequence, and the " C G G G C " box. The potential significance of these features are considered in D I S C U S S I O N section 3 . 52 100 A A T T G A C A G T C G G G C G T G T A A G T T C A A T G A T • • • * • » • • • • 191 ATCCGCGCGAC££SCACCCCCTTCATGGGTT 2957 C T G G G G T C A C T C C A T A G C C C G G C G C T C G G C T 12 C C G A A T T T G C C G C A G T G T A A G C C C G A C T T T 25 A G T G T A A G C C C G A C T T T A C A C T T G A T C G C C G 7296 C G G G C G A G A T G C C C G A G G A A C C G A C G C T T T 2399 crtA T C C T C C C C T G T G A T G T G T A A C G G G A T A T T T 2475 crtA G A C A G C G T C G A C A G T T G T A A A T C G G A A T T G • • • • 8382 crtD T C C C A T A T T C T T G G G T G T A A G T T T C A G T T T A C A A C A A C A A C A A C A A C A A C A A C G A C A 142 C A C A G G C G T 235 T T G G G T A G C 2998 C G G T G G T C T 53 C T T G A T C G C 67 C T T G G G C T C 7255 T C T G T G C G C 2440 T C T G G G T C C 2517 A C C T A T C A T * * • • • 8423 C A G G T A G G T bchCA • F i g . 11, t h i s work puf • F i g . 4, r e f . 1 puh • F i g . 4, r e f . 125 p u c 1 • F i g . 3, r e f . 126 puc 2 • F i g . 3, r e f . 126 crtC • F i g . 3, r e f . 3 crtlBK • F i g . 3, r e f . 3 crtlBK • F i g . 3, r e f . 3 crtEF • F i g . 3, r e f . 3 Figure 17. Comparison of the bchCA promoter sequence with other putative R. capsulatus photosynthesis gene promoter sequences. The genes flanking the promoter sequences are indicated to the left and right of the sequence, respectively, and are transcribed in the directions indicated by the arrows. Numbers above each sequence show the nucleotide positions as given in the cited figures. The conserved " A C A " motif is enclosed in a box. Nucleotides that match the consensus palindromic motif are shown in bold type. C G boxes are underlined. The E. coli a 7 0 -like sequences are indicated by dots (•) under the rit sequence. The asterisk (*) indicates uncertainty of whether this sequence contains the promoter for the crtA and crtlBK operons (see D I S C U S S I O N ) . 53 4. Characterization of the bchCA m R N A transcripts. 4a. R N A blot analysis of bchCA m R N A transcripts. Although several studies have provided genetic evidence that the bchC and bchA genes form an operon by analyses of mutations in the bchC gene that are polar upon the bchA genes (14,117,124), the existence of an m R N A molecule that encodes all of these genes has not yet been demonstrated. Recent D N A sequence analyses of the bchC gene (117) and the putative bchA genes (124; M . Alberti, pers. comm.) indicate that the minimum length of such a transcript would be approximately 5.5 kb. To determine if such an m R N A transcript could be detected in R. capsulatus cells, and also to test for increased levels in response to decreased oxygen concentrations, R N A was isolated from a culture of R. capsulatus BIO at 15 min intervals after shifting the culture from high to low aeration as outlined in M A T E R I A L S A N D M E T H O D S . Because preliminary experiments with R N A isolated from R. capsulatus BIO indicated that bchC and bchA m R N A was difficult to detect on an R N A blot (presumably due to its low levels, data not shown), I decided to enhance the signal by expression of additional plasmid-borne copies of the bchC and bchA genes in R. capsulatus BIO cells. Plasmid pCW2 (117) contains the BamFfl-C fragment (containing the 3' terminus of the bchC gene, the entire bchA region, and the entire puf operon) of pRPS404 inserted into the mobilizable expression vector pJAJ9. Although the promoter region and 5' portion of the bchC gene is absent in pCW2, transcription of the major portion of the bchC gene and the bchA genes is initiated on the vector by the oxygen-regulated puf operon promoter (50). Equal amounts of R N A isolated at each time point were fractionated by gel electrophoresis and transferred electrophoretically to a nylon membrane. In independent experiments, R. capsulatus R N A (from BIO and BIO [pCW2]) immobilized separately on two membranes was hybridized with a bc/iC-specific probe consisting of the radioactively labeled 0.3 kb Apal-BamHl fragment of the bchC gene (Figure 7). A bch-A-specific probe consisting of the radioactively labeled 2.0 kb AccI-EcoRl fragment of the bchA.l and bchA.3 genes (Figure 7) was hybridized with R N A immobilized on two other analogous membranes, also in independent experiments. The resultant autoradiograms (Figure 18) showed that each probe detected many m R N A species over a wide size range. The bc/tC-specific probe detected high levels of an m R N A species approximately 0.33 kb in length, and lower levels of longer transcripts that ranged in length from approximately 0.98 kb to 2.3 kb (Figure 18A and B). 54 Figure 18. Blot of R. capsulatus BIO and R. capsulatus BIO (pCW2) R N A . Each lane of the 1% agarose, 2.2 M formaldehyde gels gels contains 10 ug of R N A . Panels A and C are autoradiograms of blots of R. capsulatus BIO R N A isolated at the indicated times following a shift of the culture to low aeration. Panels B and D are autoradiograms of blots of R. capsulatus BIO (pCW2) R N A isolated at the indicated times following a shift of the culture to low aeration. In panels A and B, the 0.3 Apal-BamHl fragment of the bchC gene (Figure 7) was uniformly radioactively labeled and used as a probe. In panels C and D, the 2.0 kb AccI-EcoRI fragment of the bchA gene (Figure 7) was radioactively labeled and used as a probe. The sizes of the dominant bands are given in kb on the left margin, and were calculated from a purchased (BRL) 0.3 to 9.4 kb R N A size ladder. 55 B 1 0 (pCW2) B min 0 10 20 30 40 50 60 bchA 56 Although no discrete transcript longer than 2.3 kb was detectable, faint hybridization to molecules longer than 2.3 kb band was visible on the blot of R. capsulatus BIO (pCW2) R N A . It should be noted that the bcrtC-specific probe detected only chromosomally-encoded bchC m R N A , as the Apal-BamYQ fragment was not present in p C W 2 . Therefore, the presence of p C W 2 had no major affect on the amount of chromosomally-derived bchC m R N A . The fccnA-specific probe annealed to m R N A molecules ranging in length from approximately 0.7 kb to 11.0 kb, but not to the predominant 0.33 kb species detected by the frc/zC-specific probe (Figure 18C and D). The signal obtained by the bcnA-specific probe on the blot of R. capsulatus BIO (pCW2) R N A (Figure 18D) represents the sum of m R N A molecules initiated from the chromosomal bchCA promoter and the plasmid-borne puf promoter. Long m R N A molecules complementary to the fcc«A-specific probe were observed only on the blot of R. capsulatus BIO (pCW2) R N A (Figure 18D). Therefore, these blots suggested that the amount of full-length bchCA m R N A was very low in cells, and that the presence of p C W 2 increased the amount of the long transcripts. Because the positions of the visible bands on the R. capsulatus BIO blots corresponded reasonably well with the positions of bands on the R. capsulatus BIO (pCW2) blots, I concluded that the m R N A species detected on the R. capsulatus BIO (pCW2) blots were accurate and amplified approximations of wild-type bchCA m R N A molecules. These autoradiograms demonstrated that bchC and bchA m R N A was detectable in cells of a culture grown with high aeration and that bchC and bchA m R N A levels reached maximum levels approximately 20 min after transcription of the photosynthesis genes was induced by shifting the culture conditions to low aeration. The much more dramatic increase in bchC and bchA m R N A accumulation over time in the R. capsulatus BIO (pCW2) blots relative to the R. capsulatus BIO blots showed that transcription of the hybrid puf/bchCA m R N A molecule was more sensitive to oxygen-regulation than was transcription of the chromosomally-encoded bchCA message, and presumably reflected the different sensitivities of the bchCA and puf promoters to oxygen regulation (see D I S C U S S I O N section 5). Because both the frcnC-specific and bcnA-specific probes detected many m R N A species of various lengths, but only the frc/zA-specific probe detected discrete transcripts long enough to encode the entire bchCA operon, these results suggested that the bchCA transcript was rapidly processed, and the 3' section of this transcript was more unstable than the 5' section that encoded the 0.33 kb segment. 57 4b. 5' end site mapping of bchCA m R N A transcripts. Most of the Sl-nuclease protection experiments described below made use of a bipartite D N A probe consisting of a region of R. capsulatus D N A expected to overlap an m R N A end, plus a "tail" of heterologous (vector) D N A extending from the unlabeled end of the probe. A s noted in RESULTS section 3b, this design allows the detection of "readthrough" m R N A transcripts consisting of overlapping m R N A molecules. The bipartite probe used in a previous bchCA 5' end site mapping experiment (Figure 15) was protected by two m R N A species; a major species with a 5' end 48 nt upstream of the BchC A T G codon, and a minor species that protected the complete R. capsulatus segment of the probe. The complete R. capsulatus segment of probe could have been protected by an m R N A molecule with a 5' end that was positioned very close to the R. capsulatus/pUC junction of the probe, or by an m R N A molecule with a 5' end that was positioned well upstream of the R. capsulatus segment of the probe. To distinguish between these possibilities, two additional probes were designed to map the 5' ends of bchCA operon transcripts. One probe was radioactively 5' end-labeled at the H i n d l l l site in the bchC gene, extended 476 bp in the 5' direction to the Mlul site in the crtF gene (124), and included 181 bp of pUC13 vector D N A as a tail (Figure 19A). The other probe was 5' end-labeled at the BamHI site in the bchC gene, extended 635 bp in the 5' direction to the Mlul site in the crtF gene, and included 209 bp of pUC13 vector D N A as a tail (Figure 19A). Thus, transcripts that began on the R. capsulatus chromosome upstream of the Mlul site and extended into the coding region of the bchC gene would produce a band 476 nt in length with the H i n d l l l probe and 635 nt in length with the BamHI probe. The results of an Sl-nuclease protection experiment using these two probes and R. capsulatus BIO R N A (Figure 19B) showed that the 657 nt band for the H i n d l l l probe, and the 844 nt band for the BamHI probe, that were due to reannealing of each bipartite probe, were present in both the control lanes containing yeast t R N A and in the experimental lanes containing R. capsulatus R N A . The entire R. capsulatus-derived segment of both probes was protected by m R N A originating from upstream of the Mlul site, as was shown by the presence of a band 476 nt in length for the H i n d l l l probe, and 635 nt in length for the BamHI probe. Three minor 5' ends were detected by both probes, positioned approximately 475, 445, and 390 nt upstream of the H i n d l l l site, corresponding to the minor ends that were positioned approximately 640, 620, and 550 nt upstream of the BamHI site (the 640 and 620 nt bands were not well resolved in Figure 19). 58 Figure 19. Sl-nuclease protection 5' end site mapping of bchCA m R N A transcripts. Panel A : Representations of the double-stranded bipartite D N A probes that were 5' end-labeled (*) at either the H i n d l l l (H) or BamHI (B) sites in the bchC gene. Both probes contain R. capsulatus D N A (thick lines) extending to the Mlul (M) site in the crtF gene, and a tail of pUC13 D N A (thin lines). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-resistant hybrids were separated. Lane 5: Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. Lanes 1 to 4: Sl-resistant fragments protected by the H i n d l l l probe. Lanes 6 to 9: S l -resistant fragments protected by the BamHI probe. Lanes 1 and 6 contain 150 ng of probe hybridized with 10 u.g of yeast t R N A and digested with 500 U of Sl-nuclease. Lanes 2 and 7, 3 and 8, and 4 and 9 contain 150 ng of probe hybridized with 10 pg of R. capsulatus B10 R N A isolated 30 min following a shift of the culture to low aeration, that were digested with 500, 1000, or 1500 U of Sl-nuclease, respectively. 60 These minor 5' end sites, therefore, mapped approximately 280, 255, and 193 nt upstream of the A T G initiation codon of the bchC gene (Figure 11). The predominant 5' end site, which mapped approximately 250 nt upstream of the H i n d l l l site and approximately 405 nt upstream of the BamHl site, was identical to the previously mapped 5' end site located 48 nt upstream of the A T G initiation codon of the bchC gene (Figure 11). In agreement with the initial 5' Sl-nuclease protection experiment shown in Figure 15, these results suggested that the bchCA operon could be transcribed from more than one promoter, at least one of which is located upstream of the Mlul site in the crfF gene, and that some crtEF transcripts extend into the bchCA operon at least as far as the BamHl site in the bchC gene. Because the promoter of the bchCA operon was localized between the Smal and Bell sites in the crtF-bchC intergenic region (RESULTS section 3a), and only the predominant 5' end observed in Figures 15 and 19 mapped to this region, this major 5' end was likely to be the only 5' end that resulted from transcription initiation. Therefore, the 640,620 and 550 nt minor bands probably arose from processing of crtEF transcripts that read into the bchCA operon. The minor bands less than 250 nt in length for the H i n d l l l probe and less than 405 nt in length for the BamHl probe were probably due to decay products of the unstable bchCA m R N A transcript, as noted above. 4c. 3' end site mapping of crtEF m R N A transcripts. If transcription from the crtEF operon extends into the bchCA operon at least as far as the BamHl site in the bchC gene, then 3' end site mapping experiments of the crtEF operon transcripts should show the existence of molecules that extend past the H i n d l l l and BamHl sites in the bchC gene. The two probes designed to test this hypothesis were similar to the two used for the 5' end site mapping experiments described above. Both probes were 3' end-labeled at the Mlul site in the crtF gene. One probe extended 476 bp in the 3' direction to the H i n d l l l site in the bchC gene and included 181 bp of pUC13 vector D N A as a tail (Figure 20A). The other probe extended 635 bp in the 3' direction to the BamHl site in the bchC gene and included 209 bp of pUC13 vector D N A as a tail (Figure 20A). The results of an Sl-nuclease protection experiment using these two probes (Figure 20B) showed that the 657 nt band for the H i n d l l l probe, and the 844 nt band for the BamHl probe were present in both the control and experimental lanes, and were due to reannealing of each bipartite D N A probe. The presence of a band 476 nt in length for the H i n d l l l probe and 635 nt in length for the BamHl probe, indicated 61 Figure 20. Sl-nuclease protection 3' end site mapping of crtEF m R N A transcripts. Panel A : Representations of the double-stranded bipartite D N A probes that were 3' end-labeled (*) at the Mlul (M) site in the crtF gene, extended to either the H i n d l l l (H) or BamHI (B) sites in the bchC gene (thick lines) and included a tail of pUC13 D N A (thin lines). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, S l -resistant hybrids were separated. Lanes 1 and 6: Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. Lanes 2 to 5: S l -resistant fragments protected by the BamHI probe. Lanes 7 to 10: Sl-resistant fragments protected by the H i n d l l probe. Lanes 2 and 7 contain 150 ng of probe hybridized with 10 ug of yeast t R N A and digested with 500 U of Sl-nuclease. Lanes 3 and 8,4 and 9, and 5 and 10 contain 150 ng of probe hybridized with 10 u.g of R. capsulatus B10 R N A isolated 30 min following a shift of the culture to low aeration, that were digested with 500,1000, or 1500 U of Sl-nuclease, respectively. 63 that the entire R. capsulatus segments of both the H i n d l l l and BamHI probes were protected by an m R N A molecule. This result verified that m R N A transcripts of the crtF gene continued into the bchCA operon. Quite unexpectedly, however, the BamHI probe was also protected by an m R N A molecule with a 3' end that mapped approximately 525 nt downstream of the Mlul site. This 3' end site, which would be positioned approximately 45 nt downstream of the H i n d l l l site in the bchC gene, suggested that some crfEF transcripts either are transcriptionally terminated or are posttranscriptionally processed to give molecules that end between the H i n d l l l and BamHI sites in the bchC gene. 4d. SI nuclease protection end site mapping of the 3' end within the bchC gene. The 3' end site located between the H i n d l l l and BamHI sites in the bchC gene was mapped more precisely using a probe that was 3' end-labeled at the H i n d l l l site in the bchC gene and extended 557 bp in the 3' direction to the Fspl site in the bchC gene (Figure 21 A) . Because the probe used in this experiment was not bipartite, the 557 nt band observed in both the control and experimental lanes (Figure 21B) was due to protection of the probe by its complementary D N A strand as well as by bchCA m R N A that continued past the 3'-terminal nt of the R- capsulatus segment of the probe. The two smaller bands, approximately 57 and 60 nt in length, indicated that there were in fact two 3' end sites that mapped approximately 57 and 60 nt downstream of the H i n d l l l site. Examination of the D N A sequence of the bchC gene (Figure 11) showed that two regions of imperfect inverted symmetry could be found between the H i n d l l l and BamHI sites at approximately the same locations as these two 3' ends, although neither inverted repeat is immediately followed by T-rich sequences (Figure 21C). Because the distance from the predominant bchCA 5' end site ( R E S U L T S section 4b) to these 3' end sites was approximately 350 nt, transcription that began at the bchCA promoter and ended at the position of the inverted repeats may have given rise to the ca. 330 nt transcript detected by the bchC-specific probe in the R N A blots (Figure 18A and B). However, it was also possible that this short transcript was due instead to degradation of longer crtEF readthrough transcripts. That is, if crfEF m R N A transcripts that extend to either inverted repeat in the bchC coding region were cleaved by an endonuclease that cut at or very near the bchCA promoter, a ca. 330 ht degradation product would have been produced. Because this product would overlap the bchC gene, it would have been detected by 64 Figure 21. Sl-nuclease protection mapping of the 3' ends within the bchC coding region. Panel A : Representation of the double-stranded D N A probe that was 3' end-labeled (*) at the HmdIII site in the bchC gene and extended to the Fspl site in the bchC gene. Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, S l -nuclease resistant hybrids were separated. Lane 1: Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. Lanes 2 contains 100 ng of probe hybridized with 10 ug of yeast t R N A and digested with 500 U of Sl-nuclease. Lanes 3,4, and 5 contain 100 ng of probe hybridized with 10 |ig of R. capsulatus B10 R N A isolated 30 min following a shift of the culture to low aeration, that were digested with 500, 1000, or 1500 U of Sl-nuclease, respectively. Panel C : Nucleotide sequences of the inverted repeat regions within the bchC gene. Nucleotide positions correspond to those in Figure 11. 65 B 1 2 3 4 5 849 525 ^ —. 341 A \ c / 98 69 G - C I I C - G I I C - G I I G - C I I A - T I I A - T f \ T C X I G - C I I C - G I I T - A I I G - C I I T C G G A G A G - C G G G C G A 3 ' T C G G A G A G - C A C G T C T 3 ' 66 the bcrtC-specific probe used in the R N A blot experiments (Figure 18A and B). If the bdiC-internal 3' ends were exclusively due to termination of crtEF readthrough transcripts, then elimination of crtEF readthrough transcription should cause these 3' ends to disappear. Alternatively, if the fccfrC-internal 3' ends were due to transcriptional termination or posttranscriptional processing of m R N A molecules derived exclusively from the bchCA promoter, then elimination of crtEF readthrough transcription should not change the levels of these 3' ends. If the frcftC-internal 3' ends were due to a combination of these models, elimination of crtEF readthrough transcription would result in a decrease in the levels of these 3' ends, but would not eliminate them entirely. Therefore, an experiment was designed to test whether the internal 3' ends were located primarily on the termini of crtEF readthrough transcripts, whether they were located primarily on the termini of bchCA transcripts, or whether they were located on both m R N A species. This experiment compared the relative quantities of the bcrtC-internal 3' ends in the wild-type strain R. capsulatus SB1003 and in the crtF interposon mutant R. capsulatus DE324, in which transcriptional readthrough from the crtEF operon into the bchCA operon had been interrupted by the chromosomal insertion of an omega cartridge into the Sphl site in the crfF gene (124). The relative quantities of the bcftC-internal 3' ends were measured in an Sl-nuclease protection experiment using a bipartite probe that was labeled at the 3' end at the Bell site 5 bp downstream of the bchCA promoter, which extended 400 nt in the 3' direction to the HmdIII site in the bchC gene, and which contained 121 bp of pUC13 vector D N A as a tail (Figure 22A). To ensure that the probe would be in molar excess over the bchCA m R N A molecules, a preliminary experiment titrated various amounts of the probe with a constant input of R. capsulatus SB1003 R N A that had been extracted 30 m i n following a shift of a culture to low aeration. This time point was selected because it was likely to contain the maximum proportion of bchCA m R N A molecules, based on the results of the R N A blots (Figure 18). The results of the titration experiment are shown in Figure 22B. The band of interest in this experiment, which results from protection of the probe by m R N A with a 3' end between the HmdIII and BamHI sites in the bchC gene, is 295 nt in length. The relatively constant intensity of the 295 nt band when 50,100,200, and 400 ng of probe were used indicated that 50 ng of probe per 10 pg of R N A was sufficient to achieve a molar excess. To be certain of achieving a molar excess, 100 ng of probe was annealed to R N A samples from R. capsulatus SB1003 and R. 57 Figure 22. Titration of bc/iC-internal Sl-nuclease protection 3' end site mapping probe. Panel A: Representation of the double-stranded bipartite DNA probe that contained R. capsulatus DNA (thick line) that was 3' end-labeled (*) at the Bell (Be) between the crtF and bchC genes, extended to the Hindlll (H) site in the bchC gene, and included pUC13 DNA as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8M urea through which denatured, Sl-nuclease resistant hybrids were separated. Lane 1: Haelll-digested M13 mpll single-stranded DNA size markers, with the sizes of selected bands shown on the left margin. Lanes 2, 3,4, and 5 contain 50,100, 200, and 400 ng (respectively) of probe hybridized with 10 pg of R. capsulatus SB1003 RNA isolated 30 min following a shift of the culture to low aeration, that were digested with 500 U of Sl-nuclease. 68 capsulatus DE324. Hybrids were trimmed with SI nuclease, separated through a polyacrylamide gel, and the resultant autoradiogram is shown in Figure 23. This autoradiogram shows the presence of a 521 nt band in all lanes that was due to D N A - D N A reannealing of the probe, a 400 nt band that was due to m R N A that extends beyond the 3' most nt of the R. capsulatus segment of the probe, and a 295 nt band that was due to m R N A molecules ending at either of the two inverted repeat regions within the bchC coding region. Comparison of the relative intensities of the 400 nt and 295 nt bands in this autoradiogram showed that both bands were approximately 40% (by eye) less intense in the lanes containing R. capsulatus DE324 R N A than they were in the corresponding lanes containing R. capsulatus SB1003 R N A . Because this 295 nt band was still present in the lanes containing R. capsulatus DE324 R N A , the short transcript observed on the R N A blots cannot have been solely due to processing of crtEF transcripts, but must have resulted, in part, from transcription termination or posttranscriptional processing of m R N A molecules initiated at the bchCA promoter. Interestingly, the pattern of expression over time appeared to be slightly different in the two strains. The amount of m R N A transcripts in R. capsulatus SB1003 were low but detectable i n highly aerated cells, increased approximately 5 fold to reach a maximal level at 30 m i n after the shift to low aeration, and remained fairly constant in the remaining samples. However, the transcripts in R. capsulatus DE324 were present at much lower levels in highly aerated cells (approximately one fifth of the level observed in R. capsulatus SB1003), increased about 10 fold to reach a maximum level at 30 min (to approximately the same level observed in R. capsulatus SB1003), and decreased to approximately 30% of the maximum level by 60 min. This pattern was observed in four independent experiments. Thus, crtEF readthrough transcription in wild-type cells appeared to be necessary to give transcription of bchCA genes under high aeration, and to maintain high levels of bchCA transcription during the latter stages of this shift from high to low aeration. 4e. 3' end site mapping of bchCA m R N A transcripts. Because the puf operon promoter is buried in the coding region of the bchA.3 gene, (124), transcripts of the bchCA operon must overlap at least the promoter region of the puf operon. Direct evidence for this overlap was sought from Sl-nuclease protection experiments. A bipartite probe designed to map the 3' ends of bchCA operon transcripts was radioactively 3' 69 Figure 23. Comparison of frcnC-internal 3' ends in R: capsulatus SB1003 and R. capsulatus DE324. Panel A : Representation of the double-stranded bipartite D N A probe that contained R. capsulatus D N A (thick line) that was 3' end-labeled (*) at the Bell (Be) between the crtF and bchC genes, extended to the H i n d l l l (H) site in the bchC gene, and included pUC13 D N A as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-nuclease resistant hybrids were separated. One hundred ng of probe was hybridized to 10 pg of the indicated R N A sample and digested with 500 U of S l -nuclease prior to electrophoresis. Lanes 1 and 8 contain E. coli t R N A . Lanes 2 through 6 contain R. capsulatus SB1003 R N A isolated at 0 min (lane 2), 15 min (lane 3), 30 m i n (lane 4), 45 min (lane 5) and 60 m i n (lane 6) following a shift of the culture to low aeration. Lanes 9 through 13 contain R. capsulatus DE324 R N A isolated at 0 min (lane 9), 15 min (lane 10), 30 m i n (lane 11), 45 m i n (lane 12) and 60 min (lane 13) following a shift of the culture to low aeration. Lane 7 contains Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. crtEF bchCA.lA.2A.3 pufQBALMX Be B B 1 2 3 4 5 6 7 8 9 10 11 12 13 71 end-labeled at the EcoRl site in the bchA.3 gene, extended 865 bp in the 3' direction to the Sail site in the pufQ gene, and included 199 bp of pUC13 vector D N A as a tail (Figure 24A). A n autoradiogram of the resultant Sl-protected hybrids (Figure 24B) shows a band approximately 1060 nt in length in all lanes, due to reannealing of the bipartite D N A probe, and a band approximately 840 nt in length i n the lanes containing R. capsulatus R N A , which suggested that the probe was protected by an m R N A molecule that extends at least to the Sail site in the pufQ gene. The numerous 3' ends of smaller molecules visible in Figure 24B presumably arose from rapid degradation of the unstable bchCA m R N A , as noted above (RESULTS section 4a). 4f. 5' end site mapping of puf m R N A transcripts. A 5' end site mapping experiment was designed to confirm the overlap of bchCA and puf operon transcripts. Two bipartite probes were prepared. One probe was radioactively 5' end-labeled at the Sail site in the pufQ gene, extended 0.69 kb in the 5' direction to the Xholl site in the bchA gene, and included 2.7 kb of pUC13 vector D N A as a tail (Figure 25A). The second probe was 5' end-labeled at the EcoRl site between the pufQ and pufB genes, extended 0.885 kb in the 5' direction to the Xholl site in the bchA.3 gene, and included 2.7 kb of pUC13 vector D N A as a tail (Figure 25A). A n autoradiogram of the resultant Sl-protected hybrids (Figure 25B) shows a 3.4 kb band for the Sail probe and a 3.6 kb band for the EcoRl probe that were each due to reannealing of the respective bipartite D N A probe. The lengths of bands resulting from protection of the Sail probe by m R N A molecules were: 730 nt, due to readthrough from the bchCA operon; 475 nt, believed to be due to processing of the bchCA readthrough transcript; and 360 nt, due to initiation of transcription from the puf operon promoter (1). The lengths of bands resulting from protection of the EcoRl probe by m R N A were: 910 nt due to readthrough from the bchCA operon; 690 nt due to processing of bchCA readthrough transcription; 550 nt, due to transcription initiation from the puf operon promoter; and 300 nt and 290 nt, that are both due to processing of the puf operon m R N A (1). Because these results clearly demonstrated that m R N A from the bchCA operon extended into the puf operon at least as far as the EcoRl site between the pufQ and pufB genes, it is possible that puf operon genes may be translated from m R N A molecules that initiate at the bchCA promoter or even perhaps at the crtEF promoter. 72 Figure 24. Sl-nuclease protection 3' end site mapping of bchCA m R N A transcripts. Panel A : Representation of the double-stranded bipartite D N A probe containing R. capsulatus D N A (thick line) that was 3' end-labeled (*) at the EcoRl (E) site in the bchA.3 gene, extended to the Sail (S) site in the pufQ gene, and contained pUC13 D N A as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, S l -nuclease resistant hybrids were separated. Lane 1: Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. Lanes 2 contains 150 n g of probe hybridized with 10 |ig of E. coli t R N A and digested with 500 U of Sl-nuclease. Lanes 3,4 , and 5 contain 150 ng of probe hybridized with 10 Jig of R. capsulatus B10 R N A isolated 30 min following a shift of the culture to low aeration, that were digested with 500, 1000, or 1500 U of Sl-nuclease, respectively. crtEF bchCA.1A.2A.3 pufQBALMX E S 1 2 3 4 5 74 Figure 25. Sl-nuclease protection 5' end site mapping of puf m R N A transcripts. Panel A : Representations of the double-stranded bipartite D N A probes containing R. capsulatus D N A (thick lines) that were 5' end-labeled (*) at either the Sail (S) site in the pufQ gene or at the EcoRl (E) site between the pufQ and pufB genes, extended to the XnoII (X) site in the bchA.3 gene (thick lines) and included a tail of pUC13 D N A (thin lines). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-nuclease resistant hybrids were separated. Lanes 1 and 6: Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. Lanes 2 to 5: Sl-resistant fragments protected by the Sail probe. Lanes 7 to 10: Sl-resistant fragments protected by the EcoRl probe. Lanes 2 and 7 contain 200 ng of probe hybridized with 10 ug of yeast t R N A and digested with 500 U of Sl-nuclease. Lanes 3 and 8,4 and 9, and 5 and 10 contain 200 ng of probe hybridized with 10 ug of R. capsulatus B10 R N A isolated 30 min following a shift of the culture to low aeration, that were digested with 500,1000, or 1500 U of Sl-nuclease, respectively. 76 5. Functional significance of the overlapping crtEF. bchCA. and puf m R N A  transcripts. 5 a. Consequences of interrupting crtEF transcriptional readthrough on expression of the bchCA operon. The effects of interrupting readthrough transcription from the crtEF operon on expression of the bchCA and puf operons were evaluated using the crfF interposon mutant, R. capsulatus DE324, in comparisons to its wild-type parent, R. capsulatus SB1003. Both the levels of the BchC and B c h A enzymatic activities and the levels of bchCA m R N A were tested. Although there exists no direct assay for the activities of either the BchC or BchA enzymes, a decrease in the specific activity or in the cellular concentration of either enzyme may result in a bottleneck in the Bchl biosynthetic pathway. Such a bottleneck w o u l d be manifested by a slight accumulation of a Bchl intermediate indicative of the step in Bchl biosynthesis that is slowed. The construction and characterization of R. capsulatus DE324 has been described in a recent communication by Young et al. (124). Although the data are not shown, they note in the text of their manuscript that a small amount of the Bchl a biosynthetic intermediate, 2-devinyl-2-hydroxyethyl bacteriochlorophyllide a, is detectable in the supernatant fluids from photosynthetic and low oxygen cultures of R. capsulatus DE324 but not in the supernatant fluids from cultures of crtF point mutant strains. Because this intermediate is accumulated in strains of R. capsulatus lacking BchA activity, the results of Young et al. (124) suggest that the absence of crtEF readthrough transcripts in R. capsulatus DE324 may impair carbon flow through the Bchl biosynthetic pathway. I wished to confirm the results of Young et al. (124) by performing quantitative spectroscopic measurements of the pigments found in R. capsulatus DE324. I compared the pigment profiles of R. capsulatus SB1003 (wild-type), R. capsulatus CW100 (bchC interposon mutant), R. capsulatus BP503 (crtF point mutant), and R. capsulatus DE324 (crtF interposon mutant). Oxygen-limited cultures of each strain, and photosynthetic cultures of each strain except R. capsulatus CW100 (which is photosyhthetically incompetent), were grown. Samples were harvested both at mid-log and at stationary phases of growth. Spectra were recorded for intact cells, pigments extracted from cells with acetoneimethanol (7:2), and the culture supernatant fluids (Figure 26). The Bchl a biosynthetic intermediate, 2-devinyl-2-hydroxyethyl 77 Figure 26. Absorbency scans of R. capsulatus CW100, R. capsulatus DE324, JR. capsulatus SB1003, and R. capsulatus BP503. Oxygen-limited cultures of each strain, and photosynthetic cultures of each strain except R. capsulatus CW100 (which is photosynthetically incompetent) were grown to stationary phase. Spectra of oxygen-limited and photosynthetic cultures are shown on two consecutive pages. Equal numbers of cells from each strain were harvested and scanned over the range of wavelengths as indicated on the X axis. The absorbency values obtained are shown on the Y axis. Scan 1, pigment extracts in acetone:methanol (7:2); scan 2, culture supernatant fluids; scan 3, whole-cell spectra. Tracing a, R. capsulatus CW100; tracing b, R. capsulatus DE324; tracing c, R. capsulatus SB1003; and tracing d, R. capsulatus BP503. 80 bacteriochlorophyllide a (P668), has a red-most absorption peak at 668 n m and is accumulated in strains such as R. capsulatus CW100 that have no B c h A activity. Spectra of oxygen-limited R. capsulatus CW100 cultures detected P668 in intact cells, in the pigments extracted from cells, and in the culture supernatant fluid. Neither the wild-type strain R. capsulatus SB1003 nor the crtF point mutant strain R. capsulatus BP503 accumulated detectable levels of P668 in whole cells, pigment extracts, or culture supernatant fluids, regardless of whether the cultures were grown with low aeration or photosynthetically. Cultures of the crtF interposon mutant R. capsulatus DE324, when grown with low aeration, also d id not accumulate detectable levels of P668 in intact cells, extracted pigments, or in the culture supernatant fluid. However, small amounts of P668 were found in extracted pigments and in the spent medium of photosynthetically grown R. capsulatus DE324. The amount of P668 detected in intact cells of photosynthetically grown R. capsulatus DE324 was extremely small, but d id result in a reproducible bump between 660 and 670 n m in the recorded spectra. For each strain, the spectra obtained from mid-log phase cultures were almost identical to those obtained from stationary phase cultures (data not shown). Spectral differences in the 400-600 n m and the 750-1000 n m ranges among these strains do not affect the interpretation of these results. These results showed that under the photosynthetic growth conditions used by me, the lack of crtEF transcripts reading through the bchCA operon caused a shortage of B c h A activity, so that the flow of Bchl biosynthetic intermediates through the pathway became rate-limiting at the step catalysed by the BchA enzyme. Contrary to the results of Young et al. (124), who found P668 in the culture media of photosynthetic and oxygen-limited cultures of R. capsulatus DE324,1 found no evidence for accumulation of P668 in oxygen-limited cultures. However, this difference may have been due to subtle differences in the growth and handling of the cultures. The slight accumulation of P668 in photosynthetically grown R. capsulatus DE324 presumably reflected a reduction in the level of bchCA m R N A caused by the interposon mutation. However, there are two mechanisms whereby this interposon mutation may have caused a shortage of bchCA m R N A . If crtEF transcripts extend to the end of the bchA gene cluster, it would be possible to translate the BchC and BchA enzymes from an m R N A molecule initiated at the crtEF promoter. Therefore, elimination of the crtEF readthrough transcripts in R. capsulatus DE324 would have proportionately reduced the amount of m R N A also encoding 81 the bchCA operon. A second possibility is that elimination of the crtEF/bchCA transcriptional overlap in R. capsulatus DE324 may have impaired transcription initiation from the bchCA promoter, which also would decrease the steady-state level of bchCA m R N A transcripts available for translation. A s well, it should be noted that these hypotheses are not mutually exclusive. The possibility that the crtEF/bchCA transcriptional overlap affected the function of the bchCA promoter was tested by a quantitative Sl-nuclease protection experiment, using a bipartite probe that was 5' end-labeled at the BamHl site in the bchC gene and extended to the Mlul site in the crfF gene (Figure 27A), to measure the amounts of R N A 5' ends that result from initiation at the bchCA promoter. This probe was the same probe that was used to map the 5' ends of bchCA m R N A (Figure 19). A preliminary Sl-nuclease protection experiment was performed to guarantee that the probe would be in molar excess with respect to bchCA m R N A molecules. R N A extracted from R. capsulatus SB1003 at 30 m i n following a shift to low aeration was titrated against 50,100,200, and 400 ng of probe (Figure 27B). The band due to initiation of transcription at the bchCA promoter was 405 nt in length. The constant intensity of this band over the range of probe concentrations tested indicated that the probe was in molar excess even at a concentration of 50 n g / 10 ug of R capsulatus SB1003 R N A . To test for effects of crtEF transcriptional readthrough on bchCA promoter function, R N A from R. capsulatus SB1003 and R. capsulatus DE324 was extracted at 15 m i n intervals following a shift of the culture to low aeration and was used to protect a molar excess of the radioactively labeled BamHl bchCA 5' probe from digestion by Sl-nuclease (Figure 28). Comparison of the control lanes (where the probe was hybridized to E. coli tRNA) with the experimental lanes (where the probe was hybridized to R. capsulatus R N A ) showed the presence of a faint band approximately 415 nt in length. Although this band was positioned near the 405 nt band due to initiation of transcription at the bchCA promoter, the 405 nt band was clearly absent in the control lanes. Therefore, the 415 nt band was an artifact, but the 405 nt band was not. Further examination of this autoradiogram showed that the 640, 620, and 550 nt bands were absent in the R. capsulatus DE324 samples, which confirmed that these bands were due to protection of the probe by crtEF readthrough transcripts. Another interesting point was that the relative intensities of the 405 nt band (due to initiation of transcription at the bchCA 82 crtEF bchCA.1A.2A.3 pufQBALMX M B Figure 27. Titration of bchCA Sl-nuclease protection 5' end site mapping probe. Panel A: Representation of the double-stranded bipartite DNA probe containing R. capsulatus DNA (thick line) that was 5' end-labeled (*) at the BamHl (B) in the bchC gene, extended to the Mlul (M) site in the crtF gene, and included pUC13 DNA as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8M urea through which denatured, Sl-nuclease resistant hybrids were separated. Lane 1: Hoelll-digested M13 mpll single-stranded DNA size markers, with the sizes of selected bands shown on the left margin. Lanes 2, 3, 4, and 5 contain 50, 100, 200, and 400 ng (respectively) of probe hybridized with 10 ug of R. capsulatus SB1003 RNA isolated 30 min following a shift of the culture to low aeration, that were digested with 500 U of Sl-nuclease. 83 Figure 28. Comparison of bchCA 5' m R N A ends in R. capsulatus SB1003 and R. capsulatus DE324. Panel A : Representation of the double-stranded bipartite D N A probe containing R. capsulatus D N A (thick line) that was 5' end-labeled (*) at the BamHI (B) in the bchC gene, extended to the Mlul (M) site in the crtF gene, and included pUC13 D N A as a tail (thin line). Panel B : Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-nuclease resistant hybrids were separated. One hundred ng of probe was hybridized to 10 pg of the indicated R N A sample and digested with 500 U of Sl-nuclease. Lanes 1 and 8 contain E. coli t R N A . Lanes 2 through 6 contain R. capsulatus SB1003 R N A isolated at 0 m i n (lane 2), 15 min (lane 3), 30 min (lane 4), 45 min (lane 5) and 60 m i n (lane 6) following a shift of the culture to low aeration. Lanes 9 through 13 contain R. capsulatus DE324 R N A isolated at 0 m i n (lane 9), 15 m i n (lane 10), 30 min (lane 11), 45 min (lane 12) and 60 m i n (lane 13) following a shift of the culture to low aeration. Lane 7 contains Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. 85 promoter) over the time course differed in R. capsulatus DE324 relative to the wild-type strain JR. capsulatus SB1003. Comparison of the 0 min time points showed that the bchCA promoter was functioning when the R. capsulatus SB1003 culture was under high aeration, but its activity in R. capsulatus DE324 was approximately one fifth the level observed in R. capsulatus SB1003. A t 15 min after the shift to inducing conditions, the bchCA promoter appeared to have equal activity in both strains. A t the later time points, however, while initiation from the bchCA promoter in R. capsulatus SB1003 appeared to be relatively constant, the corresponding initiation in R. capsulatus DE324 gradually decreased over time so that the level of transcription at 60 m i n was approximately half of the maximum observed at 15 min. This pattern was reproduced in three independent experiments. These results suggested that not only did readthrough transcription from the crtEF operon seem to stimulate the activity of the bchCA promoter, but that it was necessary for maintenance of high levels of initiation during a shift of the culture from high to low aeration, as was noted in R E S U L T S section 4d and Figure 23. 5b. Consequences of interrupting crtEF transcriptional readthrough on initiation of puf operon transcription. The effects of interrupting crfEF readthrough transcription on the initiation of puf operon transcription were assessed by an Sl-nuclease protection experiment that was designed to test for changes in the levels of puf m R N A transcripts in R. capsulatus DE324 relative to R. capsulatus SB1003. The bipartite probe used in this experiment, which was 5' end-labeled at the EcoRl site between the pufQ and pufB genes and extended to the XftoII site in the bchA.3 gene (Figure 29A), had been used previously (Figure 25). A s discussed above, a preliminary experiment was performed in order to establish the correct ratio of probe to R N A to ensure that the probe would be in molar excess (Figure 29B). A t least 200 ng of probe per 10 u.g of R N A was required before the intensity of the 550 nt band, due to initiation of transcription from the puf promoter, was relatively constant. R N A extracted from R. capsulatus SB1003 and R. capsulatus DE324 cultures every 15 m i n following a shift to low oxygen conditions was used to protect a molar excess of the radioactively labeled EcoRl puf 5' probe from digestion with with Sl-nuclease. A n autoradiogram of the resultant gel is shown in Figure 30. It is important to remember that this experiment compared the level of transcription initiation at the puf promoter in R. capsulatus 86 Figure 29. Titration of the puf Sl-nuclease protection 5' end site mapping probe. Panel A : Representation of the double-stranded bipartite D N A probe containing R. capsulatus D N A (thick line) that was 5' end-labeled (*) at the EcoRl (E) between the pufQ and pufB genes, extended to the Xholl (X) site in the bchA.3 gene, and included pUC13 D N A as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-nuclease resistant hybrids were separated. Lane 1: Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. Lanes 2, 3, 4, and 5 contain 50,100, 200, and 400 ng (respectively) of probe hybridized with 10 ug of R. capsulatus SB1003 R N A isolated 30 m i n following a shift of the culture to low aeration, that were digested with 500 U of Sl-nuclease. B 1 2 3 4 5 88 Figure 30. Comparison of puf 5' m R N A ends in R. capsulatus SB1003 and R. capsulatus DE324. Panel A : Representation of the double-stranded bipartite D N A probe containing R. capsulatus D N A (thick line) that was 5' end-labeled (*) at the EcoRl (E) between the pufQ and pufB genes, extended to the XHOII ( X ) site in the bchA.3 gene, and included pUC13 D N A as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-nuclease resistant hybrids were separated. Two hundred ng of probe (Figure 29A) was hybridized to 10 ug of the indicated R N A sample and digested with 500 U of Sl-nuclease prior to electrophoresis. Lanes 1 and 8 contain E. coli t R N A . Lanes 2 through 6 contain R. capsulatus SB1003 R N A isolated at 0 min (lane 2), 15 m i n (lane 3), 30 min (lane 4), 45 m i n (lane 5) and 60 m i n (lane 6) following a shift of the culture to low aeration. Lanes 9 through 13 contain R. capsulatus DE324 R N A isolated at 0 m i n (lane 9), 15 min (lane 10), 30 m i n (lane 11), 45 min (lane 12) and 60 min (lane 13) following a shift of the culture to low aeration. Lane 7 contains Haelll-digested M13 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. 90 SB1003, in which transcripts initiated at either the crtEF and bchCA promoters could continue into the puf operon, with R. capsulatus DE324, in which only transcripts initiated at the bchCA promoter could continue into the puf operon. In general, bands resulting from protection of the probe by R. capsulatus DE324 m R N A were fainter than those resulting from protection of the probe by R. capsulatus SB1003 m R N A (Figure 30). The bands of interest are: 910, 760, and 690 nt that were due to protection of the probe by bchCA readthrough transcripts and their processing products; and 550 nt that was due to protection of the probe by m R N A initiated at the puf promoter. One minor point is that the previous puf operon 5' end site mapping experiment (Figure 25) revealed only two bands, 910 and 690 nt, that were due to protection of the probe by bchCA readthrough transcripts. The additional 760 nt band observed in Figure 30 may have been due to the different genetic backgrounds of the two strains used as the source of R N A . R. capsulatus B10 R N A was used to generate Figure 25, whereas R. capsulatus SB1003 R N A was used to generate Figure 30. Al though these strains are closely related (107a), there are significant differences in the relative amounts of B800-850 and B870 complexes accumulated in oxygen-limited or photosynthetically grown cells (data not shown) which might reflect differences in expression of photosynthesis genes. Because the 910 nt, 760 nt, and 690 nt bands on this autoradiogram were slightly less intense in the lanes containing R. capsulatus DE324 R N A relative to those containing R. capsulatus SB1003 R N A , interruption of crtEF readthrough transcription appeared to decrease the number of transcripts that extended from the bchCA operon into the puf operon. However, it is not possible to tell whether this reduction was due to the loss of the crtEF readthrough transcripts that extended into the puf operon, whether it was due to a less active bchCA promoter caused by the crtF interposon mutation, or whether it was due to a combination of these mechanisms. The intensity of the 550 nt band, due to initiation from the puf promoter, was approximately 30% (by eye) lower in lanes containing R. capsulatus DE324 R N A relative to those containing R. capsulatus SB1003 R N A , but showed the same pattern of transcription over time. This relative difference in band intensity was observed in three independent experiments. These results indicated that the interruption of crtEF transcription could affect initiation from the puf promoter almost 8 kb away. However, as above, it is not possible to tell whether this 91 effect was direct, or whether the crtEF interposon mutation affected bchCA transcription initiation, which in turn affected puf transcription initiation. 5c. Consequences of interrupting all transcriptional readthrough on initiation of puf operon transcription. Transcription initiation from the puf promoter was compared in the bchC interposon mutant R. capsulatus CW100 and its wild-type parent, R. capsulatus BIO. The interposon mutation in R. capsulatus CW100 was expected to prevent nearly all transcripts initiated at either the crtEF promoter or at the bchCA promoter from continuing into the puf operon. R N A extracted from R. capsulatus CW100 and R. capsulatus BIO at 15 min intervals following a shift to low aeration was used to protect a molar excess (Figure 29) of the radioactively labeled EcoRl puf 5' probe (Figure 31 A) from digestion with SI nuclease. A n autoradiogram of the resulting Sl-protected hybrids (Figure 31B) showed that, in general, bands resulting from protection of the probe by R. capsulatus CW100 m R N A were fainter than those resulting from protection of the probe by R. capsulatus BIO m R N A . The bands of particular importance in the interpretation of this experiment are: 910 nt and 690 nt in length, due to bchCA readthrough transcription, and the 550 nt band, due to initiation of puf operon transcription. The 910 nt and 690 nt bands were very faint, but still visible, in R. capsulatus CW100. Therefore, although there was no detectable product of the BchA-catalyzed reaction in R. capsulatus CW100, there were a few residual readthrough transcripts. It is not possible to conclude from these data whether these residual readthrough transcripts were initiated at the crfEF or bchCA promoters, or at a secondary, possibly K n R cartridge-associated, promoter. More importantly, the intensity of the 550 nt band, which resulted from initiation of transcription from the puf operon promoter, was approximately 40% (by eye) lower in lanes containing R. capsulatus CW100 R N A relative to lanes containing R. capsulatus B10 R N A , and was reproducible in four independent experiments. These results demonstrated that a large reduction in the number of readthrough transcripts into the puf operon partially impaired transcription initiation from the puf operon promoter. Said another way, readthrough transcription from the crtEF and bchCA operons appeared to stimulate initiation of transcription from the puf promoter. 92 Figure 31. Comparison of puf 5' m R N A ends in R. capsulatus BIO and R. capsulatus CW100. Panel A : Representation of the double-stranded bipartite D N A probe containing R. capsulatus D N A (thick line) that was 5' end-labeled (*) at the EcoRl (E) between the pufQ and pufB genes, extended to the Xholl (X) site in the bchA.3 gene, and included pUC13 D N A as a tail (thin line). Panel B: Autoradiogram of a 5% polyacrylamide gel containing 8 M urea through which denatured, Sl-nuclease resistant hybrids were separated. Two hundred ng of probe (Figure 29A) was hybridized to 10 ug of the indicated R N A sample and digested with 500 U of Sl-nuclease prior to electrophoresis. Lanes 1 and 8 contain E. coli t R N A . Lanes 2 through 6 contain R. capsulatus BIO R N A isolated at 0 min (lane 2), 15 m i n (lane 3), 30 min (lane 4), 45 min (lane 5) and 60 min (lane 6) following a shift of the culture to low aeration. Lanes 9 through 13 contain R. capsulatus CW100 R N A isolated at 0 min (lane 9), 15 min (lane 10), 30 m i n (lane 11), 45 m i n (lane 12) and 60 m i n (lane 13) following a shift of the culture to low aeration. Lane 7 contains Haelll-digested M l 3 m p l l single-stranded D N A size markers, with the sizes of selected bands shown on the left margin. 93 94 5d. Cis/Trans test of puf promoter stimulation by bchCA readthrough transcription. I offer two mechanisms whereby transcription of the bchCA operon could stimulate the initiation of puf operon transcription. One possibility is that R N A polymerase molecules actively transcribing the bchCA operon alter the conformation of the puf operon promoter and thus enhance puf operon transcription initiation. This mechanism would be possible only in cis. Alternatively, it is possible that transcription of the puf operon could be regulated, in part, by the availability of a bcftCA-encoded gene product. If so, the observed decrease of puf transcription initiation in R. capsulatus CW100 could have been due, for example, to the lack of the BchC or B c h A proteins, to the lack of mature Bchl a, or to the presence of large amounts of a Bchl a precursor, in the cell. Therefore, stimulation of the puf promoter in wild-type cells would be possible in trans. If the stimulation of puf transcription initiation by bchCA readthrough was cis-dependent, then the level of transcription of a plasmid-borne copy of the puf operon should be identical in the wild-type strain R. capsulatus BIO and in the bchCA readthrough-deficient strain R. capsulatus CW100. Alternatively, if the stimulation of puf operon transcription initiation was frans-dependent, then the level of transcription of a plasmid-borne copy of the puf operon should be measurably different in these two strains. Moreover, if it is Bchl a rather than the BchC or BchA proteins per se that affects puf operon transcription, these differences should also be demonstrable in other Bchl mutant strains that contain chromosomal mutations in bch genes that are not transcriptionally linked to the puf operon. The recombinant plasmid pXCA6::935 (1) was chosen as an indicator plasmid for the cis/trans experiment. This plasmid contains R. capsulatus D N A extending from the XftoII site in the bchA.3 gene to an in frame-fusion between the pufB and lac 'Z genes, and thus only the puf promoter could initiate transcription of the pufBv.lac'Z fusion. I chose to monitor puf expression by measuring the specific activity of the pufB-.-.lac'Z fusion protein rather than by directly measuring the levels of the plasmid-encoded puf m R N A 5' ends in strains harboring pXCA6::935 because the large distance between the puf promoter and the pufB::lac'Z fusion point on pXCA6::935 made the design of a bipartite probe capable of distinguishing between chromosomally-initiated and plasmid-initiated puf m R N A transcripts difficult. In this attempt to distinguish between the cis and trans mechanisms of puf transcription stimulation, pXCA6::935 was mobilized by conjugation into R. capsulatus CW100 (bchCA') and its wild-95 type parent, R. capsulatus BIO. Plasmid pXCA6::935 was also conjugated into two other bch point mutant strains, R. capsulatus BRP15 (bchH~) and R. capsulatus BRP4 (bchE~), as well as their B c h l + parent, R. capsulatus BP503. The /3-galactosidase specific activity in extracts of the exconjugant cells was determined as outlined in M A T E R I A L S A N D M E T H O D S , and the results are presented in Table IV. In extracts of cells grown with low aeration, the /?-galactosidase specific activity i n the Bchl" strains was approximately 25% less than the activity i n the appropriate B c h l + parental strain. Extracts of cells grown with high aeration showed the opposite trend, where the activity in the Bchl" strains was approximately 25% greater than that of the appropriate B c h l + parental strain. Thus, the ratio of specific activities in low O2 versus high O2 extracts indicated that initiation of transcription of the plasmid-borne puf promoter in each of the three Bchl" strains was less sensitive to regulation by oxygen than were the appropriate B c h l + strains. A t face value, these results suggested that all three bch mutations were capable of affecting expression of the puf operon in trans, and suggested that the reduction in transcription initiation from the puf promoter in R. capsulatus CW100 may have been due to the block in Bchl biosynthesis rather than the elimination of readthrough transcription per se. However, a careful analysis of the results showed that the standard deviations of these experiments almost overlapped these subtle differences in /3-galactosidase activities. Furthermore, it was noted that the protein content of the B c h l + cell extracts seemed to be slightly higher than that of the corresponding Bchl" cell extracts, even though equal volumes of the cultures were harvested at the same Klett density. These complications made the interpretation of this cis/trans test difficult (also see D I S C U S S I O N section 2). 5e. Contribution of bchCA readthrough transcription to the expression of a pufBv.lac'Z fusion. The results given above (Results section 5d) showed that secondary effects resulting from Bchl biosynthesis mutations could exert an influence on puf operon expression. Therefore, a different approach was taken to separate the direct readthrough effects from the indirect effects of bch gene mutations on transcription initiation at the puf promoter. In this approach, the contribution of bchCA readthrough transcription to expression of p« / -encoded polypeptides was evaluated by measuring the /?-galactosidase specific activity of a plasmid-borne T a b l e IV. / J - g a l a c t o s i d a s e s p e c i f i c a c t i v i t i e s o f a pufB::lac'Z f u s i o n i n B c h l + and B c h l - s t r a i n s . Low O2 H i g h O2 L / H c S t r a i n P h e n o t y p e P r o t e i n a / tea l a c t i v i t y b P r o t e i n / t e a l a c t i v i t y BIO Bch"1 3 .6 (0.4) d 2603.2 (305.5) 5 .0 (0.5) 198 .6 (18.3) 13 .1 CW100 B c h C - BchA" 2 .0 (0.2) 1990.3 (418.7) 3 .4 (0.6) 2 5 9 . 1 (29.4) 7.7 BP503 B c h 1 2 . 1 (0.3) 3120.7 (227 .8) 3 . 9 (0.3) 293 .8 (44.3) 10. 6 BRP15 BchH" 2 . 1 (0.5) 2484.7 (160.2) 3 .2 (0.4) 352 .0 (24.8) 7 . 1 BRP4 BchE" 2 .2 (0.3) 2527.4 (138.6) 2 . 6 (0.2) 322 .8 (38.7) e x p r e s s e d as mg/ml e x t r a c t e x p r e s s e d as nmoles oNPG c l e a v e d / m i n / m g p r o t e i n t h e r a t i o o f / J - g a l a c t o s i d a s e s p e c i f i c a c t i v i t y i n low O2 e x t r a c t s v e r s u s h i g h O2 e x t r a c t s s t a n d a r d d e v i a t i o n s o f f o u r i n d e p e n d e n t a s s a y s 97 pufBv.lac'Z fusion in the presence and absence of bchCA transcriptional readthrough. Each of the four plasmids used in this experiment contained the same pufBvlac'Z fusion, but each contained a different amount of R. capsulatus D N A 5' to the fusion point. T w o of the plasmids contained R. capsulatus D N A extending from the XnoII site in the bchA.3 gene to the pufBvlac'Z fusion point; they differed in that one carried the wild-type puf promoter (pXCA6::935) whereas the other contained a two nt mutation (known as A44) in the puf promoter, which had previously been shown to eliminate approximately 95% of puf promoter activity (pXCA6::935A44, ref. 1). This pair of plasmids was used to measure the activities of either the wild-type or mutant puf promoter in the absence of readthrough transcription. The other two plasmids contained R. capsulatus D N A extending from the EcoRl site in the crtF gene to the pufBvlac'Z fusion point. One plasmid (pXCA6: :EHQ) was completely wild-type in sequence, and therefore would measure the sum of the activities of both the bchCA and puf promoters. The final plasmid (pXCA6: :EHQA44) contained all wild-type R. capsulatus sequences except for the two nt puf promoter mutation, and therefore would measure the ability of the bchCA promoter alone to drive expression of the pufB::lac'Z fusion protein. The construction of plasmids pXCA6::935 and pXCA6::935A44 has been described previously (1). A flow chart for the construction of plasmids p X C A 6 : : E H Q and pXCA6: :EHQA44 is given in Figure 6 and is described in detail in MATERIALS AND METHODS. The /J-galactosidase specific activity in extracts of R. capsulatus BIO cells containing each of these plasmids was measured and the results are presented in Figure 32. Because each plasmid encoded the same pufB::lac'Z fusion protein, their resultant B-galactosidase specific activities were directly comparable. The activity of the wild-type puf promoter alone, as measured by plasmid pXCA6::935, was used as a baseline value in the interpretation of this experiment. These data showed that that transcription of the wi ld type puf promoter increased approximately 8 fold under inducing, or low O2, conditions. A s expected, the A44 mutation in the puf promoter, as measured by pXCA6::935A44, eliminated approximately 95% of its activity under low O2 conditions, and approximately 90% of its activity under high O2 conditions. The sum of the activities of the wild-type bchCA and puf promoters, as measured by plasmid p X C A 6 : : E H Q , was triple that of the wi ld type puf promoter alone under high O2 conditions, and slightly more than double that of the wild type puf promoter alone under low O2 conditions. Therefore, the presence of both promoters greatly 98 CONSTRUCT R E P R E S E N T A T I O N LOW 02 HIGH 02 L / H p X C A 6 : : 9 3 5 a p X C A 6 : : 9 3 5 A 4 4 bch'A. 3 pufQB ' X B L. I wm E E B p X C A 6 : : E H Q 1 1 <<< M l l l l crt 'F bchCA.lA.2A.3 pufQB' E E E B p X C A 6 : : E H Q A 4 4 2 5 5 1 . 2 ( 3 0 2 .4) 1 1 6 .2 ( 2 5 . 7 ) 6 6 2 4 . 8 ( 6 8 2 . 5 ) 2 4 9 7 . 1 ( 3 2 0 . 8 ) 304 .5 ( 2 8 . 8) 32 .2 ( 5 . 3 ) 1 5 4 8 . 3 ( 2 3 3 . 1 ) 854 .7 ( 9 5 . 2 ) 8 . 4 3.6 4 . 3 2.9 Figure 32. Contribution of bchCA transcription to expression of a pufBv.lac'Z fusion. Shown are the /?-galactosidase specific activities in extracts of cells containing a pufB::lac'Z fusion driven as designated by the puf and bchCA promoters. 1 Representation of the R. capsulatus segments of D N A fused to the lac 'Z gene. The structural genes encoded by the R. capsulatus inserts are shown for each construct. Restriction sites are represented as follows: EcoRl (E); XrtoII (X); and BamHI (B). The A44 mutation in the puf promoter is designated by two dots (• •). Arrows designate the locations of the active promoters in each construct. 2 /3-galactosidase specific activities are expressed as nmoles o N P G c l e a v e d / m i n / m g protein. The values in brackets are the standard deviations of four independent assays that were performed for each construct. ^ The ratio of /3-galactosidase specific activities obtained from cells grown with low aeration versus with high aeration. 99 increased the expression of the pufB::lac'Z fusion protein, but reduced its regulation by oxygen from 8 fold to 4 fold. These data suggested that either the bchCA promoter was as strong as but less sensitive to oxygen regulation than the puf promoter, or that the presence of bchCA readthrough transcription doubled the strength yet decreased the oxygen sensitivity of the puf promoter. These alternatives were distinguished by directly measuring the contribution of the bchCA promoter to this sum of activities, using plasmid pXCA6::EHQA44 in which the puf promoter was inactivated. The /J-galactosidase specific activities in extracts of cells containing pXCA6::EHQA44 was almost triple that of the wild-type puf promoter alone under high conditions, and was nearly identical to that of the wild-type puf promoter alone under low O2 conditions. Thus, the degree of oxygen-regulated transcription of the pufB::lac'Z fusion protein decreased from 8 fold to 3 fold, when driven from the bchCA promoter alone. Although the activity of the bchCA promoter was not as repressed by high oxygen concentrations as was the puf promoter, these results clearly demonstrated that the bchCA promoter was equally as capable as the puf promoter of driving the expression of a pufBv.lac'Z fusion protein. Moreover, because the /3-galactosidase activity in extracts of cells containing pXCA6: :EHQ was greater than the sum of the activities in extracts of cells containing pXCA6::935 plus pXCA6::EHQA44, the presence of readthrough transcription from the bchCA operon appeared to stimulate puf promoter activity, as had been indicated by the previous R N A studies. Therefore, I conclude that in wild-type R. capsulatus cells, readthrough transcription from the bchCA operon contributes significantly to the expression of pu/-encoded genes in the chromosomal context. 100 DISCUSSION 1. Overlapping m R N A transcripts of the crtEF, bchCA, and puf operons. The combined results of the R N A blotting and Sl-nuclease protection experiments presented in this study led to the following model for transcription and processing of the overlapping m R N A transcripts of the crtEF, bchCA, and puf operons (Figure 33). I have provided direct evidence that some crtEF m R N A transcripts end within the bchC gene, and that others extend beyond the BamHl site in the bchC gene. Similarly, I have shown that bchCA m R N A transcripts extend beyond the EcoRl site between the pufQ and pufB genes. Therefore, it is possible that transcription may be initiated at any of the crtEF, bchCA, or puf promoters, and may continue to the end of the puf operon. The 3' ends of operon-length puf transcripts map to two regions of dyad symmetry that are followed by the sequences 5'-TATT-3' and 5-TTTT-3' , respectively (20). Although it is highly probable that these regions act as efficient transcriptional terminators, the possibility of readthrough transcription extending beyond the puf operon has not been tested. Young et al. (124) have previously proposed a similar "superoperon" model of overlapping crtEF, bchCA, and puf m R N A s primarily on the basis of interposon mutagenesis, gene fusion, and sequence analysis experiments. I am indebted to Debra Young and Barry Marrs for bringing to m y attention the results of their genetic experiments which suggested the transcriptional overlap between the crtEF, bchCA, and puf operons. The R N A studies presented in this thesis directly support and extend their model. Although it is possible that m R N A molecules encoding the crrEF, bchCA, and puf operons exist in R. capsulatus, obtaining direct evidence for such long multi-operonic transcripts will probably be difficult, due to the very rapid degradation of m R N A transcribed from this region. However, the results of the blotting experiments using R. capsulatus BIO (pCW2) R N A (Figure 18D) provided some evidence for an m R N A species long enough to extend from the bchCA promoter to the puf operon 3' termination signals proposed previously (11, 20). Although it could be argued that this 11 kb transcript was an artifact due to transcription of the hybrid puf/bchCA m R N A molecule from p C W 2 , 1 believe that this is unlikely because the 101 crtEF bchCA .1A.2A.3 pufQBALMX Figure 33. M o d e l of the overlapping transcripts of the R. capsulatus crtEF, bchCA, and puf operons. The top line represents R. capsulatus chromosomal D N A encoding the crtEF operon, the bchCA operon, and the puf operon, as indicated. Horizontal arrows designate the m R N A transcripts of these three operons. Vertical arrows designate a suggested route of m R N A processing of primary transcripts. Solid lines represent authentic transcripts as have been shown by this and other work; thicker arrows represent greater amounts of m R N A transcripts. Dashed lines represent m R N A molecules that have not been identified directly, but that are likely to exist. Primary transcripts are indicated by an asterisk (*). 102 sequence of the hybrid puf/bchCA primary transcript would differ from the sequence of the chromosomally-expressed bchCA primary transcript only within a 0.5 kb segment at the extreme 5' end of the puf/bchCA primary hybrid molecule. Moreover, the BatriHl-C fragment that was inserted into p C W 2 contains the relatively well-characterized puf operon transcriptional terminators (11, 20) as well as an additional 2.25 kb of R. capsulatus D N A . Therefore, the 3' end of the hybrid puf/bchCA m R N A transcript would be expected map to the same position as would the chromosomally-expressed bchCA m R N A transcript. Each transcriptional overlap presented in Figure 33 was confirmed by at least two Sl-nuclease protection experiments, the results of which were consistent with one another. One experiment was designed to detect 3' ends plus continuation of transcription into the next downstream operon, whereas the other experiment was designed to detect 5' ends plus the presence of transcripts originating from the adjacent upstream operon. A l l except one of these experiments made use of bipartite probes, which consisted of a segment of end-labeled R. capsulatus D N A and a segment of heterologous D N A extending from the unlabeled end of the probe. The use of such bipartite probes was essential for the direct demonstration of the overlap between the m R N A transcripts of the crtEF, bchCA, and puf operons. Three different probes were used to map the 5' ends of the bchCA operon transcripts (Figures 15 and 19). A l l three detected the same major 5' end found to map 48 nt upstream of the bchC A T G initiation codon, which was the only 5' end associated with promoter activity in lac 'Z fusions. Significantly, each probe was protected by an m R N A molecule that annealed to the entire R. capsulatus segment of the probe, showing that transcripts from the previous upstream operon, the crfEF operon, extended into the bchC region. Supporting this conclusion were the observations that blockage of crtEF transcription by the insertion of an interposon into the crtF gene resulted in the disappearance of the bands on an autoradiogram that were due to the crtEF readthrough transcripts, or their processed products (Figure 28). The overlap between the bchCA and crtEF operons was confirmed by two experiments designed to test for 3' ends of crtEF m R N A transcripts that mapped to the bchCA coding region (Figure 20). The results of these experiments showed that the only detectable 3' termini of crtEF transcripts were found within bchC coding sequences. Two 3' ends were detected, and their sites were found to map near two regions with the potential to form hairpin loops in the m R N A (Figure 21B and C). A s well, both probes used to map the 3' ends of the crtEF 103 transcripts were protected by m R N A that continued beyond the 3'-most nucleotide of the R. capsulatus segment of the probe. Therefore, at least some crtEF transcripts extended even beyond the Bam¥H site located approximately 50% through the bchC gene. The presence of 3' end sites within the coding region of the bchC gene helped to explain the presence of the 0.33 kb transcript detected by the bcrtC-specific probe on the R N A blots, because m R N A molecules starting at the bchCA promoter and ending at either of the bcftC-internal inverted repeat regions would be approximately 330 nt in length. The results shown in Figure 23 demonstrated that such molecules existed. Therefore, it was possible that the frcftC-internal areas of inverted symmetry may have formed a region of secondary structure in the m R N A that resulted in premature termination of bchCA transcription to yield the short transcript. A n alternative possibility was that these regions of secondary structure may have been a barrier to m R N A degradation so that the short transcript was a stable remnant of a processed primary transcript, as has been documented previously for the region of secondary structure found between the puf A and pufL genes in R. capsulatus (11, 20). Chen et al. (20) have observed that at least 75% of R N A polymerase molecules transcribing the puf operon read through the pufA/pufL intercistronic hairpin, which consists of a G C - r i c h region of dyad symmetry followed by the sequence 5 ' -CATA-3 ' . They have shown that the efficiency of transcriptional termination at m R N A hairpins in R. capsulatus increases as the number of thymidine residues immediately following such structures increases. Because the sequences following the two inverted repeat regions in the bchC gene are 5 ' -GGGC-3 ' and 5 ' -ACGT-3 ' , respectively (Figure 21C), it seems unlikely that either inverted repeat effectively terminates transcription. Although I favor the hypothesis that the two inverted repeat regions are more likely to act as decay barriers than transcriptional terminators, some transcriptional termination by these secondary structures cannot be excluded. The overlap between the transcripts of the bchCA and puf operons was similarly characterized by both 5' and 3' Sl-nuclease protection experiments. A n experiment designed to map the 3' ends of the bchCA m R N A transcripts showed that the majority of bchCA transcripts extended into the pufQ coding region (Figure 24). The numerous other 3' ends also detected on this autoradiogram were likely to be degradation intermediates of bchCA m R N A . This overlap was confirmed by two experiments designed to detect the 5' ends of the puf m R N A transcripts. In addition to the 5' ends previously mapped (1), each of the two probes used for 104 these experiments detected readthrough transcripts that protected the entire R. capsulatus segment of each probe, as well as other 5' ends that were likely to be processing products of these readthrough transcripts (Figure 25). Confirmation that these additional bands resulted from protection of the probes by readthrough transcripts was obtained in analogous puf 5' end site mapping experiments using R. capsulatus CW100 R N A , where insertion of an interposon into the chromosomal copy of the bchC gene resulted in a drastic reduction of the intensity of the bands that correspond to the readthrough transcripts or their processing products (Figure 31). The model shown in Figure 33 also contains a suggested route of decay of these primary transcripts, although it should be stressed that more work is required to test for proposed precursor-product relationships between the primary transcripts of the crtEF and bchCA operons and their suggested degradation products. Of these three operons, degradation of the puf mRNA transcript is the best understood. It is believed that the extreme 5' end of the puf primary transcript, encoding the pufQ gene, is very rapidly removed, leaving a more stable 2.7 kb m R N A intermediate encoding the pufB, puf A, pufL, pufM, and pufX genes (1). The 3' end of this intermediate is then degraded to leave a very stable 0.5 kb segment encoding the pufB and puf A genes, which is protected against further rapid degradation by a hairpin loop at its 3' end (1, 11, 20). Although processing of the bchCA primary transcript is less well characterized, the data presented in the R N A blotting experiments (Figure 18) and the 3' S l -nuclease protection experiments (Figures 23 and 24) suggest that the 3' or central region of the bchCA primary transcript may be the most unstable, presumably because there is at least one endonucleolytic cleavage site in this region. Following endonucleolytic cleavage(s), rapid 3' to 5' exonucleolytic digestion could then degrade the bchCA m R N A until the bcwC-internal secondary structures are reached. The resultant remnants would encode only a 5' portion of the bchC gene. Because transcription of the bchCA operon extends well into puf operon coding regions, and probably continues to the end of the puf operon, it seems likely that the portion of the bchCA primary transcript that overlaps the puf operon would be subject to the same decay processes as the puf transcript. Although degradation of the crfEF m R N A transcript has not yet been studied, I assume that the portion of the crtEF primary transcript that overlaps the bchCA and puf operons can give rise to some of the same processing products as the bchCA and puf transcripts. The model of overlapping crtEFi bchCA, and puf operons presented in this work raises an interesting point of semantics. O n one hand, the crtEF, bchCA, and puf operons can be 105 viewed as separate, but overlapping operons. O n the other hand, they can also be viewed as one large operon with two internal promoters; one that promotes transcription of the bchCA and puf genes, and another that promotes transcription of only the puf genes. The grouping of cotranscribed genes into units called operons, therefore, is somewhat arbitrary when a model such as this is considered. Superoperonic clustering was first described in Pseudomonas species, when it was found that functionally related genes were tightly clustered on the chromosome (46,64,118). However, these conclusions were drawn from numerous classical genetic mapping studies, most of which used cotransduction frequencies to map specific markers. Because little is known about the m R N A transcripts of these gene clusters, the "superoperonic" clustering observed in Pseudomonas may simply mean that the separate operons are located reasonably close to each other on the chromosome, but are not transcriptionally linked. Therefore, the Pseudomonas "superoperons" may have little in common with the R. capsulatus crtEF/bchCA/puf "superoperon". There are, however, a few other examples of true "superoperons" in the literature. For example, the promoter of the E. coli ampC gene (encoding /J-lactamase) is found within the coding region of the last gene in the frdABCD operon (encoding the fumarate reductase complex), so that the carboxy-terminal 12 codons of the frdD gene are also contained on an m R N A transcript initiated from the ampC promoter (42). This 12 codon sequence also forms an attenuator that has been proposed to moderate growth-rate-dependent transcription of the ampC gene (49). Furthermore, the ampC attenuator also acts as a transcriptional terminator for the frd operon (42). A second example is the E. coli ptsH, ptsl, err operon that encodes three proteins of the phosphoenol pyruvate-dependent phosphotransferase system (HPr, enzyme I, and enzyme I I I G l c , respectively; ref. 29). Transcription of this operon may be initiated at either of two promoters; one (PI) that promotes transcription of the entire operon, and the other (P2) that is located in the ptsl coding region and that promotes transcription of only the err gene. Approximately 85% of err m R N A transcripts are due to initiation at P2, which is regulated differently than PI. Whereas PI is stimulated by growth on glucose (by the c A M P -C R P complex), P2 is unresponsive to this condition and promotes constitutive transcription of err. This region may alternatively be considered to contain two overlapping transcriptional units. One unit, encoding err, is constitutively expressed: However, its expression can be increased 106 under certain conditions by transcriptional readthrough from the upstream unit encoding ptsH and ptsl. Although phage lambda can also be considered to use transcriptional readthrough, this system is controlled through the N and Q antitermination systems (93). Expression of lambda intermediate and late genes is dependent on the synthesis of the N and Q antiterminator proteins because these genes do not have their own promoters. Therefore, antitermination differs from the overlapping m R N A s described in this work for two major reasons. Each of the R. capsulatus crtEF, bchCA, and puf operons have their own individually regulated promoters (1, 8, 40,117), and thus can be expressed without transcriptional readthrough. Also, because there do not seem to be transcriptional terminators between the crtEF, bchCA, and puf operons equivalent to the lambda arrangement, the production of these overlapping m R N A s is not dependent on the synthesis of antiterminator proteins. 2. Functional significance of overlapping crtEF, bchCA, and puf m R N A transcripts. The crtF interposon mutant R. capsulatus DE324 was used to test whether transcriptional readthrough from the crtEF operon was important for normal expression of the bchCA and puf operons. Absorption spectra of photosynthetically grown R. capsulatus DE324 showed that a small amount of the Bchl a biosynthetic intermediate 2-devinyl-2-hydroxyethyl bacteriochlorophyllide a (P668) could be readily detected in pigment extracts and in the culture supernatant fluids (Figure 26), indicating that loss of crfEF transcriptional readthrough in R. capsulatus DE324 resulted in a bottleneck in Bchl a biosynthesis. However, the amount of P668 accumulated in photosynthetically grown R. capsulatus DE324 was less than 10% of that accumulated in oxygen-limited cultures of R. capsulatus CW100. The photosynthetic growth conditions employed in these experiments used an incident light intensity of 55 W / m ^ , when relatively low levels of Bchl a must be made. Under these conditions, R. capsulatus DE324 grew at the same rate as d id the wild-type R. capsulatus SB1003 and the crtF point mutant R. capsulatus BP503 (data not shown). Because a decrease in incident light intensity requires a corresponding increase in the amount of Bchl a synthesized for growth, it would be interesting to determine if the bottleneck in Bchl a biosynthesis in R. capsulatus DE324 would be significant enough to slow its photosynthetic growth rate relative to either R. capsulatus SB1003 or R. capsulatus BP503 under growth-rate-limiting light 107 intensities. If so, then the transcriptional overlap between the crfEF and bchCA operons in wild-type cells may have evolved to provide high levels of Bchl biosynthesis under conditions of weak illumination. The bottleneck in Bchl a biosynthesis in R. capsulatus DE324 was presumably due to a reduction in the amount of m R N A molecules encoding the bchCA operon. Although the simplest explanation for this reduction was that the shortage was due to the elimination of m R N A molecules that simultaneously encoded the CrtE, CrtF, BchC, and BchA enzymes in R. capsulatus DE324, it was also possible that the bchCA promoter was less active in R. capsulatus DE324. Comparison of the 5' ends due to initiation of transcription from the bchCA promoter in R. capsulatus DE324 relative to R. capsulatus SB1003 (Figure 28), showed that elimination of the transcriptional overlap between the crtEF and bchCA operons d i d indeed appear to influence the activity of the bchCA promoter such that there was less accumulation of bchCA m R N A under high aeration, and a failure to maintain high levels of expression during the shift to inducing conditions. It should be stressed that this experiment measured steady-state amounts of m R N A molecules whose levels are a function of transcription initiation as well as m R N A turnover. If the assumption that decay of bchCA m R N A occurred at equal rates in the wild-type and mutant strains is made, the reduction in total bchCA m R N A transcripts in R. capsulatus DE324 was most likely due to the combination of a less active bchCA promoter as well as elimination of the readthrough transcripts initiated at the crtEF promoter that encoded the BchC and BchA enzymes. Similar 5' end site mapping experiments showed that the puf promoter appeared to be approximately 40% less active in the crtF interposon mutant, R. capsulatus DE324 (Figure 30) . However, it is not possible to conclude from these data whether this effect was direct, or whether the crtF interposon mutation decreased the activity of the bchCA promoter, which in turn decreased the activity of the puf promoter. Similarly, the puf promoter appeared to be approximately 30% less active in the bchC interposon mutant, R. capsulatus CW100 (Figure 31) . It is difficult to compare these different values, because transcripts initiated at the bchCA promoter overlap the puf operon in R. capsulatus DE324 but not in R. capsulatus CW100. Additionally, the interposon mutant strains used in these experiments originated from different genetic backgrounds. R. capsulatus CW100 was derived from the wild-type strain, R. capsulatus BIO; whereas R. capsulatus DE324 was derived from a different wild-type strain, R. capsulatus SB1003. 108 In each of the above R N A studies, 5' Sl-nuclease protection experiments were used to generate autoradiograms in which a band due to transcription initiation at a particular promoter could be observed. A decrease in band intensity was interpreted to mean that the particular promoter being studied was less active. However, it could also be argued that the observed decrease in band intensity was due instead to losing the transcriptional readthrough into the region delimited by the probe. For example, if the readthrough m R N A was cleaved by an endonuclease that cut at or very near the location of the transcription start site, a processed 5' end would be produced at the same place as the 5' end due to transcription initiation. In wild-type cells, both types of molecules (those with 5' monophosphate ends and those with 5' d i - or triphosphate ends) would be detected by the probe in these experiments, whereas in interposon mutant cells, only molecules with 5' d i - or triphosphate ends would be detected. Although this is possible, I believe it is unlikely because such processing would have to occur very near both the bchCA and puf transcription start sites to fit the data. If m y interpretation that the absence of readthrough transcription reduces the activity of the downstream promoter is correct, then it follows that the presence of readthrough transcription stimulates the activity of the downstream promoter. It is interesting to speculate how this stimulation might occur. One possible mechanism is that the promoter is responding to local changes in conformation that would be produced when R N A polymerase transcribes a nearby region. These conformational changes would presumably enhance promoter recognition or transcription initiation. The possibility that changes in the level of D N A supercoiling occur when R. capsulatus is shifted from high to low aeration was recently investigated by Cook et al. (25), who have developed an assay capable of detecting stable changes in superhelicity in any genomic fragment of interest. Their system uses a photoreactive psoralen derivative that intercalates into D N A and forms interstrand crosslinks when irradiated with U V light. The extent of crosslink formation has been shown to be a function of D N A superhelicity both in vivo and in vitro (48,104). The assay consists of in vivo crosslinking of genomic D N A to low levels followed by purification and alkali denaturatiori of the D N A . Because the crosslinked strands are melted but are held in position by the psoralen adduct, they renature rapidly when neutralized. Therefore, changes in superhelicity result in changes in the rate of renaturation of the modified D N A . Double-stranded D N A is separated from single-stranded D N A by non-109 denaturing gel electrophoresis, and the proportion of crosslinked to uncrosslinked D N A is measured by using any restriction fragment of interest as a probe in a Southern blot. Interestingly, their results have shown that there is no significant change in the superhelicity of any tested R. capsulatus D N A fragment encoding photosynthesis genes upon a shift to inducing conditions (25). However, their assay is designed to test for stable changes in chromosomal superhelicity within the length of the fragment used a probe. Therefore, it remains possible that the mechanism whereby the activity of the puf and bchCA promoters appeared to be stimulated by transcriptional readthrough was that they were responding to conformational changes that spanned distances too short or that occurred over time scales too small to be detected by the method of Cook et al. (25). Alternatively, because the proteins encoded by the crtEF, bchCA, and puf operons are each involved in the production of the photosynthetic apparatus, it is also possible that regulatory circuits may exist to coordinate the expression of these functionally related operons. For example, the observed influence of the interposon mutation on puf promoter activity in R. capsulatus CW100 may have been caused by the BchC", BchA" phenotype the cells, rather than the elimination of overlapping transcripts. If so, then the activity of the puf promoter should be lower in strains with mutations in bch genes that are not linked transcriptionally to the puf promoter. This was tested by measuring the j3-galactosidase specific activity of a plasmid-borne pufB::lac'Z fusion in the bchC interposon mutant strain as well as in two other strains each carrying a point mutation in a bch gene that maps far from the puf operon (Table IV). In each case, the average /3-galactosidase specific activity in each of the oxygen-limited bch mutant strains was approximately 25% lower than the activity observed in the appropriate bch+ parental strains. These results correlated well with the observed 30% decrease in puf transcription initiation by direct measurement of m R N A levels in R. capsulatus CW100 (Figure 31). Additionally, in each case the average )3-galactosidase specific activity in each of the highly aerated bch mutant strains was approximately 25% higher than the activity observed in the appropriate bch+ parental strains. Therefore, transcription from the puf promoter appeared to be less sensitive to oxygen regulation in strains that cannot synthesize Bchl a. These results suggest that the observed decrease in transcription from the puf promoter in the R N A experiments could be explained by the inability of R. capsulatus CW100 to synthesize Bchl a. 110 However, this model is less satisfactory for several reasons. One is that because the differences in j3-galactosidase activities are small and the standard deviations in these assays are large, it is difficult to tell whether these differences are meaningful. Additionally, it is possible that regulation of the multiple copies of the pufBvlac'Z fusion may not have reflected the regulation of the single-copy puf promoter on the chromosome. A s noted previously, the differences in protein content of the cell extracts further complicates interpretation of the results. Also, Bauer et al. (7) have measured the /?-galactosidase specific activity of a pufMv.lac'Z fusion in several Bch" strains relative to a B c h + control, and found differences even smaller than the ones reported here. Furthermore, it is difficult to explain the influence of the crtF interposon mutation on puf promoter activity by this model, unless the puf promoter is also sensitive to the presence of carotenoids. Therefore, acceptance of the idea that precise levels of transcription initiation of the crfEF, bchCA, and puf operons are coordinated by trans-active regulatory circuits will require much more supporting evidence. The obvious tests w o u l d include measuring the /3-galactosidase specific activity of a puf viae'Z fusion in several Crt" strains relative to a C r t + control, measuring the /J-galactosidase specific activity of a bchv.lac'Z fusion in several Crt" strains relative to a C r t + control, and measuring the galactosidase specific activity of a pucvlac'Z fusion in several Bch" strains relative to a B c h + control. A s well, these measurements should be done with single-copy gene fusions inserted into the chromosome. Regardless of whether the puf promoter is responding to conformational changes, to a block in Bchl a biosynthesis, or a combination of these mechanisms, transcriptional readthrough was found to be important to obtain normal levels of puf m R N A . The importance of readthrough transcription to the production of normal levels of pw/-encoded polypeptides was assessed by measuring the /?-galactosidase specific activity of a pufBvlac'Z fusion in four contexts: i) the wild-type puf promoter without bchCA transcriptional readthrough, ii) the mutant puf promoter without bchCA transcriptional readthrough, iii) the wild-type puf promoter with bchCA transcriptional readthrough, and iv) the mutant puf promoter with bchCA transcriptional readthrough (Figure 32). These results showed that transcriptional readthrough from the bchCA operon was equally capable as transription initiated at the puf promoter in driving expression of the pufBvldc 'Z fusion gene under low oxygen conditions, and that the bchCA promoter was more active than the puf promoter alone in driving expression of the fusion under high oxygen conditions. Although the simplest interpretation of these results is I l l to assume that the transcriptional readthrough originated from the bchCA promoter, it is possible that there are additional promoters between the bchCA and puf promoters. In any case, if these plasmid fusions are indicative of the normal chromosomal context, transcriptional readthrough from the bchCA operon would result in twice the amount of pu/-encoded polypeptides than would be expressed from the puf promoter alone. It was particularly satisfying to note that the /J-galactosidase specific activity observed when both the bchCA and puf promoters were active in the same construct ( p X C A 6 : : E H Q ) was approximately 24% greater than the activity calculated by adding the respective individual contributions from the bchCA and puf promoters (pXCA6: :EHQA44 plus pXCA6::935, see Figure 31). Thus, having both promoters on the same molecule resulted in slightly more transcription than d i d having both promoters on separate molecules. Therefore, the activity of puf promoter appeared to be stimulated by the presence of transcriptional readthrough from the bchCA operon, exactly as predicted by the R N A results (Figure 31). The differences in oxygen-dependent regulation of expression of the lac'Z fusions correlated very well with the observed differences in oxygen-dependent accumulation of m R N A molecules, as measured by quantitative Sl-nuclease protection experiments, for both the puf and bchCA promoters. However, the levels of m R N A accumulation in these experiments would reflect different promoter activities only if the decay rates of the puf and bchCA m R N A segments detected in these experiments were similar. The puf promoter appeared to show approximately 10-15-fold greater activity in cells grown with low aeration than in cells grown with high aeration. This difference was observed in studies of the /3-galactosidase specific activities of a pufBv.lac'Z fusion (Figure 32 and Table IV), as well as in studies following the accumulation of m R N A initiated from the puf promoter after shifting the cultures from high to low aeration (Figures 18B, 18D; 30 and 31). Similarly, the bchCA promoter appeared to show an approximately 5-fold greater activity under inducing conditions, both in studies with various bchCv.lac'Z fusions in p C W l , pJPl , and pJPlOO (Figure i4), as well as with a pufBr.lac'Z fusion that was expressed almost exclusively from the bchCA promoter (pXCA6: :EcoHQA44 in Figure 32). A 5-fold increase in bchCA m R N A accumulation was also observed in several of the R N A studies, including the SI experiments (Figure 28) and the R N A blots containing R. capsulatus BIO R N A (Figure 18A and C). It will be very interesting to extend this analysis to include the contribution of 112 crtEF transcription to puf operon expression. In order to construct the complete picture, however, it will first be necessary to create a defined bchCA promoter mutation analogous to the A44 puf promoter mutation. This would allow the three components of crtEF/bchCA/puf superoperon to be separated into modules, so that various combinations of active promoters can be mixed and matched. It would also be useful to correlate these results to measurements of the appropriate m R N A levels when testing various combinations of wild-type and mutant promoter sequences. 3. Analysis of the bchCA promoter. Although the precise nucleotides that make up the bchCA promoter have not yet been identified, I have localized the bchCA promoter to a 91 bp segment of D N A between the Smal and Bell sites in the crtF-bchC intergenic region. This was done by measuring the )3-galactosidase specific activities in extracts of cells containing one of three bchC::lac'Z fusion plasmids (Figure 14), as well as by directly mapping the 5' end sites of bchCA m R N A molecules in Sl-nuclease protection experiments using three different bipartite probes (Figures 15 and 19). These independent approaches gave consistent results. Plasmid pJPl contained sequences that gave rise to both the major and minor 5' ends, and gave wild-type /?-galactosidase specific activity. Because deletion of sequences to which the minor 5' ends mapped did not appreciably alter the resultant /J-galactosidase specific activity, these minor 5' ends most likely arose from endonuclease-mediated cleavage of m R N A molecules that were initiated from an upstream, possibly crtEF, promoter. However, because deletion of sequences to which the major 5' end mapped resulted in an approximately 98% decrease in the resultant /J-galactosidase specific activity, the major 5' end appeared to result from transcription initiation at the bchCA promoter. It is possible to distinguish between 5' m R N A ends that result from transcription initiation from those that arise from cleavage of preexisting molecules by using the enzyme guanylyltransferase to add a 5' "cap" onto m R N A molecules that have d i - or tri-phosphate termini (82, 83). Despite repeated attempts, the major 5' end associated with promoter activity could not be detected in capping experiments, although control experiments with puc m R N A gave positive results (data not shown). I attribute this failure to be due to the low sensitivity of the capping assay in conjunction with the relatively low levels of bchCA m R N A molecules. It 113 is noteworthy that the puf mRNA primary transcript could not be capped unless certain sequences, believed to be R N A s e cleavage substrates, were deleted (1). However, even though I was unable to cap the bchCA major 5' end, m y conclusion that this end represents the start site for bchCA transcription rests on the evidence that this end mapped to sequences that are required for promoter function as discussed above, and on the evidence of high resolution mapping experiments that positioned the major 5' end to sequences that are similar to other photosynthesis gene promoters. Alignment of the bchCA and other photosynthesis gene promoter regions showed that the proposed bchCA promoter sequence contained several features also found in the proposed promoter sequences of other photosynthesis genes (Figure 17). The most striking of these feature is the conserved " A C A " motif that has been found in all but one of the putative promoters sequenced to date (1,3,117,124,125,130). I believe that the one exception (marked by an asterisk in Figure 17) is unlikely to represent a true promoter sequence because it appears to be positioned within the crtlBK m R N A transcript (compare the results in refs. 3 and 40). This " A C A " motif has been proposed to be an R N A polymerase contact site in the puf promoter, based on the observation that changing this " A C A " sequence to " G C G " resulted in an approximately 95% reduction in puf promoter activity, as measured by lac 'Z fusion expression as well as the levels of the puf m R N A primary transcript (1). Although similar fine-structure analyses have not been performed with the puc promoter, two 5' triphosphate m R N A ends have been mapped near two repeats of the " A C A " motif in the puc promoter sequence (130). Similarly, the bchCA major 5' end is located 8 nt downstream of this " A C A " motif. Although this motif seems likely to be important for the function of the bchCA and other photosynthesis gene promoters, verification of this hypothesis will depend upon experimental evidence obtained from studies of the effects of site-directed mutations. A second feature that may be significant to bchCA promoter function is a palindromic sequence that overlaps the " A C A " motif. The high degree of conservation of this palindromic sequence, " T G T A A - N g - T T A C A " , among many photosynthesis gene promoters, and the similarity of this sequence to the consensus binding sequences for several prokaryotic regulatory transcription factors led Armstrong et ah (3) to suggest that the R. capsulatus palindromic sequence may bind a protein involved in oxygen regulation. This palindromic sequence is found in the proposed promoter regions of the bchCA, crtA, crtlBK, crtD, crtEF, crtC, and puc operons (Figure 17 and ref. 3). Support for the idea that this palindromic 114 sequence does bind such a regulatory factor will require experimental evidence from gel mobility shift and D N A footprinting assays performed on both wild-type and specific mutant templates. It is noteworthy that the oxygen-regulated puf and puh promoter sequences contain only the the " T T A C A " promoter-proximal half of the R. capsulatus consensus palindromic sequence (1,124,125), and therefore may bind only one equivalent of the proposed transcription factor. However, both the puf and puh promoters contain dissimilar inverted repeat regions located upstream of the " T T A C A " sequence. So far, only the two puf promoter inverted repeat regions have been studied, and the results have shown that only the promoter-proximal inverted repeat region has regulatory significance. Deletion of the entire promoter-proximal inverted repeat sequence ( 5 ' - C C C G G C G C G G C G A T C C G C C G C G C G A C G G G - 3 ' ; base positions 179-207 in Figure 4 of ref. 1) not only resulted in a dramatic decrease in overall puf promoter activity, but the residual transcription was much less sensitive to oxygen regulation (1). It is very interesting that Taremi and Marrs (109) have recently isolated a protein (called PPBP) that is proposed to bind specifically to this inverted repeat region, and have observed that the binding characteristics of this protein change depending on its state of phosphorylation. They have proposed a model whereby PPBP may act as the response regulator in an oxygen-sensing "two-component" regulatory system similar to that proposed for other prokaryotic systems that respond to environmental changes (94, 108, 109). Whether the PPBP binds to the " T G T A A - N g - T T A C A " palindromic sequence identified by Armstrong et al. (3) remains to be established. If there are two different proteins that bind these two sequences, then it would seem that there is more than one regulatory system controlling expression of these photosynthesis genes. These systems would not necessarily be mutually exclusive, since it is possible that the puf promoter may be capable of binding one equivalent of Armstrong's proposed transcription factor at the " T T A C A " half of the palindromic sequence in addition to binding the PPBP protein to the regulatory inverted repeat region. Comparison of the bchCA and puf promoter sequences showed the presence of a conserved " C G G G C " box located 19 nt upstream of the bchCA transcriptional start site, and 28 nt upstream of the puf transcriptional start site. This result was interesting because the puf promoter " C G G G C " box is almost completely contained within the regulatory inverted repeat. However, the bchCA promoter region appeared to have no analogous inverted repeat. Furthermore, a comparison of the bchCA promoter to several other possible photosynthesis gene 115 promoters showed little conservation of the " C G G G C " box, with it being found only in the puf, bchCA, and crfC promoter regions. Because the " C G G G C " boxes of the bchCA and crtC promoter regions are not part of an inverted repeat region similar to that found upstream of the puf promoter, the significance of the bchCA and crtC " C G G G C " boxes is questionable. O f potential significance is the presence of an E. coli a^O-like R N A polymerase recognition sequence in the bchCA promoter region (76,124). Such a ^ - l i k e recognition sequences were found in the proposed promoter sequences of three of the putative crt promoters, and were found to overlap the " T G T A A - N 3 - T T A C A " palindromic sequences (3). This suggests that binding of the putative regulatory transcription factor to the palindromic sequences may be required for proper promoter recognition or transcription initiation by an R N A polymerase holoenzyme that binds to a cj^U-like sequence on the other face of the helix. It will be interesting to test such R. capsulatus photosynthesis gene promoters for activity in E. coli, which contains o7® as its major R N A polymerase a subunit (76) However, failure to obtain promoter activity in E. coli may be due to the lack of an analogous activator protein. It must be stressed that none of these E. coli a^ u - l ike recognition sites have been shown to be required for the function of the putative promoter sequences in which they have been found. Therefore, additional experimentation with these specific mutant derivatives of these proposed promoter sequences will be required to support this model. 4. Comparison of Bch protein predicted sequences. Currently, six bch genes from R. capsulatus have been sequenced. The BamHl-F fragment of pRPS404 was sequenced by Youvan et al. (123), who found several ORFs that map upstream of the puhA gene. T w o of these ORFs were recently identified as bch genes by Yang and Bauer (121), who used interposon mutagenesis to show that O R F F108 encodes a subunit of the BchB enzyme, and that O R F F1025 encodes a subunit of the BchE enzyme. Similarly, the contiguous EcoRI-H and EcoRI-Q fragments of pRPS404 were entirely sequenced by Marie Alberti (pers. comm.). Not only did this analysis independently confirm m y bchC sequence, but also showed that three ORFs could be found within the bchA region; bchA.l, bchA.l, and bchA.3. Although the identities of these ORFs have not yet been confirmed by systematic mutagenesis experiments, it is possible that they encode three subunits of the BchA enzyme. Other Bch enzymes may also be composed of multiple subunits, as implied by the 116 complementation analysis of Yang and Bauer (121). The availability of additional bch gene sequences made possible a comparative analysis between the proposed structure of the BchC enzyme and the postulated structures of these other Bch enzymes. While it should be stressed that each of these proposed coding regions have yet to be confirmed by obtaining the amino acid sequences of the Bch enzymes themselves, certain observations about the structures of these enzymes can be made and interpreted with appropriate caution. Each of the R. capsulatus bch genes sequenced to date encode mid-sized polypeptides that range in length from 290 amino acids (BchB) to 497 amino acids (BchA.2), and have an average size of 360 amino acids. None of these putative Bch polypeptides sequenced so far have particularly large hydrophobic regions that would compellingly indicate that they are integral membrane proteins, although each appears to have several hydrophobic segments that may be membrane associated, as was shown by the hydropathy plot of the putative BchC amino acid sequence (Figure 13A). This is interesting because the only Bch enzyme to be purified so far from a purple prokaryote, the R. sphaeroides S-adenosyl-L-methionine:magnesium protoporphyrin methyltransferase, was found to be very tightly associated with the membrane (45), suggesting that the protein itself may have been largely hydrophobic. Alternatively, the S-adenosyl-L-methionine:magnesium protoporphyrin methyltransferase may have been anchored to the membrane by many possible mechanisms, including the posttranslational addition of hydrophobic moieties or by association with an integral membrane protein. It will be interesting to purify other Bch enzymes, especially those for which sequence information is available, to determine if there is a correlation between the water solubility of enzymatic activity and the predicted solubility of the protein based upon sequence data. Although there were no extensive sequence similarities among the various Bch enzymes, one very interesting feature that was conserved in each of the six Bch sequences so far examined was at least one copy of the putative Bchl a binding sequence A l a / G l y - X - X - X - F f i s (Table V , refs. 19, 119). The presence of this motif in the Bch polypeptides raises the possibility that the Bch enzymes themselves are capable of binding Bchl a, or that these motifs may instead be binding sites for the substrates of the Bch enzymes. Because Bchl a biosynthesis is likely to be regulated primarily at the level of Bch enzymatic activity (see INTRODUCTION), it is particularly attractive to imagine that Bchl a may bind to this motif and act to control Bch enzymatic activity by a classical end-product inhibition mechanism. A n alternative view, 117 T a b l e V . A l i g n m e n t o f A l a / G l y - X - X - X - H i s s e q u e n c e s i n B c h p r e d i c t e d p r o t e i n s e q u e n c e s ( i n s i n g l e l e t t e r c o d e ) . B c h p o l y p e p t i d e amino a c i d p o s i t i o n a n d s e q u e n c e 131-A L A A T A R H A L A-141 2 7 6 - S L D G L I T H R R P - 2 8 6 A . l 5 0 - F T L A N L S H M M A - 60 6 0 - P G F G V P T H A E A - 7 0 A . 2 b A . 3 b 2 7 2 - V G S A P V G H D G T - 2 8 2 46-Y T D G L P P H E L P - 56 169-F N M A S D L H E I R-179 167-G F A A P L Q H A D R-177 158-T E L G K R T H S A 1-16* a b c t h i s work , F i g u r e 11 M . A l b e r t i , p e r s o n a l c o m m u n i c a t i o n Yang a n d B a u e r (121) 118 where this A l a / G l y - X - X - X - H i s motif may instead be the binding site for the Bchl a biosynthetic intermediates, seems less likely because one must also then consider w h y Bchl biosynthetic intermediates do not seem to be bound by the L H polypeptides, which also contain this motif. It is possible that other structural differences between L H and Bch polypeptides account for this difference. In any event, the conservation of the A l a / G l y - X - X - X - H i s sequence in all of the Bch enzymes suggests that this motif may indeed have some purpose. It will therefore be of considerable interest to direct mutations to this sequence as part of a study of the mechanism and regulation of Bch enzymatic function. 5. Concluding remarks. The purple non-sulfur bacteria are good model systems for studies of such things as photosynthetic membrane differentiation, electron transfer reactions, and Bchl biosynthesis. M a n y of the results presented in this thesis improve our understanding of the regulation of photosynthesis gene expression in R. capsulatus, and others provide a framework for additional experimentation. The sequence analysis of the bchC gene presented in this work was the first structural investigation of a chlorophyll biosynthesis gene to be reported, and the sequences of several other bch genes are now available (121; M . Alberti, pers. comm.). The current availability of primary sequence data and cloned D N A fragments bearing bch genes is expected to aid in the design of purification strategies for the Bch proteins, and possibly in the development of specific enzymatic assays. These developments, in turn, may be used to overcome some of the great difficulties encountered in the study of the late steps of chlorophyll biosynthesis, such as the cellular location of the Bch enzymes, regulation of Bch enzymatic activity, and the control of carbon flow through the M g + + branch of the tetrapyrrole biosynthetic pathway relative to the F e + + ahd C o + + branches. Investigations of these biological problems fundamentally depend upon the ability to detect and assay specific Bch enzymes. The recent results of photosynthesis gene promoter studies have led to the proposal of several experimentally testable models of the regulation of photosynthesis gene transcription. The bchCA promoter region, localized to the 91 nt segment between the Smal and Bell sites in the crtEF/bchCA intergenic region, contains the " A C A " motif that is 119 absolutely conserved in all photosynthesis gene promoters sequenced so far; the " T G T A A - N g -T T A C A " palindromic sequence that has been proposed to be a binding site for a transcription factor involved in oxygen regulation; the E. coli a ^ - l i k e recognition sequence that may be important for initiation of transcription; and the " C G G G C " box. Because the bchCA promoter region contains all of these features and has been partially characterized, it will be a particularly good candidate for a detailed analysis of R. capsulatus photosynthesis gene promoter fine structure. I have no doubt that such an investigation will yield useful information. The results of the R N A studies shown in this thesis complement and extend the genetic results of Young et al. (124) to demonstrate the transcriptional overlap between the crtEF, bchCA, and puf operons, and elucidate quantitative aspects of the regulatory significance of this genetic arrangement. I believe it is likely that several other examples of overlapping m R N A s similar to these operons will be found, as more detailed maps of transcriptional units are assembled. Because overlapping m R N A s can be detected by simply using bipartite probes in routine Sl-nuclease protection experiments, demonstration of transcriptional overlap between adjacent operons need not be complicated. It is interesting to contemplate the evolutionary significance of transcriptionally linked operons. It has been proposed that such arrangements may facilitate the lateral transfer of groups of functionally related genes (118). Alternatively, the coordinate regulation of groups of functionally related genes may be enhanced by linking them together transcriptionally. It is possible that during evolution of prokaryotic genomes different operons may initially be brought together by chance. If this arrangement d id not disrupt the normal functioning of the separate operons, transcriptional linkage would be tolerated. Some such operons then may have evolved a dependency upon the transcriptional linkage, especially if the genes encode products that functioned together or provided cells with a selective advantage under certain conditions. Regardless of the evolutionary forces that resulted in transcriptional linkage of the crtEF, bchCA, and puf operons, the data presented in this thesis clearly demonstrate that this arrangement is necessary for the normal levels of expression of the genes encoded by these transcriptional units. 120 REFERENCES 1. Adams, C. W v M . E. Forrest, S. N . Cohen, and J. T. Beatty. 1989. Structural and functional analysis of transcriptional control of the R. capsulatus puf operon. J. Bacteriol. 171: 473-482. 2. Allen, J. P., G . Feher, T . O. Yeates, H . Komiya, D. C. Rees, J. Deisenhofer, H . Michel, and R. Huber. 1986. Structural homology of reaction centers from Rhodopseudomonas sphaeroides and Rhodobacter viridis as determined by X-ray diffraction. Proc. Natl . Acad . Sci. U S A 83: 8589-8593. 3. Armstrong, G . A., M . Alberti, F. Leach, and J. E. Hearst. 1989. Nucleotide sequence, organization, and nature of the protein products of the carotenoid biosynthesis gene cluster of Rhodobacter capsulatus. M o l . Gen. Genet. 216: 254-268. 4. Baccarini-Melandri, A., R. Casadio, and B. A . Melandri. 1981. Electron transfer, proton translocation, and A T P synthesis in bacterial chromatophores. Curr . Top. Bioenergetics 12:197-257. 5. Barr, P. J., R. M . Thayer, P. Laybourn, R. C. Najarian, F. Seela, and D. R. Tolan. 1986. Deaza-2'-deoxyguanosine-5'-triphosphate: enhanced resolution in M13 dideoxy sequencing. BioTechniques 4:428-432. 6. Bartley, G. E., and P. A . Scolnik. 1989. Carotenoid biosynthesis in photosynthetic bacteria. J. Biol. Chem. 264: 13109-13113. 7. Bauer, C. E . , M . Eleuterio, D. A. Young, and B. L. Marrs. 1987. Analysis of transcription through the Rhodobacter capsulatus puf operon using a translational fusion of pufM to the E. coli lacZ gene, p. 669-705. In J. Biggens (ed.), Progress in Photosynthesis Research, vol . IV. Martinus Nijhoff Publishers, Dordrecht-Netherlands. 8. Bauer, C. E. , D. A . Young, and B. L. Marrs. 1988. Analysis of the Rhodobacter capsulatus puf operon: location of the oxygen-regulated promoter region and the identification of an additional pu/-encoded gene. J. Biol. Chem. 263: 4820-4827. 9. Bauer, C. E . , and B. L. Marrs. 1988. Rhodobacter capsulatus puf operon encodes a regulatory protein (PufQ) for bacteriochlorophyll biosynthesis. Proc. Natl . Acad . Sci. U S A 85: 7074-7078. 10. Beatty, J. T., and H . Gest. 1981. Generation of succinyl-coenzyme A in photosynthetic bacteria. Arch . Microbiol. 129: 335-340. 11. Belasco, J. G. , J. T. Beatty, C. W. Adams, A. von Gabain, and S. N . Cohen. 1985. 121 Differential expression of photosynthesis genes in R. capsulatus results from segmental differences in stability within the polycistronic rxcA transcript. Cel l 40: 171-181. 12. B i b b , M . J., and S. N . Cohen. 1982. Gene expression in Streptomyces: construction and application of promoter-probe plasmid vectors in Streptomyces lividans. M o l . Gen. Genet. 187: 265-277. 13. B ie l , A . J. 1986. Control of bacteriochlorophyll accumulation by light in Rhodobacter capsulatus. J. Bacteriol. 168: 655-659. 14. Bie l , A . J., and B . L . Marrs . 1983. Transcriptional regulation of several genes for bacteriochlorophyll biosynthesis in Rhodopseudomonas capsulata in response to oxygen. J. Bacteriol. 156: 686-694. 15. Bie l , A . J., and B. L . Marrs. 1985. Oxygen does not directly regulate carotenoid biosynthesis in Rhodopseudomonas capsulata. J. Bacteriol. 162: 1320-1321. 16. Brent, R., and M . Ptashne. 1980. The lexA gene product represses its own promoter. Proc. Natl . A c a d . Sci. U S A 77:1932-1936. 17. Breton, J., and A . Vermeglio. 1982. Orientation of photosynthetic pigments in vivo, p. 153-194. In Govindjee (ed.), Photosynthesis, vol . 1. Academic Press, Inc., N e w York. 18. Brunisholz, R. A . , and H . Zuber. 1988. Primary structure analyses of bacterial antenna polypeptides: - correlation of aromatic amino acids with spectral properties. - structural similarity with reaction center polypeptides, p. 103-113. In H . Scheer and S. Schneider (ed.), Photosynthetic Light-Harvesting Systems: Organization and Function. Walter de Gruyter & Co. , Berlin-New York. 19. Byl ina , E. J., S. J. Robles, and D . C . Youvan. 1988. Directed mutations affecting the putative bacteriochlorophyll-binding sites in the light-harvesting I antenna of Rhodobacter capsulatus. Israel J. Chem. 28: 73-83. 20. C h e n , C-Y. A . , J. T . Beatty, S. N . Cohen, and J. G . Belasco. 1988. A n intercistronic R N A stem-loop structure functions as a decay terminator necessary but insufficient for puf m R N A stability. Cell 52: 609-619. 21. Clark, W. G . , E . Davidson, and B . L . Marrs. 1984. Variation in levels of m R N A coding for antenna and reaction center polypeptides in Rhodopseudomonas capsulata in response to changes in oxygen concentration. J: Bacteriol. 157: 945-948. 22. Cogdel l , R. J., a n d j . P. Thorriber. 1980. Light-harvesting pigment-protein complexes of purple photosynthetic bacteria. FEBS Lett. 122: 1-8. 23. Cogdell , R. J. 1978. Carotenoids in photosynthesis. Phil. Trans. R. Soc. Lond. B. 284: 569-122 579. 24. Cohen-Bazire, G . , W. R. Sistrom, and B. Y. Stanier. 1957. Kinetic studies of pigment synthesis by non-sulfur purple bacteria. J. Cell . C o m p . Physiol. 49: 25-68. 25. Cook, D . N . , G . A . Armstrong, and J. E . Hearst 1989. Induction of anaerobic gene expression in Rhodobacter capsulatus is not accompanied by a local change in chromosomal supercoiling as measured by a novel assay. J. Bacteriol. 171: 4836-4843. 26. Coomber, S. A . , M . Chaudhri , A . Connor, G . Britton, and C . N . Hunter. 1990. Localized transposon Tn5 mutagenesis of the photosynthetic gene cluster of Rhodobacter sphaeroides. M o l . Microbiol. 4: 977-989. 27. Daldal , F., S. Cheng, J. Applebaum, E . Davidson, and R. Prince. 1986. Cytochrome C2 is not essential for photosynthetic growth of Rhodopseudomonas capsulata. Proc. Natl . Acad. Sci. U S A 83: 2012-2016. 28. Davidson, E . , and F. Daldal . 1987. Primary structure of the bc\ complex of Rhodopseudomonas capsulata; Nucleotide sequence of the pet operon encoding the Rieske, cytochrome b\, and cytochrome b/c\ apoproteins. J. M o l . Biol. 195: 13-24. 29. D e Reuse, H . , and A . Danchin . 1988. The ptsl, ptsl, and err genes of Escherichia coli phosphoenolpyruvate-dependent phosphotransferase system: a complex operon with several modes of transcription. J. Bacteriol. 170: 3827-3837. 30. Deisenhof er, J., O . E p p , K . M i k i , R. Huber , and H . M i c h e l . 1985. Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3A resolution. Nature (London) 318: 618-624. 31. Diers te in ,R. 1983. Biosynthesis of pigment-protein complex polypeptides in bacteriochlorophyll-less mutant cells of Rhodopseudomonas capsulata YS. FEBS Lett. 160: 281-286. 32. Drews, G . 1985. Structure and functional organization of light-harvesting complexes and photochemical reaction centers in membranes of phototrophic bacteria. Microbiol. Rev. 49: 59-70. 33. Drews, G . 1986. Adaptation of the bacterial photosynthetic apparatus to different light intensities. Trends Biochem. Sci. 11: 255^257. 34. Drews, G . , and J. Oelze . 1981. Organization and differentiation of membranes of phototrophic bacteria. A d v . Microb. PhysioL 22: 1-92. 35. Feick, R., and G . Drews. 1978. Isolation arid characterization of light harvesting 123 bacteriochlorophyll-protein complexes from Rhodopseudomonas capsulata. Biochim. Biophys. Acta. 501: 499-513. 36. Feinberg, A. P., and B. Vogelstein. 1983. A technique for radiolabeling D N A restriction fragments to high specific activity. A n a l . Biochem. 132: 6-13. 37. Feher, G., and M. Y. Okamura. 1978. Chemical composition and properties of reaction centers, p. 349-396. In R. K . Clayton and W . R. Sistrom (ed.), The Photosynthetic Bacteria. Plenum Press, N e w York. 38. Forrest, M. E., A. P. Zucconi, and J. T. Beatty. 1989. The pufQ gene product of Rhodobacter capsulatus is essential for formation of B800-850 light-harvesting complexes. Curr . Microbiol. 19:123-127. 39. Gamier, J., D. J. Osguthorphe, and B. Robson. 1978. Analysis of the accuracy and implications of simple methods for predicting the secondary structures of globular proteins. J. M o l . Biol. 120: 97-120. 39a. Gibson, J., E. Stackebrandt, L. B. Zablen, R. Gupta, and C. R. Woese. 1979. A phylogenetic analysis of the purple photosynthetic bacteria. Curr . Microbiol. 3: 59-64. 40. Giuliano, G., D. Pollack, H. Stapp, and P. A. Scolnik. 1988. A genetic-physical map of the Rhodobacter capsulatus carotenoid biosynthesis gene cluster. M o l . Gen. Genet. 213: 78-83. 41. Gribskov, M., J. Devereux, and R. R. Burgess. 1984. The codon preference plot: graphic analysis of protein coding sequences and prediction of gene expression. Nucleic Acids Res. 12: 539-547. 42. Grundstrom, T., and B. Jaurin. 1982. Overlap between ampC and frd operons on the Escherichia coli chromosome. Proc. Natl . Acad. Sci. U S A 79: 1111-1115. 43. Hattori, M., and Y. Sakaki. 1986. Dideoxy sequencing using denatured plasmid templates. A n a l . Biochem. 152: 232^238. 44. Hinchigeri, S. B., J. C-S. Chan, and W. R. Richards. 1981. Purification of S-adenosyl-L-methionine:magnesium protoporphyrin methyltransferase by affinity chromatography. Photosynthetica 15: 351-359. 45. Hinchigeri, S. B., D. W. Nelson, and W. R. Richards. 1984. The purification and reaction mechanism of S-adenosyl-L-methiohihe:magnesium protoporphyrin methyltransferase from Rhodopseudomonas sphaeroides. Photdsynthetica 18: 168-178. 46. Holloway, B. W., and A . F. Morgan. 1986. Genome organization in Pseudomonas. A n n u . Rev. Microbiol. 40: 79-105. 124 47. H i i d i g , H . , N. Kaufmann, and G. Drews. 1986. Respiratory deficient mutants of Rhodopseudomonas capsulata. A r c h . Microbiol. 145: 378-385. 48. H y d e , J. E. , and J. E . Hearst. 1978. Binding of psoralen derivatives to D N A and chromatin: influence of the ionic environment on dark binding and photoreactivity. Biochemistry 17: 1251-1257. 49. Jaurin, B., T . Grundstrom, T . E d l u n d , and S. Normark. 1981. The E. coli /3-lactamase attenuator mediates growth-rate-dependent regulation. Nature (London) 290: 221-225. 50. Johnson, J. A., W . K . R. Wong, and J. T . Beatty. 1986. Expression of cellulase genes in Rhodobacter capsulatus by use of plasmid expression vectors. J. Bacteriol. 167: 604-610. 51. Jones, O. T . G. 1978. Biosynthesis of porphyrins, hemes, and chlorophylls, p. 751-777. In R. K . Clayton and W . R. Sistrom (ed.), The Photosynthetic Bacteria. Plenum Press, N e w York. 52. Jones, O. T . G. 1979. Chlorophyll biosynthesis, p. 179-233. In D . Dolphin (ed.). The Porphyrins, vol. VI . Academic Press, Inc., N e w York. 53. K i l e y , P. J., and S. Kaplan . 1988. Molecular genetics of photosynthetic membrane biosynthesis in Rhodobacter sphaeroides. Microbiol. Rev. 52: 50-69. 54. Kirmaier , C , and D . Holten. 1987. Primary photochemistry of reaction centers from the photosynthetic purple bacteria. Photosynthesis Research 13: 225-260. 55. K l u g , G v N. Kaufmann, and G. Drews. 1984. The expression of genes encoding proteins of B800-850 antenna pigment complex and ribosomal R N A of Rhodopseudomonas capsulata. FEBS Lett. 177: 61-65. 56. K l u g , G., N. Kaufmann, and G. Drews. 1985. Gene expression of pigment-binding proteins of the bacterial photosynthetic apparatus: Transcription and assembly in the membrane of Rhodopseudomonas capsulata. Proc. Natl . Acad. Sci. U S A 82: 6485-6489. 57. K l u g . , G., R. Liebetanz, and G. Drews. 1986. The influence of bacteriochlorophyll biosynthesis on formation of pigment-binding proteins and assembly of pigment protein complexes in Rhodopseudomonas capsulata. Arch . Microbiol. 146: 284-291. 58. K r i n s k y , N. I. 1968. The protective function of carotenoid pigments, p. 123-195. In A . C . Giese (ed.), Photophysiology, vol. 3. Academic Press, N e w York-London. 59. K w a n , L . Y - M . , D . L . Darl ing , and W . R. Richards. 1986. Affinity chromatographic purification of two enzymes of the latter stages of chlorophyll synthesis, p. 57-62. In G . Akoyunoglou and H . Senger (ed.), Regulation of Chloroplast Differentiation, "Plant 125 Biology" Series, vol. 2. A l a n R. Liss, Inc., N e w York. 60. Kyte, J., and R. F. Doolittle. A simple method for displaying the hydropathic character of a protein. J. M o l . Biol. 157:105-132. 61. Lascelles, J. 1966. The accumulation of bacteriochlorophyll precursors by mutant and wild-type strains of Rhodopseudomonas sphaeroides. Biochem. J. 100: 175-183. 62. Lascelles, J. 1978. Regulation of pyrrole synthesis, p. 795-808. In R. K. Clayton and W . R. Sistrom (ed.), The Photosynthetic Bacteria. Plenum Press, N e w York. 63. Lee, J. K., P. J. Kiley, and S. Kaplan. 1989. Posttranscriptional control of puc operon expression of B800-850 light-harvesting complex formation in Rhodobacter sphaeroides. J. Bacteriol. 171: 3391-3405. 64. Leidigh, B . J., and M . L. Wheelis. 1973. The clustering on the Pseudomonas putida chromosome of genes specifying dissimilatory functions. J. M o l . Evol . 2: 235-242. 65. L i l b u r n , Tim. 1990. The role of the pufX gene product of Rhodobacter capsulatus. M.Sc. Thesis. University of British Columbia, March 1990. 66. Lowry, O. H . , N . J. Rosebrough, A . L . Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265-275. 67. Madigan, M . T., J. C. Cox, and H . Gest. 1980. Physiology of dark fermentative growth of Rhodopseudomonas capsulata. J. Bacteriol. 142: 908-915. 68. Madigan, M . T., and H . Gest. 1978. Growth of a photosynthetic bacterium anaerobically in darkness, supported by "oxidant-dependent" sugar fermentation. Arch. Microbiol. 117: 119-122. 69. Madigan, M . T., and H . Gest 1979. Growth of the photosynthetic bacterium Rhodopseudomonas capsulata chemoautotrophically in darkness with H2 as the energy source. J. Bacteriol. 137: 524-530. 70. Maniatis, T., E . F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. C o l d Spring Harbor Laboratory, C o l d Spring Harbor, N e w York. 71. Marck, C. 1988. ' D N A Strider': a ' C program for the fast analysis of D N A and protein sequences on the A p p l e Macintosh family of computers. Nucleic Acids Res. 16:1829-1836. 72. Marrs, B. 1981. Mobilization of the genes for photosynthesis from Rhodopseudomonas capsulata by a promiscuous plasmid. j . Bacteriol. 146: 1003-1012. 73. Marrs, B. L . , and H . Gest. 1973. Genetic mutations affecting the respiratory electron-126 transport system of the photosynthetic bacterium Rhodopseudomonas capsulata. J. Bacteriol. 114: 1045-1051. 74. Marrs , B. L . , C . L . Stahl, S. L i e n , and H . Gest. 1972. Biochemical physiology of a respiration deficient mutant of the photosynthetic bacterium Rhodopseudomonas capsulata. Proc. Natl . Acad. Sci. U S A 69: 916-920. 75. Masepohl , B., W. K l i p p , and A. Punier. 1988. Genetic characterization and sequence analysis of the duplicated nifA/nifB region of R. capsulatus. M o l . Gen. Genet. 212: 27-37. 76. M c C l u r e , W. R. 1985. Mechanism and control of transcription initiation in prokaryotes. A n n u . Rev. Biochem. 54:171-204. 77. Meinhardt , S. W., P. J. K i l e y , S. Kaplan , A. R. Crofts, and S. Harayama. 1985. Characterization of light-harvesting mutants of Rhodopseudomonas sphaeroides. 1. Measurement of the efficiency of energy transfer from light-harvesting complexes to the reaction center. A r c h . Biochem. Biophys. 236: 130-139. 78. M e l a n d r i , B. A., Baccarini-Melandri, A., A. San Pietro, and H . Gest. 1971. Interchangeability of phosphorylation coupling factors in photosynthetic and respiratory energy conversion. Science 174:514-516. 79. M i c h e l , H . , O. Epp , and J. Deisenhofer. 1986. Pigment-protein interactions in the photosynthetic reaction center from Rhodopseudomonas viridis. E M B O J. 5: 2445-2451. 80. M i l l e r , J. H . 1972. Experiments in molecular genetics. C o l d Spring Harbor Laboratory, C o l d Spring Harbor, N e w York. 81. M i z u s a w a , S., Nishimura , S., and F. Seela. 1986. Improvement of the dideoxy chain termination method of D N A sequencing by use of deoxy-7-deazaguanosine triphosphate in place of d G T P . Nucleic Acids Res. 14:1319-1324. 82. M o s s , B. 1977. Utilization of the guanylyltransferase and methyltransferase of vaccinia virus to modify and identify the 5'-terminals of heterologous R N A species. Biochem. Biophys. Res. C o m m u n . 74: 374-383. 83. M o s s , B. 1981. 5'-End labelling of R N A with capping and methylating enzymes, p. 253-266. In J. G . Chirikjian and T. S. Papas (ed.), Gene amplification and analysis, vol. 2. Elsevier /North-Holland Publishing Co. , Amsterdam. 84. Nei th , K . F., G . Drews, and R. Feick. 1975. Photochemical reaction centers from Rhodopseudomonas capsulata. Arch . Microbiol. 105: 43-45. 127 85. Oelze, J., and K. Arnheim. 1983. Control of bacteriochlorophyll formation by oxygen and light i n Rhodopseudomonas sphaeroides. F E M S Microbiol . Lett. 19: 197-199. 86. Okamura, M . Y., G . Feher, and N . Nelson. 1982. Reaction centers, p. 195-274. In Govindjee (ed.), Photosynthesis, vol. 1. Academic Press, Inc., N e w York. 87. Parson, W. W., and B. Ke. 1982. Primary photochemical reactions, p. 331-385. In Govindjee (ed.), Photosynthesis, vol. 1. Academic Press, Inc., N e w York. 88. Peters, J., and G . Drews. 1983. The B870 pigment-protein complex of Rhodopseudomonas capsulata contains two different pigment-binding polypeptides. F E M S Microbiol. Lett. 17: 235-237. 89. Pudek, M . R., and W. R. Richards. 1975. A possible alternate pathway of bacteriochlorophyll biosynthesis in a mutant of Rhodopseudomonas sphaeroides. Biochemistry 14: 3132-3137. 90. Richards, W. R., M . Fung, A. N . Wessler, and S B. Hinchigeri. 1987. The purification and properties of three latter-stage enzymes of chlorophyll synthesis, p. 475-482. In J. Biggens (ed.), Progress in Photosynthesis Research, vol. IV. Martinus Nijhoff Publishers, Dordrecht-Netherlands. 91. Richards, W. R., and J. Lascelles. 1969. The biosynthesis of bacteriochlorophyll. The characterization of latter stage intermediates from mutants of Rhodopseudomonas sphaeroides. Biochemistry 8: 3473-3482. 92. Richards, W. R., R. B. Wallace, M . S. Tsao, and E . Ho. 1975. The nature of a pigment-protein complex excreted from mutants of Rhodopseudomonas sphaeroides. Biochemistry 14: 5554-5561. 93. Roberts, J .W. 1988. Phage lambda and the regulation of transcription termination. Cell 52: 5-6. 94. Ronson, C . W., B. T. Nixon, and F. M . Ausubel. 1987. Conserved domains in bacterial regulatory proteins that respond to environmental stimuli. Cell 49: 579-581. 95. Schmidhauser, T. J., and D. R. Helinski. 1985. Regions of broad-host range plasmid RK2 involved in replication and stable maintenance in nine species of gram-negative bacteria. J. Bacteriol. 164: 446^455. 96. Schumacher, A., and G . Drews. 1978. The formation of bacteriochlorophyll-protein complexes of the photosynthetic apparatus of Rhodopseudomonas capsulata during the early stages of development. Biochihi. Biophys. Acta. 501: 183-194. 97. Schumacher, A. , and G . Drews. 1979. Effects of light intensity on membrane 128 differentiation in Rhodopseudomonas capsulata. Biochim. Biophys. Acta. 547: 417-428. 98. Schumann, J. P, G . M . Waitches, and P. A . Scolnik. 1986. A D N A fragment hybridizing to a nif probe in Rhodobacter capsulatus is homologous to a 16 S r R N A gene. Gene 48: 81-92. 99. Scolnik, P. A . , and B. L . Marrs. 1987. Genetic research with photosynthetic bacteria. A n n u . Rev. Microbiol. 41: 703-726. 100. Scolnik, P. A . , D . Zannoni , and B. L . Marrs. 1980. Spectral and functional comparisons between the carotenoids of the two antenna complexes of Rhodopseudomonas capsulata. Biochim. Biophys. Acta. 593: 230-240. 101. Shiozawa, J. A . , P. A . Cuendet, G . Drews, and H . Zuber. 1980. Isolation and characterization of the polypeptide components from light harvesting pigment-protein complex B800-850 of Rhodopseudomonas capsulata. Eur. J. Biochem. I l l : 455-460. 102. Shiozawa, J. A . , W . Welte, N . H o d a p p , and G . Drews. 1982. Studies on the size and composition of the isolated light-harvesting B800-850 pigment-protein complex of Rhodopseudomonas capsulata. A r c h . Biochem. Biophys. 213: 473-485. 103. S imon, R., U . Prief er, and A . Punier. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Bio/Technology 1: 37-45. 104. Sinden, R. R., J. O . Carlson, and D . E . Pettijohn. 1980. Torsional tension in the D N A double helix measured with trimethylpsoralen in living E. coli cells: analogous measurements in insect and human cells. Cell 21: 773-783. 105. Smith, A . H . J . 1980. D N A sequence analysis by primed synthesis. Methods Enzymol. 166: 560-580. 106. Sockett, R. E. , T . J. Donohue, A . R. Varga, and S. Kaplan. 1989. Control of photosynthetic membrane assembly in Rhodobacter sphaeroides mediated by puhA and flanking sequences. J. Bacteriol. 171: 436-446. 107. Sojka, G . A . , H . H . Freeze, and H . Gest. 1970. Quantitative estimation of bacteriochlorophyll in situ. Arch . Biochem. Biophys. 136: 578-580. 107a. Solioz, M . and B. Marrs. 1977. The gene transfer agent of Rhodopseudomonas capsulata. A r c h . Biochem. Biophyz. 181: 300-307. 108. Stock, J. B., A . J. N i n f a , and A . M . Stock. 1989. Protein phosphorylation and regulation of adaptive responses in bacteria. Microbiol. Rev. 53: 450-490. 129 109. Taremi, S. S v and B. L . Marrs. 1990. Regulation of gene expression by oxygen: phototrophic bacteria. In G . Hauska and R. Thauer (ed.), The Proceedings of the 41st Mosbach Colloquium; The molecular basis of bacterial metabolism. Springer-Verlag, Berlin Heidelberg N e w York Tokyo. In press. 110. Taylor, D . P., S. N . Cohen, W. G . Clark, and B. L . Marrs. 1983. Alignment of genetic and restriction maps of the photosynthetic region of the Rhodopseudomonas capsulata chromosome by a conjugation-mediated marker rescue technique. J. Bacteriol. 154:580-590. 111. Tichy, H . V . , B. Oberle, H . Stiehle, E . Schiltz, and G . Drews. 1989. Genes downstream from pucB and pucA are essential for formation of the B800-850 complex of Rhodobacter capsulatus. J. Bacteriol. 171: 4914-4922. 112. Thornber, J. P., R. J. Cogdell , B . K . Pierson, and R. E . B. Seftor. 1983. Pigment-protein complexes of purple photosynthetic bacteria: an overview. J. Cell . Biochem. 23: 159-169. 113. v a n N i e l , C . B . 1944. The culture, general physiology, morphology, and classification of the non-sulfur purple and brown bacteria. Bacteriol. Rev. 8:1-118. 114. Vieira , J., and J. Messing. 1982. The p U C plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19: 259-268. 115. v o n G a b a i n , A . , J . G . Belasco, J. L . Schottel, A . C . Y . C h a n g , and S . N . Cohen. 1983. Decay of m R N A in Escherichia coli: investigation of the fate of specific segments of transcripts. Proc. Natl . Acad. Sci. U S A 80: 653-657. 116. Weaver, P. F., J. D . W a l l , and H . Gest. 1975. Characterization of Rhodopseudomonas capsulata. Arch . Microbiol . 105: 207-216. 117. Wellington, C . L . , and J. T . Beatty. 1989. Promoter mapping and nucleotide sequence of the bchC bacteriochlorophyll biosynthesis gene from Rhodobacter capsulatus. Gene 83: 251-261. 118. Wheelis, M . L. 1975. The genetics of dissimilatory pathways in Pseudomonas. A n n u . Rev. Microbiol. 29: 505-524. 119. Weissner, C , I. Dunger, and H . M i c h e l . 1990. Structure and transcription of the genes encoding the B1015 light-harvesting complex B and a subunits and the photosynthetic reaction center L , M , and cytochrome c subunits from Rhodopseudomonas viridis. J. Bacteriol. 172: 2877-2887. 120. Wil l iams, J. C , L . A . Steiner, and G . Feher. 1986. Primary structure of the reaction center from Rhodopseudomonas sphaeroides. Proteins: Structure, Function, and Genetics. 1: 312-325. 130 121. Yang, Z . , and C . E . Bauer. 1990. Rhodobacter capsulatus genes involved in early steps of the bacteriochlorophyll biosynthetic pathway. J. Bacteriol. In press. 122. Yanisch-Perron, C , J. Viera, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of M13 and p U C vectors. Gene 33:103-119. 123. Yen, H - C . , and B. Marrs. 1976. M a p of genes for carotenoid and bacteriochlorophyll biosynthesis in Rhodopseudomonas capsulata. J. Bacteriol. 126: 619-629. 124. Young, D . A . , C . E . Bauer, J. C . Will iams, and B. L . Marrs. 1989. Genetic evidence for superoperonal organization of genes for photosynthetic pigments and pigment-binding proteins in Rhodobacter capsulatus. M o l . Gen. Genet. 218: 1-12. 125. Youvan, D . C , E . J. Byl ina , M . Alberti , H . Begusch, and J. E. Hearst 1984. Nucleotide and deduced polypeptide sequences of the photosynthetic reaction-center, B870 antenna, and flanking polypeptides from R. capsulata. Cell 37: 949-957. 126. Youvan, D . C , and S. Ismail. 1985. Light harvesting II (B800-850 complex) structural genes from Rhodopseudomonas capsulata. Proc. Natl . A c a d . Sci. U S A 82: 58-62. 127. Z h u , Y. S., D . N . Cook, F. Leach, G . A . Armstrong, M . Alberti , and J. E . Hearst 1986. Oxygen-regulated m R N A s for light-harvesting and reaction center complexes and for bacteriochlorophyll and carotenoid biosynthesis in Rhodobacter capsulatus during the shift from anaerobic to aerobic growth. J. Bacteriol. 168: 1180-1188. 128. Z h u , Y. S., and J. E. Hearst. 1986. Regulation of expression of genes for light-harvesting antenna proteins L H - I and LH-II; reaction center polypeptides R C - L , R C - M and R C - H ; and enzymes of bacteriochlorophyll and carotenoid biosynthesis in Rhodobacter capsulatus by light and oxygen. Proc. Natl . Acad. Sci. U S A 83: 7613-7617. 129. Zsebo, K . M . , and J. E . Hearst. 1984. Genetic-physical mapping of a photosynthetic gene cluster from R. capsulata. Cell 37: 937-947. 130. Zucconi , A . P., and J. T . Beatty. 1988. Posttranscriptional regulation by light of the steady-state levels of mature B800-850 light-harvesting complexes in Rhodobacter capsulatus. J. Bacteriol. 170: 877-882. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0101048/manifest

Comment

Related Items