. x} u, . 1 . : .4». .1211}: . ii 1 .- fininghwflwfiwu . . . :32 .1432! A \ s! can“; . leftist! NEH... L9£rtrummflflm 2 2 Icyix‘ghnl {21.9.00 Jig? . ilnobnflw’ra; . 3... 50438: . Prof. 2 t .28 ..,.\$:L.\.u..4.H.222”H jfiififi‘ 2 $1.}! , i: gift; . I . grazicri L . A! 2232:.rf2rfh: . ’13.... If... , . _ , fizrfrrv Shun: v. 35.5"! .2 . A! 5! . 21(flit22f...;¢1 v.2. . . . .209 s .1. s MAM... -..:2l+,tns..rnfinr:3 IO . . . . ILIJJQER: afrr g .5 2.va 2 Pa: 1. airfimafifihnffimfift Wu. raffle: ., 2.. I: . E. 2 ..2 ffirfxzirél . : 1?... . 2 2 I. 22} Aw:!(§.)i :2 I. . 2 nfllfl.»n,uulfifil$§ujflnt$n r01: “5:31:31? H2... 5.221.? . .gf . A 2. 2.2.2:?» 7‘ If? ‘ u 0. p . 2 2c in.“ 31’s-?!» .. it . .. .354}??? y f. uflg‘Q .I‘v‘: \l I]! n v.“ a 1.53.0.2. 252.2%. 5.1.x» 3 i? I, : 3531:»! 11! 13g 2.11).}: quay-Win.“ ,.- givgibi fan. «on Y . 24.. : gig}... §§. 2:3... at :t ‘ $Vu U T 2?.‘hm ._ . v? 1-2:th g; ’T g 3.3.: x}: €231.25): .n‘v . . )l'fsl gunman..." gim- . . u , u :2 1 , 2:2 , .ff.‘ . s. »_v; _ ', ' a I" y ”gr-Mun: 1)....” in): \饻‘1.\<1r . ~ . r4.5:1.ufl..l.t}.. 12121.1“? . fifafigzu MN (’31.. SI.\21{’J.§ 2 5g! 1.. u u if ji.l.§z?§ Vzgfi $§3§§¥iaflflk§lhfih1§i§fl .§.x\¢li..x$. 6124:9004!» . . A: . 8”), if? 5'21 . g3“! §;?§~;H}a . 12'. 1 DR. PIA If. 2 . If... §é§75d95€23h§ lg 31.. . . 1.6..J‘nnv11ral. 7: ”fig Full-DWI. 0, u. . {(315.24}! 0! I!) .thJsdflnuf? \nJH-dwflfli 3\1 ?§.b{ . . . Hui: ’r... . 1.15.! 3 A. . g; I I... , 4 as? , .11.; 53." . izttfflfir H: :2. 2. {2:2 2 13%: .fiflfiflfftfirtttft cf. ftfiafu 2: 3:331: . a if}??? .h. At: (III! e: at erafluprruruuqtfirl}u3if L1: r. tit! g‘uvr? II. 35:12 .l¢ .r: H}..§t:0£é .1 Tryxcrfi n I .11).: XVI, Id‘qf t t '0‘?! 523 . MM.M§22.. 211213.” 2!..fdufl....un.2pmrltri§.+12%}. 2 gféilgfis .1. f . .2 5:11.): I EALHIV‘IP..L -?H..r! (J‘jir. ~P‘26ablltlv... fight‘llflr. a”; ”r... unit. _ it’llfi’t 2 I . .v tfllfllfllltitftJii a I!!! 215;“ gnaw...» alsfinkxuflllrlton: sit; ant l , 3 2. 2 3.1 .fA. ‘1‘... l :1. . E: “(stiff "52.13% fluu.-i.fiumw»nw.fltiu :4. 2.?! i f .nuviiilifflflfifluif . $1139 . 2:: (1.112. 2 ,2. fast»... “4.15.1...1 ....€2.Iltlll:2:r.n.flrl unit}: 2 .2 r . 2!-rrtatnflyt Java... 2 Inf}; , {a}... $12!»)? 205 . . it u . "3.2.5.1.! . r. 1)\l:x.r:u. . 5qu..;. 2 2 ,: . I! Y..~ 2 x J 4 .5 5. . . - «22:5; Iv." (Nata: “mania . 48!. n2. [g t' 2 ‘. $3, . Ii... 36% ? 1.. ‘5 $1? $15.; Lg“: . 313,214.52. . gfifgg 2 .. 25242.2)nfififlifis . if guy-ul‘igfiirt; 1. 2 . ._ . .2 :2 ii. (A, , , v ££.I§JN.ZIP.3§P r]. I. 1 ii A ,' .2511 4.1). J..- . . » IfQ- i‘vlfl‘Iw—z.)v ‘IVS‘. \4 {31“ 1.1.531. 2;. V.“ {#313410}. .51. \ §um . r . .1 f;“~n’|¢ y \tsg L41 311¥18 (I. Cit «fix! a? Rafi”? r... z: i rmu P .2 Ivt'. 2 :U... n. 1.5.”? v , 3‘ 2 1,, 1“]: 12¢; . a 313.0) 153%“? 36.3 in“; . 2,1!- 3.13,): . .. a flaming»? Haw; u finish». . 2 g .1» .f.» 11‘ at; 1 5’" . . «1.57:2. ‘1’...) 533122;: . ‘2 1 gg’nlgfig innuwvimuzflbvguéi 13.1%.) 2:529:33: .‘Iiizaun 1 $3.3..likzallx!‘ In)? \Vfiifdfqgi’é K gag . Q. Sc. {Rye}... “Ti? IINRVLBE rfiggtt. 9. 2.}...- Nflg _. . .v.‘ . . .. 248;...3»: jihafluu‘uvlup .. .. 141853;. 2322 .. gé‘Jhfiagssiihsibfiwuufiuiggg362 :1 2 2 . . V, . . . it»; .I i 3...!F 2 LI .u‘l‘nr .rflzgrgnfiir: fig‘nflmmmcugjgv‘d’? A? . . $V§Llrgghhr§ :J‘IVHMW’P‘." ilvrzl.b . f . I; . v . n 2 2 .421 . “.2 (.1 a) it 1 31.3%: 2 1:75 .. airmftu,d::zfi..;. 2:392:23} graft: . q 533.2% 5 2 , unflvflrflflfluv ,2J.Jufiuufhuuoi.‘fg. a; , i2 )1" . . , tr “Uh“: 11:42.2 If!!! 2. (Pnfirhdvf 5,210.4 f i ‘ n . I r - I V 0‘ , .. :2 ., . t w: xw‘gmtiza - I It .i "‘ 2y 2 6"! v If. g.’ 7‘53) (6’3ng ii. 2 gut. , . 4 list . .. , 2d,! {Hr-52701... utlv\tg5: ., _ -. Ndflriat'c‘; 4.. . . (TIJUI‘ try... . It! .51 .1300»... 1.1.0.: {OJ-{1.11:} :5}: Al. '.V‘vfo\ r?» g! In, I . . ”HM... ..)HT. f .t‘ if“: 2 .VIf: . A \\< {.31} 0.. . 2 it {fill’l‘ . $52}. $35.. £2 a .10 . I .11 . . . . .. . Pvmnuumgti..nufi.m,§w.bf . 03”}..u-Aé3ioutlvr7 {hurl}: 31o: 3...»Hfiofltfi§§? begun}: .. Vaiyiedwdumuiuufidi 2- . n .... 2.4.!.uwzwnflefylitf.ihl¥-§w-r . Tu .lwaflnflfliié - £92202! fur‘fvvf . I! . .{IPIA 11.: f- ‘51.“: MINIMUM [171117 1351 LIBRARY Michigan State University This is to certify that the dissertation entitled TRANSCRIPTIONAL REGULATION OF NITROGEN FIXATION AND NITROGEN ASSIMILATION GENES IN BRADYRHIZOBIUM JAPONICUM presented by Thomas H. Adams has been accepted towards fulfillment of the requirements for DOCTORAL degree in MI CROB IOLQGY 'or professor 10/15 86 Date / MCIIi-n— Arr ,; A ' 1‘ 'fl , , ' ' ' ' 0-12771 MSU LIBRARIES .——— RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. W TRANSCRIPTIONAL REGULATION OF NITROGEN FIXATION AND NITROGEN ASSIMILATION GENES IN BRADYRHIZOBIUM JAPONICUM by Thomas H. Adams A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1986 ABSTRACT TRANSCRIPTIONAL REGULATION OF NITROGEN FIXATION AND NITROGEN ASSIMILATION GENES IN BRADYRHIZOBIUM JAPONICUM by Thomas H. Adams I describe the isolation and characterization of three transcription units (nifH, nifDK, and nifB) that are required for nitrogen fixation by Bradyrhizobium japonicum. The promoters for transcription of each of these three 5. japonicum gene units share a high degree of DNA sequence homology with each other and with flif gene promoters from K. pneumoniae and other diazotrophs. Transcription from these 5. japonicum promoters is induced both during symbiotic association with soybean plants and during axenic free-living growth in microaerobic or anaerobic culture. I have isolated a regulatory gene, termed gggA, that is required for this nif transcription. The protein encoded by this regulatory locus (ODCA) has sequence homology to the product of the g. pneumoniae nifA gene. The gdgA gene is also necessary for the microaerobic and symbiotic expression of the gene encoding glutamine synthetase II, 91311. The wildtype gggA allele somehow limits the final viable cell density achieved in microaerobicaly grown B. japonicum cultures. Finally, mutation of QQQA results in defective nodulation of soybeans by B. japonicum. gdgA mutant bacteria are released at a reduced efficiency into plant cells of the nodule cortex, Thomas H. Adams and infected plant cells rapidly enter a degenerative phase. Mutant bacteria continue to divide in the intercellular space of the nodule cortex. The abnormal developmental phenotype observed for this mutant strain cannot be attributed to nitrogen starvation due to the lack of nitrogen fixation, since nodules induced by a nitrogenase deficient strain (gifgg') differentiate normally. These results indicate that an important aspect of pr0per nodule development is the bacterial localization in the oxygen limited plant tissue resulting in the expression of a number of bacterial functions. One or more of these functions must be involved in signalling the plant of the symbiotic nature of the invading bacteria. Three other B. japonicum nifA-like genes (finag, anag, and hflafi) are described. None of these genes are required for transcriptional control of gif or 913 gene expression under any condition tested. These genes are discussed in terms of their potential roles in regulating other aspects of cellular physiology in B. japonicum. ACKNOWLEDGEMENTS First, I would like to thank Barry Chelm for his support and guidance and for allowing me to work in his laboratory these last few years. I would also like to thank all of the people that I’ve worked with in Barry’s lab; Mary Lou Guerinot, Rob McClung, Todd Carlson, Annette Tumulo, Bryan Rahe, John Scott-Craig, Elizabeth Verkamp, Bill Holben, John Somerville, Todd Cotter, Prudy Hall, and Greg Martin. The willingness of all these people to help with problems and to discuss ideas has made graduate school a very enjoyable experience. I would also like to thank all of the people that have served on my guidance committee at various times; P. T. Magee, Arnold Revzin, Lee McIntosh, Bob Uffen, and Larry Snyder. I gratefully acknowledge Stuart Pankratz for his help in ultrastructural analysis of mutant bacteria and Frank Dazzo for helpful discussions towards interpreting these experiments. I would also like to thank my parents for all that they have done for me. Finally, and most importantly, I would like to thank Erin Bell for everything else. TABLE OF CONTENTS List of Tables List of Figures CHAPTER I INTRODUCTION CHAPTER 2 The nifH and nifDK Promoter Regions from Bradyrhizobium japonicum Share Structural Homologies with Each Other and with Nitrogen-Regulated Promoters from Other Organisms INTRODUCTION RESULTS Isolation of the Bradyrhizobium japonicum nifH Gene DNA Sequence of the nifH 5’ Terminus DNA Sequence of the nifDK 5’ Terminus Promoter Mapping Promoter and Leader Region Sequence Comparisons DISCUSSION EXPERIMENTAL PROCEDURES Bacterial Strains Soybean Growth and Nodule Bacteria Isolation RNA Isolation DNA Techniques 51 Nuclease Protection Analysis CHAPTER 3 Physical Organization of‘the Bradyrhizobium japonicum Nitrogenase Gene Region INTRODUCTION RESULTS Identification and Characterization of flif-Specific Cosmid Clones Localization of a B. japgniggm giffi Gene niffi Promoter Mapping DISCUSSION EXPERIMENTAL PROCEDURES Bacterial Strains and Media DNA Techniques Construction and Maintenance of a B. japonicum Cosmid Clone Bank Hybridization Procedures CHAPTER 4 Bradyrhizobium japonicum Genes With Sequence Homology to the Klebsiella pneumoniae nifA Gene INTRODUCTION RESULTS Identification and Cloning of B. japonicum nifA—like Regions DNA Sequence Analyses Construction and Characterization of B. japonicum h_a Deletion Strains . DISCUSSION EXPERIMENTAL PROCEDURES Bacterial Strains Bacterial Media, Growth Conditions, and Strain Construction PAGE 46 48 48 58 63 66 67 67 68 68 69 7O 72 72 75 75 86 9O 90 91 PAGE Nucleic Acid Techniques 91 CHAPTER 5 Microaerobic Induction of nif and gln Gene Expression in Bradyrhizobium japonicum INTRODUCTION 92 RESULTS 95 Effects of O2 Concentration on B. japonigum Gene Expression 95 Effect of Ammonia on Microaerobic Growth and Gene 'Expression 98 Growth Phase Dependence of Microaerobic nif and glfl Transcription 100 Transcription of pit and gin Genes During Symbiotic Development 103 DISCUSSION 106 ’ EXPERIMENTAL PROCEDURES 111 Bacterial Strains, Media, and Growth Conditions 111 Isolation of Bacteria from Soybean Root Nodules 112 Nucleic Acid Techniques 112 CHAPTER 6 Characterization of a Bacterial Gene Required for the Normal Differentiation of the Bradyrhizobium japgnigum / Soybean Symbiotic Interaction INTRODUCTION 113 RESULTS 116 Identificationand Mutagenesis of the gdgA Gene 116 gggA Control of Bacterial Growth and Gene Expression 123 gggA Control of the Plant-Bacterial Interaction 131 DISCUSSION 140 EXPERIMENTAL PROCEDURES 145 CHAPTER 7 REFERENCES Bacterial Strains Recombinant Plasmids, Cosmids, and Phage Bacterial Media, Growth Conditions and Strain Construction Plant Tests Nucleic Acids Techniques Protein Analysis Microscopy SUMMARY AND CONCLUSIONS vi PAGE 145 145 146 147 148 149 150 151 161 LIST OF TABLES TABLE PAGE 1 Summary of nif and gln Gene Expression data for B. japonicum hfla Deletion Strains 84 2 Acetylene Reduction by Soybean Root Nodules 2 and 4 Weeks After Inoculation With BJllO or BJ1011 85 3 Effects of Varying O2 Concentrations on Transcription of Eli and gln Genes 99 4 31: and 913 Gene Transcription During Growth at 0.2% Oxygen 104 5 mRNA Abundance of 31f and gin Genes During Symbiotic Development 105 6 Abundance of nifH, nifDK, and glnII Transcripts in B. japonicum Grown Under VArious Conditions 126 7 Abundance of nifH, nifDK, and glnII Transcripts in B. japonicum Isolated From Soybean Nodules 139 vii FIGURE 1 mNOtU'I 10 11 12 13 14 15 16 17 LIST OF FIGURES Klebsiella pneumoniae and B. japonicum nifH genomic regions DNA sequence of the 5’ end of the B. japonicum nifH gene DNA sequence of the 5’ end of the B. japonicum nifD gene Determination of 5’ ends of nifH and nifD mRNAs by 51 protection and DNA sequence analysis Comparisons of nucleotide sequences of nif promoters Analyses of nifH mRNA leader region Summary of hybridization probes Hybridization of nick-translated pRJcosZ-63 to restriction endonuclease digests of pRJcosI-62 Restriction endonuclease map of the B. japonicum nif gene cluster Verification of the B. japonicum nif genomic structure 'by Southern hybridization to restriction endonuclease digests of total genomic DNA Hybridization of the K. pneumoniae nifAB genes to BgQRI digested recombinant cosmids from the nif gene cluster Nucleotide sequence of the 5’ end of the B. japonicum gifB gene Identification of the B. japgnicum nifB promoter sequence nifA-homologous sequences in B. japonicum, B. meliloti, and K. pneumoniae Genomic restriction endonuclease maps for B. japonicum strain BJllO nifA-homologous genes hna2, hna3, hna4, and hna5 Nucleotide sequence of the B. japonicum nifA-like region hnag Nucleotide sequence of the B. japonicum nifA-like region Baa; viii PAGE 21 24 26 29 32 40 so 52 55 57 60 62 65 74 77 78 79 FIGURE 18 19 20 21 22 23 24 25 26 27 Partial nucleotide sequenc of the B. japonicum nifA-homologous region hna4 Comparison of derived amino acid sequences from the K. pneumoniae nifA and ntrC genes with amino acid sequences for hna2, hna3, and hna4, from B. japonicum Abundance of nifH, nifDK, glnII, and glnA transcripts in B. japonicum grown under a variety of atmospheric oxygen concentrations Growth properties of B. japonicum cultured under 0.2% oxygen with and without ammonia Restriction map of the B. japgnicum strain BJIlO och genomic region Genomic hybridization analyses of the B. japonicum strains BJ110 and 802101 Abundance of nifH, nifDK, and glnII transcripts in B. japgnigum from a variety of growth conditions Growth properties of B. japonicum strains 80110 and 302101 under limiting oxygen conditions SOS-PAGE of plant protein isolated from root nodules incited by B. japonicum strains 802101, BJ702, and 80110 Typical electron micrographs of nodule tissue incited by B. japonicum strain BJIIO, BJ2101, or 80702 ix PAGE 80 82 97 102 119 122 125 130 134 137 CHAPTER 1 INTRODUCTION The biological reduction of atmospheric dinitrogen into organic compounds accounts for more than two-thirds of global nitrogen fixation. The majority of the remaining ammonia production is by chemical means (Burris, 1980). Biological nitrogen fixation, as catalysed by the nitrogenase enzyme complex, is carried out by a number of diverse prokaryotes including obligate anaerobes (e.g. Clostridium pasteurianium), facultative anaerobes (e.g. Klebsiella pneumoniae), and obligate aerobes (e.g. Azgtobacter vinlandii). Nitrogenases from all these different bacteria have similar properties (Eady gt al., 1974). The nitrogenase complex consists of two oxygen-labile components. Component I (dinitrogenase) has two copies each of two unique subunits and contains the site for substrate binding and reduction. Component II (dinitrogenase reductase) functions in reducing component I and is a homodimer (Mortenson and Thornley, 1979). The molecular genetics of nitrogen fixation is best understood in the free-living diazotroph K. pneumoniae. Here, a cluster of 17 genes, grouped in 7 operons encode the enzymes specifically involved in nitrogen fixation (nif genes; Roberts and Brill, 1981). Three of these genes, nifH, nifD, and nifK, encode dinitrogenase reductase and the two subunit types of dinitrogenase, respectively. These three genes comprise a single operon that is transcribed in the order nifHDK (Merrick gt gl., 1978; Kaluza and Hennecke, 1979). The genes gift, gitfl, nitB, gjtg, gift, and gifg, all encode products involved in the synthesis of the FeMo cofactor of nitrogenase component I (Roberts and Brill, 1981). The products of the giffl, g1£§, gift, and nitB genes are required for modification or activation of nitrogenase component 11 (Roberts and Brill, 1981). Finally, 31:5 and gift encode git-specific regulatory molecules (Roberts and Brill, 1981). F The transcription of K. pnggmgnigg git genes is modulated in response to oxygen and fixed nitrogen by two distinct regulatory systems each of which has positive and negative control elements. One links . nitrogen fixation to the general nitrogen control pathway (Ntr) known for several enteric bacteria (Magasanik, 1982). The second is specific for git gene control (Dixon, 1984). The general nitrogen control (Ntr) system found in enteric bacteria has three regulatory components; nttA (also termed glgfi or tpgN), ntrB (also termed glut), and gttt (also termed glgB) (Leonardo and Goldberg, 1980; de Bruijn and Ausubel, 1981; Epsin gt g1., 1981; 1982; CW and Ausubel, 1983; Merrick, 1983; Drummond gt gl., 1983). The products of these loci (NTRA, NTRB, NTRC; gene products for these and other. regulatory genes will be referred to with capitals) interact to control the expression of genes for nitrogen assimilation, glnA (glutamine synthetase), gmt (ammonia transport) (Jayakumar gt gl., 1986), and git, as well as those genes involved in the utilization of secondary nitrogen sources including histidine (hgt), proline (pgt), and arginine (ggt). All of the loci that comprise the Ntr regulon have a similar promoter —;_Jl. structure characterized by the consensus sequence CTGGYAYR-N4-TTGCA (where C, T, G, and A represent the four standard nucleotides; Y=C or T, RsA or G, and N-C, T, G, or A) in the region -26 to -10 bp relative to the transcriptional initiation site (Beynon gt g1., 1983; 0w gt gl., 1983; Sundaresan gt gl., 1983). The identification of this novel consensus sequence for Ntr promoters led to the hypothesis that transcriptional initiation of these genes is mediated by a novel RNA polymerase holoenzyme (0w gt g1., 1983). Subsequent experiments using in vitro transcription systems have demonstrated that NTRA is a sigma factor that is required for transcription of promoters in the Ntr regulon by bacterial core RNA polymerase (Hirschman gt l., 1985; Hunt and Magasanik, 1985). Transcription of gttA is not controlled by nitrogen regulation (Costano and Bastacchea, 1984; de Bruijn and Ausubel, 1983). Instead, global Ntr control is mediated by modulation of the activity and expression of NTRC, a positive regulator of Ntr promoters. Ntr mediated activation of gene expression involves both the NTRA and NTRC proteins. NTRC functions as a dimer in binding double stranded DNA at the consensus sequence TGCACY-NSAGGTGCA (Ueno-Nishio gt g1., 1984; Hirschman gt g1. I985; Reitzer and Magasanik, 1986; MacFarlane and Merrick, 1985; Ames and Nikaido, 1985). Deletion of NTRC binding sites upstream of the B. ggli glnA promoter eliminates Ntr-mediated transcriptional activation of this gene (Reitzer and Magasanik, 1986). The function of these upstream binding sites in transcriptional activation is at least somewhat independent of their position and orientation with respect to the RNA initiation site (Reitzer and Magasanik, 1986). These properties ' suggest that NTRC binding sites may function similarly to eukaryotic enhancer elements (Khoury and Gruss, 1983). The activity of NTRC is modulated via a covalent modification mechanism involving NTRB (Ninfa and Magasanik, 1986). NTRB is a kinase/phosphatase and phosphorylates NTRC in response to nitrogen limitation. Dephosphorylation of NTRC occurs when nitrogen availability no longer limits growth. The activity of NTRB is controlled by an unknown mechanism requiring the glgB and glgB products. It has been proposed that these proteins sense the relative levels of glutamine and 2-ketoglutarate in the cell (Magasanik, 1982; Bueno gt g1., 1985). Thus, when nitrogen limits growth, the glutamine to 2-ketoglutarate ratio is low causing the conversion of NTRB to a kinase which then phosphorylates NTRC. This phosphorylated NTRC (NTRC-P) is then able to activate transcription of promoters under Ntr control. The manner in which NTRC phosphorylation affects its activity as a transcriptional regulator is not known. Nonphosphorylated NTRC is apparently able to bind DNA. In addition, ntrB deletion mutants do not abolish the ability of NTRC to activate transcription of Ntr promoters (Chen gt gl., 1982; Bueno gt gl., 1985). These gtLB mutants are slightly delayed in their response to nitrogen limitation. Perhaps phosphorylation affects the affinity of NTRC for DNA binding and/or interacting with NTRA. General nitrogen regulation of git gene expression in B. pgggmggigg occurs through the action of NTRA, RNA polymerase, and NTRC on the expression of gift and nifA (Leonardo and Goldberg, 1980; de Bruijn and Ausubel, 1981; 1983; Epsin gt gl., 1981; 1982; 0w and Ausubel, 1983; Drummond gt g1., 1983; Merrick, 1983). These genes comprise a single operon in B. pngumoniag and are transcribed from a Ntr promoter (0w and Ausubel, 1983). The gift and nifA gene products (NIFL and NIFA) in turn control expression from other git promoters (Buchanan-Wollaston gt g1., 1981a; 1981b; Dixon gt gl., 1980, Hill gt g1., 1981, Merrick gt g1., 1982). NIFA serves as a positive activator for transcription from nifLA and other nit operons. As with NTRC, NIFA mediated transcriptional activity requires a functional NTRA protein. When oxygen or intermediate levels of ammonia are present, NIFL prevents the activation of other nit gene transcription. The mechanism of NIFL mediated repression of git gene transcription is not understood. It is likely that NIFL does not function as a repressor by binding DNA since it inhibits NIFA- but not NTRC-mediated activation of nifLA transcription (Cannon gt gl., 1985). In view of the NTRB-NTRC interaction described above, one might propose that NIFL exerts its effect through some covalent modification of NIFA. In addition to their joint requirements for NTRA, NIFA and NTRC are structurally and functionally related. The amino acid sequences for both proteins are homologous throughout a large central region (Buikema gt 31., 1985; Drummond gt 31., 1986; Nixon gt gl., 1986; Gussin gt gl., In Press). Promoters for git genes have the same -26 to -10 consensus sequence CTGGYAYR-N4-TTGCA aS'Ntr regulated promoters (Beynon gt g1., 1983; 0w gt gl., 1983). Finally, under certain conditions, some of these promoters can be activated by either regulatory system (0w and Ausubel, 1983; Merrick, 1983; Buck gt g1., 1985). The mechanism of NIFA mediated transcriptional activation is not as well understood as that for NTRC. However, genetic studies coupled with characterization of nif gene promoter sequences suggest certain analogies to NTRC dependent regulation. As with Ntr promoters, nit promoters usually have a characteristic upstream consensus (TGT-NlO-ACA) that is required but functions independantly of position or orientation with respect to transcription initiation (Buck gt g1., 1986). This sequence is similar to the proposed consensus binding sequence (TGTGT-NG-lO-ACACA) for a variety of DNA binding protiens including LACI, CRP, ARAC, LEXA, and GALR (Buck gt gl., 1986; Giequel-Sanzey and Cossart, 1982). However, it has not yet been possible to detect binding of purified NIFA protein to this upstream region. Studies of regulatory mechanisms for controlling nitrogen fixation by diazotrophs other than 5. pneumoniag have relied on the models described above. This approach has proven to be particulaly useful in studying the regulation of nitrogen assimilation and nitrogen fixation by bacteria of the genera Rhizobium and Bradyrhizobium. Rhizobia and bradyrhizobia reduce atmospheric dinitrogen primarily during symbiotic association with their leguminous host plants. The successful establishment of a nitrogen fixing symbiosis is complex, requiring the coordinate differentiation of both the plant and the bacterial cells. For the bacterium, this developmental process culminates upon differentiation into a morphologically and functionally distinct nitrogen fixing endosymbiont termed a "bacteroid". Nitrogen fixation by rhizobial bacteroids differs from nitrogen fixation by B. pgggmggigg and most other free-living diazotrophs in that it is not coupled to nitrogen assimilation. Very little of the ammonia produced from dinitrogen reduction by bacteroids is used to support bacterial growth but instead, most of this ammonia is exported (Bergerson and Turner, 1967) into the plant cytoplasm where it is assimilated by plant enzymes (Miflin and Lea, 1976). Enzyme assays indicate the glutamine 7 synthetase activity is repressed in these nitrogen fixing cells (Brown and Dilworth, 1975; Stripf and Werner, 1978; Werner gt gl., 1980). In return for this fixed nitrogen, the plant supplies reduced carbon compounds to the bacteria to fuelthe respiration needed to meet the high energy demands of nitrogen fixation. The symbiotic mechanisms controlling plant and bacterial differentiation in root nodules are not well understood. One physiological factor that has been experimentally implicated in developmental control is oxygen limitation. Low oxygen levels presumably exist during the early stages of nodule development due to the presence of multiple diffusion barriers and the active respiration of both the plant and the bacterium. Later in nodule development, the plant encoded oxygen binding protein leghemoglobin maintains a very low concentration of free oxygen in the nodule, protecting the bacteroid nitrogenase from 02 inactivation (Appleby, 1984). In response to these conditions rhizobia induce a terminal oxidase activity with a very high affinity for oxygen (Bergerson and Turner, 1980). In addition to its 02 binding activity, leghemoglobin acts as an oxygen carrier protein in nodules, transporting oxygen to the bacteroids by facilitated diffusion. In this manner, bacteroids are provided a low oxygen environment to protect the oxygen labile nitrogenase complex from inactivation, coupled to a high oxygen flux for the vigorous respiration required to supply the energy necessary for nitrogen fixation. Oxygen limitation has been shown to affect gene expression in both plants and bacteria. A small but detectable induction of uricase, a nodule specific enzyme, can occur in non-nodule soybean tissue when oxygen is limiting (Larsen and Jochimsen, 1986). In addition, many Bradyrhizobium strains can be made to fix nitrogen in free-living culture under specialized conditions that include microaerobiosis (Keister, 1975; Pagan gt gl., 1975; Kurz and Larue, 1975; McComb gt g1., 1975; Tjepkema and Evans, 1975). Nitrogen metabolism during asymbiotic nitrogen fixation by bradyrhizobia is similar to that observed in bacteroids since most of the fixed nitrogen is not used to support growth but is instead exported into the medium (O’Gara and Shanmugan, 1976; Bergerson and Turner, 1978; Ludwig, 1980). For at least some Bradyrhizotium strains this failure to assimilate fixed nitrogen is partly due to a decrease in glutamine synthetase activity (Ludwig, 1980; Bergerson and Turner, 1978). There is also evidence that low oxygen induces an ammonia export system that could account for the observed accumulation of fixed nitrogen in the medium (Gober and Kashkett, 1983). Ammonia export by bacteroids is "logical” in terms of their role as symbiotic nitrogen fixing "organelles”. However, it is difficult to understand why such apparently altruistic behavior should have evolved. Two hypothesis have recently been put forward to explain the nitrogen fixation behavior of rhizobia. Ludwig (1984) proposed that nitrogen fixation coupled to ammonia export has evolved as a mechanism for syntrophic growth. This model implies that the differentiation of metabollically specialized nonviable cells occurs within a growing cell population that is limited for nitrogen. Thus, under conditions of low 02 with a limiting supply of fixed nitrogen, some population of cells in a culture would switch their metabolic state to one in which nitrogen is fixed but growth is repressed. Nitrogen assimilation would be blocked in these cells and ammonia export activated (Ludwig, 1980; Gober and Kashkett, 1983). This exported ammonia could then be utilized by non-nitrogen fixing members of the population (and by the plant) that are still active in ammonia assimilation. An alternative but perhaps related model for symbiotic nitrogen fixation was proposed by Kahn gt g1. (1985). According to this model the plant is able to induce the bacterial dinitrogen reduction by feeding amino acids. In this way, nitrogen might be used as a carbon carrier into the bacteroids. Nitrogen fixation would then occur to replace ammonia utilized by the plant. This would insure a continued flow of carbon to the bacteria. The mechanism for these models differ in that one requires bacterial differentiation while the other simply exploits known bacterial physiology. It is possible that both proposed models could play roles in controlling symbiotic nitrogen fixation. Several laboratories have begun to characterize regulatory mechanisms for controlling the expression of nitrogen fixation and nitrogen assimilation genes in rhizobia and bradyrhizobia. These studies have been helped by the fact that DNA and amino acid sequences among nitrogenase genes from several organisms are highly conserved (Ruvkin and Ausubel, 1980; Chen gt gl., 1973; Emerich and Burris, 1978). Many genes required for nitrogenase activity have been identified in both Rhingiug and Bradyghizogigg strains using cloned 5. ggggmggigg git genes as hybridization probes. The genes encoding the three nitrogenase polypeptides (giiB, B, and B) have been isolated from B. ggiiigti (Ruvkin and Ausubel, 1980; Ditta gt gi., 1980), R. trifgiii (Scott gt g1., 1983a), B. leguminogarum (Ma gt gi., 1982; Schetgens gt gi., 1984; B. pgggggii (Quito gt gi., 1982), B. japonicum (Hennecke, 1981; Adams and Chelm, 1984), and several cowpea Bradyrhizobium strains (Scott gt gi., 1983b, Weinmann gt gi., 1984). In addition, genes homologous to 10 the 5. gngumoniae gitA (Szeto gt gl., 1984; Adams gt gi., 1984; Fischer gt gi., 1986; Donald gt gi., 1986), gitB (Fuhrmann gt gi., 1985, Rossen gt g1., 1984; Noti gt gl., 1986) gitt (Hennecke gt gi., 1985; Norel gt a1., 1985; Donald gt g1., 1986) and gitB (Fischer gt g1., 1986b) genes have been identified in both Bhingium and Brgdxrhizggigg strains. Other genes (tixA, B, and B) that are required for rhizobial nitrogen fixation but are not found in Klgbsiglla have been identified by transposon mutagenesis (Ruvkin gt gi., 1982; Corbin gt gi., 1983, Fuhrmann gt gi., 1985).. Despite the high degree of interspecies nucleotide sequence homology observed among git genes, their location and organization is varied. All Bgizgtigg ggg. studied contain symbiotic plasmids (Banfalvi gt 31., 1981). These plasmids carry clusters of genes that are required for nitrogen fixation and nodule development (ggg genes; Long gt gi., 1982; Corbin gt gi., 1983; Downie gt gi., 1983; Schofield gt gi., 1983; Kondorosi gt gi., 1984). Although git and ggd genes are also clustered in Bragyrgizobium strains (Kaluza gt g1, 1983; Adams and Chelm, 1984; Fuhrmann gt gi., 1985; Lamb and Hennecke, 1986), there is no evidence for symbiotic plasmids in these bacteria (Haugland and Verma, 1981) indicating a probable chromosomal location. Much of what is known about the regulation of git genes in rhizobia and bradyrhizobia comes from comparisons to and extrapolations from the detailed understanding of git gene regulation in B. gggggggigg (see above). Promoters for git and ti; genes in both Bgiggtigg and Biggyggizgtigm closely resemble those identified for git operons in B. ,pggggggigg (0w gt gl., 1983; Adams and Chelm, 1984; Alvarez-Morales and liennecke, 1985; Alvarez-Morales gt gi., 1986). The similarities include . 11 homology to both the -26 to -10 CTGGYAYR-N4-TTGCA Ntr promoter consensus sequence and the git gene specific TGT-NlO-ACA upstream activator consensus sequence. Sundaresan gt g1. (1983a,b) demonstrated that the NIFA and NTRC proteins from B. gngumonjae could each activate transcription from the B. mgljlgtj gitB promoter in Escherichia 9911 and in B. ggliigti. In E. ggii, this transcriptional activation required a functional gtgA gene product (Sundaresan gt gi., 1983a). Such heterologous expression experiments have also been performed using promoters from the B. jagonigum gitB and gitBB transcription units (Alvarez-Morales and Hennecke, 1985; Alvarez-Morales gt gi., 1986). Here, transcription could be activated by NIFA but not by NTRC. As with the 3. ggliloti gitB promoter, the transcriptional activation of B. . jggggiggg git promoters in a heterologous t. ggii system requires a functional gtgA gene product. Finally, gitA-homologous genes have now been described in B. ggljloti (Szeto gt al., 1984), B. leguminosarum (Rossen gt gi., 1985), and B jagonicum (Adams gt gi., 1984; Fischer gt gi., 1986). Mutations in at least some of these genes yield rhizobial strains that are unable to induce git gene transcription. All of these data support the hypothesis that git gene expression is controlled in rhizobia by a system that is somewhat analogous to the git-specific regulatory pathway found in 5. gggggggigg. In contrast to Blgtgigiig, It seems unlikely that the expression of the rhizobial gitA-like regulatory genes are controlled by a central nitrogen regulatory system. As described above, nitrogen fixation by the rhizobia and bradyrhizobia differs from nitrogen fixation by B. gggggggigg in that it is not coupled to nitrogen assimilation. In addition, rhizobial nitrogenase activity is usually not repressed by the 12 presence of a fixed nitrogen source (Bergerson and Turner, 1967; Keister, 1975; Scott gt g1., 1979). Finally, a gene with gth-like properties has been described in B. ggljlotj (Ausubel gt gi., 1985). Mutations in this gene have no effect on the development of symbiotic nitrogen fixation. This differential control of nitrogen fixation and nitrogen assimilation must be central to effective symbiotic nitrogen fixation. ' In free-living rhizobia, most ammonia assimilation occurs by the coordinate activity of glutamine synthetase (GS) and glutamate synthase (GOGAT) (Brown and Dilworth, 1975; Kondorosi gt gi., 1977; Vairinhos gt gl., 1983). This pathway is primarily controlled by regulating the levels of GS activity. GS regulation is complicated in rhizobia and bradyrhizobia by the fact that these bacteria contain at least two unique forms of the GS enzyme, 651 and GSII (Darrow and Knotts, 1977). These two distinct nitrogen assimilatory enzymes are differentially regulated in response to a number of environmental factors including nitrogen source, carbon source, oxygen concentration, and symbiotic development (Cullimore gt gi., 1983; Darrow gt gi.,1981; Fuchs and Keister, 1980; Ludwig, 1980b; Rao gt gi., 1978). Rhizobial 681 is similar to the single GS enzyme found in most other Gram negative bacteria. It is a polymeric enzyme with 12 identical subunits of 59,000 daltons each. Like the GS in t. ggii and other enterics (Magasanik, 1982), the activity of this enzyme is modulated by a reversible adenylylation cascade system (Darrow and Knotts, 1977; Darrow, 1980). The gene encoding GSI (gigA) has been isolated from both 3. ggiiigti (Somerville and Kahn, 1983) and B. jggggiggg (Carlson gt gi., 1985). The transcriptional control of the B. 13 jagonicum gigA gene is different from that for B. ggii gigA. In B. ggii, gigA transcription is regulated from two tandem promoters, gigApl and gigApZ (Reitzer and Magasanik, 1985). gigApl is active under nitrogen excess conditions and provides the cell with basal levels of glutamine synthetase. gigApZ is controlled by the Ntr system described above. When nitrogen availability limits growth, gigApZ is activated to increase glutamine synthetase levels and thereby the cell’s ability to assimilate nitrogen. In contrast, the B. jagonicum gigA gene is transcribed from a single promoter and the amount of transcription is independent of nitrogen availability (Carlson gt gi., 1985). Thus, at least for B. jagggigum, GSI activity would seem to be primarily controlled by adenylylation. The gene encoding GSII in B. jaggniggm (gigii) has recently been described (Carlson and Chelm, 1986). This enzyme is unlike any known prokaryotic glutamine synthetases but is closely related to the analogous eukaryotic enzymes both in subunit structure (Darrow, 1980) and amino acid sequence (Carlson and Chelm, 1986). The level of GSII activity is regulated by nitrogen availability (Darrow gt gi., 1981; Ludwig, 1980b) and oxygen concentration (Rao gt gi.,1978). Since no post-translational control of GSII activity is known, it is likely that the activity of GSII is primarily modulated by transcriptional control. The promoter for gigii is similar in sequence to Ntr controlled promoters from enteric bacteria in the -26 to -10 region (Carlson, 1986). In addition, under aerobic growth conditions, gigii gene transcription is regulated by a mechanism that interprets the relative availabilities of carbon and nitrogen (Carlson, 1986). This Ntr-like control is similar to that observed for glnApZ in B. coli. 14 The molecular mechanisms for transcriptional control of rhizobial git and gig genes remain unclear. In this dissertation I describe the isolation and characterization of B. jagonicum genes that are homologous to the t. gneumoniae gitB, gitB, and gitB genes (Chapters 2 and 3). I have examined the effects exerted by environmental factors (including both nitrogen and oxygen limitation) and symbiotic development on the expression of these git genes (Chapter 5). In addition, I have isolated four separate DNA regions with homology to both the K. pneumoniae gitA and gttt genes (Chapter 4). The effects of mutations in these genes on both git and gig gene expression are described (Chapters 4 and 6). Finally, a model will be presented for the control of nitrogen fixation and assimilation in B. jagonicum. The materials in Chapter 2 and most of the information in Chapters 3 and 4 has been presented elsewhere (Adams and Chelm, 1984; Adams gt gi., 1984). Chapter 6 has recently been submitted as a manuscript for publication (Journal of Bacteriology). CHAPTER 2 The nifH and nifDK Promoter Regions from Bradyrhizobium jagonicum Share Structural Homologies with Each Other and with Nitrogen-Regulated Promoters from Other Organisms INTRODUCTION The biological reduction of atmospheric dinitrogen, catalyzed by the nitrogenase enzyme complex, is carried out by a number of procaryotic organisms. The nitrogenase complex is composed of two enzymes, dinitrogenase (component I) and dinitrogenase reductase (component 11). Component 1 itself is a multimeric enzyme composed of two a-subunits and two p-subunits encoded by the gitB and gitB genes. The reduction of component I is carried out by component 11 which is composed of two identical subunits encoded by the gitB gene. The conservation of DNA and amino acid sequences among nitrogenase genes from a variety of nitrogen-fixing bacteria (including ngdyggitgtigg iggggiggg) is well documented (Ruvkin and Ausubel, 1980; Chen gt gi., 1973; Emerich and Burris, 1978). This interspecies conservation, however, does not extend to include a preservation of git . operon structure. Whereas nifH, B, and t genes are transcribed as a 15 16 single operon in Blebgiglla gngumonigg (Reidel gt gi., 1979) and several Bgizggigg spp., such as B. meliloti (Banfalvi gt gi., 1981), they occur as two separate transcription units gitBB and gitB in Anabaeng 7120 (Rice gt g1., 1982) and as gitB and gitBB in B. jaggniggm and two cowpea Bradyrgingjgm spp. (Kaluza gt al., 1983; Scott gt gi., 1983). The expression of nitrogenase activity is under complex and varied regulation. The control mechanisms are best studied in B. gneumoniae where transcriptional regulation of nitrogenase activity has been demonstrated to occur in response to concentrations of 02, fixed nitrogen, and molybdenum (Roberts and Brill, 1981; Buchanan-Wollaston gt g1., 1981a; Dixon gt gl., 1980). In most cases, derepression of git genes in Bhizobium species requires symbiotic association with a host legume leading to extensive differentiation of both organisms (Corbin gt gi., 1982; Paau and Brill, 1982). Symbiotic nitrogen fixation is affected by environmental factors; fixed nitrogen has been shown to inhibit this developmental process (Wong, 1980). Among Bradyrhigobium spp. some strains do not require symbiosis for expression of nitrogenase activity and can be induced to derepress git genes in free-living culture (Keister, 1975; Scott gt gi., 1979). For these organisms, environmental control of nitrogenase expression appears to be somewhat similar to that seen for B. gggggggigg except that the presence of ammonia in the growth medium does not repress git gene expression Scott gt gi., 1979) . Transcriptional control of git genes in B. gggggggigg occurs in part by the general nitrogen regulatory system common to all enteric bacteria (Magasanak, 1982). The primary regulatory proteins involved in this process are encoded by ntrA (glnF), ntrB (glnL), and ntrC (glnG). 17 These proteins work together to control expression of genes involved in nitrogen assimilation, gigA (glutamine synthetase) and git, as well as those involved in degradation of the amino acids histidine (ggt), arginine (ggt), and proline (ggt). The current model for regulation of the Klgbsiellg git genes involves a concerted action of the gttA, and gth, and gttt gene products on the transcription of the gittA operon (Leonardo gt gl., 1980; de Bruijn gt gi., 1981; Epsin gt gi;, 1981; Epsin gt gi., 1982; Ow and Ausubel, 1983; Drummond gt gi., 1983; Merrick, 1983; de Bruijn and Ausubel, 1983). The gitL and gitB gene products in turn control expression from other git operons (Buchanan—Wollaston gt gi., 1981a; Dixon gt gi., 1980; Roberts and Brill, 1980; Hill gt gi., 1981; Buchanan-Wollaston gt gi., 1981b; Merrick gt gi., 1982). Thus, under conditions when nitrogen limits growth, gtgg and gtgt combine to activate transcription of the gittB genes. The gitA gene product then serves as a positive activator of transcription from other git operons by a mechanism that also requires gt_A. In addition, it has recently been demonstrated that the gitA gene product can substitute for gttt in activation of its own expression as well as expression from other nitrogen assimilatory promoters (Ow and Ausubel, 1983; Drummond gt gi., 1983). The function of the gitt gene product is to repress git transcription as a consequence of rising concentrations of oxygen and ammonia. Finally, when fixed nitrogen is in excess, the gth and gttt products combine to eliminate git expression entirely through repression of the nifLA operon. In order to begin to evaluate the molecular mechanisms involved in regulating the transcription of genes controlled by nifA, ntrC, and ntrA, recent work has centered on the determination of promoter 13 sequences for nitrogen assimilatory genes from Klebsiella. This has led to the establishment of a consensus sequence for promoters controlled by this pathway. Interestingly, the promoter for B. mglilgtj gitB shares these sequences with gtgt/gitA regulated promoters for Blgbsigllg (0w gt 31., 1983; Beynon gt g1., 1983). In addition, transcription of the 1gg1 gene, when fused to the B. mgliloti gitB promoter, is controlled in Egggggiggig ggli via the general gtgt/gitA nitrogen regulatory pathway, suggesting similar control mechanisms may be involved (Sundaresan gt 11., 1983; Sundaresan gt g1., 1983). The genus Bgigobigm has traditionally been divided into two groups based on growth rate (Vincent, 1982). The "slow-growing“ bradyrhizobia are apparently only distantly related to the "fast-growing" rhizobia. In fact, when many slow growers were tested, all had less than 10% overall DNA homology with fast growers such as B. mglilgti or B. trjtglij. In addition the slow growers are themselves a very diverse group with overall DNA sequence homologies between different strains often being only in the 25% range (Hollis gt 1., 1981). In this paper we present the nucleotide sequences of the promoter regions for the gitB and gitBB operons from the slow-growing species B. jgpggiggg strain 110. The comparison of these promoters with each other and with git promoter regions from Bgizgtigg species shows striking structural homologies. In addition, conserved sequences are found with git promoter sequences from i ll . 19 RESULTS Igglgtjgn of tgg Brggyggizobigm jaggniggm gifH Bgng, In B. jagggigum strain 110 the nitrogenase complex polypeptides are encoded by two separate operons, gitB and gitBB (Kaluza gt a1., 1983). The gitBB operon had previously been isolated as a cloned BingII fragment by Hennecke (1981). We therefore isolated the gitB gene from a library of B. japogicum USDA strain 110 DNA constructed in the E. ggii phage lambda vector BFlOl as described previously (Carlson gt g1.,'1983). The lambda library was screened by hybridization to the 730 bp tggRI-BglII fragment isolated from the plasmid pSA30 (Cannon gt g1., 1979). This probe fragment contains a portion of the B. gneumoniae gitB gene and no gitB or gitB sequences (Fig. 1). This library screen resulted in the isolation of a single phage, XNH-l. A restriction map of the B. jagogigum genomic DNA cloned in XNH—l is illustrated in Fig. 1. The approximate location of the B. jagonicum gitB gene was determined by Southern hybridization analysis (Southern, 1975) of NH-l DNA using the B. gneggonigg gitB fragment described above as probe (data not shown) allowing the placement of the'structural portion of the gitB gene within the region indicated in Fig. l. Transcription of the gitB gene was determined to proceed in the direction indicated by 51 protection analysis of separated strands of the completely internal 180 bp Bgiii;fligg111 fragment as described in Experimental Procedures (data not shown). DNA Sequence of the nifH 5’ Terminus. The nucleotide sequence of the 5’ terminal region, identified by Southern hybridization and SI 20 .Am. “Haw new .sz -_uQHm .Amv H: awn .Aamv _~Hmm .Amv macaw mam mmu_m «mampuacoucm copaopgammm .mcwp Pao_ugm> cognac mg» mg umaaowucw m. ammamdmmw .m =.=u_z covmmc mcpuoo mHHa we“ mo ego .m mg» mo =o_pauop muaspxocaaa mg» .mmwmdaqum .m =_;»_3 mama mwflm mg» mo copuumgww pacovquLUmcmga can copupmca mg» mmumo_ccw soggm mgh .m:o_mmg omsocmm gnaw amwflmmmmn .m tam mmqmqamuqm MHwawmmmm .fi mgzmmg 1 2 Engage—d moEoEzmca .v. 22 protection analyses described above, was determined in order to verify the presence of nitrogenase reductase coding capacity (Fig. 2). Beginning at nucleotide 202 and proceeding to the end of the sequenced region is an open reading frame encoding 94 amino acids which have 74% homology to the amino terminus of B. gneumoniag nitrogenase reductase (Sundaresan and Ausubel, 1981; Scott gt g1., 1981). This open reading frame can actually be extended four amino acids in the 5’ direction (to the.ATG marked by ?). However, we favor the downstream start site based on homology with the N-terminal ends of gitB from other organisms. In addition, a purine-rich region precedes the ATG suggested to encode the N-terminal methionine. This sequence falls in a position analogous to that designated by Shine and Dalgarno (1975) as the ribosome binding site which precedes translataional initiation in enteric bacteria. DNA Beguengg of the nifDK 5’ Tgrminus, The nifDK operon was isolated previously within a 7.6 kbp HindIII fragment by Hennecke (1981). Southern hybridization experiments with a DNA fragment containing the B. gneumoniae gitB gene as probe allowed the placement of the gitB amino terminus as illustrated in Fig. 3 (data not shown). The nucleotide sequence of this region was determined in order to confirm this placement (Fig. 3). An open reading frame encoding at least 115 amino acids begins at nucleotide 151 and proceeds to the end of the region sequenced. This region has 51% amino acid homology to the B. gngumoniae gitB amino terminal region (Scott gt g1., 1981). As described for the amino terminus of the gitB gene, the initiation ATG codon of gitB is preceded by a purine-rich region in a position appropriate to be the Shine and Dalgarno ribosome binding site 23 Figure 2. DNA sequence of the 5’ end of the B. jagonicum nifH gene. The strategy used in sequencing the 5’ end of the B. jagonicum nifH gene is illustrated in the upper part of the figure. The approximate position and direction of transcriptional initiation are indicated by the arrow labelled "nifH". The sequenced DNA strands are illustrated by arrows below the restriction map. Restriction endonuclease sites are B9111 (Bg), HindIII (H), ngl (0), Hian (Hf), ngl (X), and ngHI (B). The nucleotide and amino acid sequence of this region are presented in the lower portion of the figure. The question mark indicates a second possible translational initiation codon as discussed in the text. 24 nifH --——> HJ DIE Di 89 l Hf Hf l Hf B 1 ‘ IOObpl AGATCTTGTC AGATCCAAAA CAGCCTACGA TCGCGCGCCG GCTGGTTGCT TTTGGAAACG TAATCAGAAG CTTAAGGTGC CGGGTTAGAC CTTGGCACGG CTGTTGCTGA TAAGCGGCAG CAACACTGAG TGAGGGCTGA GTGCACGCCG ACGTGTAAGG CGAGCGATGC GCTCCTTCCC TTGAACCCGT GTGCCCCCGT TTCTGCGAGG GAAGCAAAGC TCGCAAAAGA AGCGCGCAAC GTTTGGCAAA TCGGTTGATG GAGAGCAGC iiE iii EEE 11% SEE EEK Iii 8% Y8 6C iii 55% iii Gtt tit iEE Iti tEE tii iii tit tit ttt iii tti Ski 63E IE8 Iii tit iEi tii t Sit tit tit YC GT tit ttt the tit tti 818 lit Eti tit tit lit tit Iii tit 3t 1 U6 66 9| PC V.. C 66 YT 66 RC ST U6 GG 25 Figure 3. DNA sequence of the 5’ end of the B. jagonicum nifD gene. The strategy used in sequencing the 5’ end of the B. jagonicum nifD gene is illustrated in the upper portion of the figure. The arrow labelled "nifD" indicates the approximate position of transcriptional initiation as well as the direction in which it proceeds. The sequenced DNA strands are illustrated by arrows below the restriction map. Restriction endonuclease sites are B9111 (Bg), Hian (Hf), ngl (D), and tggRI (E). The nucleotide and amino acid sequences for this region are shown in the lower part of the figure. 26 ----> nifD Hf 100 bp- GATTCGCAAC AACAGCCCGT CACCGTACAA GTCGCGCTAA GAAACTGTTG TTGTTCTAGT TTTAGTGCTC ATGAGACCCT GGCATGCCGG TTGCAAAGTC TTGGATCAAG AAGCCGCCCT CCCAACAGCT AACCTTTTAA AGGACACCAG tit tti tit lit 13% tti GEE “st iii tit tit lit Ett Sit tit EEE kit tit iii EEE tit iii I . 27 (1975). The comparison of these two genes would suggest that the B. jagonigum USDA strain 110 ribosome binding site sequence can be approximated by the sequence 5’-ANGGANA-(4 or 5 nucleotides)-ATG. We do not detect any significant secondary structure possibilities around the ATG initiator codon of either of these genes. Promoter Mapping. In order to determine the precise locations of the initiation sites for both git operons, Sl nuclease protection analyses (Berk and Sharp, 1977) were initiated in the region 5’ to the translational initiation sites of gifH and nifDK, respectively. All RNA used in these experiments was prepared from bacteria isolated from within soybean root nodules induced by infection with B. japonicum strain 110 (see Experimental Procedures). A 245 bp Bigtl fragment containing the region upstream from the gitB N-terminal methionine indicated in Fig. 2 as well as the capacity to code for the first 15 amino acids of gitB was 5’-end-labelled as described. The strands were separated electrophoretically and hybridized separately to nodule bacterial RNA. The partially Sl nuclease-resistant mRNA coding DNA strand was run opposite a DNA sequencing ladder of the same DNA strand (Fig. 4a). There are two major protected fragments separated by about 35 nucleotides and designated PHI and PHZ. The positions of these RNA 5’ ends are indicated in Fig. 5 but must be considered to be i 2 nucleotides due to the inherent ambiguity in the SI nuclease digestion. These $1 protection products could correspond to either transcriptional initiation sites or sites of RNA processing or specific degradation. They have both been detected in several independent RNA preparations. However, the relative abundance 28 . .vmuauww=_ mam muavgomcmca am new :8 map .p+u any ecu .p Amy .UA< Amy .w Aqv .mcopuummg m=_u:m:cmm <2: pau.5mso A~-ev “mammpuzc Hm sap: coaummmwu ca seven ammflmeMfi .m m:_x_m- 2 seem <2: o» =o_u-_upgazg seem m:_up=mms z:~mo Amnoan mL mL.. Hmmwv. man a. m:_*doa Amczamxmmma mL mL.. Homuv mam mwsmznmg "o amxaaa~m e:m*1 :oaodomk 2‘”: arm smuox mucosanca aim: usosonmw vN H: maaaaaos. firm R. nsmcaosamm mDLH nosmmzmcm mmncmznmm mamncmmmg ck wm<:o= mL mL. ammwv mam iandcamg woe noaumxawoa. qusmnxacflaozmd flzmfiimwmos aw asaanmflma ck arm mkacod .. zcn‘moeaamm trans mxm soaodococm ”o firm a. humozanca Hwo cecaoamx v N ~10 uxmmmswma mm madmwcmq Emeamxm. arm coxmm azadnmam rammdk nozmm1wawpmc mo mcowmmc mcm new: umucmmmca ww mmocmscmm :owmmc Lmuwm— ILL: AxmzoF ”mmmfi...flm MM “mommy mwwsz< Ezvao~wzc vwcm wmazoo ccw Acmaqsv oflH amwflmmmmw .m mgw mo comwcquou u< .=o_mmc cmcwm— olypeptides of the nitrogenase complex, the synthesis of an active 46 47 nihogmmse in the free-living nitrogen-fixing bacterium B. gneumoniae t al., 1979; remflresthe expression of at least 14 other genes (Reidel Robmts mm Brill, 1980; Roberts and Brill, 1981). These nitrogen fixafion(git) genes are arranged in seven or eight operons clustered mthhia 24-kilobase-pair (kbp) section of the chromosome. The expression of these git genes is coordinately controlled Via the iroducts of the nifLA operon (Buchanon-WollaSton gt 91., 1981; Filser gt 1.,1983; Hill _t _1., 1981; MacNeil and Brill, 1980; Merrick gt 91., 982; Ow and Ausubel, 1983; Roberts and Brill, 1980), which is in turn ontrolled by the general nitrogen regulatory system (Ntr) common to all iteric bacteria studied (Magasanak, 1982). I have begun to characterize the organization and expression of the nes encoding the nitrogen fixation process in the soybean symbiont B. gonicum USDA 110. In this chapter I report the determination of the vsical arrangement of the nifH and nifDK transcription units on the B. ionicum genome. In addition, I have localized a region that is Iologous to another B. pneumoniae git gene, nifB. This B. jagonicum B gene is transcribed from its own promoter and, as with nifH and mg, there is considerable sequence homology between this promoter and consensus for Klebsiella git promoters (Beynon gt 91., 1983). 48 RESULTS Identification and characterization of nif—sgecific cosmid clones. determine whether B. jagonicum USDA 110 git genes are clustered in a nner similar to that seen for other nitrogen-fixing organisms, we have ed DNA sequences from the gitB and gitB genes of B. jagonicum (see 9. 7) as hybridization probes to screen a library of B. jagonicum snomic DNA cloned into the broad-host-range cosmid vector pLAFRl 'riedman gt 91., 1982). This library is maintained as an ordered array ’ 1,426 individual transformants. The average insert size per :combinant cosmid is 24 kbp, therefore, with a genome size of 10,000 1p (T. Caspar and B. K. Chelm, unpublished), any genomic sequence has a '% probability of being represented at least once. In this manner, we 1ve identified two cosmids which hybridize to the gitB probe >RJcos1-62 and pRJcosl4-58) and one cosmid which hybridizes to gitB )RJcosz-63). Ethidium bromide staining patterns for restriction endonuclease igests of DNA from pRJcosl-62 (gitB) and pRJcosZ-63 (gitB) suggested hat these cosmids may represent overlapping regions of the B. japonicum enome. To verify this observation, radioactively-labelled pRJcosZ-63 gitB) was used to probe Southern transfers of various restriction igests of the gitB cosmid pRJcosl—62 (Fig. 8). Since these recombinant osmids were constructed as ngRI partial digestion products (see bove), the full extent of the overlapping region can most easily be een in the ngRI digest of pRJcosI-62 (Fig. 8a and b, lanes 4). There ire four ngRI fragments totalling 13 kbp in length in pRJcosl-62 that beridize to pRJcosZ-63. 49 .wwoca =o_pwwcwwen»; wwww-wwww _wcwcwm w aw we“: aw: Amwmfi ..Hw mm _wwawmv Nwwxwa seem BewEmwcc Hwamuwwmm age w.H mch .sz so: mcmm a mfi mEEwucoo acmsmwt Ha mf mo mcoponzm w 3 mmwma 3.:me mi .wzogcw mgw >2 vmawuvucw mcw .AHmmH .mxomccmzv ~-o~cwma new mmwma Eocw umuwpomw mpcmemwce mnoca msa cwgawz mmcmm mHHm ucw 2m»: ameQQQMfl .m msa so» mmuwm cowww_wwc? chowuawcomcwcm mo mcowfimoa msL .mmnoca :oSwNEIEE mw mm: .3“— vmuwpofl mucmsmwé 333385 553.58; 8.3.6.: 23mm.» umxom .mmnoca 5:35.282 05 \Cwsszm .L mesa: 50 —UJ L—m Umgma Umgmwmbu commow 51 Figure 8. Hybridization of nick-translated pRJcosZ-63 to restrictiort endonuclease digests of pRJcosl-62. (a) Ethidium bromide-stained agarose gel of restriction endonuclease-digested pRJcos 1-62. Lanes : 1, HindIII; 2, HindIII plus ngHI; 3, HindIII plus ngRI; 4, ngRI; 5, ngRI plus ngHI; 6, ngHI. (b) Hybridization of nick-translateci pRJcosZ-63 to cellulose nitrate transfer of the gel in a. Lane designations are the same as in a. The top band in each lane conta1' n5 the vector DNA, pLAFRI. All other bands represent B. jagonicum DNA.. 52 123456 123456 53 The overlapping cosmids pRJcosl-62 and pRJcosZ-63 were used to construct a restriction endonuclease map representing approximately 33‘ kbp of B. japonicum DNA (Fig. 9). The second M cosmid clone (pRJcosl4-58) exhibits an _E_chI restriction pattern that is nearly identical to pRJcosl-62 and thus has not been further analyzed. The positions and directions of w and gitB genes and the small gene "egion of unknown function marked (?) were defined previously within' the LS kbp Hi_ndIII fragment by using the plasmid clone pRJ676 (Fuhrmann and . ennecke, 1982; Hennecke, 1981). The M gene was previously isolated s a recombinant lambda phage, XNHI (Adams and Chelm, 1984) and its isition within this region is shown. The approximate position of a B. mm gene is also illustrated on the map in Fig. 3. The entification and localization of this region is discussed below. To verify that these cosmid clones accurately represent the true 10mic arrangement of the _rfl gene cluster in B. jagonicum, total omic DNA was digested with the restriction enzymes used to map this ion, separated on agarose gels, and blotted to cellulose nitrate. resulting blots were hybridized to either the 1.2 kbp Xh_oI fragment I pBJ86 (Fig. 10a) or to pBJ87 (Fig. 10b), each of which had been olabelled by nick translation (see Fig. 9 for map positions of as). In each case the hybridization pattern predicted from the map I in Fig. 3 was in agreement with the results. This, coupled with act that B. jagonicum DNA regions that were independently cloned in and pRJ676 are unaltered in these cosmid clones, indicates that :1-62 and pRJcosZ-63 accurately reflect the physical structure of .j apon i cum genome . 54 .2: as; we... .3: Sex .sz Humuaflm .Amv _zmwm mew mmu'm mmwmpuneonem eoppuveummm .souaoe me» uw nmaweamzrpv mew uxmu me» e_ nmmmeuwwn mew new xeoz mmea e. nmme memz awe» eowmme m_eu mo mmeopu <20 uewevesoumz .wwwa nmppmewp zonew nmeuuwe me» x: nmuwo_new m, meme nmeFLUmmn memm mwflm me» we eo_eomewn Pweowaeweomewe» new eoru_moe meL .mzoeew nmeouwe me» he nmawu_ne_ mew Ammmfi ..HM MM eewseezm ”Hmmfi .mxomeem: “Nmmfi .mxomeem: new eewseezm memmfi .Epmeu new menmee nmeweommn mmemm mflw new Hflm mo meowuomLPn FweopueLLUmeweu new meo—u_moe meL .emamepo memm Hfld ammflmmmMfi m me» mo awe mmwm—ozeonem eowauweummz .m mesmmm 55 m ran T|1 I w ImI m m: I II w I m w I _ _ = _ _ _ _ _ u. _ _ _ _ _ _ _ _ _ __ _ __ __ _ _ _ _ _ __ _ m x mx m xx m mmkmx m m x m m xm x I§Oeiw 21 :fiw n :fix:fio nmeoomeimm L UILOOmMIaw . >ZII r L nwemw owemm UILmnm [ _ . _ _ _ nmemn omemumbe [IL .11 56 Figure 10. Verification of the B. japonicum git genomic structure by Southern hybridization to restriction endonuclease digests of total genomic DNA. Genomic DNA from B. japonicum was digested with restriction endonucleases and separated electrophoretically on 1% agarose gels. Cellulose nitrate transfers of these gels were then hybridized to radioactively labelled (a) pBJ87 or (b) the 1.2 kbp B991 fragment from pBJ86 (Fig. 3). (a) Lanes: 1, ngHI; 2, B99RI; 3, HindIII. (b) Lanes: 1, B991; 2, HindIII; 3, B99RI; 4, ngHI. 57 1 2 3 4 kbp kbp 8.6 _ .‘ _8.6 4. _ Q 9 _4.7 3.9 _ O _3.7 _2.5 _1.4 1.2 _ o a _ 1'. 58 Localization of a B. japonicum nifB gene. In all species of Rhizobium examined thus far, at least some of the genes responsible for nitrogen fixation (including the nifHDK operon) and nodulation of host legumes are closely linked on large plasmids (Banfalvi gt 91., 1981; Hirsch gt 91., 1980). In B. meliloti, the gitBBB operon and at least two other transcription units essential for nitrogen fixation are found within 15 kbp of one another (Buikema gt 91., 1983; Corbin gt 91., 1982; Corbin gt 91., 1983; Ruvkin gt 91., 1982). One of these transcription units encodes a positive regulator of symbiotic nitrogen fixation that is similar in both structure and function to the B. gneumoniae gitB gene product (Zimmerman gt 91., 1983; Buikema gt 91., 1986). We were interested in determining whether a similar nifA-like gene is closely linked to the B. japonicum gitB, B, and B gene region. With this in mind, a 1.8 kbp ngI-ngl fragment from pGR397 (see Fig. 7, Chen gt 91., 1982) which carries most of the B. pneumoniae gitB gene and a portion of the amino-terminus of gitB was used to probe Southern transfers of B99RI digested pRJcosl-62 and pRJcosZ-63. This probe hybridizes predominantly to the 4.9 kbp B99RI fragment common to both pRJcosl-62 and pRJcosZ-63 that is located approximately 11 kbp downstream of the nifDK promoter (Fig. 11). Subsequent analysis (not shown) of this nifAB hybridization Ias localized the region of homology to the 1.2 kbp Bail-59H fragment :hown in Fig. 9. The nucleotide sequence of the B. japonicum DNA that hybridizes to he B. pneumoniae nifAB probe was determined in order to confirm the resence of nifA- or nifB genes in this region. Although no homology to he K. pneumoniae nifA gene was observed, an open reading frame xtending for at least 193 amino acids and with 49% amino acid homology 59 Figure 11. Hybridization of the B. pneumoniae nifAB genes to B99RI digested recombinant cosmids from the git gene cluster. (a) Ethidium bromide-stained agarose gel of B99RI~digested cosmid clones. Lanes: 1, pRJcosl-62; 2, pRJcosZ-63. (b) The ngI-ngl fragment of pGR397 (Reidel gt 91., 1983), that contains portions of the B. pneumoniae nifA and nifB genes, was nick-translated and hybridized to a cellulose nitrate transfer of the gel shown in a. Lane designations are the same as in a. 60 .. . 61 Figure 12. Nucleotide sequence of the 5’ end of the B. japonicum nifB gene. A DNA and amino acid sequence for the N~terminal portion of a nifB open reading frame from B. japonicum is presented. Amino acid homologies to the B. leguminosarum nifB gene are indicated by *. The approximate transcriptional inititiation site as determined by S1 nuclease analyses (Fig. 7) is indicated by the arrow. Nucleotide sequences in this region with homology to the B. japonicum nifH promoter (Chapter 2) are underlined. 62 GATCGGGCCA ATTGAGAGCG GCAGCGCGCG AAAGCCCGTT CAGATGAGCT GGATCGGGGC GAGCCCCGCG AGAGTGAAAA GCACACGAGA TGGAGCCCCG AGTGAGAATG GCGACGCATC TGTCGCTTTG CCAACAAGCT CATCATTGCT GTCTTGCAAG TGCCAACGCT ATCGCTGCGC ~~~~~~ > CTTQCTGAAG CGCGCTCTAG GATGATCTGT GGGCGTCGAT GCTTGCAGGG GAGTGATCTC M Q GGTCGTGGAG CGCGGAAAAT ATATCAAAGC AGCGGTCAAT AGCGGGAAGA TC ATG CAG s 1 T E H K c c R A s A K T c R A ch ATA ACC GAG CAT AAG ccc ch ccc GCT ch cco AAG ACC can can ace 5 c e s o A c R c o L P 'v E 1‘ R E AGC ch ccc TCG CAG ccc coo CGA occ GAT CTG ccc GTC GAA ATC AGG GAA * R V K N H P C Y S E D A H H H Y A AGG GTG AAA AAC CAT CCC TGT TAC AGC GAG GAT GCG CAC CAT CAT TAC GCT * * * * * * * * * * * * 'k * R M H V A V A P A C N I Q C N Y C CGC ATG CAT GTC GCG GTC GCA CCT GCC TGC AAT ATC CAG TGC AAC TAC TGC * * * * * * * * * * * * N R K Y D C A N E S R P G V V S E AAC CGA AAA TAC GAC TGC GCC AAT GAA TCG CGT CCG GGT GTG GTG AGC GAG * K L T P E Q A V R K V I A V A T T AAG CTC ACC CCT GAG CAG GCA GTG AGA AAA GTG ATC GCG GTC GCG ACG ACC * * * * * * I P 0 M T V L G I A G P G D A L A ATT CCG CAG ATG ACG GTA CTT GGC ATC GCT GGT CCC GGC GAT GCC CTG GCC' * 'k 'k 'k * N ' P A K T F K T L A L V T E A A P AAT CCA GCA AAG ACG TTC AAA ACG CTC GCG TTG GTC ACC GAG GCT GCT CCT * * * * * * * * * * * * * * . D I K L C L S T N G L A L P D V V GAC ATC AAG CTG TGT CTG TCA ACC AAC GGA CTA GCG CTG CCA GAC TAT GTC * * 'k * * * * * * * D T I V R A K V D H V T I T I N M GAT ACC ATC GTG AGG GCC AAA GTT GAC CAC GTC ACC ATC ACC ATC AAC ATG * * * * * * * * * 'k * * * * V D P E I G A K I Y P W I F F N H GTC GAT CCT GAA ATC GGA GCC AAG ATT TAT CCA TGG ATC TTC TTC AAC CAC * * K R Y T AAG CGA TAC AGG 63 to the B. leguminosarum gitB gene (Rossen gt 91., 1984) was identified (Fig. 12). As described for the amino termini of the B. japonicum gitB and gitB genes (Chap. 2), the proposed gitB initiation ATG codon is preceded by a purine rich region in a position appropriate to be the Shine and Dalgarno ribosome binding site (Shine and Dalgarno, 1975) suggesting that this is indeed the translational start for this gene. nifB Promoter Mapping. In order to determine the precise transcriptional initiation site for the gitB gene, SI nuclease protection analyses (Berk and Sharp, 1977) were initiated in the region 5’ to the proposed translational start for the gitB open reading frame. A 302 bp ngI-Aigl fragment that contains the first 16 amino acids of gitB coding sequence as well as about 250 bp upstream of the N~terminal methionine was 5’ end-labelled using T4 polynucleotide kinase. The DNA strands were separated electrophoretically and hybridized separately to bacterial RNA from soybean root nodules. As expected from the DNA sequence, only the ngl 5’ end-labelled strand showed partial Sl nuclease-resistance confirming that this is the mRNA coding strand. To precisely define the gitB transcriptional initiation site, the protected DNA strand was separated electrophoretically next to a DNA sequencing ladder of the same DNA strand (Fig. 13). A single group of partially protected fragments was obServed approximately 105 (i 2) bp upstream from the putative gitB translational initiation site indicating that a gitB RNA 5’-end is located at the position shown in Fig. 6. We tentatively consider this to be the sole initiation site for nifB transcription in root nodules. As with nifH and nifDK (Chap. 2), the promoter region for this transcription unit shares extensive DNA 64 Figure 13. Identification of the B. japonicum nifB promoter sequence. 51 protection and DNA sequence analysis was carried out using a 302 bp ngI-AigI fragment containing the 5’ end of the B. japonicum nifB gene. Only the DNA strand which hybridizes to bacterial RNA isolated from soybean root nodules is shown. Lane designations are as follows: (1) C+T, (2) T, (3) A>C, (4) G, (5-7) S1 protection analysis using either 2 :9 (5), 1 ug (6), or 0.5 ug (7) of RNA isolated from soybean nodule acteria. 65 1234567 66 sequence homology in the ~10 to ~27 region with other B. a onicum _it promoters (Adams and Chelm, 1984; Alvarez-Morales gt 91., 1985) and with git promoters from other nitrogen fixing species (Beynon gt 91., 1983; Ow gt 91., 1983). DISCUSSION Unlike many other diazotrophs where the constituent polypeptides of nitrogenase are encoded by a single operon in the order nifHDK (Reidel gt 91., 1979; Banfalvi gt 91., 1981; Krol gt 91., 1982), these genes occur as two separate transcription units, gitB and gitBB, in B. j9ponjcug strain USDA 110 (Kaluza gt 91., 1983; Chap. 2). We have isolated overlapping cosmid clones of B. japonicum genomic DNA carrying either the nifB or nifDK genes. Using these clones, a partial restriction endonuclease map representing approximately 33 kbp of the B. jappnjgum genome has been constructed. The transcriptional initiation sites for the nifH and nifDK genes, which have been mapped previously (Chap. 2), are separated by approximately 20 kbp, and all three nitrogenase genes are transcribed in the same direction (Fig. 9). About 13 kbp downstream from the gitBB Operon we have identified an open reading frame which has extensive homology to the gitB genes of B. pngumogiae and B. leguminosarum. This gene is also transcribed in the same direction as gitB, B, and B. Mutational analyses of-the B. pneumoniae and B. leguminosarum gitB genes (Roberts and Brill, 1981; Rossen gt 91., 1984) indicates that their products are required for synthesis of the MoFe cofactor of nitrogenase component 1. Although the 67 exact function of the gitB product is unknown, analysis of the predicted amino acid sequence of the B. leguminosarum gitB gene (Rossen gt 91., 1984) indicated the presence of clustered cysteine residues in the N~terminal region of the protein. This data has been used to suggest that the gitB gene product functions in the binding of [Fexzsx] clusters (Rossen gt 91., 1984). We have observed similar groupings of cysteine residues in the N~terminus of the B. japonigum gitB gene (Fig. 12). Using 51 protection analyses, we have mapped a putative transcriptional inititiation site for the B. japonicum gitB gene. As expected, the sequences 5’ to this initiation site share several blocks of highly conseved sequence with the major gitB and gitB promoters described in chapter 2. A genetic analysis of the role these sequences play in the coordinate control of this gene set will be useful in understanding the symbiotic control of git gene expression. EXPERIMENTAL PROCEDURES Bagterial strains and media. The Bscherichig coli K~12 strain E08654 (991 ggt gggBk 999B 999B) was used for general plasmid cloning and maintenance as wgll as for maintenance of the B. japonicum genomic DNA library cloned into the broad-host-range cosmid cloning vehicle‘ pLAFRl (Friedman gt 91., 1982). B. japonicum USDA 110 was grown in YEX medium (0.04% yeast extract, 0.3% xylose, 3 mM KZHPO4, 0.8 mM MgSO4, 1.1 mM NaCl). 68 DNA teghnigugs. Genomic DNA from B. japonicum was purified by phenol extraction (Marmur and Doty, 1962). Relaxed replicon plasmid DNA was isolated by the method of Clewell and Helinski (1972). Isolation of plasmid DNA from the low-copy-number cosmid clone band was modified as described by Friedman gt 91. (Friedman gt 91., 1982). Isolation of DNA restriction endonuclease fragments for use as probes was as described previously (Adams and Chelm, 1984). All other restriction endonuclease mapping and enzymatic cloning techniques were standard (Maniatis gt 91., 1982). S1 nuclease protection analysis was done as described (Adams and Chelm, 1984). DNA sequencing was by the methods of Maxam and Gilbert (1981). All sequences were determined using both DNA strands and all restriction sites were overlapped. Congtrugtion and maintenance of a B. japonicum cosmid gene bank. Total DNA from B. jappnigum 110 was partially digested with B99RI and subjected to centrifugation for 8 hr at 4°C and 155,000 x gav through a 12-ml 5 to 20% (wt/vol) sucrose gradient dissolved in l M sodium acetate, 10 mM Tris-hydrochloride, l‘mM EDTA, 0.01% sodium lauryl sarcosinate (pH 8). Fractions from this gradient that contained DNA molecules greater than 15 kbp in size were pooled and used for construction of a gene bank in the broad-host-range cosmid vector pLAFRl (Friedman gt 91., 1982). B99R1~cleaved, phosphatase-treated vector DNA was ligated to the partially digested B. japonicum DNA at a molar ratio of 5:1 (vector/insert) as described by Maniatis gt 91. (1982). After t al., ligation, DNA was packaged as described previously (Maniatis 1982) and used to transduce B. coli K-12,ED8654 to tetracycline 69 resistance. Transductants were replicated in an ordered array to both agar stabs and liquid medium in polycarbonate microtiter dishes. For screening, the cosmid-containing isolates were transferred with the aid of a steel prong replicator to sheets of cellulose nitrate lying on a dish of agar-solidified medium and then allowed to grow to colonies. Hybridizgtjon procedures. B.coli E08654 colonies harboring the ordered cosmid library were lysed, and their DNA was bound to nitrocellulose filters as described by Grunstein and Hogness (1975). Before hybridization, filters were incubated for l h at 65°C in 5 x Denhardt solution (Denhardt, 1966), 5 x SSPE (1 x SSPE is 0.18 M NaCl, 10 mM NaPO4 [pH 7.7], 1 mM EDTA), and 200 ug of sheared and denatured salmon sperm DNA per ml. Y32P~labelled hybridization probes were prepared by nick translation (Maniatis gt 91., I975). Hybridization ,reactions were carried out for 12 to 24 h at 65°C in 5 x SSPE, 1.5 x Denhardt solution, and 100 ug of salmon sperm DNA per ml. After hybridization, filters were washed twice for 15 min each at room temperature in-2 x SSPE, 1% sodium dodecyl sulfate followed by two washes in 0.1 x SSPE, 0.1% sodium dodecyl sulfate. Hybridization signals were detected by autoradiography. CHAPTER 4 Bradyrhizobium japonicum Genes With Sequence Homology to the Klebsiella pneumoniae nifA Gene INTRODUCTION The free-living nitrogen-fixing bacterium Klebsiell9 pneumoniae contains a cluster of at least 17 genes arranged in 7 or 8 operons which are required for nitrogen fixation (git genes; Reidel gt 91., 1979; Roberts and Brill, 1981). This complex regulon is controlled by two separate systems. One links nitrogen fixation to the general nitrogen control pathway (Ntr) known for several enteric bacteria (Magasanik, 1982). The second is specific for git gene control (Dixon, 1984). The git-specific aspect of this regulation requires the gitB gene product (NIFA). NIFA works in conjunction with the gtgA gene product (NTRA), a putative sigma factor for recognition of gtg~activated . promoters (Hirschman gt 91., 1985; Hunt and Magasanik, 1985), to turn on transcription from other git operons (Dixon gt 91, 1980; Merrick, 1983; Ow and Ausubel, 1983). The gitB gene itself is the site at which the more general Ntr system acts to control git expression. Under nitrogen limiting conditions, transcription of the nifA gene is activated through the concerted action of NTRA and the product of another gene, ntrC 70 71 (NTRC; Ow and Ausubel, 1983; Dixon, 1984). Under these conditions NTRC and NTRA are also necessary for the activation of genes involved in nitrogen assimilation including those for glutamine synthetase, histidine utilization, proline utilization, and arginine utilization (Magasanik, 1982). In addition to their mutual requirement of NTRA for activity, NIFA and NTRC are structurally and functionally related.. The amino acid sequences for both proteins are homologous throughout a large central region (see Fig. 19; Buikema gt 91., 1985; Drummond gt 91., 1986; Nixon gt 91., 1986; Gussin gt 91., In Press). Also, the promoters for genes under NIFA or NTRC control share a common consensus sequence (Beynon gt 91., I983; 0w gt 91., 1983). Under certain conditions, some of these promoters can be activated by either regulatory system (Ow and Ausubel, 1983; Merrick, 1983; Buck gt 91., 1985). For Bragyrhigobium japonicum, the soybean symbiont, several genes required for nitrogen fixation have been localized in two distinct clusters (Adams gt 91., 1984; Fuhrmann gt 91., 1985). Promoter sequences for these genes share a high degree of homology with those identified for NIFA/NTRA and NTRC/NTRA regulated promoters in B. pggggpgigg (Adams and Chelm, 1984; Fuhrmann gt 91., 1985; Chap. 2). In addition, when overexpressed in a heterologous B. 9911 system, the B. pggpg9gigg gitB gene product is able to activate transcription from the ,B. japonicum nifH and nifDK promoters (Alvarez-Morales and Hennecke, 1985; Alvarez-Morales gt al., 1986), supporting the hypothesis that B. ,igppgiggg git genes may be controlled by similar regulatory mechanisms. With this in mind, we have used the B. pneumoniae gitB gene to screen libraries of B. 199991999 genomic DNA. In this chapter, I report on the 72 isolation and structural characterization of four separate genes from B. japonicum with sequence homology to B pneumoniae nifA. RESULTS Identification and cloning of B. japonicum nifA-like regions. A 950 bp BgtI~B99RV DNA fragment which is entirely internal to the B. pneumoniae gitB gene (see Chap. 3, Fig. 7; Buikema gt 91., 1985) was radioactively labelled and hybridized to Southern transfers of B99RI digested B. japonicum genomic DNA. As shown in Figure 14, B. japonicum contains several B99RI fragments with homology to the gitB gene of B. pneumoniae. Rhizobium meliloti (Fig. 14, lanes 2 and 3) and Bscherichia 9911 (Fig. 14, lanes 4 and 5) also contain multiple restriction fragments with homology to gitB. In order to isolate the B. japonicum gitB-homologous regions a phage library of B. japonicum genomic DNA was screened with the gitB-specific hybridization probe described. To reduce the background observed from hybridization to the host chromosome, phage were plated on an B. 9911 deletion strain which lacks the single known gitB-like B. 9911 gene, gttt (Fisher gt 91., 1981). Several recombinant phage with gitB-like B. japonicum sequences were isolated. These phage were then used to isolate larger genomic DNA regions by directly screening the ordered cosmid library described previously (Adams gt 91., 1984). Cosmids identified in this manner were isolated and grouped into four distinct classes based on restriction site patterns. gitB-specific hybridization to Southern transfers of B99RI digested DNA from cosmids 73 Figure 14. nifA-homologous sequences in B1 japonicum, B. meliloti, and B. pneumoniae. Southern transfers of total cellular B1 japonicug DNA digested with B99RI (lane 1); B. meliloti DNA digested with HindIII (lane 2) or B99RI (lane 3); B. coli DNA digested with HindIII (lane 4) or B99RI (lane 5); and cosmid clones of B. japonicum DNA, pRJcosZ-43 (lane 6), pRJcosl-62 (lane 7), pRJcos7-36 (lane 8), pRJcosl4-28 (lane 9), or pRJcos3-44 (lane 10) were hybridized with a probe specific for internal sequences from the nifA gene of B1 pneumoniae as described in the Experimental Procedures. 74 78910 6 vi . Jm. . .,.. i . V 9.693 3.... ,M. e, :01... a 75 representing each of these classes is shown in Fig. 14 (lanes 6, 8, 9, and 10). Restriction endonuclease maps for each of the gitB-homologous regions (designated 9992-9999 for 9omology to gifA) are shown in Fig. 15. By inspection of the restriction map for the region labelled 9991, this gene was found to be linked with another gene required for nitrogen fixation termed tiBA (Fig. 15; Fuhrmann gt 91., 1985). This localization of tiBA was confirmed by DNA sequence analyses (see below). This DNA region also contains several genes required by B. japonicum for nodulation (999 genes) of soybean plants (Lamb gt 91., 1986). None of the other gitB-like regions are closely linked to any B. japonicum genes that have been described. DNA geguence analyses. To determine whether the B. japonicum regions descussed above have the capacity to ehcode NIFA—like proteins, partial DNA sequences for three of the four gitB—like regions have been ‘ determined (Figs. 16-18). In each case, open reading frames with extensive amino acid sequence homology to the B. pneumoniae gitB and gth (Buikema gt_91., 1985) genes were found (Fig. 19). This homology is most apparent through the central domain of the Klgbgiella genes. No sequence data is available for the gitB-like cloned region 19999). However, this region does hybridize to each of the other three B. japonicum regions as well as to nifA (not shown). Construction and characterization of B. japonicum 999 deletion strains. To determine whether any of the NIFA-like regions described above are involved in regulating nitrogen fixation or nitrogen assimilation genes, we have constructed four B. japonicum mutant strains 76 Figure 15. Genomic restriction endonuclease maps for B. japonicum strain BJ110 LifA- ~homologous genes Lna2 (A), Lna3 (B), Lna4 (C), and hLa5 (0). Also shown are genomic _maps for B. japonicum hLa2, hLa3, hna4, and Lna5 deletion strains BJ1011, BJ2221, BJ2101, and BJ2111 respectively. In each case, a portion of the LifA- like gene region was deleted and replaced with the nptII gene. Arrows indicate the orientation and approximate position of each nifA-like gene region. The solid line portion of each arrow indicates the gene region that has been sequenced (Figures 16-18). Dashed lines indicate approximate dimensions for each 999 gene assuming that their size and structure are like that of the B. pneumoniae nifA gene. For hna5 (0), the dashed lines represent the DNA region that hybridizes to the B. pneumoniae gitB gene as well as to other B. japonicum 999 regions. The position and orientation of fixA within the hna4 genomic region (C) has been described previously (Fuhrmann gt 91., 1985). Restriction sites indicated are H, HindIII; C, 9191; S, 9911; B, B99HI; E, B99RI; X, B991; and 89, B9111. BJ 110 BJ1011 ’77 n m m n4 1 2 J IV 3 8 HI. HJ BIL BIL EJ\1\ m o. ..... $1 Bl B Q 0.. El EL fixA :2) I) h_n_a_‘1« BJ110 XI‘ El X1 XII SL Hal. Ell BJ 2101 E l BJ110 El EaL a; u....:a.o<¢¢4 r . ‘ ‘| “ “ ‘ BJ 2111 61 121 241 301 361 421 481 601 661 721 781 841 Figure 16. m1- TGACTAGGCT CAATGCTGGT GTGCCGGGCA ACGGCATTGG TTTCCGAGGG GCAGCATCAT TCGCGCTGGC TCGAAAGAAT ACGGCGTGGT CTCGCTGGCG GTCGCGAAGG AGCGGATCGA CGAAGCCGCG TGTCGGCAAA TTGAGCGGCG AGGAACTGTT TTGCCTTCAA CGCCGGGCGC ATGGCGGCGT TGCTCCGTGT TGGACGTTCG GTTTCCGCAA GCGATCGGCT TCTACAGCAC CCTGGCCGGG TCACCGGTAT AACCTCCCTC TCGCGCGCTT GCAATATTGC TGCACCAATA AACAAAAGGG GCGCCGACCG CGACATTTCA CATCCGCTAC AGGGTATCGG CTGCGCGGTT GCAACACTTC CCAGCGCCCA CATCACCGAC CGGTGTTCGC CACGGCGCTC CATCAAGGCC AGGCATCATC GGTGATCGCG ACGCCTGCTC GATCCTGGAT GCGGATCACG CCTGTTGGGC GCCACTGATC AACCCATCCT GGAAGCGATC GGCCCTCGGC CGCCCGCCTG TTGCGGTGCG TTGTACACCG GCTCAGCTTA GCTGGAAGAG GCTGGTGGCG GGATCTGTAC GGGCGATGTC CCAGCCGCTG AAACGTTCGC CGTCACTCCC AATCTCCGTG TGACGAGATG CGCTGCCGCG CCGTTCCCGG CGGGGAAGGT CGACGTCACT TCATGTTCAT TGCGCCCTTT CTTGCGGCAT GCTGAGGGTC GTATCCGATG 78 GATTTCCGCC GCGTTGCGGA AGTACTTGGC AGGAATGGTG TCGTGCTTGG TGTCGGGGGC CTCGAAGTGC GCCGCCAGGG AGAGACGGTG GTCACAGGAA AGCCGGTTCA TGTCCACGCC TGGACCTGCG TCGGTAGCCC GATTTTTCCG GCCCCCAGGC GCCAATCATA TTGAGCTTGC GAGGCAAGCA TTTGTCGAAT CGGTTTGGTC GGGTCGTCCA CAGTCTCGTG AGTTGAGCAT GAAGACTTGG CCGAGGGGTT GTTGACGGGG TCGTGAAGGG GCTGCATCCG AGATGTCGCC GTCGGATGCA GCGAGGCACT CGGACCGCGG TCCTGCTCGA GTTCATGCCG CCAGTCTGAA GTGTCGAAAG AGTTGCTCGG GCTACGCGCG AGGGAAGGCC GACGAGATCG GCGAGATGCC CGCGCGGTGT ACAGACTGGG TCGACCAATC GGAATTTGAA TTCCGGATCG GCGTCGTCAA GAGATCCTGA TCGATCACTT CGGTTCGAGG CGGCCACTAT GAGCTGCGTA ATCTGATTGA AGGGACCTGC CAGAGGATTT CAGTCGGAGA ATAGCGCGAC GAACGTCGTG CCATCGAGAG GAAAAGCTGG GGCTTTCGCG ACCTAGCGAT CGGGGGCTGG CCGGCCCCCG CGGCGACGCT TTCGTGACGC GCCTGGTTGG CGAGACCGAG CTCGCAAAGC CGTCGAACGC ATGCACGCGA CAGTTCGGCC TCCAGACCAT GATTTGAGGG ACTCGCTGCA CCCGGGGCCT CCAGCGGATC CATTCGTAGC GATCTTTCAT GCTCTCGGAG GCCAGGGATG TCACAACGAT TGGGTCACAG GCCCGAGGAG CGCAATGCTC GACAGCCCGC GCTGGACGTG GGGCGAGACG GACGGGCAAG CGGCGAATTG CGGGCGGTTC GATGGACCTG TGACAGCAAG GCAGGAAGTG GTTCACCATC GATGCGGCCA GATCTGAGGA ACCGGCACCA GCGGAGCATT CCGATCGACG CGTCACAATG AAGATTCGGC GGCAGTGCCG ATGGCGTCGG ATCCTTCCAA CTGCGGGAGC ‘ GTGCTTGCTT ACTGCCCTGA GCGCAGAAAG GGCGTCGGCA GAGCCGTTCG TTCGGGCATG GAATTTGCCA CAACCCTACC GCGCGCCCGG GCCGAGGGAC CCGCCGCTCC CAACAGGCAA TTCGCCACGA GGAGTTGCTG CGCCGCTATT GAACCTCGTG CTGATGACGA CGTGGAGGTG GCGCCGGCTC GGGATCGGAG AGTGACGAGA AGCCGTCGCA AACGCCGGCG CGCTACGCTG TACCGCAAGT CATGGCGGCT GTCTGGGATG GGCCTCGCAG ATCCTTGCGC CGAAAGCAAG CTCGCTGGTG GCGGCATGAG CTACAGCGTA GCTGTCGGAT TTCGTATTGC CGAGACGATG GCGGGCGATC ATCTCACGTT AATGACGCGA CTGGATCAAA TCGAATCGAG CGCATGAGTG TCGATCCGGA AGTCCTTGTC GAAATGCTGA AGGAGCGGTT GTTCGTCGTT CAGCAGATAT CGGCCGCTCA ATCGTGGAAC CTTCTGAACC GTCTGCTTGC TGGT Nucleotide sequence of the B. japonicum nifA-like region A1 A61 A121 A181 A241 81 861 8121 8181 8241 B301 B361 79 GTCGACGTCA CGATCTATGG TGAGACCGGA ACCGGCAAGG ACGTCGTCGC GCGCTGCCTG CACAATCACA GCGGCCGCCG CGGCGAGCAT TATGTCGCGG TGAATTGCGG CGGTCTGCCG GAATCGCTGG TCGAGAGCGA GCTGTTCGGC CACGAGGTGG GCGCGTTCAC CGGCGCCACC CTGTTCGGCC ACGAGGTGGG CGCGTTCACC GGCGCCACCC GTCAGCGGAT CGGAAAGATC GAATACGCCA GTGGCGGAAC GCTATTCCTC GACGAGATCG AGAGCGTGCC GCTGA CTGAGCGAAA AGCGTCTGTT TCGTGCCGAT CTGTACTATC GCCTCGGCGT CGCTTTCATC GAGCTCCCGC CGTTGCGCGA ACGGCGGGAG AATATCCCGC TGCTGTTCGA GCACTTCACC CTGGATGCGG CGAGGCGGTT CGAGCGTGAC GCCCCGATTC TGGACGAACA GACCATGTCT CACCTGCTGG CCTATTCGTG GCCGGGCAAC GTACGCGAGC TTCGCAACGT CGCCGACCGT TTCGTGCTCG GTGTCCTCGA CAGCAAGGCG GTCAATAAAT CCGGTTCGGT ACGTCGGATG TGTCATTGCC GCGGCAACTC GAGCGATCAT CGAGGACCGC TCCGGCGCAA GCAGGGCGAC GTCCAGGCAA CGAGCCGCGC ATCTCA Figure 17. Nucleotide sequence of the B. japonicum nifA—like region hna3. The two blocks (A and B) of sequence shown are from separate regions of this nifA-like gene. ' 61 121 181 241 301 361 421 481 541 601 661 721 781 80 CTCGAGTACG ATGCGCGGCT GCTCGCCATG GTCGCGAACG TGATAGGACA AACGATCAAG TTACATCGCT TGTTCGCCGG CGATCGCGAA CAATCGTTGG TGGACAAGGA CCGGCTAGAG AAACAGACAG TTGATCGTGG GCCGCCCGCT CGCGAGCGCA AGCAGCTTCA GGCACACGGG ATCATTGGCG ACAGCCCGGC GCTGAGCGCA CTGCTTGAGA AGATTGTCGT TGTAGCCAGA TCAAACAGCA CGGTTCTGCT GCGTGGCGAA TCCGGTACCG GGAAGGAGCT GGTAGCCAAG GCCATTCACG AGTCGTCCGT TCGTGTAAGC GGCCGTTCCG TTAAGCTGAA TTGCGCGGCG CTCCCCGAGA CGGTCCTGGA ATCGGAATTG TTTGGCCATG AGAAAGGAGC CTTTACCGGT GCTGTCAGGC GCCCGCAAGG GCGCTTCGAG CTTGCTGACA AAGGGACGCT ATTTCTTGAC GAGATCGGAG AGATCTCACC TCCGTTCCAG GCGAAGTTGC TGCGAGTTCT GCAAGAGCAG GAGTTCGAGC GCGTCGGCAG CAATCACACG ACAAAGTCGA TGTTCGGGTG ATAGCTGCGA CCAACAGGAA CCTTGAAGAG GCTGTGGCAA GGAGCGAATT CCGCGCGGAC CTCTACTATC GTATTAGCGT AGTTCCCTTG TTGTTGCCGC CGCTTCGCGA AAGACGCAGT GATATTCCGC TGCTCGCAAG AGAGTTCCTC AGAAAGTTTA ACAGCGAGAA CGGCCGCTCT CTTACTCTGG AGGCGAGTGC GATCGATGTA CTGATGAGCT GTAAATTTCC GGGAAATGTC CGCG Figure 18. Partial nucleotide sequence of the 5. jagonicum nifA-homologous region hna4. 81 Figure 19. Comparison of derived amino acid sequences from the K. pneumoniae (Kp) nifA and ntrC (Buikema gt al., 1985; Drummond gt al., 1986) genes with amino acid sequences for hna2, hna3, and hna4 from B. japonicum (Bj). Sequences are aligned to maximize homology with the K. pneumoniae nifA gene. Amino acids homologies to this gene are indicated by underlining. 82 MIHKSDSDTTVRRFDLSQQFTAMQRISVVLSRATEASKTLQEVLSVLHNDAFMQHGMICLYD ...TVTGTNGIGIALVTGKPVHVHAAEHF§EGIKAWICVGSPIRSPIDGSIIGIIDFSQPQAIFH S EILSIEAL TED TLPGST IRYRPGEGLVGTVLA G SLVLPRVADD RFLDRLSL MQRGIAWIVDDDSSLBWVLERALTQAGLSCTTFESGNEMLQALTTKTPDVL RHNVALAVIAANHIELALSEKIRLERIBLLEASICRMPGMQSADGVVILDRFGBVVHHNDM YDYDLPFIAVPLMGPHSRPIGVLAAHAMAROEERLPACTRFLETVANLIAQTIRLMILPTS LSDIRMPGMDGLALLKQIKQRHPMLPVIIMTAHSDLDAAVSAYQQGAFDYLPKPFDlDEAV ASARWRRLTQSRELSIGSQLLPSREGLLGEDLAEGLPEELREQRIEPLLVDGVVKGAMLVL .............................. LEYDVRLLAMVANVIGQTLKLHRLFAGDREQ AA AP SPRIERPRACTPSRGFGLENMVGKSPAMR IMDIIR VSRWDTTVLVRGESGTG ALVDRAI§HYQ§QQQPRNAPINSPTADIIQEREAMQDVFRLLGRL§BSSISMLINGESGTG ASKPRTHPAASEMSPTARTALMSAKEAIM§C§EALLDVAQKVERRALGRIAMLLEQETQVQ ................................................. VDMTIYQLTQIQ SLVDKDRLETVDRGPPARERKQLQAHGII§D§EALSALLEKlVVMABSNSTVLLRGESGTG KELIANAIHHNSPRA-AAAFVKFNCAALPDNLLESELFGHEKGAFTGAVRQRK-GRFELADG KELVAHALHRHSPRA-KAPfiIALNMAAIPKDLIESELFGHEKGAFTGANTVBQ-GRFEQAQ§ EELFARLVHAA§LKTGKEPFVAFNCGAVSKELLGGELFGHAPQACIPATBEGRPGRFEFANQ fiDVVARLLflNH§GBR~GEHYMAVEQGGLEESLVESELFGHEVGAFTGATBQBI-§KI§YAS§ GTLFLDEIGESSASFQAKLLRILQEGEMERVGGDETLRV-NVRIIAATNRHLEEEVRLGHFRE GTLFLDEIGDMPLDVQTRLLBVLADQQFYRVGGYAPVKM-DVRIIAATHQNLELRMQEQKEBE QVLSLDEIGEMPMDLQPYLLRVLELRAVYBLQDSKARP1-DMBLVASIEBNLKQEMAEQREBK GTLFLDEIA§§VPL ........................................ LSEKRLEBA GTLFLDEIGEI§PPFQAKLLRVLQ§Q§FERVGSNHHTSKVDVRIIAATNRNLEEAMARSEEBA DLYYRLNVMPIALPPLRERQEDIAELA-HFLVRKIAHSQGRTLRISDGAIRLLMEYSWPGNV QLFHRLNVIRVHLPPLRERREQLPRLARHELQIAARELGVEAKQLHPETEMALTRLAWPGNV QLXFBIGMVKFTIPPLRDBLGQVEILIDflENRQFATIYSTQPLBFEAATMELLRRYSWPGNV DLYYRLGMAFLELPPLRERRLNLPLLFEHETLDAARRFERDAPILDEQTMSHLLAYSWPGNV DLYYRISMVELLLPPLRERRSQLPLLAREEL-BKFNSENQBSLTLEASALDVLMSCKFPGNV RELENCLERSAVLSESGLIDRDVILFNHRDNPPKALASSGPAEDGWLDNSLDERQRLIAALE BQLEflTCRWLTMMAAGQEVLTQDLPSELFETAIPDNPTQMLPDSWATLLGQWADRALRSGHQ BfiLRflLIENLVLMTITfiIVTPRDLPEDFVEVAEAQPP§ISVQSENSATG§ESDEIARFDEME BELRfl ......................................................... REL ........................................................... KAGWVQAKAARLLGMTPRQVAYRIOIMDITMPRL NLLSEAQPEMERTLLITALRHTQGHKQEAARLLGWGRNTLTRKLKELGME RRAIERAVANAGGNIAAAAEKLGLSRATLYRKLHQYRSRT 83 each of which contains a deletion in a single 3i£A-like open reading frame. For purposes of manipulation, each deleted region has been replaced by a DNA fragment carrying the 33111 gene from the transposon 13§ which confers kanamycin resistance to g. japonicum. Restriction endonuclease maps for the B. japonicum genomic regions carrying these deletions are shown in Fig. 15. Each deletion strain was assayed for the expression of 31: genes,' the 91311 gene, and nitrogen fixation activity in 3. japonicum grown under a variety of conditions. These results are summarized in Table 1. Gene expression data was obtained using the quantitative S1 protection assay described in Chapter 6. In this way, we have assayed for aspects of 3i£A-like (3113 and 31133 transcription) and 3139-like (91311 transcription; Carlson,1986) control. A detailed analysis of the 3331 deletion strain, BJ2101, is reported in Chapter 6. This mutation affects the microaerobic and symbiotic expression of a number of genes, including 3113, 31133, and 91311. In addition, a wildtype copy of this gene is required for the normal development of soybean root nodules. Aerobic, Ntr-like control of 91311 is normal in this mutant strain. We believe that this gene is part of a regulatory system (termed Odc) that is required for oxygen and developmentally mediated control of gene expression. The gene encoded by the 3331 region has been termed 9933 as the first gene in this system to be described. No changes in the expression of nifH, nifDK or 91311 have been observed for the 339;, 333;, or 3331 mutant strains (BJ1011, BJ2221 and 'BJ2111 respectively). The gene encoded by the 333; region does, however, appear to have an effect on development of the symbiosis c3 88 0 -869 two mm 6.5:. 6338; §~Bm8 .3 Beam... 83 3333b 53%: 58:2 .. .3338 mm: 9339.5»..- 0: amen 38:63 TV 33.? A8388»; 333m .83 Egg 333350 m5 gm?! acmmmt mm: 8335... memm m5 Low 32:. whoa—~33 don—Sm At < a I .1889 + I + + + + + :32. Ag + I I I I I I SEE Ag + I + + + + + gm $85. + I + + Bfifig + . DUP— ~40de 33. + . I + . + . + + + 0338 soon 53.. z 2 383056 38300335 0339.5. 300333... A .37: 308% 839.30% 93302 909. 3 533% 393m 3ch ebflmxfi Snob? who flu. ”.682 III -gbmcbfloama g a .m ht 58 1.398th 38 film can “I: up Sea 4 £3» 85 .vmumuvccw mesa—g we gmnssc mgg mcwms cmcwacmpwu mmmmcm>m mcm pcmwmz zmwgm mrzvo: cam .Lmnssc m—svoc .zpw>wpum wmmcmmocuw: Lo» vacmmmca mama mm.o He 0H m mxmoz v HHonm ow.o mm NH m mxmwz v oHHcm mmmo.o . mH Huo.o . OH mxwwz N aflofiwm moH.o mH ¢N.H 0H mxmmz N oHHdm 3:33:95: “arse; 2838123335: :25: 5:380: £23 cmmcm w_:coz mpsuoz >u_>wpu< mmmcmmocpwz u=~_a Lmum< wewp .HHoHnm Lo oHHdm ;u_3 :omumpsuocH cmumm mxmmz ¢ new N mmF=woz uoom camaxom xn :owuuzuom mcw—5umu< .N m_nmp 86 (Table l). Soybean root nodules harvested two weeks after infectio with BJ1011 reduced approximately twenty-five-fold less acetylene p: hour than did wildtype infected nodules of the same age (Table 2). four weeks after inoculation and sowing, nodules induced by both th wildtype and mutant strains reduced acetylene at nearly equal rates addition to this symbiotic phenotype, BJ1011 typically grows more 5 than wildtype under all media conditions that have been tested. Th slow growth phenotype is very unstable in liquid culture in that on growth is initiated, it proceeds at a rate similar to wildtype. Th return to normal growth rate is retained when cells are transferred fresh media. Since the mutation in this strain is a deletion of th ‘tflgz region, the high frequency of phenotypic reversion observed mu arise through genetic changes at a second site. This tendency towa pseudoreversion has complicated efforts to determine the actual rol this gene in B. japonicum. DISCUSSION Bradyrhizobium japonicum, Rhizobium meliloti, and Escherichia genomic DNAs all contain multiple restriction endonuclease fragment \Nthh hybridize to a Klebsiella pneumoniae flifiA DNA probe. At pres (only one nifA-like gene has been described in g. gglt (Buikema gt g 1985) and two in B. mgliloti (Szeto gt_gl., 1984; Buikema gt gl., 1 (Sussin gt gl., In Press). It will be interesting to see whether th (different nitA:homologous regions encode a family of NIFA-like leegulatory proteins responsible for controlling different gene syst 87 Recently, it has been pointed out that the E, gglj_nifA homc gene ELLE is related to a number of other regulatory genes in its N-terminal domain (Drummond gt gl., 1986; Nixon gt g1., In Press} of these genes, with the exception of the B. leguminosarum gth g (Gussin gt_g1., In Press), share any homology with the central dc gtgt. This middle portion is the region of ELLE that is closely to the K. pneumoniae gifA gene. It has been suggested that this is involved in the interaction of these proteins with NTRA and/oi polymerase (Drummond gt g1., 1986; Nixon gt gl., In Press). Sin: was this internal region of nifA which we used as the hybridizati probe in these experiments (see Chap. 3, Fig. 7), the multiple homologies described here may represent a group of as yet unideni regulatory genes which function in tandem with NTRA. We have isolated 4 and determined partial DNA sequence for E .5. japonigum nifA-like regions. In each case, we have found thai .gitA hybridizing DNA from g. japonicum potentially encodes an ope reading frame with extensive amino acid sequence homology to the ;Qneumoniae nifA and gttg genes. Four separate mutant strains of .japonicum, each containing a deletion in a single hug region, ha‘ Iconstructed and are described here. These mutant strains have bl (examined for alterations in the Nif-like (transcription of nitfl : tltfgg) and Ntr-like (transcription of 91311; Carlson, 1986) conti systems. A detailed discussion of the hflgg mutant is presented in Ch; 'The gene encoded by this region has some properties that are sim' ‘the 5.123ggmgnigglnttA gene. In addition, this 5. jgggfliggm gen: Inore general role in controlling gene expression in response to i 88 limitation and symbiotic development than expected for gifA. We that this gene is part of a general regulatory system that is re for oxygen availability and developmentally mediated control. T2 regulatory system has been termed Odc. We call the nifA-like gen in the h9g5 region gggA as the first gene identified in this reg pathway. The h3g1 deletion strain, BJ1011, has been characterized as a delayed fixation phenotype in soybean root nodules. The mecha which this delay occurs is not yet clear. Since at least some n (gitfl and nifQK) are expressed normally during this lag in the o nitrogen fixation, it is unlikely that the delay in nitrogen fix due to a general alteration in git gene control. One characteri the hug; deletion strain in free-living culture iS a tendency fo reversion from a very slow growing phenotype to a more normal gr rate. It is possible that the delay observed in nitrogen fixati activity results from a similar requirement for phenotypic rever Ausubel gt g1. (1985) described a 3, meliloti mutant strain ‘which a transposon, 135, had been inserted into a nth—homologou reading frame. Like the tug; deletion, this 3. meliloti mutatio results in a delayed nitrogen fixation phenotype. In addition, meliloti mutation results in an altered response to nitrogen lim such that a nitfl-lggz fusion is no longer activated when 3. mglj starved for ammonia. No such nitrogen limitation response has b [observed in the 5. jgpgniggm mutant strain, BJ1011. A better understanding of the relationship between these two mutants may helpful in defining the factors necessary for early nodule devel 89 No phenotypic differences from wildtype have yet been observed for the bag} or flag; mutant strains, BJ2221 and 802111 respectively. Expression of genes in response to microaerobic or nitrogen limited growth is unperturbed in these strains. In addition, BJ2111 and BJ2221 grow like wildtype on a variety of carbon sources, including the dicarboxylic acid succinate, indicating that neither hag; nor tag; are likely to encode a B. japgnjgum analog of the 3. leggmingsarum‘giffi homologous ggtn gene which regulates the uptake of dicarboxylic acids (Gussin gt_gl., In Press). It is possible that these nifA homologous regions are simply genetically silent copies of nifA or related genes. Pseudogene copies of nit genes in other bacteria have been described previously (Scolnik and Haselkorn, 1984). In this case, Nif’ mutations in the active gene copy were found to revert at high frequency, presumably due to an activation of one of the previously silent genes. Such an event might explain the high reversion frequency observed with the h5g2 deletion strain. In addition, the lack of any obvious phenotype for the flag; and flag; deletion strains might be explained in such a way. Alternatively, flag; and tag; may represent two gene copies both of which are always active for regulating the same system. A g. Japonigum strain in which both hug; and hflgt have been deleted would address this possibility. Finally, one must consider that bug; and nag; could encode 'regulatory genes for some as yet uncharacterized gene systems. A number (of gt_t- and/or nifA-homologous genes have been described. In many <:ases, these genes regulate expression of other genes in response to environmental stimuli such as altered osmdlarity (fi. gofi gmpfi; [Ramakrishnan gt gl., 1986), phosphate limitation (E, coli phoB; Makino 90 gt g1., 1986), the presence of dicarboxylic acids (3. leguminosarum ggtfl; Gussin gt gl., In Press), the presence of detergents (fi. gglt gftA; Drury and Buxton, 1985), or the presence of plant exudates (A. tumefagigng yttg; Gussin gt gl., In Press) in the growth media (Drummond gt g1,, 1986; Gussin gt g1., In Press; Nixon gt g1., In Press). An examination of the B. japonicum mutants described here for altered responses to environmental stimuli other than oxygen or nitrogen limitation would be useful. .3. japgnigum contains other nifA-homologous regions which have not been described in this paper. Recently, two of these regions have been isolated (6. Martin, unpublished) in our laboratory. .These regions are homologous to a gene isolated from g. parasponig (Nixon gt gl., In Press). No mutation in this gene has yet been described but the g. nargsponia gene has a very similar sequence to the 5. pngumoniag ntgt gene and is linked to a gene which is homologous to gttfi, another gene in the general nitrogen regulatory pathway (Magasanik, 1982). It will be interesting to see the effects of mutations in these genes and other nifA-like sequences on bradyrhizobial gene expression. EXPERIMENTAL PROCEDURES Bagterial strains. The wildtype B. japonicum strain used in all lexperiments is a small-colony forming derivative of g. japonicum USDA .3311b110 (BJ110) isolated as described (Kuykendall and Elkan, 1976; lWeyer and Pueppke, 1980). BJ1011, BJ2111, BJ2101, and BJ2222 are B. .japonicum strains in which a nifA-like region has been deleted and 91 replaced by the ngttt gene. The g. ggli strain ED8654 (gglfi, ggll, tggg, hggg, gggfi, gggfi, lggl) was used for construction and maintenance of plasmids and cosmids (Adams gt gl., 1984), as well as for preparation of E. ggli genomic DNA. The 5. japonicum genomic DNA library maintained in lambda phage BF101 was screened for 31:5 homologous sequences using the g. ggli host ET8051 ( (rug-glgA), hgttk, ttg, nglr) which contains a deletion of the entire glnAntrBC operon (Fisher gt git, 1981). Bacterial media, growth conditions. and strain construction. All growth conditions, media, and gene replacement techniques used are described in Chapter 6. Nucleic acid technigues. Preparation of genomic DNAs, library screening, Southern hybridization, and DNA sequencing were all as described (Adams and Chelm, 1984; Adams gt gl., 1984). The quantitative 51 protection assay used for evaluating gene expression levels is described in Chapter 6. CHAPTER 5 Microaerobic Induction of nit and Q15 Gene Expression in Bradyrhizobium japonicum INTRODUCTION The expression of both nitrogen fixation (git) and nitrogen assimilation (e.g. glnA, the gene encoding glutamine synthetase) genes in the free-living diazotroph K. pneumoniae is mediated by the same general nitrogen regulatory (Ntr) system (Magasanik, 1982; Dixon, 1984). Under nitrogen-limiting anaerobic conditions, the transcription of genes encoding both the nitrogen fixation and nitrogen assimilation processes is activated. In this way, K. pneumoniae is able to utilize atmospheric dinitrogen to support growth. Rhizobia and bradyrhizobia reduce atmospheric dinitrogen primarily during symbiotic association with their leguminous host plants. The development of symbiotic nitrogen fixation is a complex process requiring the coordinate differentiation of both the plant and the bacterial cells. For the bacterium, this developmental process culminates upon differentiation into a morphologically and functionally distinct nitrogen fixing endosymbiont termed a "bacteroid". Very little 92 93 of the ammonia produced from dinitrogen reduction by bacteroids to support bacterial growth since glutamine synthetase activity repressed in these cells (Brown and Dilworth, 1975; Stripf and h 1978; Werner gt gl., 1980). Instead, most of this ammonia is ex (Bergerson and Turner, 1967) into the plant cytoplasm where it i assimilated by plant enzymes (Miflin and Lea, 1976). In return, plant feeds reduced carbon coumpounds to the bacteria to meet th energy demands required by nitrogen fixation. Some bradyrhizobial strains can be made to fix nitrogen in free-living culture under specialized conditions (Keister, 1975; gt g1., 1975; Kurz and Larue, 1975; McComb gt gl., 1975; Tjepken Evans, 1975). These conditions include requirements for both a of fixed nitrogen (preferably glutamate) and the presence of oxy an electron acceptor for oxidative phosphorylation. The require oxygen in these systems presents a difficulty since the nitrogen enzyme is inactivated by oxygen. Oxygen partial pressures must maintained at a level which is high enough to allow the oxidati' phosphorylation levels necessary to provide the 12-15 ATPs reqU' N2 reduced (Ljones and Burris, 1972), and low enough to prevent irreversible inactivation of nitrogenase (Mortenson and Thornle; Nitrogen metabolism during asymbiotic nitrogen fixation by bradyrhizobia is similar to that observed in bacteroids since mu the fixed nitrogen is not used to support growth but is instead into the medium (O’Gara and Shanmugan, 1976; Bergerson and Turn Ludwig, 1980). For at least some bradyrhizobial strains this f assimilate the fixed nitrogen is partly due to a decrease in gl synthetase activity (Ludwig, 1980; Bergerson and Turner, 1978). 94 regulation of glutamine synthetase activity in rhizobia and bradyrhizobia is complicated by the fact that these bacteria contain two unique forms of the enzyme, 651 and GSII (Darrow and Knotts, 1977). These two distinct nitrogen assimilatory enzymes are encoded by separate genes that are regulated by different mechanisms. Transcription of the gene encoding GSI (glnA) is relatively constant under all conditions (Carlson gt g1., 1985). This enzyme is regulated post-translationally via adenylylation such that activity is decreased during growth in nitrogen rich media (Bishop gt gl., 1976; Ludwig, 1978). No post-translational modification of GSII has been described. Transcription of the gene encoding GSII (91311) is controlled by a mechanism that resembles the Ntr system (Carlson, 1986) as described for enteric bacteria (Magasanik, 1982). Under nitrogen limiting growth conditions transcription of the 91311 gene is increased presumably to increase the nitrogen assimilatory capacity of the cell. We are interested in understanding the mechanisms through which 913 and nit gene expression are controlled by Bradyrhizobium japonicum, the soybean symbiont, during symbiotic nitrogen fixation as well as in free-living cultures. We have begun to examine this regulation by quantitatively determining the relative steady-state mRNA levels for the nttfl, nifDK, glnA, and 91311 transcription units in bradyrhizobial cells grown under a variety of conditions. For the nifH, nifDK, and glgtt transcription units, expression is coordinately induced in response to both symbiotic development and microaerobic growth. This response is not mediated through fixed nitrogen supply but instead is primarily controlled either by oxygen availability or by developmentally-specific signals. The physiological mechanisms by which these signals are 95 transduced to change gene expression will be discussed. RESULTS Effects of 02 concentration on B. japonicum gene expression. To study the role of oxygen in regulating git and glg gene expression for g. japonicum, we have examined total cellular RNA isolated from 5. japonicum cultures grown in rich media under a variety of oxygen concentrations. Results for one such set of experiments are shown in Figure 1. The activities of the nifH, nifDK, glnII, and glnA genes were monitored using a quantitative Sl protection assay system. Equal quantities of four single-stranded, 5’ end-labelled DNA probes specific to the nifH, nifDK, glnII, and glnA promoters were mixed and hybridized to total cellular RNA under conditions where the probes were in excess to the respective RNAs (Experimental Procedures). After hybridization, the mixtures were treated with $1 nuclease and the protected fragments separated by gel electrophoresis. Hybridization probes were designed such that the protected fragments would be well resolved on the gel. This allows the simultaneous measurement of transcript abundance for all four genes using a single RNA sample. The radioactive content for each band on the gel was determined directly (Experimental Procedures). ’ When 5. japonicum is grown in a standard rich medium (YEMN) which is being continuously sparged with air, no transcript for the gitfl, nifDK, or glnII genes is observed (Fig. 20, Lane 1; Tab. 3). The glnA gene, in contrast, is transcriptionally active. As reported previously (Carlson gt gl., 1985), we have observed only small variations between 96 Figure 20. Abundance of nifH, nifDK, glnII, and glnA transcripts in g. japonicum grown under a variety of atmospheric oxygen concentrations (with a gas balance of N ) Quantitative analyses were by the 51 protection method descriged in the Experimental Procedures. The migration positions of the four protected fragments are indicated by P P , P , and P . The other major radioactive products observed represant undi ested probe DNA. The total cellular RNAs (2 ug RNA in each experiment) used in the hybridization reactions were isolate from .5. japonicum cultured under the following conditions. 20% O in YEMN (lane 1); o. 2%0 in YEMNA (lane 2); 5%0 in YEMN (lane 3), 3%0 YEMN (lane 4); 1?0 in YEMN (lane 5L 0. 3%0 02 in YEMN (lane 6L 06. 2%0 in YEMN (lane7); anfi 0.17.02 in YEMN (lane 8). All media and growth conditions are described in the Experimental Procedures. 0’ 2 97 A 98 any two cultures in gigA transcript levels. This property of gigA makes it a good control for the integrity of an RNA preparation. If E. japonicum is cultured in this same medium but continuously sparged with a gas mixture of nitrogen and oxygen such that oxygen availability limits growth (5% 02 or less in these experiments), then transcription of all four gene units is observed. (Fig. 20, Table 3). Similar cultures have been grown either with (YEMN) or without (YEN) added KN03. Under microaerobic conditions, KNO3 helps to stimulate growth by serving as an alternative electron acceptor to oxygen (Daniel gt g1, 1982; M. L. Guerinot and B.K. Chelm, unpublished). Very little difference in the relative levels of transcript for any of the genes monitored was observed under any of the conditions tested (Table 3). However, a 3- to 4-fold decrease in mRNA abundance for gifH, gifQK, and gigii was observed as 02 levels were increased above 1% in YEMN medium or above 0.1% in YEM medium. Despite the high levels of the two git mRNAs observed in these cultures, no nitrogen fixation activity, as measured by the acetylene reduction assay (Hardy gt gl., 1968), was detectable. Effect of ammonia on microaerobic growth and gene expression. Under aerobic growth conditions, gigii transcription is regulated by a mechanism that senses the relative availabilities of carbon and nitrogen (Carlson, 1986). When growth is limited by nitrogen supply, gigii transcription is activated. Presumably this activation is an attempt by the cells to maximize their capacity to assimilate nitrogen. Conversely, carbon limited growth leads to a repression of gigii transcription since increased nitrogen assimilation would not increase growth. This control is Similar to the Ntr control system which 99 Table 3. Effects of Varying 02 Concentrations on Transcription of git and gig Genes. oxygen concentration Mediuma Probe 0.1 0.2 0.4 1.0 2.0 - 5.0 10.0 20.0 YEM giffl 380b 56 36 78 - - - - gifH 295 22 20 23 - - - - glnII 101 27 9 45 - - - - glnA 44 12 11 18 - - - - YEMN gifQ 161 175 231 211 110 59 0c 0 nifH 133 72 157 151 59 75 0 0 g1n11 51 43 72 63 24 25 0 0 glnA 34 23 34 37 29 35 31 18 YEMNA 3110 - 62 - - - - - - nifH - 42 - - - - - - lnI - 11 - - - - - - glnA - 21 - - - - - - a media and growth conditions were as described in Experimental Procedures with carbon and nitrogen sources abbreviated as follows: YE, yeast extract; M, mannitol; N, nitrate; A, Ammonia. cpm of hybridization probe protected from $1 nuclease digestion per ug of input total cellular RNA. This indicates the relative steady state mRNA level for each specific gene. Data represent a mean value determined from two separate experiments. c 0 implies that no transcript was detected 100 operates in the enteric bacteria (Magasanik, 1982). For enteric diazotrophs, Ntr regulation is also involved in git gene induction (Dixon, 1984). Some Ntr-like control of nitrogenase expression has also been described for bradyrhizobia (Ludwig, 1980; Bergerson and Turner, 1978). For E. Japonicum, ammonia is an especially good nitrogen source, .being capable of supporting the highest rates of growth. Under most aerobic culture conditions, addition of ammonia to the medium leads to a repression of gigii transcription (Carlson, 1986). To determine whether the low oxygen induction of gigii and git gene transcription is under Ntr-like control, 5. japonicum was grown under microaerobic conditions in YEMN medium with 10 mM NH4Cl. Comparison of growth rate and total growth for this culture with a second culture grown under precisely the same conditions except without added ammonia indicates that in this microaerobic culture medium, growth can be stimulated by addition of nitrogen in the form of ammonia (Fig. 21). This demonstrates that the microaerobic cultures are nitrogen limited. Microaerobic induction of gigii as well as gitfl and gitQE could therefore be due to an Ntr-like response similar to aerobic gigii induction. When total cellular RNA was isolated from these cultures and assayed for gigii, gitfl, gitQE and gigA expression, transcript levels for all but gigA were significantly reduced in cultures with ammonia (Table 3) supporting the conclusion that microaerobic induction of these genes may be under some level of Ntr control. Growth phase dependence of microaerobic nif and qln trgnscription. Previous studies on the role of microaerobiosis in nonsymbiotic bradyrhizobial git gene induction have relied upon the expression of 101 Figure 21. Growth properties of E. japonicum cultured under 0.2% oxygen (balance N ) with and without ammonia. The data represent optical density measured at 600 nm. 102 Days 103 nitrogenase enzyme activity. These studies indicate that the presence of nitrogenase activity in free-living cultures of bradyrhizobia usually requires that the cells have stopped growing (Ludwig, 1984). For this reason, the cultures used for the experiments described above were grown for two days into stationary phase prior to harvesting (with the notable exceptions of the 1, 2, and 5% 02 cultures which were harvested earlier; see Experimental Procedures). To determine whether the microaerobic transcriptional induction of gigii and git genes has a growth phase requirement like that for nitrogenase activity, cells were harvested at different times during microaerobic growth and total cellular RNA prepared. As shown in Table 4, nifH, nifDK, and glnII transcripts are all observed within 24 h of inoculation. There is no discernible pattern to the variation in transcription levels observed throughout the growth curve for any of the genes monitored. Transcription of nif and gln genes during symbiotic development. In order to evaluate the role of bacteroid development in git and gig transcription, RNA was isolated from the bacteria within 5. japonicum induced soybean root nodules. Two weeks after sowing and inoculation, transcript levels for both git operons are about 2 to 4-fold higher in the nodule bacteria population than under any of the microaerobic growth conditions tested (Table 5). Transcript abundances for the gitfl and nifDK genes are approximately equal under these conditions. In contrast, bacterial RNA from four week old soybean root nodules has an increased ratio of nifH to nifDK message with nifH transcript about twice as plentiful as nifDK. This is accomplished mainly through an increase in the steady-state mRNA levels for nifH. Whether these 104 . Table 4. git and gig Gene Transcription During Growth at 0.2% Oxygen. hours after culture inoculation a Probe 24 48 72. 96 120 gifD 93b 48 80 83 63 gifH 101 43 118 78 36 glnII 42 23 39 35 27 glnA 67 12 23 14 14 a see figure 2 for growth curve. cpm of hybridization probe protected from $1 nuclease digestion per ug of input total cellular RNA. Data represent mean values obtained in duplicate experiments. 105 Table 5. mRNA Abundance of git and gig Genes During Symbiotic Development. mRNA Abundancea RNA Source ni H nifDK glnII glnA 2 week old nodule bacteria 498 243 63 29 4 week old nodule bacteria 636 205 66 15 "Nodule Bacter1a"b 579 181 34 23 ”Transforming Bacteria"b 767 166 34 25 ”Bacteroids"b 442 171 19 31 a cpm of hybridization probe protected from $1 nuclease digestion per ug of input total cellular RNA. Data represent mean values determined from two identical experiments. "developmental" cell types isolated from the total population of bacterial cells within a soybean root nodule as described in Exper- imental Procedures. 106 differences represent changes in transcription rate or changes in mRNA stability has not yet been determined. gigii transcript levels in these same nodule bacterial cell populations are approximately equal to those seen in microaerobic culture. A method for fractionating the mixed populations of bacteria isolated from within soybean root nodules has been described (Ching gt gi., 1977). This technique relies on the change in buoyant density which occurs for bacteroids due to the accumulation of polyhydroxybutyrate granuoles. The expression of certain other "bacteroid-specific" properties (including nitrogenase enzyme activity) parallel this PHB accumulation (Ching gt gi., 1977 ). RNA was isolated from the three bacterial populations distinguished by this method (Experimental procedures). No significant differences in gifH, nifDK or glnII gene expression patterns were observed (Table 5). DISCUSSION Under specialized environmental conditions, several bradyrhizobial strains, including Bradyrhizobium japonicum USDA 110, are able to reduce dinitrogen to ammonia in free-living culture (Keister, 1975; Pagan gt gi,., 1975; Kurz and Larue, 1975; McComb gt gi., 1975; Tjepkema and Evans, 1975). In this paper we have demonstrated that the requirements for transcriptional induction of the genes encoding the nitrogenase structural polypeptides are not as stringent as for achieving enzyme activity. Optimal conditions for asymbiotic nitrogen fixation include requirements for organic acids such as succinate or malate as carbon and 107 energy sources; the presence of a growth promoting nitrogen source; and most importantly, extreme microaerobiosis. Microaerobic conditions are required for nitrogen fixation because the nitrogenase enzyme is inactivated by intracellular oxygen. Some oxygen is required for growth however, since rhizobia are obligate aerobes. The highest free-living nitrogen fixation rates by bradyrhizobia are achieved at dissolved oxygen concentrations of about 1 uM or less (Ludwig, 1984). Little or no nitrogenase activity is observed above 4 uM oxygen (Rao gt al., 1978). For E. japonicum, transcription of the nifH and nifDK genes is observed in cultures which are being continuously sparged with gas mixtures containing as high as 5% oxygen (2 mM, balance N2) at a rate of 30 liters/hr. The transcript levels for nifH and nifDK were only slightly different in cultures grown under these intermediate 02 levels than in similar cultures grown with a 0.1% 02/99.9% N2 atmosphere. No expression was observed when cultures were sparged with air at the same rate. Nitrogen fixation activity was never observed with any of the media or aeration conditions used. Induction of git gene transcription takes place less than 24 hours after inoculation of the microaerobic culture medium described. This contrasts with a four day microaerobic incubation requirement for expression of translational fusions between the gitfl or gitQ genes from B. japonicum and the E. ggii iggl gene (Alvarez-Morales gt gi., 1986). Some differences exist between the cultural conditions described by Alvarez-Morales gt gi. (1986) and those described here. This might account for the large time lag necessary for expression of the translational fusions described. Alternatively, the difference in the times required for transcriptional or translational activity of nifH and 108 gitDE might result from a translational control mechanism for nitrogenase expression in E. jappgigum. No translational control of nitrogen fixation processes has ever been reported, however, post-translational regulation of nitrogenase activity has been described in the Bhpdospitillgceag (Ludden and Burris, 1979). A related regulatory mechanism in B. jappnicum would not resolve the conflict between our results and those of Alvarez-Morales gt gi. (1986). In addition to inducing git gene transcription, the low 02 conditions described here result in the accumulation of high levels of gigii messenger RNA, the gene encoding the nitrogen assimilatory enzyme glutamine synthetase 11. As with the git genes, gigii was not transcribed in the same medium under fully aerobic conditions. Thus expression of gigii appears to be activated coordinately with gitfl and gitQE in response to low oxygen. Previous studies by Rao gt gi. (1978) indicate that glutamine synthetase II activity is decreased during microaerobiosis for some bradyrhizobia. This inhibition of glutamine synthetase II activity occurs at the same oxygen tension that nitrogenase activity first becomes apparent (about 0.4 % O2 in the experiments described; Rao gt gi., 1978). Since no nitrogenase activity was observed in the experiments we describe, it is possible that at lower oxygen concentrations gigii transcription would have been repressed. However, we observed similar levels of gigii transcript in total cellular RNA populations isolated from the bacteria in soybean root nodules and bacteria grown in microaerobic non-nitrogen fixing culture. We therefore have no explanation for the discrepancy between our results and those of Rao gt gi. (1978). 109 The observation that transcription of the gene encoding the nitrogen assimilatory enzyme glutamine synthetase II is activated coordinately with git gene expression both microaerobically and in bacteroids is surprising. Other workers have proposed that bradyrhizobial nitrogen-fixation is uncoupled from nitrogen assimilation (O’Gara and Shanmugan, 1976; Bergerson and Turner, 1967). The large majority of the ammonia produced by bradyrhizobial nitrogenase is exported (Bergerson and Turner, 1967; O’Gara and Shanmugan, 1976; Bergerson and Turner, 1978; Ludwig, 1980). Why then is gigii transcribed at high levels in cultures which express git? Ludwig (1984) proposed that asymbiotic nitrogen-fixing cultures of Eradyrhizobium sp. 32HI contain mixed populations of cell types that grow cooperatively. One population terminally differentiates into nonviable nitrogen fixing cells while the other grows utilizing the ammonia produced from nitrogen fixation by the first population. It is possible that the gigii transcription observed here occurs in a different cell subpopulation than does git expression. If this were the case, one might expect some change in relative transcript levels for glnII, niffl and nifDK during the growth curve since the population of nonviable cells increases as the culture enters stationary phase (Ludwig, 1984). No such change was observed (Tab. 4). In addition, gigii transcript levels were the same in all three of the "developmental" cell types isolated from nodules leaving us to conclude that either the export of ammonia does not require complete GS repression or that GSII is repressed by some as yet unknown post-transcriptional mechanism. It has been noted previously that the promoter for the gigii gene shares sequence homology in the -10 to -25 region with several git 110 promoters in E. Japonicum . This region is also related to promoters 1from other bacteria that are controlled by nitrogen availability (Ntr control; Magasanik, 1982). In aerobic culture, transcriptional activity of gigii is mediated by an Ntr-like system (Carlson, 1986; Adams gt gi., In Press). When 5. japonicum is grown aerobically under nitrogen .limiting conditions, transcription of the gigii gene is induced and when nitrogen nutrition does not limit growth, gigii is not transcribed (Carlson, 1986). The relationship between oxygen availability and nitrogen limitation in bradyrhizobia is complex. Addition of ammonia to ' microaerobic cultures of E. japonicum in YEMN media increases both growth rate and final cell density. Similar changes in growth. properties can be achieved by increasing atmospheric oxygen levels indicating that the amino acids present in yeast extract do not limit aerobic growth. These data imply that growth limitation during microaerobiosis results from nitrogen limitation brought about through changes in the cells ability to utilize amino acids as a source of nitrogen. Oxygen limitation has been shown to induce an ammonia export system in Bradyrhizobium gp. 32HI(Ludwig, 1980; Gober and Kashkett, 1983). If such a system were induced here, nitrogen limitation could occur even if amino acids were catabolized normally since active ammonia excretion would prevent its incorporation. Addition of ammonia to these cultures might then partially relieve nitrogen limitation and stimulate growth by shifting the intracellular/extracellular ammonia equilibrium concentrations. We find that transcript levels for git genes and for gigii are significantly reduced by addition of ammonia, indicating that nitrogen limitation might be at least partially responsible for causing an Ntr-like induction of glnII and/or git genes. Similar results have 111 been observed by others in studying ammonia effects on nitrogenase activity (Ludwig, 1980; Bergerson and Turner, 1978). Transcript levels for the nifH and nifDK genes differ quantitatively between microaerobically grown cells and bacteria isolated from soybean root nodules. Steady-state mRNA levels for the gitEE operon are approximately 2 to 3-fold higher during symbiotic nitrogen fixation than in microaerobic culture. For gitfl this increase in message level is closer to 5-fold. The differences in transcript levels observed might result from either changes in transcription rates or varied rates of mRNA stability under the different conditions. Recently we described a E. japonicum gene (och) that is required for transcription of ifH, gitEE and gigii both microaerobically and during symbiosis (Chapter 6). Mutations in this gene have severe effects on symbiotic development for both the plant and the bacterium even at very early times after nodule initiation indicating that the gene is required long before the onset of nitrogen fixation. Since nodules induced by the gggA mutant strain fail to develop normally, it is impossible to tell whether this gene is responsible for git transcription late in nodule development. It is possible that a second regulatory gene is required to give the increased transcript levels observed during symbiotic nitrogen fixation. EXPERIMENTAL PROCEDURES Bacterial strgins. media. and growth conditions. All of the experiments in this paper used a small-colony derivative of E. japonigum 112 311b110 isolated described (Kuykendall and Elkan, 1976; Meyer and Pueppke, 1980). Bacteria were grown in YEM (0.04% yeast extract, 1% mannitol, 3 mM KZHPO4, 0.8 mM M9304, 1.1 mM NaCl), YEMN (YEM with IOmM KNO3), or YEMNA (YEMN with 10 mM NH4Cl) media. 10 liter cultures of E. japonicum were grown using Microferm Fermenters (New Brunswick Scientific, Edison, New Jersey) agitated at a rate of 200 rpm and sparged with gas mixtures of oxygen and nitrogen at rates of 500 ml/minute. Gas flow rates were controlled using thermal mass flowmeters (Brooks Instruments, Hatfield Pennsylvania, Model 5850 C) to give the atmospheric oxygen concentrations reported. Isolation of bacteria from soybean root nodules. Total bacterial populations from frozen soybean nodules were prepared as described previously (Chapter 2, Adams and Chelm, 1984). Separation of these bacteria into the three developmental fractions discussed was by centrifugation through a discontinuous sucrose gradient using a zonal ultracentrifuge rotor (Beckman 14 Ti, Beckman Instruments, Fullerton, California) as described (Carlson gt gi., 1985). Nucleic acid Tgchnigues. RNA isolations and S1 nuclease protection experiments were as described previously (Chapters 2 and 6; Adams and Chelm, 1984). Probes for the quantitative 51 protection experiments I described were synthesized by primer extension (Chapter 6; Carlson, 1986). The partially protected fragments from $1 nuclease protection experiments with these probes are 120 bp (gigA), 150 bp (gitfl), 170 bp (gigii), and 200 bp (gitEE). CHAPTER 6 Characterization of a Bacterial Gene Required for the Normal Differentiation of the Bradyrhizobium jappnicum / Soybean Symbiotic Interaction INTRODUCTION Rhizobia and bradyrhizobia can associate with leguminous plants to establish a symbiotic relationship in which photosynthetic products made by the plant provide the energy required for bacterial nitrogen fixation and, in exchange, the bacteria provide ammonia to the plant. The successful outcome of this interaction requires a series of developmental steps for both the bacterium and the host plant, and results in the development of root nodules, Specialized organs in which both plant and bacterial cells are highly differentiated. This coordinate differentiation requires extensive regulation of gene expression, the best characterized examples of which are the induction of bacterial genes involved in nodule initiation (ggg) and nitrogen fixation (git) (Mulligan and Long, 1985; Corbin gt gi., 1982; Paau and Brill, 1983) and the activation of plant genes involved in respiratory control (e.g. leghemoglobin) and nitrogen assimilation (e.g. glutamine 113 114 synthetase) (Verma and Bal, 1976; Lara gt gi., 1983). The symbiotic mechanisms controlling this differentiation are not understood, although some bacterial mutants which are affected in this complex process have been identified (Vincent, 1978; Noel gt gi., 1982; Chua gt gi., 1985; Stanley gt gi., 1986; Regensburger gt gi., 1986). One factor that has been experimentally implicated in the control of differentiation is oxygen limitation. Low oxygen levels presumably exist during the early stages of nodule development due to multiple diffusion barriers and the active respiration by both the plant and the bacteria. Oxygen limitation has been shown to cause a small but detectable induction of uricase, an enzyme which is normally nodule-specific, in non-nodule soybean tissue (Larsen and Jochinsen, 1986). In addition, git gene expression can be induced in a number of bradyrhizobial strains by growth under oxygen limiting conditions (Keister, 1975; Scott gt gi., 1979). The importance of this oxygen-mediated gene regulation to symbiotic differentiation has not, however, been directly tested. Efforts to understand the symbiotic coordination of gene expression have initially focused on the analysis of bradyrhizobial genes whose expression is induced during nodule development. The transcriptional initiation sites for several git operons from Egggytgitppigg igpggiggg, the soybean symbiont, have been determined (Adams and Chelm, 1984; Fuhrmann gt l., 1985), and these sequences closely resemble those identified for git operons in other diazotr0phs including Eigpgigiig pgggmggigg and Egitppigg ggiiipti. In these other organisms a positive regulatory gene, gitA, has been shown to be necessary for the transcriptional activation of several git operons (Dixon gt gi., 1980; Szeto gt gi., 1984; Ausubel, 1984). For E. pneumoniae, where the 115 regulation is best characterized, the expression of gitA is dependent upon the more general nitrogen regulation system (Ntr) through the action of the gttE gene product (for review see Magasanik, 1982; Dixon, 1984). gitA and gttE have many structural and functional similarities (Ow and Ausubel, 1983; Buikema gt gi., 1985; Drummond gt gi., 1986). In addition, the promoters for genes normally controlled by either gitA or gttg share a common consensus sequence (Ow gt gi., 1983; Beynon gt gi., 1983) and under certain conditions some of these promoters can be activated by the product of either regulatory gene (Ow and Ausubel, 1983; Merrick, 1983; Buck gt gi., 1985). Physiologically, however, these genes can be distinguished by their role in either the global control of genes involved in nitrogen assimilation (gttE) or the specific control of genes encoding the nitrogen fixation process (gitg). The E. pngumoniag gitA gene product is able to activate transcription from the E. japonicum nifH and nifDK promoters when overexpressed in the heterologous Ei gpii system (Alvarez-Morales and Hennecke, 1985; Alvarez-Morales gt gi., 1986), supporting the idea that E. japonicum git gene expression may be modulated via a similar regulatory gene product. I previously reported the identification of two separate regions in the E. japonicum genome which hybridize to the K. pneumoniae gitA gene and suggested that these regions might encode proteins having either gttt- or gitE-like properties (Adams gt gi., 1984). Further characterization of these regions has demonstrated that the first of these homologous regions (designated gggi for homology to gitA) was incorrectly identified due to the unexpected presence of gitE sequences in the hybridization probe utilized (Fuhrmann et al.,1985; my unpublished results). Sequence analysis of the second region (hna2) 116 has confirmed that this region does indeed contain an open reading frame having extensive gitA homology (Chapter 4). Here I report on the identification of additional gitE-homologous regions in E. japonicum DNA and present a detailed phenotypic analysis of one of these. This gitA-like gene has a pleiotropic effect on the symbiotic differentiation of both the plant and bacterium. The defective interaction becomes apparent after the bacteria are released into the host plant cells of the nodule cortex. Further analysis showed that this gene is necessary for the regulation of expression of both the E. japonicum git genes and the gene encoding glutamine synthetase II, gigii (Carlson and Chelm, 1986). This effect was seen both in association with the plant host and in response to oxygen-limited growth gt piggtg, supporting the proposed importance of oxygen limitation in the E. japonicum / soybean symbiosis. The wild type allele of this locus limits the final cell density achieved in cultures grown under controlled oxygen limitation, indicating that growth control might also play an important role in the plant-bacterial interaction process. I propose to term this gene pggA to indicate its pleiotropic regulatory role in oxygen and developmental control. RESULTS Identification and Mutagenesis of the och Gene. E. japonicum contains several different regions with sequence homology to the nifA gene of E. pneumoniae. A 950 bp EgtI-EngV fragment which is entirely internal to the E. pneumonigg nifA gene (Buikema gt gi., 1985) was 117 radioactively labelled and hybridized to Southern transfers of Eng1 digested E. japonicum genomic DNA (see Fig. 14, Chapter 5). Several restriction fragments in addition to the one which was previously identified and tentatively labelled gggE,(11 kbp; Adams gt gi., 1984) hybridize to the probe. The region which was previously identified as gggi (Adams gt gi., 1984) does not hybridize in these experiments (Chapter 5, Fig.14, lane 9) confirming that the isolation of gggi was actually due to contamination of the probe used in those experiments with gitE sequences (Fuhrmann gt gi., 1985) and that the probe utilized here is free of gitE contamination. In order to isolate the additional gitA-homologous regions described here, I screened a cosmid library (Adams gt gi., 1984) with the probe described above. Several recombinant cosmids which contain gitE-homologous sequences were identified and further characterized into distinct classes based on restriction site patterns. A cosmid from one such class, pRJcos7-36, contains two EngI restriction fragments of 3.2 and 3.8 kbp with sequences homologous to gitA (Chapter 5, Fig. 14, lane 8).. A restriction endonuclease map of the region containing these two fragments is illustrated in Figure 22a. The gitA-like gene localized in this region has been termed pggA for the reasons elaborated below. To verify that the two EngI fragments in pRJcos7-36 that have homology to gitA exist as a contiguous stretch of DNA in the bacterial genome, these fragments were individually recloned into plasmid vectors, radioactively labelled and used to probe Southern transfers of various restriction digests of E. igppgiggg genomic DNA (data not shown). All of the hybridization data is in complete agreement with the restriction map presented in Figure 22a. The approximate position of the region 118 . .mcmm Hammm as“ new: nmumpamg van vmumrmv cams m>mg mcwm mwfld mg» mo Fpm cam wcmm duflm mza mo wean gown: cw Noxqm :.acum =o_umpmc deflm one wo nos oweocmm msa m_ :zocm omp< .Aemm~ ..Hm mm msmw>m.z>>aa¢l . u o . u o a o a o 0 c u n o o . o o u u . . o u o o u o c o 0=U:._._.._._<_..!>za=>>._axzs=oau>xh 0<0000°<°O°OOb00000000h000¢0hu>a>l.hszwD>IMhuOu°d . a u o n . c u o o o o g n n o o I n u o . o o a o o c o ua=22h<<——I>8>B.—hn699>lhtu°l°a <¢°Uh0h<0000":<00<0000000h¢0huhusa«<0: scou>; mg“ c. vo~...p= m3 kumuwvcw mew mangmme umuompoga 889;. 8;. mo m:o_..moa :o.»m.m.g use .A. m:a.. mcm— cmxxme as. c. ems. Am.-~ mmcm.. mmcm. .macms.cmaxw mg. :. cmmz.mm3 mnoga nose mm mwswa cub .2..: can xomwc ...:.m we umpauwucw w.“ A. mamFV monoga mmgsw msa mo mcowawmon comamsmms msh .mmcszoocm Pmpcmswgmaxm mg» :. can.96mmn vogpme cowuumpocn .m cm a; wee: mmmachm m>mpmpwuc~=o .mcompwccou zpzosm .o mammcw> a sock E:o.:ommn 4m c. mug..0mcacu wwmflm can mmmfld .mwflq eo mocmvcsn< .¢~ mesmP. 125 _m aha m... mma: 333533.05 O—aul -I’ Il’)”)”’ll-”\l\u 3:03... 9 :21... I no... 7 .681 I Table 6. Abundance of nifH, nifDK and glnII Transcripts in E. japonicum Grown Under Various Conditions b mRNA Abundancea Strain Medium Aeration nifH nifDK glnII BJIIO x0 Air 0c 0 100 XGA Air 0 O O YEMN Air 0 O 0 YEMN 0.1% 02 300 154 39 BJ2101 XG Air 0 0 50 XGA Air 0 O O YEMN 0.1% O2 0 0 O BJ702 YEMN 0.1% 02 135 92 19 a cpm of hybridization probe protected from $1 nuclease digestion per ug of inputtotal cellular RNA media and growth conditions are as described in Experimental Procedures with carbon and nitrogen sources as follows: nitrate; YE, yeast extract; M, mannitol X, xylose; G, glutamate; A, ammonium; N, 0 implies that no transcript was detectable 127 sample. The radioactive content of each protected fragment band reflects the steady state mRNA level for a given gene in the total cellular RNA sample. When E. japonicum is grown under aerobic conditions, no nitrogen fixation activity is detected and, as expected, no transcription of the two git operons is observed (Fig. 24 and Table 6). Under these aerobic growth conditions, the gene encoding glutamine synthetase II, gigii, is controlled in response to the quality of the available nitrogen source, being expressed with glutamate, a poorer nitrogen source, but not with ammonia or the mixture of amino acids found in yeast extract, better nitrogen sources. This control is similar to the Ntr control system found in the enterics (Magasanik, 1982). This aerobic Ntr control in E. japonicum is not dependent on the functional pggg gene as indicated by the normal pattern of aerobic gigii expression seen in BJ2101. Similar experiments were performed using a hybridization probe specific for the E. jappnigum gigA gene (Carlson gt gi., 1985), which was reported to be transcribed at a relatively constant level under the conditions used in this study (Carlson gt gi., 1985). The level of expression of gigA was also constant in the BJ2101 mutant strain (data not shown). When E. japonicum BJIIO is grown in a standard rich medium with limiting oxygen (0.1 % oxygen in these experiments), transcription of the gitE and gitEE operons is induced (Fig. 24 and Table 6), and this induction is enhanced when 10 mM KNO3 is included in the media. Nitrate also helps to stimulate growth under these conditions by serving as a terminal electron acceptor for respiratory metabolism (Daniel gt gi., 1982; M.L. Guerinot and B.K. Chelm, unpublished). Despite the high levels of nifH and nifDK mRNA seen in these cultures, no nitrogen 128 fixation activity, as measured by the acetylene reduction assay (Hardy gt gi., 1968), is detectable. In addition to git gene expression, the gigii gene is transcribed under these conditions. No expression is observed for any of these genes when E. japonicum is grown under well aerated conditions in this same medium, indicating that the expression of these genes requires the microaerobic condition. When the gggA mutant strain 802101 was grown under limiting oxygen in this medium, gitE, gitEE or gigii transcripts were not detectable (Fig. 24 and Table 6). These defects are not observed for the nifDK deletion strain 80702, indicating that these results are not due to a secondary effect of the lack of nitrogenase expression. The pggA mutant 802101 also has modified growth properties under oxygen limiting culture conditions. Data for cultures of the wild type 80110 and the gggA deletion 802101 grown at 0.1% oxygen are illustrated in Figure 25. At low oxygen the gggA mutant grows to similar turbidity and protein concentrations as the wildtype strain. However a large difference is found for the viable cell density, with 802101 reaching about a three-fold higher viable cell density than 80110 under these conditions. These differences indicate that the microaerobically grown B02101 cells have a modified control of cell division or viability when compared to wildtype cells grown under oxygen limited conditions. No differences have been detected between 802101 and 80110 in normal aerobic cultures or between 80702 and 80110 in either environment. It is unlikely that the observed regulatory phenotype of the pggg 3 deletion strain 802101.is due to polar effects of the gptii insertion usedin its construction since the tigA gene and a £118 promoter have been mapped just past och (Fig. 22a; Fuhrmann gt gi., 1985) indicating 129 Figure 25. Growth properties of E. japonicum strains 80110 (solid figures) and 802101 (open figures) under limiting oxygen conditions. The data represents the optical density at 600 nm (circles), protein concentration (squares), and viable cell density (triangles) determined by measurement of colony forming units. 130 0.01 110 —--— 2101 100 - 10 —- l 1.5%.... .22.... =8 1.... 3...... 55...... Days 131 that pggA is at the 3’ end of its transcription unit. In addition, tizg mutant strains have been reported to have a normal developmental phenotype (Fischer gt gi., 1986). Therefore, these experiments indicate that gggA encodes a product that functions directly in a positive control system for git and gigii, and in some manner on low oxygen growth control. chA Eontrol of the Plant-Bacterial Interaction. The gggA mutation has a dramatic effect on the interaction between E. japonicum and its normal symbiotic host, the soybean plant. Surface sterilized soybean seeds were inoculated with wildtype 80110, the Eggg mutant 802101 or the gitEE mutant 80702 and grown in modified Leonard jars (Vincent, 1970) with a nitrogen-free nutrient solution. After four weeks of growth the plants inoculated with either of the two mutant strains are small and chlorotic when compared to those inoculated with wildtype E. japonicum, and no nitrogen fixation activity can be detected by the acetylene reduction assay (Hardy gt gi., 1968) for plants inoculated with either of the mutant bacterial strains. Examination of the nodules formed by infection with 802101, 80702, or the parental strain 80110 suggests that the effects produced by the deletion of gggA are more pleiotropic than one would expect if this gene simply regulated bacterial git or gig gene expression. The number of nodules induced by 802101 is always at least twice as high as with the wildtype 80110 or the nifDK' mutant 80702. In addition, plants inoculated with a mixture of 80110 and 802101 form many more nodules than plants inoculated with the wildtype strain alone, indicating that the pch mutant may override the plant’s internal regulation of effective nodulation events (Pierce and Bauer, 1983). The 132 nodules formed following infection with 802101 are also visually different than those induced by either wildtype or nifDK' bacteria, being smaller in size and lacking the normal pink color of the wildtype and nifDK" induced nodules. To investigate whether plant gene expression is different in 802101-induced nodules than in those induced by 80110 or 80702, the plant nodule proteins were examined for each nodule type. The plant. protein profiles in nodules induced by 80110, 802101 or 80702 are similar but not identical when examined by one dimensional denaturing polyacrylaminde gel electrophoresis. The most notable difference observed in the plant proteins from 802101-induced nodules is the complete absence of the leghemoglobin protein. This is evident in coomassie brilliant blue stained gels (data not shown) but since this protein does not stain effectively with silver, it is not apparent in Figure 26. Leghemoglobin is the major protein in the 80702- and 80110-induced nodules, in agreement with the pink color of these nodules. There are several protein bands which are present in 802101 but absent in both the 80110- and 80702-incited nodules (Fig. 26). An additional difference between these tissues is that the 802101-induced nodule extracts contain less than 10% of the glutamine synthetase activity found in the wildtype 80110-induced nodule extracts; however since the gitEE' 80702-induced nodule extracts contain only 25% of the glutamine synthetase activity found in wildtype nodules, this decrease might not be pggA specific. The nodules induced by 802101 are also abnormal at the ultrastructural level, as illustrated in representative electron micrographs shown in Figure 27. Two weeks after sowing and inoculation, 133 Figure 26. SOS-PAGE of plant protein isolated from root nodules incited by E. japonicum strains 802101 (lanes 1-4), 80702 (lanes 5-8), and 80110 (lanes 9-12). 35 ug (odd numbered lanes) or 70 ug (even numbered lanes) of protein were loaded. Nodules were harvested at 19 days (lanes 1,2,5,6,9,10) or 29 days (lanes 3,4,7,8,11,12) after inoculation and sowing of seeds. Molecular weight markers are shown in lanes M. The heavier staining of the protein from 802101-induced nodules is due to the loading of equal total protein and the absence of a major protein, leghemoglobin, in those lanes (not shown). 134 Z _ N w A mm.wl mmhl ID .3 1 "-1 u_ 1 1.. n ‘ " N_.ml 3.5.1 . g a 3‘- A. mqum 11 manna 1-1-1 43.? 11' 1N_.m 1:... 135 the nodules induced by the wildtype 80110 contain the normal nodule cortex in which a high percentage of the plant cells have been infected and are densely packed with bacteroids contained within peribacteroidal envelopes (Fig. 27a). In contrast, the cortical regions of gggg mutant 802101-induced nodules contain few infected plant cells, and the densely packed bacteria found in these infected cells are not surrounded by peribacteroidal envelopes (Fig. 27b). These 802101-infected plant cells appear to be undergoing a degradative response, having lost the integrity of their cell walls and organelles and their distinct shape. Four weeks after sowing and inoculation the nodules induced by 80110 have enlarged, and their ultrastucture continues to appear normal; the only difference from the younger tissue is the increased accumulation of polyhydroxybutyrate granules in the bacteroids (Fig. 27c). After four weeks the 802101-induced nodules have not enlarged further, and the infected cells continue to degenerate (Fig. 27d). In addition, the four week old 802101-induced nodules contain a large number of bacteria which are not within plant cells but are instead embedded within a darkly staining matrix in intercellular spaces (Fig. 27f). Bacteria in the process of cell division can easily be found within these spaces. The nifDK deletion strain 80702 serves as a control in these experiments, confirming that the effects of the och mutation are not due to the lack of expression of nitrogen fixation. No significant developmental differences are found between nodules induced by wildtype 80110 or nifDK- 80702 (Fig. 27e). The only structural difference detected is the increased accumulation of glycogen granules in the 80702-induced . nodules. This is likely to be a result of the energy excess condition 136 Figure 27. Typical electron micrographs of nodule tissue incited by E. japonicum strain 80110 (a and c), 802101 (b, d and f) or 80702 (e). The tissue was harvested either 2 weeks (a, b and e) or 4 weeks (c, d and f) after seed inoculation and sowing. All micrographs were taken at the same magnification with the bar in panel f representing 1 um. 137 . Q _. , x. ;. .0111... 5.9 . .1... .,,..T I .337 138 resulting from the lack of energy utilization by the nitrogen fixation process. Bacterial gene expression is also abnormal in nodules incited by the odgA mutant 802101 when compared to the wildtype 80110 and the nifDK mutant 80702 (Fig. 24 and Table 7). Bacteria isolated from nodules incited by 80702 contain the nifH, nifDK and glnII transcripts, although their relative abundances are slightly modified from those seen in nodules incited by 80110. In contrast, no transcripts from either gitE, gitEE or gigii can be detected in bacteria isolated from nodules incited by the gggg mutant 802101. It is interesting to note the quantitative effects of these mutations on the expression of the gigii gene, which is expressed at approximately three-fold higher levels in nodules incited by the nitrogen fixation deficient gitEE mutant than in nodules incited by the wildtype bacteria. This effect may be analogous to the Ntr regulation system observed for E. japonicum ex planta and may indicate a nitrogen starved condition within the nodule resulting from the lack of nitrogen fixation. These results agree with measurements of glutamine synthetase activity in other Nif' rhizobial strains (Werner gt gi., 1980). The complete absence of gigii transcription in the 802101-induced nodules suggests two alternative explanations. Under oxygen limiting conditions, the Odc system, acting through the pggA product, rather than the Ntr system, controls the expression of the gigii gene. This could also be the case in nodules. Alternatively, gigii expression in nodules may require some aspect of the proper nodule environment which the 802101-induced nodules lack due to the earlier block in nodule development. 139 Table 7. Abundance of nifH, nifDK and glnII Transcripts in E. japonigum Isolated From Soybean Nodules Bacterial Strain mRNA Abundancea nifH nifDK glnII 1450 276 49 468 169 146 0b 0 0 80110 80702 802101 a . as 1n Table 6 b 0 implies that no transcript was detected 140 DISCUSSION In this paper we report the isolation, physical description and phenotypic characterization of a regulatory gene from the bacterium Eradyrhizobjum japonicum. This gene and the protein it encodes have sequence homology to both the gitA and gttE genes of Klebsiella pneumpniae, although functionally it is not strictly analogous to either of these genes. The Klebsiella gitA gene product functions only as a positive transcriptional\regulator of other git genes, while the gttE-encoded protein has a more general role in controlling the expression of gene systems involved in nitrogen assimilation (for review see Magasanik, 1982; Dixon, 1984). Phenotypically, E. pngumoniag mutants in either gene are Fix', but only gttg mutations have additional phenotypes, being unable to utilize a number of nitrogen sources. The E. japonicum gene described in this report is required for the transcription of nitrogen fixation genes, at least one gene necessary for ammonia assimilation, and for some aspect of the control of cell division when cultures are grown under oxygen limiting conditions. This gene is also essential for proper cellular differentiation during the symbiotic association of E. japonicum with its soybean host. Plants which have been inoculated with a E. igppgiggg strain defective in this gene initiate root nodule development normally, but the interaction becomes abnormal upon release of the bacteria into the host cortical cells. These abnormal nodules lack some nodule-specific plant proteins, such as leghemoglobin, and contain some additional proteins not normally found in nodule tissue. Recently, mutations in this gene were described by Fischer gt gi. 141 (1986), who have labelled it nifA due to its homology with the Klebsiella gene as well as its requirement for git gene expression. These criteria can also be met by the Klebsielig ntrC gene which has a more general function than gitA. Since our results demonstrate that the E. japonicum gene is also of more general function than 5. pngumoniae ifA, we propose to call this new gene gggA in Order to indicate its general regulatory role under oxygen limiting conditions and during symbiotic development. Bacterial mutants have previously been used to study the rhizobial and bradyrhizobial interactions with their host plants. One class of mutants that have been extensively characterized are completely defective in nodulation at the earliest stages in the interaction with the host plant (Kondorosi gt gi., 1984; Downie gt gi., 1985; Jacobs gt gi., 1985). The genes which have been identified in those studies are termed ggg genes, and their mechanism of action is currently under investigation (Mulligan and Long, 1985). Other studies, notably those of Noel gt gi. (1982), Chua gt gi. (1985) and Regensburger gt gi. (1986), have identified genetic loci by mutagenesis which are necessary for the proper progession of the later stages of nodule differentiation. This report, however, is the first to correlate a defect in nodule differentiation with a bacterial deficiency in a specific physiological regulatory system. It Was not completely unexpected that the regulation of the bacterial response to oxygen limited growth conditions would be important for nodule function, since the induction of nitrogen fixation had previously been described during oxygen limited growth in some bradyrhizobia (Keister, 1975; Scott gt gi., 1979). Oxygen limitation 142 has also been Shown to cause a small but detectable induction of uricase, a nodule=specific enzyme, in soybean tissue (Larsen and Jochinsen, 1986). It is obvious that the proper control of oxygen availability is essential to nodule function since molecular oxygen irreversibly inhibits the nitrogenase reaction and yet is essential for the continued oxidative respiration of both the plant and bacteria. This is accepted as the reason for the accumulation of the oxygen binding protein leghemoglobin, which presumably supplies a high flux of molecular oxygen at low partial pressure (for review see Appleby, 1983). However, it was not expected that the elimination of a bacterial low oxygen response system would have the dramatic effect on nodule differentiation described in this report. The results presented here imply a crucial role for the response to a lowered oxygen availability at a stage well before the induction of leghemoglobin accumulation, possibly even before the release of bacteria from the infection thread into the host cells. It is likely that the bacteria are oxygen limited at this stage in the developmental process since they are actively respiring while embedded in the mucilaginous matrix of the infection thread. Ludwig (1984) has presented evidence that, upon a shift to low oxygen conditions, Eradyrhizobium sp. RC3200 can induce the ability to fix nitrogen and coordinately differentiates into nongrowing cells which lose their colony-forming capacity. The limited oxygen growth control phenotype described here for 802101 when compared to 80110 is reminiscent of that reported by Ludwig (1984). This might indicate that the Odc regulatory system we describe is responsible for the differentiation to non-viability described by Ludwig (1984) and thereby iS crucial for proper nodule development. 143 The results presented here give rise to several further questions. What is the mechanism by which gggA controls the expression of E. japonicum genes in response to oxygen limitation? The gggA gene product could act directly as a positive regulator of genes such as nif, glnII, and genes involved in the growth control and developmental phenotypes. Alternatively, the gggA product could act indirectly through the induction Of a regulatory cascade involving at least one secondary regulator. The major precedent for a cascade system are the Nif and Ntr systems of the enterics (for review see Magasanik, 1982; Dixon, 1984). In that case the gitA gene product acts directly on the promoters of the genes it activates, and the gttg product acts indirectly by inducing the expression of the secondary regulator gitA. Since the gttE, gitA and gggA genes and products share sequence homology (Buikema gt gi., 1985; Figure 22) both the direct activation and cascade activation mechanisms remain plausible possibilities. Perhaps knowledge of the regulation of the expression of gggA itself will shed light on these alternatives. It has been shown that the E. japonicum git promoters can be activated by the E. pneumoniae gitA gene product but not by the E. pggggpgigg gttE protein in E. gpii (Alvarez-Morales and Hennecke, 1985). However these heterologous system experiments are known to be capable of yielding specificities which do not directly represent those that occur in the homologous system (e.g. Sundaresan gt gi., 1983; Better gt gi., 1985). A direct genetic analysis of the possible gggA product / git or gigii promoter interaction is certainly warranted. Several aspects of the plant-bacterial symbiotic interaction are perturbed in the pggA mutant strain. In a wildtype interaction, the plant controls the number of effective nodulation events through a 144 negative feedback response termed autoregulation (Pierce and Bauer, 1983) which apparently occurs after the infection of root hairs at the stage of nodule emergence (Calvert gt gi., 1984). Both the wildtype strain (80110) and a Nif' derivative (80702) seem to elicit such a response and control the number of nodules formed, but the gggg mutant strain (802101) does not. This ability to bypass the internal nodulation control mechanisms even during a mixed infection with 80110 and 802101 indicates that the pggA product is required for the expression of some aspect of recognition which is necessary for this autoregulation. The ultrastructural analyses indicate that during the later stages of infection, 802101 is not recognized as a symbiont, but instead appears to be treated more like a pathogen. The infected plant cells exhibit a degradative response which is in some ways reminiscent of the hypersensitive reponses known for many plants during invasion by pathogenic bacteria (for review see Keen and Holliday, 1982). The inability to be recognized as a symbiont could be due either to a lack of expression of some factor(s) or the lack of repression of some free living function(s) which the plant identifies as a characteristic of pathogens. The abnormal oxygen-limited growth control exhibited by the gggA mutant might relate to the latter of these possibilities, with the mutant apparently unable to repress cell division and become non-viable. Along this line, it is interesting to note the identification of a gene which can cause the inhibition of exopolysaccharide production under certain circumstances and which is required for effective nodulation in Egitppigg pggggpii (Borthakur gt gi., 1985). The gggA gene might be involved in regulating the expression of similar functions during nodulation by E. japonicum. An understanding of the mechanisms through 145 which och mediates plant-bacterial recognition will certainly shed light on the control of this process. EXPERIMENTAL PROCEDURES Bacterial Strgins. The wildtype E. japonicum strain used in all experiments is a small-colony derivative of E. japonicum USDA 311b110 isolated as described (Kuykendall and Elkan, 1976; Meyer and Pueppke, 1980), and is designated here as 80110. E. japonicum strains 802101 and 80702 are gggA and nifDK deletion derivatives of 80110 respectively. The Escherichia gpii strains ED8654 (ggiE, ggii,‘ttpE, ggtE, gggE, gng, gng, iggi) and H8101 (Boyer and Roulland-Dussoix, 1969) were used for routine plasmid construction and maintainance. The E. japonicum genomic DNA cosmid library described previously (Adams gt gi., 1984) was maintained in the E. gpii glutamine auxotroph ET8051 03(ggg-gigA), ggtEk, Egg, ggir) which contains a deletion of the entire glnAntrBC oper0n (Fisher gt gi., 1981). E. gpii 0M103 (Messing gt gi., 1981) was used for manipulations of recombinant M13 phage. Recombinant Plasmids, Cosmids, and Phage. pR0cos7-36, pR0c033-44, pR0cos14-28 and pR0cos2-43 are cosmids which were identified in the library described above on the basis of their hybridization to the E. pneumoniae gitA probe. pR0cos1-62 contains the gitE and gitE region of the 80110 genome (Adams gt gi., 1984). p80150 and p80152 are derivatives of pBR328 carrying either the 3.8 or 3.2 kbp gitE-homologous EggRI fragment from pR0cos7-36 respectively. These plasmids were used 146 in making the pBR322 derivative p80210, which was used for construction of the pggg deletion strain 802101 (Fig. 22). The 300 bp EngI-Egpl fragment from p80150 was deleted and replaced with the 1.6 kbp gptii containing EngI-Egpl fragment from pKC7 (Rao and Rogers, 1979). p80210 was then constructed by combining the Eng1 inserts from this p80150 derivative and from p80152 in pBR322. p8070 is a pBR322 derivative which was constructed by deleting 3.6 kbp of E. japonicum DNA between the Eng1 and EggHI restriction sites of pRJ676 (Hennecke, 1981) and replacing this with the 2 kbp gptii containing EngI-EggHI fragment from pKC7 as shown in figure 22c. The E. pneumoniae gitA internal hybridization probe used in these experiments was isolated from p80240 which carries a 947 bp EgtI-EngV fragment from the E. pneumoniae nifLA plasmid pGR397 (Reidel gt gi., 1983) in pIC-19R (Marsh gt gi., 1984). The following recombinant M13 phage were constructed for use in synthesizing single-stranded DNA hybridization probes for S1 protection analyses. M13DK contains a 417 bp EgtI-Epgl restriction fragment from the promoter region of the nifDK operon (Adams and Chelm, 1984) cloned into the polylinker of M13mp19 (Yanisch-Perron gt gi., 1985). M13H has a 704 bp EggHI-EiggIII fragment from gitE and its upstream regions (Adams and Chelm, 1984) cloned into m13mp19. M1311 has a 2.1 kbp EgiI fragment from the E. jgponicum qlnII gene (Carlson and Chelm, 1986) carried in M13mp18. Bagterial Medig. Growth Conditions and Strain Construction. E.' japonicum was routinely cultured on broth and agar plates in YEX medium (Adams gt gi., 1984). Derivatives of wildtype strain 80110 with chromosomal deletions and insertions were constructed by the method 147 described (Guerinot and Chelm, 1986). Gene-directed mutagenesis was carried out by triparental matings with 80110, H8101/pRK2013, and either H8101/p8070/pDS4101 (gitEE) or HBlOl/pBJZlO/pDS4101 (gggA) by the methods of Guerinot and Chelm (1986) with the one exception that chloramphenicol (25 ug/ml) was used in place of rifampicin for the counterselection against E. gpii. Cultures for aerobic gene expression in E. japonicum strains were grown either in formate medium (Manian and O’Gara, 1982) with 0.2% xylose and 10 mM of the designated nitrogen source or on YEMN (0.04% yeast extract, 1% mannitol, 3 mM K HPO 2 4’ M9504, 1.1 mM NaCl, 10 mM KNOB). Oxygen limited cultures of E. 0.8 mM japonigum were grown in YEMN medium continuously sparged with 0.1 % O2 in N2 at a rate of 500 ml/minute. Gas flow rates were controlled using thermal mass flowmeters (Brooks Instruments, Hatfield Pennsylvania, Model 5850C). E. goli strains were routinely grown on LB medium or M9 minimal salts medium with glucose and thiamine (Miller, 1972). Plant Tests. Soybean [Eiygigg mg; (L.) Merr. cv. Amsoy 71] seed sterilization, inoculation, and plant growth were all done as previously described (Guerinot and Chelm, 1986) except that the daily light period was 16 hr. Unless otherwise designated, nodules were harvested 28 days after planting. Nitrogen fixation rates were estimated using the acetylene reduction assay (Hardy gt gi., 1968). No acetylene was reduced to ethylene during a three hour period for strains labelled Fix'. Following the assay, nodules were immediately frozen in liquid nitrogen and stored at -70°C for later use. Prior to freezing, random nodules 148 were chosen, surface sterilized (Guerinot and Chelm, 1986) and the bacteria isolated for strain verification. Nugleic Acids Technigues. Methods for Southern hybridizations, colony hybridization, plasmid and total genomic DNA isolations, and RNA isolations have all been described (Adams and Chelm, 1984; Adams gt gi., 1984). Single-stranded 5’ end-labelled probes for S1 protection analyses were synthesized using the primer extension method as follows. 80 ng of a gene-specific oligonucleotide was labelled by reaction with T4-polynucleotide kinase at 37°C for 1 hr in 100 mM tris(hydroxymethyl)aminomethane (Tris)-HCl, pH 7.5, 10 mM MgCl 6 mM 2’ dithioerythritol and 100 uCi of (1L32P)ATP (3000 Ci/mmole). In experiments with mixed probes, the specific activity of each 5’ end-labelled oligomer was adjusted to approximately 250 cpm/pg with unlabelled oligonucleotide prior to the elongation reaction. The labelled primers were combined with 20 ug of single-stranded M13 recombinant phage, ethanol precipitated, suspended in 50 ul of 10 mM Tris-HCl, pH 8.5 and 10 mM MgCl heated for 5 minutes at 90°C, and 2, allowed to hybridize for 1 hr at 37°C. The hybridized oligonucleotide primer was extended using 3 units (Bethesda Research Laboratory) of the large fragment of E. gpii DNA polymerase (Klenow fragment) for 1 hr at 37°C in 100 ul of the same buffer with 0.6 mM each of ATP, CTP, GTP, and TTP. The reaction was stopped by heating at 65°C for 5 minutes and the double stranded DNA was ethanol precipitated and resuspended in the appropriate restriction endonuclease buffer. The probes specific for the nifH and nifDK promoters were prepared by digesting the product of the extension reaction with EgtNI for 3 hr at 58°C, ethanol 149 precipitated, and suspended in 200 ul of 80 % formamide with tracking dyes. The gigii probe was prepared by digestion with EgiI for 3 hr at 37°C. Extended primers were denatured by incubation at 100°C and purified by electrophoresis on 8.4 M Urea, 4 % polyacrylamide gels. The labelled probe was detected by autoradiography and eluted by the crush and soak method (Maxam and Gilbert,1980). Oligonucleotides for primer extension were synthesized and purified as described previously (Carlson and Chelm, 1986). The sequences of each primer were as follows: 5’-TGCTCTCCATCAACCGA (gitE); 5’-CTTCACCCCGCAGTCCG (0118); and 5’-CGACGCGAATTCCTTGA (gigii). The partially protected fragments from S1 protection experiments with each of the probes synthesized with these primers are 150 bp, 200 bp, and 170 bp respectively. For mixed probe experiments, approximately 6 fmoles of each 5’ end-labelled single-stranded fragment was included in the hybridization. For the driver RNA quantities used in these experiments, this is an excess of probe since increasing RNA concentrations results in proportional increases in the amount of hybrid protection (e.g. Fig. 24). Hybridization, S1 digestion, and identification of partially protected products were as described (Adams ad Chelm, 1984). DNA sequence analysis of the pggA gene region utilized the methods of Maxam and Gilbert (1980). The sequence of the region presented was determined from analysis of both DNA strands. Protein Anglysis. Soluble cytoplasmic proteins from root nodules were prepared as described by Legocki and Verma (1980). Approximately 0.2 g of nodule tissue was homogenized in 500 ul of buffer containing 50 mM Tris-HCl, pH 6.8 and 5 % (v/v) B-mercaptoethanol. Bacteria and plant 150 cell debris were removed by centrifugation at 20,000 x g. Protein concentration was determined by a modification of the Lowry procedure (Markwell gt gi., 1978). Proteins were separated by SOS-polyacrylamide gel electrophoresis (Laemmli, 1970) on 10 to 15 % polyacrylamide gradient gels and were visualized by silver staining (BioRad). -Microscopy. Nodules were harvested from roots, rinsed in deionized water, sliced with a razor to pieces of approximately 1 mm in the largest dimension and then immediately fixed in 4 %.glutaraldehyde in 0.15 M sodium cacadylate, pH 7.2 for 2 hr under vacuum and then at 4°C overnight. Following fixation, the tissue was rinsed three times over a period of 3 hr at room temperature with Buffer A (Maupin and Pollard, 1983) containing 0.2% (w/v) tannic acid and then for 15 min in Buffer A alone. Postfixation was in 1% (w/v) osmium tetroxide in Buffer A for 1 hr at room temperature. The tissue was then dehydrated through a graded ethanol series, cleared in propylene oxide and embedded in VCD/HSXA ultra low viscosity medium (Ladd Research Industries, Burlington, VT). Thin sections were cut on an LKB Ultrotome III with a Dupont diamond knife. Sections were subsequently stained with uranyl acetate and lead citrate and examined in a Philips EM3OO electron microscope. For light microscopy, 2-3 um sections were stained with 0.5% toluidine blue. CHAPTER 7 SUMMARY AND CONCLUSIONS For many nitrogen-fixing bacteria, the control of the expression of genes for nitrogen fixation and nitrogen assimilation are linked through overlapping regulatory systems (Magasanik, 1982; Dixon, 1984). In this way, the atmospheric dinitrogen reduced via nitrogenase to ammonia is utilized to support growth. For rhizobia and bradyrhizobia, nitrogen fixation and nitrogen assimilation are apparently not so tightly coupled since most of the ammonia produced by nitrogenase is exported from the bacteria (Bergerson and Turner, 1967; O’Gara and Shanmugan, 1976). For at least some rhizobial strains, this failure to assimilate fixed nitrogen is partly due to a decrease in glutamine synthetase activity (Bergerson and Turner, 1967; O’Gara and Shanmugan, 1976; Ludwig, 1980; Brown and Dilworth, 1975) and to the induction of an ammonia transport system (Gober and Kashkett, 1983). These properties must be related to the roles of Rhizobium and Bradyrhizobium in symbiotic nitrogen fixation. In this dissertation, I have described the isolation of three transcription units (nifH, nifDK, and nifE) that are required for nitrogen fixation by Eradyrhizobium japonicum strain USDA 110 (80110). I have examined the physiological and developmental parameters required for transcription of these genes. In addition, I have characterized the transcriptional regulation of E. japonicum glutamine synthetase genes 151 152 (gigA and gigii) under conditions that induce git gene transcription. I found that in contrast to enzyme activity data reported previously, transcription of the gene encoding GSII, gigii, is induced coordinately with git gene transcription. Finally, I have described four separate E. japonicum genes that have sequence homology to both the gitA and gttt genes of E. pneumoniae. One of these genes (gggA) is required for both microaerobic and symbiotic induction of E. japonicum git and gigii gene expression. This gene, however, is not required for the aerobic induction of gigii in response to nitrogen limitation. Much of what is known about the regulation of nitrogen fixation and nitrogen assimilation genes in E. japonicum comes from comparison to the well characterized Ntr and Nif regulatory systems of E. pneumoniae. The transcription of Klebsella git genes is controlled by each of these two distinct regulatory systems. This links the expression of nitrogen fixation activity to nitrogen assimilation through the general nitrogen control pathway (Ntr) known for several enteric bacteria (Magasanik, 1982). The addition of a second regulatory mechanism that is Specific for git gene control prevents the expression of the oxygen labile nitrogenase under aerobic conditions. The currentmodel for regulation of E. pneumoniag git genes involves the concerted action of the gttA, gttE, and gth gene products (NTRA, NTRB, and NTRC) on the transcription of the gittA operon (Dixon, 1984). The gitE and gitE products (NIFL and NIFA) in turn control expression from other git operons. Thus, in nitrogen limited culture, NTRC works in conjunction with NTRA, a putative sigma factor for recognition of Ntr-activated promoters (Hirschman gt gi., 1985; Hunt and Magasanik, 1985), to induce transcription from the nifLA promoter. Under these conditions, NTRC and 153 NTRA are also needed to activate the transcription of genes involved in nitrogen assimilation including gigA (glutamine synthetase), pgt (proline utilization genes), ggt (arginine utilization genes), and ggt (histidine utilization genes). Under anaerobic, nitrogen-limited conditions, NIFA then serves to activate transcription from other git operons by a mechanism that also requires NTRA. The function of NIFL is to repress git transcription when oxygen is present. Finally, NTRB acts to modulate the activity of NTRC in response to the availability of ' fixed nitrogen. The promoter sequences for git genes in E. japonigum share a high degree of nucleotide sequence homology with those identified for the NIFA/NTRA and NTRC/NTRA regulated promoters in E. pneumoniae (Chapters 2 and 3) supporting the hypothesis that E. japonicum git genes may be controlled by similar regulatory mechanisms. 1 have described the isolation and characterization of four different E. japonigum genes that have sequence homology to both the E. pneumoniae gitA andgttE genes. One of these potential regulatory genes has been shown to have an effect on the transcription of git genes both symbiotically and in microaerobic culture. This gene is also required for the microaerobic and symbiotic expression of the gene encoding glutamine synthetase II (gigii), and for some aspect of controlling cell division or cell viability in oxygen limited cultures. This same gene is also needed for normal development of the symbiotic association between E. igppgiggg and its soybean host. I suggest that this gene is part of a general regulatory system that is necessary for oxygen availability and developmentally mediated control of gene expression. This regulatory system has been termed Odc and the 154 gitE-homologous gene required for activity of this system has been called gggA as the first gene identified for this regulatory pathway. Interpreting the effects of the gggA deletion mutation on I microaerobic gene expression is complicated by the complex relationship between oxygen availability and nitrogen limitation in E. japonicum. Growth rate and final cell density of microaerobic cultures in YEMN medium can be stimulated by the addition of ammonia. Similar changes in growth properties can be achieved by increasing oxygen levels indicating that the amino acids present in yeast extract do not limit aerobic growth. This implies that growth limitation during microaerobiosis results, at least in part, from nitrogen limitation brought about through changes in the cells ability to utilize amino acids as a source of nitrogen. Thus, the microaeroabic induction of git and gigii gene transcription might be due to nitrogen limitation, oxygen limitation, or a combination of the two. It is possible that the expression of pggA activity is induced in microaerobic culture by another regulatory protein in response to the nitrogen limiting growth conditions observed. This would be analogous to the NTRC mediated activation of the gittA promoter under nitrogen limited growth conditions in E. pneumoniae. If this occurred in E. japonicum, one might expect that microaerobic gigii transcription, like aerobic gigii transcription, would be independent of the gggA gene. This is not the case, however, clarification of this question awaits the isolation of‘a E. igppgiggg gttE-like mutant strain. Alternatively, ODCA activity might be the cause rather than the effect of microaerobic nitrogen limitation. Low oxygen tensions have been shown to induce ammonia exp0rt in some wildtype Bradyrhizobium strains (O’Gara and 155 Shanmugan, 1976; Ludwig, 1980; Gober and Kashkett, 1983). Active ammonia excretion might result in nitrogen limited growth even if cultures were supplied with a utilizable nitrogen source like amino acids. Addition of ammonia to these cultures might partially relieve nitrogen limitation and stimulate growth by shifting the intracellular/extracellular ammonia equilibrium concentrations. In the extreme case, nitrogen limitation brought about by the export of fixed nitrogen, might lead to a loss of cell viability as the cells deplete their internal nitrogen stores such as protein. Ludwig (1984) observed that cells grown microaerobically with a growth limiting nitrogen supply export ammonia and lose cell viability. This loss of cell viability is most dramatic at oxygen tensions that are too high for nitrogen fixation. From these experiments Ludwig (1984) suggested a model for development of bradyrhizobial nitrogen fixation activity. This model implies that during microaeobiosis, a subpopulation of cells in a culture switch their metabolic state to one in which nitrogen is fixed but growth is repressed. This metabolic change could be accomplished if the induction of git genes were directly coupled to the induction of ammonia transport genes. I propose that ODCA is responsible for the induction of the ammonia export system as well as the transcription of git genes. Such a role for ODCA could explain the effect of gggA mutation on cell viability in microaerobic culture. The effects of gggA mutation on ammonia export should be examined. It is difficult to underStand why Etggytgitppigg should have evolved a system for ammonia export that is active even under nitrogen limiting growth conditions. Perhaps this is a colonial survival mechanism in which some bacteria export their nitrogen reserves to 156 supply other bacteria in the colony with fixed nitrogen thus resulting in syntrophic growth. This behavior becomes most useful when, as described by Ludwig (1984), these non-growing ammonia-exporting bacteria also fix nitrogen. For the model described above, gigii expression could be explained by either the direct action of ODCA in inducing the gigii promoter, or the indirect effect of ODCA on ammonia limitation which then results in a Ntr-like response. A E. mgliloti mutant strain that carries a transposon insertion in the gigii gene is unable to fix nitrogen in root nodules (M. Kahn, personal communication). In contrast, the E. meliloti gigA gene is not required for symbiotic nitrogen fixation (Somerville and Kahn, 1983). The need for gigii during symbiotic development could relate to the observation that the gene is induced during microaerobiosis by a mechanism requiring ODCA. High concentrations of ammonia present in nitrogen fixing root nodules would repress the Ntr-like activation of the gflgii promoter and lead to the adenylylation of GSI. By providing a second mechanism for the activation of gigii transcription that does not necessarily require nitrogen limitation, GSII activity could be induced in bacteroids and allow some ammonia assimilation for bacterial growth and differentiation. Alternatively, it has been suggested that the bradyrizobial GSII may have some alternative function beyond its role in glutamine synthesis (Ludwig, 1980b). gigii and gigA mutant strains of E. jgppgiggg will be helpful for understanding the respective roles of GSII and 031 in soybean root nodules. The function of gggA in control of normal root nodule development is not yet clear. The results presented in Chapter 6 indicate that och 157 is essential even at very early stages in nodule differentiation. Ultrastructural analyses demonstrate that 802101 (the pggA deletion strain) is treated more as a pathogen than a symbiont when in association with the plant. Infected plant cells exhibit a degenerative response that is reminiscent of the hypersensitive responses described ,for several plants during invasion by pathogenic bacteria (Keen and Holliday, 1982). The inability of 802101 to be recognized as a symbiont could be due either to a lack of expression of some factor(s) or the lack of repression of some free-living function(s) that the plant identifies as characteristic of pathogens. One possibility is that the ammonia export system described above is required for normal symbiosis. Perhaps Bradyrhizobium evades the hypersensitive response by feeding the plant back some of the nitrogen released from amino acid degradation (Kahn gt gi., 1985). If gggA is required for induction of the ammonia export system then 802101 would be unable to accomplish this evasive mechanism. ODCA is required by E. japonicum for the expression of several genes during microaerobic growth. The mechanism of ODCA action as a positive regulator of genes such as git and gigii could be either direct at the promoters for these genes, or indirect through the induction of a regulatory cascade involving at least one secondary regulator. The second possibility would be analogous to the situation in E. pggggpgigg where NTRC controls git gene expression indirectly by activating the transcription of gitA. Some evidence for a direct role of ODCA in activating transcription from E. igppgiggg git promoters comes from t al. (1986) in which they demonstrated that experiments of Fischer ODCA can activate the E. japonicum nifD promoter in a heterologous E. 158 gpii expression system that requires the E. gpii,gttA gene. However, these heterologous system experiments are known to be capable of yielding specificities that do not directly represent those that occur in the homologous system (Better gt gi., 1985; Sundaresan gt gi., 1983). A more direct genetic analysis of the possible ODCA / git or gigii promoter interaction is needed. The mechanism by which oxygen limitation is sensed in E. japonicum and how this stress is transmitted to ODCA activity is unknown. In E. pnepmoniae the activities of both NTRC and NIFA are controlled through other regulatory components that somehow sense relevant aspects of cellular physiology. Hennecke and co-workers (personal communication) have found that ODCA mediated activation of transcription from the E. japonicum gitE promoter in E. gpii is sensitive to oxygen. No Such oxygen sensitivity has been observed for the E. pneumoniae NIFA protein. This indicates that the ODCA protein may sense oxygen stress, directly. However, these experiments are complicated by the complex physiological differences between aerobic and microaerobic growth. The development of ig yittp systems for ODCA mediated transcriptional control should be useful for elucidating this regulatory pathway. The functions of other gitE-like gene regions in E. jgppgiggg are unknown. Escherichia gpii and Rhizobium meliloti also appear to contain multiple gitE-like genes indicating that this property is likely to be generally found in many bacteria. At present, only one gitE-like gene has been described in E. gpii (Buikema gt gi., 1985) and two in E. ggiiipti (Buikema gt_gi., 1985; Ausubel gt gi., 1985). It is possible that these different gitE-homologous regions encode a family of NIFA-like regulatory proteins responsible for controlling different gene 159 systems. The homologies described here might then represent sequences that are required for interaction with NTRA or NTRA-like proteins. The four separate E. jappgicum mutant strains (each with a deletion in a single Egg region) described in Chapter 4 have been examined for alterations in Nif-like or Ntr-like transcriptional control systems. Only the gggi (gggA) mutant discussed above has an easily discernible regulatory phenotype. Of the other ggg mutant strains, only 801011, the gggg deletion strain, has any obvious alterations in growth and symbiotic development. Nitrogen fixation activity in soybean root nodules formed by infection with wildtype E. japonigum (80110) usually begins about 12 days after sowing and inoculation. The onset of nitrogen fixation in soybean nodules induced by infection with 801011 is delayed at least one week in comparison to wildtype infected nodules. Some git genes are expressed normally during this lag in the initiation of nitrogen fixation suggesting that a general alteration in git gene control has not occurred. 801011 is also perturbed in its free-living growth characteristics and grows more slowly than 80110 under all media conditions tested. The slow growth phenotype of 801011 might account for the observed lag in development for nodules induced by this strain. Alternatively, Egg; might encode a regulatory product that is required for the early establishment of nitrogenase activity. This function would either not be required in older nodules or some other gene product (perhaps HNA3 or HNAS) might be able to replace HNA2. The construction of E. japonicum strains with deletions in more than one ggg gene region will be useful in understanding the role of these genes. Finally, E. igpggiggg contains other gitE-homologous regions that I have not described in this dissertation. Two of these genes were 160 isolated recently (G. Martin, unpublished) in our laboratory. These two E. japonicum regions were identified by their homology to a gene isolated from E. parasponia (Nixon gt gi., In Press). No mutation in the E. paragponja gene has yet been described but this gene does have a very similar derived amino acid sequence to the E. pneumoniag gttE gene and is linked to a gene that is homologous to gth, another gene in the general nitrogen regulatory pathway (Magasanik, 1982). It will be interesting to see the effects of mutations in these genes and other gitE-like and gttE-like sequences on bradyrhizobial gene expression. The isolation of a E. japonicum mutant strain that is deficient in Ntr-like control of gene expression will be invaluable in answering many of the questions raised by this dissertation. REFERENCES Adams, T. H., and Chelm, B. K. (1984). The gth and gitEE promoter regions from Rhizobium japonicum share structural homologies with each other and with nitrogen-regulated promoters from other organisms. 0. Mol. Appl. Genet. 2, 392-405. Adams, T. H., McClung, C. R., and Chelm, B. K. (1984). Physical organization of the Bradyrhizobium japonicum nitrogenase gene region. 0. Bacteriol. 159, 857-862. Adams, T. H., Pankratz, H. S., and Chelm, B. K. (In Press). Characterization of a bacterial gene required for the normal differentiation of the Bradyrhizobium japonicum / soybean symbiotic interaction. J. Bacteriol. Alvarez-Morales, A., and Hennecke, H. (1985). Expression of Rhizobium japonicum nifH and nifDK operons can be activated by the Klebsiella pneumoniae nifA protein but not by the product of ntrC. Mol. Gen. Genet. 199, 306-314. Alvarez-Morales, A., Betancourt-Alvarez, M., Kaluza, K., and Hennecke, H. (1986). Activation of the Etgdyrhizobiqg japonicum nifH and nifDK operons is dependent on promoter-upstream DNA sequences. Nucleic Acids Res. 14, 4207-4227. Ames, G. F.-L., and Nikaido, K. (1985). Nitrogen regulation in Salmonella typhimurium. Identification of an ntrC protein-binding site and definition of a consensus binding sequence. EMBO J. 4, 539-547. Appleby, C. A. (1984). Leghemoglobin and Rhizobium respiration. Ann. Rev. Plant Physiol. 35, 443-478. Ausubel, F. M. (1984). Regulation of nitrogen fixation genes. Cell 37, 5-6. Ausubel, F. M., Buikema, W. 0., Earl, C. D., Klingensmith, 0. A., Nixon, B. T., and Szeto, W. W. (1985). Organization and regulation of Rhizobium meliloti and Parasponia Bradyrhizobium nitrogen fixation genes. in Nitrogen Fixation Research Progress, H. 0. Evans, P. 0. Bottomley, and W. E. Newton eds. (Martinus Nino f Publishers: Dordrecht, The Netherlands) pp. 165-171. Banfalvi, Z., Sakanyan, V., Konez, C., Kiss, A, Dusha, 1., and ' Kondorosi, A. (1981). Location of nodulation and nitrogen fixation 161 162 genes on a high molecular weight plasmid of E. meliloti. Mol. Gen. Genet. 184, 318-325. Benton, W. D., and Davis R. W. (1977). Screening gt recombinant clones by hybridization to single plaques ig situ. Science 196, 180-182. Bergerson, F. 0., and Turner, G. L. (1967). Nitrogen fixation by the bacteroid fraction of breis of soybean root nodules. Biochim. Biophys. Acta 141, 507-515. Bergerson, F. 0., and Turner G. L. (1978). Activity of nitrogenase and glutamine synthetase in relation to availability of oxygen in continuous cultures of a strain of cowpea Rhizobium sp. supplied with excess ammonium. Biochim. Biophys. Acta 538, 406-416. Bergerson, F. L., and Turner, G. L. (1980). Properties of terminal oxidase systems of bacteroids from root nodules of soybean and cowpea and of N -fixing bacteria grown in continuous culture. 0. Gen. MicrobioI. 118, 235-252. Berk, A. 0., and Sharp, P. A. (1977). Sizing and mapping early adenovirus mRNAs by gel electrophoresis of S1 endonuclease-digested hybrids. Cell 12, 721-732. Better, M., Ditta, G., and Helinski, D. R. (1985). Deletion analysis of Rhizobium melilpti symbiotic promoters. EMBO J. 4, 2419-2424. Beynon, 0., Cannon, M., Buchanan-Wollaston, V., and Cannon, F. (1983). The git promoters of Klebsiella pneumoniae have a characteristic primary structure. Cell 34, 665-671. Bishop, P. E., Guevara, 0. G., Engelke, 0. A., and Evans, H. 0. (1976). On the relation between glutamine synthetase and nitrogenase activities in the symbiotic association between Rhizobium japonicum and Glycine ggE. Plant Physiol. 57,542-546. Bisseling, T., van den Bos, R. C., Weststrate, M. W., Haddaart, M. 0. 0., and van Kammen, A. (1979). Development of the nitrogen-fixing and protein-synthesizing apparatus of bacteroids in pea root nodules. Biochim. Biophys. Acta 562, 515-526. Borer, P. M., Dengler, 8., and Tinoco, I. (1974). Stability of ribonucleic acid double-stranded helices. 0. Mol. Biol. 86, 843-853. Borthakur, D., Downie, 0. A., Johnston, A. W. 8., and Lamb, 0. W. (1985). pgi, a plasmid-linked Rhizobium phaseoli gene that inhibits exopolysaccharide production and which is required for symbiotic nitrogen fixation. Mol. Gen. Genet. 200, 278-282. Boyer, H. W., and Roulland-Dussoix, D. (1969). A complementation analysis of the restriction and modification of DNA in Escherichia coli. 0. Mol. Biol. 41, 459-472. 163 Brown, C. M., and Dilworth, M. 0. (1975). Ammonia assimilation by Rhizobium cultures and bacteroids. 0. Gen. Nicrobiol. 86, 39-48. Buchanan-Wollaston, V., Cannon, M. C., and Cannon, F. C. (1981a). The use of cloned git (nitrogen fixation) DNA to investigate transcriptional regulation of git expression in Klebsiglla pneumoniag. Mol. Gen. Genet. 184, 102-106. Buchanan-Wollaston, V., Cannon, M. C., Beynon 0. L., and Cannon, F. C. (1981b). Role of the nifA gene product in the regulation on git expression in Klgpsiella pneumoniae. Nature 294, 776-778. Buck, M., Khan, H., and Dixon, R. (1985). Site-directed mutagenesis of the Klgbsiglla pneumoniae nifL and nifH promoters and ig vivo analysis of promoter activity. Nucleic Acids Res. 13, 7621-7637. Buck, M. (1986). Deletion analysis of the Klebsiella pneumoniae nitrogenase promoter: importance of spacing between conserved sequences around positions -12 and -24 for activation by the nifA and ntrC (glnG) products. J. Bacteriol. 166, 545-551. Bueno, R., Pahel, G., and Magasanik, B. (1985). Role of glnB and glnD gene products in regulation of the glnAlG operon of Eschgrichig goli. J. Bacteriol. 164, 816-822. Buikema, W. 0., Szeto, W. W., and Ausubel, F. M. (1985). Nitrogen fixation specific regulatory proteins of Klebsiella pneumoniae and Rhizobium meliloti share amino acid homology with a general nitrogen regulatory protein of K. pneumoniae. Nucleic Acids Res. 13, 4539-4555. Burris, R. H. (1980). The global nitrogen budget - Science or seance? in Nitrogen Fixation, Vol.1: Free-living systems and chemical models, W. E. Newton and W. H. Orme-Johnson, eds. (University Park Press: Baltimore, MD) pp. 7-16. Calvert, H.E., Pence, M.K., Pierce, M., Malik, N.S.A. and Bauer, W.D. (1984). Anatomical analysis of the development and distribution of Rhizobium infections in soybean roots. Can. 0. Bot. 62, 2375-2384. Cannon, F. C., Reidel, G. E., and Ausubel, F. M. (1979). Overlapping sequences of Klebsiella pneumoniae git DNA cloned and characterized. Mol. Gen. Genet. 174, 59-66. Cannon, M., Hill, 5., Kavanaugh, E., and Cannon, F. (1985). A molecular genetic study of git expression in Klebsiella pneumoniae at the level of transcription, translation and nitrogenase activity. Mol. Gen. Genet. 198, 198-206. Carlson, T. A., Guerinot, M. L., and Chelm, B. K. (1983). Isolation of Rhizobium japonicum glutamine synthetase genes. in Plant Molectular Biology, UCLA Symposia on Molecular Biology, R. B. Goldberg ed. (Alan R. Liss: New York) New series vol. 12, 291-302. 164 Carlson, T. A., Guerinot, M. L., and Chelm, B. K. (1985). Characterization of the gene encoding glutamine synthetase I (glnA) from Eragyrhizobigm japonicum. 0. Bacteriol. 162, 698-703. Carlson, T. A., and Chelm, B. K. (1986). Apparent eucaryotic origin of glutamine synthetase II from the bacterium Bradyrhizobium japonicum. Nature, in press. Carlson, T. A. (1986). Characterization of the genes encoding glutamine synthetase I and glutamine synthetase II from Bradyrhizobium japonicum. Phd dissertion, Michigan State University Castano, I. and Bastarrachea, F. (1984). glnF-lac; fusions in Escherichig goli studies on glnF expression and its chromosomal orientation. Mol. Gen. Genet. 195, 228-233. Chelm, B. K., and Hallick, R. B. (1976). Changes in the expression of the chloroplast genome of Euglgna gracilus during chloroplast development. Biochemistry 15, 593-599. Chen, 0. S., Multani, 0. S., and Mortenson, L. E. (1973). Structural investigation of nitrogenase components from Clostridium pasteurianum and comparison with simialar components of other organisms. Biochim. Biophys. Acta 310, 51-59. Chen, Y.-M., Backman, K., Magasanik, B. (1982). Characterization of a gene, glnL, the product of which is involved in the regulation of nitrogen utilization in Egcherichig coli. 0. Bacteriol. 150, 214-220. Ching, T. M., Hedke, S., and Newcomb, W. (1977). Isolation of bacteria, transforming bacteria, and bacteroids from soybean nodules. Plant Physiol. 60, 771-774. Chua, K.-Y., Pankhurst, C. E., MacDonald, P. E., Hopcroft, D. H., Jarvis, B. D. W., and Scott, D. B. (1985). Isolation and characterization of transposon Tn5-induced symbiotic mutants of Rhizobium loti. J. Bacteriol. 162, 335-343. Clewell, 0. 8., and Helinski,‘D. R. (1972). Effect of growth conditions on the formation of the relaxed complex of supercoiled ColEl deoxyribonucleotic acid and protein in Escherichia coli. J. Bacteriol. 110, 1135-1146. Corbin, D., Ditta, G., and Helinski, D. R. (1982). Clustering of nitrogen fixation (tit) genes in Rhizobium meliloti. J. Bacteriol. 149, 221-228. Corbin, D., Barran, L., and Ditta, G. (1983). Organization and expression of Rhizobium meliloti nitrogen fixation genes. Proc. Natl. Acad. Sci. USA 80, 3005-3009. 165 Cullimore, J. V., and Miflin, B. J. (1983). Glutamine synthetase from the plant fraction of Phaseolus root nodules: purification of the mRNA and ig vitro synthesis of the enzyme FEBS lett. 158, 107-112. Daniel, R. M., Limmer, A. W., Steele, K. W., and Smith, I. M. (1982). Anaerobic growth, nitrate reduction and denitrification in 46 Rhizobium strains. 0. Gen. Microbiol. 128, 1811-1815. Darrow, R. A., and Knotts, R. R. (1977). Two forms of glutamine synthetase in free-living root-nodule bacteria. Biochem. Biophys. Res. Comm. 78, 554-559. Darrow, R. A. (1980). Role of glutamine synthetase in nitrogen fixation, in Glutamine synthetase: metabolism, enzymology and regulation, 0. Nora and R. Palacios eds. (Academic Press, Inc.:N. Y.) pp. 139-166. Darrow, R. A., Crist, D., Evans, W. R., Jones, 8. L., Keister, D. L., and Knotts, R. R. (1981). Biochemical and physiological studies on the two glutamine synthetases of Rhizobium, in Current perspectives in nitrogen fixation, A. H. Gibson and W. E. Newton, eds. (Australian Academy of Science: Canberra) pp. 182-185. Davis, R. W., Botstein, D., and Roth, 0. R. (1980). Advanced bacterial genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. Davis, R. W., and Thomas, M. (1974). Studies on the cleavage of bacteriophage lambda DNA with Eng1 restriction endonuclease. J. Mol. Biol. 91, 315-328. de Bruijn, F. 0., and Ausubel F. M. (1981). The cloning and transposon Tn5 mutagenesis of the glnA region of Klebsiella pneumoniae: Identification of glnR, a gene involved in the regulation of the git and ggt operons. Mol. Gen. Genet. 183, 289-297. de Bruijn, F. 0., and Ausubel, F. M. (1983). The cloning and characterization of the qlnF (ntrA) gene of Klebsiella pneumoniae: role of qlnF (ntrA) in the regulation of nitrogen fixation (git) and other nitrogen assimilatory genes. Mol. Gen. Genet. 192, 342-353. Denhardt, D. T. (1966). A membrane filter technique for the detection of complementary DNA. Biochem. Biophys. Res. Commun. 23, 641-646. Ditta, G., Stanfield, S., Corbin, D., and Helinski, D. R. (1980). Broad host range DNA cloning system for gram-negative bacteria: Construction of a gene bank of Rhizobium meliloti. Proc Natl. Acad. " Sci. 77, 7347-7351. Dixon, R. A. (1984). The genetic complexity of nitrogen fixation. J. Gen. Microbiol. 130, 2745-2755. Dixon, R., Eady, R. R., Espin, G., Hill, 5., Iaccarino, M., Kahn, D., and Merrick, M. (1980). Analysis of regulation of Klebsielig pneumoniae 166 nitrogen fixation (git) gene cluster with gene fusions. Nature 286, 128-132. Donald, R. G., Nees, D. W., Raymond, C. K., Loroch, A. 1., and Ludwig, R. A. (1986). Characterization of three genomic loci encoding Rhizobium sp. strain ORSS71 N2 fixation genes. J. Bacteriol. 165, 72-81. Downie, J. A., Ma, Q.-S., Knight, C. D., Hombrecher, G., and Johnston, A. W. B. (1983). Cloning of the symbiotic region of Rhizobium lgguminosarum: the nodulation genes are between the nitrogenase genes and a nifA- like gene. EMBO J. 2, 947- 952. Downie, J. A., Knight, C. D., Johnston, A. W. 8., and Rossen, L. (1985). Identification of genes and gene products involved in the nodulation of peas by Rhizobium leguminosarum. Mol. Gen. Genet. 198, 255-262. Drummond, M., Clements, 0., Merrick, M., and Dixon, R. (1983). Positive control and autogenous regulation of the nifLA promoter in Elebsiella pneumoniae. Nature 301, 302-307. Drummond, M., Whitty, P., and Wootton, J. (1986). Sequence and domain relationships of ntrC and nifA from Klebsiella pneumoniae: homologies to other regulatory proteins. EMBO J. 5, 441-447. Drury, L. S., and Buxton, R. S. (1985). DNA sequence analysis of the gyg gene of Escherichia coli reveals amino acid homology between the Dye and OmpR proteins. J. Biol. Chem. 260, 4236-4242. Eady, R. R., and Postgate, J. R. (1974). Nitrogenase. Nature 249, 805-810. Emerich, D. W., and Burris, R. H. (1978). Complementary functioning of the component proteins of nitrogenase from several bacteria. J. Bacteriol. 134, 936-943. Espin, G., Alvarez-Morales, A., and Merrick, M.' (1981). Complementation analysis of glnA-linked mutations which affect nitrogen fixation in Klebsiellg pneumoniae. Mol. Gen. Genet. 184, 213-217. Espin, G, Alvarez- Morales, A, Cannon, F. Dixon, R., and Merrick, M. (1982). Cloning of the glnA, ntrB and ntrC genes of Klebsiella pneumoniae and studies of their role in regulation of the nitrogen fixation (nif) gene cluster. Mol. Gen. Genet. 186, 518- 524. Filser, M., Merrick, M., and Cannon, F. (1983). Cloning and characterisation of nifLA regulatory mutations from Klebsiella pneumoniae. Mol. Gen. Genet. 191, 485-491. Fischer, H. M Alvarez- Morales, A. , and Hennecke,H . (1986). The pleiotropic nature of symbiotic regulatory mutants: Bradyrhizobium japonicum nifA gene is involved in control of nif gene expression and formation of determinate symbiosis. EMBO 0. 5,1165- 1173. 167 Fischer, H.-M., Alvarez-Morales, A., Ebeling, S., Thony, B., Ramseier, T., Hahn, M., and Hennecke, H. (1986). Organization and regulation of symbiotic nitrogen fixation genes in Bradyrhizobium japonicum. Abstract for the Third International Symposium on the Molecular Genetics of Plant-Microbe Interactions. p. 184. . Fisher, R., Tuli, R., and Haselkorn, R. (1981). A cloned cyanobacterial gene for glutamine synthetase functions in Escherichia coli, but the enzyme is not adenylylated. Proc. Natl. Acad. Sci. USA 78, 3393-3397. Friedman, A. M., Long, S. R., Brown, S. E., Buikema, W. J., and Ausubel, F. M. (1982). Construction of a broad host range cosmid cloning vector and its use in the genetic analysis of Rhizobium mutants. Gene 18, 289-296. Friedman, D. I., and Gottesman, M. (1983). Lytic mode of lambda development. in Lambda II, R. W. Hendrix, J. W. Roberts, F. W. Stahl, and R. A. Weisberg eds. (Cold Spring Harbor Laboratory: Cold Spring Harbor, New York) pp. 21-52. Fuchs, R. L., and Keister, D. L. (1980). Comparative properties of glutamine synthetases I and II in Rhizobium and Agrobacterium spp. J. Bacteriol. 144, 641-648. Fuhrmann, M., and Hennecke, H. (1982). Coding properties of cloned nitrogenase structural genes from Rhizobium japonicum. Mol. Gen. Genet. 187, 419-425. Fuhrmann, M., Fischer, H.-M., and Hennecke, H. (1985). Mapping of Rhizobium japonicum nifB-, fixBC- and fixA-like genes and identification of the fixA promoter. Mol. Gen. Genet. 199,315-322. Gicquel-Sanzey, B., and Cossart, P. (1982). Homologies between different procaryotic DNA-binding regulatory proteins and between their sites of action. EMBD J. 1, 591-595. Gober, J. W., and kashkett, E. R. (1983). Methylammonium uptake by Rhizobium sp. strain 32H1. J. Bacteriol. 153, 1196-1201. Guerinot, M. L. and Chelm, B. K. (1986). Bacterial 5-aminolevulinic acid synthase activity is not essential for leghemoglobin formation in the soybean / Bradyrhizobium japonicum symbiosis. Proc. Natl. Acad. Sci. USA 83, 1837-1841. Gussin, G. N., Ronson, C. W., and Ausubel, F. M. (In Press). Regulation of nitrogen fixation genes. Ann. Rev. Genet. Hardy, R. W. F., Holsten, R. D., Jackson, E. K., and Burns, R. C. (1968). The acetylene-ethylene assay for N2 fixation: laboratory and field evaluation. Plant Physiol. 43, 1185-1207. Hennecke, H. (1981). Recombinant plasmids carrying nitrogen fixation genes from Rhizobium japonicum. Nature 291, 354-355. 168 Hennecke, H., Alvarez-Morales, A., Betancourt-Alvarez, M., Ebeling, S., Filser, M., Fischer, H.-M., Gubler, M., Hahn, M., Kaluza, K., Lamb, J. W., Meyer, L., Regensburgere, B., Studer, D., and Weber, J. (1985). Organization and regulation of symbiotic nitrogen fixation genes from Bradyrhizobium japonicum. in Nitrogen Fixation Research Progress, H. J. Evans, P. J. Bottomley, and W. E. Newton eds. (Martinus Nijhoff publishers:Dordrecht, The Netherlands) pp.157-163. Hill, S., Kennedy, C., Kavanaugh, E., Goldberg, R. B., and Hanau, R. (1981). Nitrogen fixation gene (nifL) involved in oxygen regulation of nitrogenase synthesis in K. pneumoniae. Nature 290, 424-426. Hirschman, J., Wong, P.-K., Sei, K., Keener, J., and Kustu, S. (1985). Products of nitrogen regulatory genes ntrA and ntrC of enteric bacteria activate glnA transcription in vitro: Evidence that the ntrA product is a crfactor. Proc. Natl. Acad. Sci. USA 82, 7525-7529. Ho, Y. S., Wulff, D. L., and Rosenberg, M. (1983). Bacteriophage protein cII binds promoters on the opposite face of the DNA helix from RNA polymerase. Nature 304, 703-708. Hollis, A. 3., Kloos, w. E., and Elkan, o. H. (1981). DNAzDNA hybridization studies of Rhizobium japonicum and related Rhizobiaceae. J. Gen. Microbiol. 123, 215-222. Hunt, T. P., and Magasanik,B. (1985). Transcription of glnA by purified E§cherichia coli components: Core RNA polymerase and the products of glnF, glnG, and glnL. Proc. Natl. Acad. Sci. USA 82, 8453-8457. ' Jacobs, T. W., Egelhoff, T. T., and Long S. R. (1985). Physical and genetic map of a Rhizobium meliloti nodulation gene region and nucleotide sequence of nodC. J. Bacteriol. 162, 469-476. Jayadumar, A., Schulman, I., MacNeil, D., and Barnes, E. M. Jr. (1986). Role of the Escherichia coli glnALG operon in regulation of ammonium transport. J. Bacteriol. 166, 281-284. Johnson, G. V., Evans, H. J., and Ching, T. M. (1966). Enzymes of the glyoxylate cycle in rhizobia and nodules of legumes. Plant Physiol. 41, 1330-1336. Kahn, M. L., Krauss, J., Somerville, J. E. (1985). A model of nutrient exchange in the Rhizobium-legume symbiosis. in Nitrogen Fixation Research Progress, H. J. Evans, P. J. Bottomley, W. E. Newton eds. (Martinus Nijhoff Publishers: Dordrecht, The Netherlands) pp. 193-199. Kaluza, K., and Hennecke, H. (1982). The nitrogenase genes of Klebsiella pneumoniae are transcribed into a single polycistronic mRNA subsequent to nifLA expression. FEMS Microbiol. Letters 15, 57-60. 169 Kaluza, K., Fuhrmann, M., Hahn, M., Regensburger, B., and Hennecke, H. (1983). In Rhizobium japonicum the nitrogenase genes nifH and nifDK are separated. J. Bacteriol. 155, 915-918. Keen, N. T., and Holliday, M. J. (1982). Recognition of bacterial pathogens by Plants. in Phytopathogenic Prokaryotes vol.2, M. S. Mount and G. H. Lacy, eds. (Academic Press: New York) pp.179-217. Keister, D. L. (1975). Acetylene reduction by pure cultures of Rhizobia. J. Bacteriol. 123, 1265-1268. Khoury, G., and Gruss, P. (1983). Enhancer elements. Cell 33, 313-314. Kondorosi, A, Svab, 2., Kiss, G. B. , and Dixon, R. A. (1977). Physical and genetic analysis of a symbiotic nitrogen fixation in Rhizobium meliloti. Mol. Gen. Genet. 151, 221- 226. Kondorosi, E., Banfalvi, Z., and Kondorosi, A. (1984). Physical and genetic analysis of a symbiotic region of Rhizobium meliloti: identification of nodulation genes. Mol. Gen. Genet. 193, 445-452. Krol, A. J. M., Hontelez, J. G. J., Roozendaal, B., and van Kammen, A. (1982). On the operon structure of the nitrogenase genes of Rhizobium leguminosarum and Azotobacter vinelandii. Nucleic Acids Res. 10, 4147-4157. Kurz, W. G. W. , and Larue, T. A. (1975). Nitrogenase activity in rhizobia in absence of plant host. Nature 256, 407-408. Kuykendall, L.D., and Elkan, G. H. (1976). Rhizobium japonicum derivatives differing in nitrogen fixing efficiency and carbohydrate utilization. Appl. Envir. Microbiol. 32, 511-519. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature New Biol. 227, 680-685. Lamb, J. W., and Hennecke, H. (1986). In Bradyrhizobium japonicum the common nodulation genes, nodABC, are linked to nifA and fixA. Mol. Gen. Genet. 202, 512-517. Lara, M., Cullimore, J. V., Lea, P. J., Miflin. B. J., Johnston, A. W. B., and Lamb, J. W. (1983). Appearance of a novel form of plant glutamine synthetase during nodule development in Phaseolus vulgaris L. Planta 157, 254-258. Larsen, K., and Jochimsen, B. U. (1986). Expression of nodule-specific uricase in soybean callus tissue is regulated by oxygen. EMBO J. 5, 15-19. 170 Legocki, R. P., and Verma, D. P. S. (1980). Identification of "nodule-specific" host proteins (nodulins) involved in the development of Rhizobium-legume symbiosis. Cell 20, 153-163. Leonardo, J. M., and Goldberg, R. B. (1980). Regulation of nitrogen metabolism in glutamine auxotrophs of Klebsiella pneumoniae. J. Bacteriol. 142, 99-110. Ljones and Burris, (1972). Biochim. Biophys. Acta. 275, 93-101. Long, S. R., Buikema, W. J., and Ausubel, F. M. (1982). Cloning of B. meliloti nodulation genes by direct complementation of Nod mutants. Nature 298, 485-488. Ludden, P. W., and Burris, R. H. (1979). Removal of an adenine-like molecule during activation of dinitrogenase reductase from Rhgdgspirillum rubrum. Proc. Ntl. Acad. Sci. USA 76, 6201-6205. Ludwig, R. A.(1978). Control of ammonium assimilation in Rhizobium 32HI. J. Bacteriol. 135, 114-123. Ludwig, R. A. (1980). Regulation of Rhizobium nitrogen fixation by the unadenylylated glutamine synthetase I system. Proc. Natl. Acad. Sci. USA 77, 5817-5821. Ludwig, R. A. (1980). Physiological roles of glutamine synthetases I and II in ammonium assimilation in Rhizobium sp. 32Hl. J. Bacteriol. 141, 1209-1216. Ludwig, R. A. (1984). Rhizobium free-living nitrogen fixation occurs in specialized nongrowing cells. Proc. Natl. Acad. Sci. USA 81, 1566-1569. Ma, Q-S., Johnston, A. W. B., Hombrecher, G., and Downie, J. A. (1982). Molecular genetics of mutants of Rhizobium leguminosarum which fail to fix nitrogen. Mol. Gen. Genet. 187, 166-171. MacFarlane, S. A., and Merrick, M. (1985). The nucleotide sequence of the nitrogen regulation gene ntrB and glnA-ntrBC intergenic region of Klebsiella pneumoniae. Nucleic Acids Res. 13, 7591-7605. MacNeil, D., and Brill, W. J. (1980). Mutations in nif genes that cause Klebsiella pneumoniae to be derepressed for nitrogenase synthetase in the presence of ammonium. J. Bacteriol. 144, 744-751. Magasanik, B. (1982). Genetic control of nitrogen assimilation in bacteria. Ann. Rev. Genet. 16, 135-168. Makino, K., Shinagawa, H., Amemura, M., and Nakata, A. (1986). Nucleotide sequence of the phoR gene , a regulatory gene for the phosphate regulon of Escherichia coli: homology between the PhoR and EnvA proteins. J. Mol. Biol. In press 171 Manian, S.S., and 0’Gara, F. (1982). Induction and regulation of ribulose bisphosphate carboxylase activity in Rhizobium japonicum during formate-dependent growth. Arch. Microbiol. 131, 51-54. Maniatis, T. M., Jeffrey, A., and Kleid, D. G. (1975). Nucleotide sequence of the rightward operator of phage . Proc. Natl. Acad. Sci. USA 72, 1184-1188. Maniatis, T. E., Fritsch, E. F., and Sambrook, J. (1982). Molecular cloning. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. Markwell, M. A. K., Haas, S. M., Bieber, L. L., and Tolbert, N. E. (1978). A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal. Biochem. 87, 206-210. . Marmur, J., and Doty, P. (1962). Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J. Mol. Biol. 5, 109-118. Marsh, J.L., Erfle, M., and Wykes, E.J. (1984). The pIC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional inactivation. Gene 32, 481-485. Maupin, P. and Pollard, T.D. (1983). Improved preservation and staining of HeLa cell actin filaments, clathrin-coated membranes, and other cytoplasmic structures by tannic acid-glutaraldehyde-saponin fixation. J. Cell Biol. 96, 51-62. Maxam, A. M., and Gilbert W. (1977). A new method for sequencing DNA. Proc. Natl. Acad. Sci. USA 74, 560-564. Maxam, A. M., and Gilbert, W. (1980). Sequencing end labelled DNA with base-specific chemical cleavages. Methods Enzymol. 65, 499-580. McComb, J. A., Elliot, J., and Dilworth, M. J. (1975). Nitrogenase activity in rhizobia in absence of plant host. Nature 256, 407-410. Merrick, M., Hill, S., Hennecke, H., Hahn, M, Dixon, R., and Kennedy, C. (1982). Repressor properties of the nifL gene product in Klebsiella pneumoniae. Mol. Gen. Genet. 185, 75-81. Merrick, M. J. (1983). Nitrogen control of the flif regulon in Klebsiella pneumoniae: involvement of the ntrA gene and analogies between ntrC and nifA. EMBO J. 2, 39-44. Messing, J., Crea, R., and Seeburg, P. A. (1981). A system for shotgun DNA sequencing. Nucleic Acids Res. 9, 309-321. Meyer, M. C., and Pueppke, S. G. (1980). Differentiation of Rhizobium japonicum strain derivatives by antibiotic sensitivity patterns, lectin binding, and utilization of biochemicals. Can. J. Microbiol. 26, 606-612. 172 Miflin, B. J., and Lea, P. J. (1976). The pathway of nitrogen assimilationin plants. Phytochemistry 15, 873-885. Miller, J.H. (1972). Experiments in molecular genetics. (Cold Spring Harbor Laboratory: New York). Mortenson, L. E., and Thornley, R. N. F. (1979). Structure and function of nitrogenase. Ann. Rev. Biochem. 48, 387-418. Mulligan, J. T., Long, S. R. (1985). Induction of Rhizobium meliloti nodC expression by plant exudate requires nodD. Proc. Natl. Acad. Sci. USA 82, 6609-6613. Ninfa, A. J. , and Magasanik, B. (1986). Covalent modification of the glnG product, NR by the glnL product, NR , regulates transcription of the glnALG Goperon in Escherichia coli. Prdé Natl. Acad. Sci. USA 83, 5909- 5913. Nixon, B. T., Ronson, C. W., and Ausubel, F. M. (In Press). Two-component regulatory systems responsive to environmental stimuli share strongly conserved domains with the nitrogen assimilation regulatory genes ntrB and ntrC. Proc. Natl. Acad. Sci. USA Noel, K. 0., Stacey, G., Tandon, S. R., Silver, L. E., and Brill, W. J. (1982). Rhizobium japonicum mutants defective in symbiotic nitrogen fixation. J. Bactereiol. 152, 485-494. Norel, F., Desnoues, N, and Elmerich, C. (1985). Characterization of DNA sequences homologous to Klebsiella pneumoniae nifH, D, K, and E, in the tropical Rhizobium ORSS71. Mol. Gen. Genet. 199, 352-356. Noti, J. D., Folkerts, 0., Turken, A., and Szalay, A. A. (1986). Organization and Characterization of genes essential for symbiotic nitrogen fixation from Bradyrhizobium japonicum 1110. J. Bacteriol. 167, 774-783. O’Gara, F., and Shanmugan, K. T. (1976). Regulation of nitrogen fixation by Rhizobia export of fixed N2 as NH4. Biochim. Biophys. Acta 437, 313-321. Ow, D. W., and Ausubel, F. M. (1983). Regulation of nitrogen metabolism genes by nifA gene product in Klebsiella pneumoniae. Nature 301, 307-313. Ow, D. W., Sundaresan, V., Rothstein, D. M., Brown, S. E., and Ausubel, F. M. (1983). Promoters regulated by the qlnG (ntrC) and nifA gene products share a heptameric consensus sequence in the -15 region. Proc. Natl. Acad. Sci. USA 80, 2524-2528. Paau, A. S., and Brill, W. J. (1982). Comparison of the genomic arrangement and the relative transcription of the nitrogenase genes in Rhizobium meliloti during symbiotic development in alfalfa root nodules. Can. J. Microbiol. 28, 1330-1339. 173 Pagan, J. D., Child, J. J., Scowcroft, W. R., and Gibson, A. H. (1975). Nitrogen fixation by Rhizobium cultured on a defined medium. Nature 256, 406-407. Ptashne, M., Backman, K., and humayun, M. Z. (1976). Autoregulation and function of a repressor in bacteriophage lambda. Science 194, 156—161. Pierce, M., and Bauer, W. D. (1983). A rapid regulatory response governing nodulation in soybeans. Plant Physiol. 73, 286-290. Queen, C., and Rosenberg, M. (1981). Differential translation efficiency explains discoordinate expression of the galactose operon. Cell 25, 241-249. Quinto, C., de la Vega, H., Flores, M., Fernandez, L., Ballado, T., Soberon, G., and Palacios, R. (1982). Reiteration of nitrogen fixation gene sequences in Rhizobium phaseoli. Nature 200, 724-726. Ramakrishnan, G., Comeau, D. E., Ikenaka, K., and Inouye, M. (1986). Transcriptional control of gene expression: the osmoregulation of porin protein synthesis. in Bacterial Outer Membranes as Model Systems, ed. M. Inouye. New Yorsz. Wiley and Son. In Press. Rao, R. N., and Rogers, S. G. (1979). Plasmid pKC7: a vector containing ten restriction endonuclease sites suitable for cloning DNA segments. Gene 7,, 79-82. Rao, V. R, Darrow, R. A, and Keister, D. L. (1978). Effectn of oxygen tension on nitrogenase and on glutamine synthetases I and 111 Rhizobium japonicum 61A76. Biochem. Biophys. Res. Comm. 81, 224-231. Regensburger, B., Meyer, L., Filser, M., Weber, J., Studer, D., Lamb, J. W., Fischer, H.-M., Hahn, M., and Hennecke, H. (1986). Bradyrhizobium japonicum mutants defective in root-nodule bacteroid development and nitrogen fixation. Arch. Microbiol. 144, 355-366. Reidel, G. E., Ausubel, F. M., and Cannon, F. C. (1979). Physical map of chromosomal nitrogen fixation (31f) genes of Klebsiella pneumoniae. Proc. Natl. Acad. Sci. USA 76, 2866-2870. Reidel, G. E., Brown, S. E., and Ausubel, F.'M. (1983). Nitrogen fixation by Klebsiella pneumoniae is inhibited by certain multicopy hybrid nif plasmids. J. Bacteriol. 153, 45-56. Reitzer, L. J. and Magasanik, B. (1985). Expression of glnA in Escherichia coli is regulated at tandem promoters. Proc. Natl. Acad. Sci. USA 82,1979- 1983 Reitzer, L. J., and Magasanik, B. (1986). Transcription of glnA in E. coli is stimulated by activator bound to sites far from the promoter. Cell 45, 785-792. 174 Rice, 0., Mazur, B. J., and Haselkorn, R. (1982). Isolation and physical mapping of nitrogen fixation genes from the cyanobacterium Anabaena 7120. J. Biol. Chem. 257, 13157-13163. Roberts, G. P., and Brill, W. J. (1980). Gene-product relationships of the nif regulon of Klebsiella pneumoniae. J. Bacteriol. 144, 210-216. Roberts, G. P., and Brill, W. J. (1981). Genetics and regulation of nitrogen fixation. Ann. Rev. Nicrobiol. 35, 207-235. Rossen, L., Ma, Q.-S., Mudd, E. A., Johnston, A. W. B., and Downie, J. A. (1984). Identification and DNA Sequence of fixZ, a nifB-like gene from Rhizobium leguminosarum. Nucleic Acids Res. 2, 7123-7134. Ruvkin, G. B., and Ausubel, F. M. (1980). Interspecies homology of nitrogenase genes. Proc. Natl. Acad. Sci. USA 77, 191-195. Ruvkin, G. B., Sundaresan, V., and Ausubel, F. M. (1982). Directed transposon Tn5 mutagenesis and complementation analysis of Rhizobium meliloti symbiotic nitrogen fixation genes. Cell 29, 551-559. Schetgens, T. M. P., Bakkaren, G., van Dun, C., Hontelez, J. G. J., van den Bos, R. C., and van Kammen, A. (1984). Molecular cloning and functional characterization of Rhizobium leguminosarum structural nif-genes by site-directed transposon mutagenesis and expression in Escherichia coli minicells. J. Mol. Appl. Genet. 2, 406-421. Schofield, P. R., Djordjevic, M. A., Rolfe, B. G., Shine, J., and Watson, J. M. (1983). A molecular linkage map of nitrogenase and nodulation genes in Rhizobium trifolii. Mol. Gen. Genet. 192, 459-465. Scolnik, P. A., and Haselkorn, R. '(1984). Activation of extra copies of genes coding for nitrogenase in Rhodopseudomonas capsulata. Nature 307,289-292. Scott, D. B., Hennecke, H., and Lim, S. T. (1979). The biosynthesis of -nitrogenase MoFe protein polypeptides in free-living cultures of Rhizobium japonicum. Biochim. Biophys. Acta 565, 365-378. Scott, K. F., Rolfe, B. G., and Shine, J. (1981). Biological nitrogen fixation: primary structure of the Klebsiella pneumoniae nifH and nifD genes. J. Mol. Appl. Gen. 1, 71-81. Scott, K. F., Rolfe, B. G., and Shine, J. (1983a). Biological nitrogen fixation: Primary structure of the Rhizoabium trifolii iron protein gene. DNA 2, 149-155. Scott, K. F., Rolfe, B. G., and Shine, J. (1983b). Nitrogenase structural genes are unlinked in the nonlegume symbiont Parasponia Rhizobium. DNA 2, 141-148. 175 Shaw, 8. D. (1983). Non-coordinate regulation of rhizobium nitrogenase synthesis by oxygen: studies with bacteroids from nodulated Lupinus angustifolius. J. Gen. Microbiol. 129, 849-857. .Shine, J., and Dalgarno, L. (1975). Determination of cistron specificity in bacterial ribosomes. Nature 254, 34-38. Somerville, J. E., Kahn, M. L. (1983). Cloning of the Glutamine synthetase I gene from Rhizobium meliloti. J. Bacteriol. 156, 168-176. Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503-517. Stanley, J., Longtin, D., Madrzak, C., and Verma, D. P. S. (1986). Genetic locus in Rhizobium japonicum (fredii) affecting soybean root nodule differentiation. J. Bacteriol. 166, 628-634. Stripf, R., and Werner, D. (1978). Differentiation of Rhizobium japonicum. II. Enzymatic activities in bacteroids and plant cytoplasm during the development of nodules of Glycine max. Z. Naturforsch. 33c, 373-381. Sundaresan, V. and Ausubel, F. M. (1981). Nucleotide sequence of the gene coding for the nitrogenase iron protein from Klebsiella pneumoniae. J. Biol. Chem. 256, 2802-2812. Sundaresan, V., Ow, D. W., and Ausubel, F. M. (1983a). Activation of Klebsiella pneumoniae and Rhizobium meliloti nitrogenase promoters by gln (31;) regulatory proteins. Proc. Natl. Acad. Sci. USA 80, 4030-4034. Sundaresan, V., Jones J. D. G., Ow, D. W., and Ausubel, F. M. (1983b). Klebsiella pneumoniae nifA product activates the Rhizobium meliloti nitrogenase promoter. Nature 301, 728-732. Szalay, A. A., Grohmann, K., and Sinsheiner, R. L. (1977). Separation of complementary strands of DNA fragments on polyacrylamide gels. Nucleic Acids Res. 4, 1569d1578. Szeto, W. W., Zimmerman, J. L., Sundaresan, V., and Ausubel, F. M. (1984). A Rhizobium meliloti symbiotic regulatory gene. Cell 36, 1035-1043. Tinoco, 1., Borer, P., and Dengler, B. (1973). Improved estimation of secondary structure in ribonucleic acids. Nature New Biol. 246, 40-41. Tjepkema, J., and Evans, H. J. (1975). Nitrogen fixation by free-living Rhizobium in a defined liquid medium. Biochim. Biophys. Res. Comm. 65, 625-628. Ueno-Nishio, S., Mango, S., Reitzer, L. J., and Magasanik, B. (1984). Identification and regulation of the glnL operator-promoter of the complex glnALG operon of Eschrichia coli. J. Bacteriol. 160, 379-384. 176 Vairinhos, BHandari, B., and Nicholas D. J. D. (1983). Glutamine synthetase, glutamate synthase and glutamate dehydrogenase in Rhizobium japonicum strains grown in cultures and in bacteroids from root nodules of Glycine max. Planta 159, 207=215. Verma, D. P. S., and Bal, A. K. (1976). Intracellular site of synthesis and localization of leghemoglobin in root nodules. Proc. Natl. Acad. Sci. USA 73, 3843-3847. Vincent, J. M. (1970). A manual for the Practical Study of Root-Nodule Bacteria. (Blackwell, Oxford). Vincent, J. M. (1978). Factors controlling the legume-Rhizobium symbiosis. in Nitrogen Fixation Vol.11, W. E. Newton and W. H. Orme-Johnson, eds. (University Park Press: Baltimore, Maryland), pp. 103-129. Vincent J. M. (1982). Nitrogen fixation in legumes. New York: Academic Press. Walz, A., Pirrotta, V., and Ineichen, K. (1976). Repressor regulates the switch between PR and PRM promoters. Nature 262, 665-669. Weinman, J. J., Fellows, F. F., Gresshoff, P. M., Shine, J., and Scott, K. F. (1984). Structural analysis of the genes encoding the molybdenum-iron protein of nitrogenase in Parasponia rhizobium strain ANU289. Nucleic Acids Res. 12, 8329-8344. Werner, D., Morschel, E., Stripf, R., and Winchenbach, B. (1980). Development of nodules of Glycine max infected with an ineffective strain of Rhizobium japonicum. Planta 147, 320-329. Wong, P. P. (1980). Nitrate and carbohydrate effects on nodulation and nitrogen fixation activity of lentil. Plant Physiol. 66, 78-81. Yannisch-Perron, C., Vieira, J., and Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103-119. Yanofsky, C., and Kolter, R. (1982). Attenuation in amino acid biosynthetic operons. Ann. Rev. Genet. 16, 113-134.