5 K5 5|. .I. .5555. 5 «55555.55 ; 55 5.55.. I. “I.“ZrI-II-cz: ” I-x‘g‘i. ”155' 5.53117. K~ 22:... 555. 5.5.2:? It".-. ~ . I- . . . . IEIZ “? 3.1.3.51 $7.5m“... “ . . m .. If“ $5.22: I“ I? KER. n. . .r 'I 353.3. 1:3.» a x ‘Q 5.1.9212. 12.3.5.3} jawcts I53; _ .1. 5:72“ - 5th :5 . “:5 .,_ VIII-LK“ ~955“ minim“); . .. arm 3.: V L scan! Lq:1:‘;.'~\ .- . WEN“: ‘ o 11:: iii :3: 3.“ C‘ .. I» as III-I .5 _...' I "‘2‘. I53“ Tia» ‘rdcfi ggéfifi‘. -. "c‘ “1* -7731: ’t amwg‘gwuah 3»th ‘- s“. “€213: ' L‘s-I I. I. I I. . .,- 5.... 71*“ I.“ curl} 5.5a Ai‘fi £3??? mmfiq h I " . n N . .I. miwgmlg.‘ '1‘ 1w 13:.» :‘Ln awk‘i‘ 5.553. in u“ KM,‘ 5.. 3;~*I$&"”~"? 91‘3“ ‘Efi. a..I."“I.?c w- - (Etna ' R2... I... {W .1; I,.....L I“ 1K5? 517g U51: IN: All 51;, K"; :5: E31. 3'3"}: ,5 a??? . yufiuk“ ‘um 12:”;- 3‘31; ‘I ' ‘ SS 5 I}? 1.1. as; Iguana. 5.9%.. . . .. 1.3wa as ‘“- “K“ {:3 “£4? 3% “*fix % \m ,3 {ski “I “‘ ,QQiQ-fi‘kmggx‘: 3% - I I. :1‘35‘: upmé. 5:11;?“ u-u'I.~I'5~.“I. #:31-1} NE}: *3» I031? Ira-c . . I. :“I‘mI . £113.“ 55:... “ ' \ :I‘E‘ MFR».- II «,3 “IV\‘ GI‘MW gfii .‘ MW. #398“ :55 a. 5% 3:31.“: ‘79‘1‘0: ‘I-an 5'5 " man‘sx‘WI. ”I; «L‘k S :"‘ *Egrdfl‘ .A ‘QL: . cdfir: 3‘52va4 “ “3‘3: ”33“}; LU-w um“. i‘ xhwhq‘ n. 5.4:, 32,7, 53:4 ugh“: “1.35m mini; L 5-}. “\n‘ .. wag?“ Keir Wu. \- {:1 k:IESL . 3"_ Min 713‘ ywmtmggs. I L H-‘E‘E‘éz‘f: 2‘13“ 'E‘Iwr-‘E I14 I35. $335k W} NJ}: :91‘13 , "fig. ”y‘a- 0“. n.2,: ‘ Qt. Phi-Q... ‘15? ‘3" 31:: W 3:; ' 1.533.. 55‘. ' I3 x, ‘ ‘91 I- If, . C331. , "”‘ E‘M‘w‘m‘fi‘g ~ Eu. :‘II‘LIfiF 1" 133" utdxww‘nfla‘ . II. a In. I. in“ 9. full?! I. Ii max”? ‘ “3‘13““ “Lit?! " mam I23; . . In. $- .2: .9" ‘1 13‘ x I; I 53%;" .. A . I?! 9333‘.“ Z . Er' Cfi-LIAN“ “I... I“? . 5 555.3551“. .5 :‘LCL~ “ 555m. 4‘. .. . . - 2.55-- “.55 ‘1 NH“ 5555 55" "5 55-. '7 5555115"; .5555 , NC5555“; 7 515 ?5‘5£Y:.5‘1"\““‘55“".'¢! NI»: 155251.»:lr5'wfi.‘ SLR-5‘3 -n5K5-I55I.'5.'ZT 5., :71? ”EEL- .I. .5555K5... 'v, ~ M—c‘w -. “$23. 5551... 5 5 55.5.“ 555 ”5.4.53.5 5 “31.51%;- “b5 T‘fiéfl‘t‘m‘ Eli’i‘fgt‘ I}: W «I 5“ V‘k‘::\‘:: \V‘L‘x‘ :\;$L§ :‘fumlf;xu «5", “9‘. ’ «In-1“ fi‘wfi’; Isa-2 5‘15“}. M' “V\“§Lm ”i 0% xx: :kh‘fillxtcd‘ ‘7th ‘31" "FE“: fiwkktfll‘ 45"“ ”ii" ":78‘1 ' 3' .._~$ -. 4E1 - . “h T ‘n" ’ :11 _. cg”. fir“. £315?“ F" {'44: x £5. ii." 25:. 3-35 ‘I. Y. I -4375 ii? . "I This is to certify that the dissertation entitled CAMOClZaCEZaé/_J)[ a; tZL W5 5146001443, flufimm Syn/la AseZ'MA é/o/amme ‘ $7m/ietma 2L {IUBQSCéfbH‘fl l’lAJUéh/u janMIZawx has been accepted towards fulfillment of the requirements for @— degree in 5564:4723 a y (4 Major professor : urn: 112.....- A m. .Arr . . . . 0-1 1 Date «fly/gé MSU LIBRARIES an RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wiIl be charged if book is returned after the date stamped below. CHARACTERIZATION OF THE GENES ENCODING GLUTAMINE SYNTHETASE I AND GLUTAMINE SYNTHETASE II FROM BRADYRHIZOBIUM JAPONICUM By Todd A Carlson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1986 Q fit (A A m bx ABSTRACT CHARACTERIZATION OF THE GENES ENCODING GLUTAMINE SYNTHETASE I AND GLUTAMINE SYNTHETASE II FROM BRADYRHIZOBIUM JAPONICUM By Todd A Carlson Bacteria of the genera Rhizobium and Bradyrhizobium are able to fix nitrogen in symbiotic association with leguminous plants. The repression of bacterial glutamine synthetase (GS) allows newly fixed nitrogen to be exported from the symbiotic bacteroids to the host plant. Therefore, the regulation of GS is an important part of the development of an effective symbiosis. In this dissertation, I describe the isolation and characterization of the genes encoding the two GS isozymes, G51 and G811, from Bradyrhizobium japonicum, the soybean symbiont. The gene encoding GSI, glnA, was identified in a genomic library using a fragment of the Escherichia coli glnA gene as a hybridization probe. The 5. japonicum glnA gene was found to have extensive homology to E. coli glnA, and when expressed in E. coli, produced large amounts of a GS with properties identical to that of GSI of B. japonicum. SI nuclease protection experiments indicate that glflA is constitutively transcribed from a single promoter. The gene encoding GSII, 91311, was identified in a cosmid library using a mixed oligonucleotide hybridization probe based on a partial amino acid sequence determined from purified GSII. Amino acid sequence comparisons revealed extensive homology between GSII and eucaryotic glutamine synthetases, suggesting that 6511 evolved as the result of a eucaryote to procaryote gene transfer event. Quantification of glnII mRNA in Todd A Carlson nutrient limited cultures demonstrates that glnII is transcriptionally activated in response to nitrogen starvation or nitrogen limited growth. The site of transcriptional initiation upstream of glgll was mapped by $1 nuclease protection. The 91311 promoter has no homology to the glnA promoter but has homology to B. japonicuni flif promoters in the RNA polymerase binding region. Promoter structure and gene expression are discussed with respect to possible models of glnII and nif regulation during symbiotic nitrogen fixation. ACKNOWLEDGMENTS I would like to thank a number of people for their advice and assistance. My foremost thanks go to my advisor, Barry Chelm, for providing expert advice and guidance when needed, and for not providing advice and guidance when not needed. I also thank the members of my guidance committee for their suggestions and criticisms. I would also like to acknowledge all the members of the Chelm lab for freely supplying advice, assistance, and RNA samples. I am grateful to Larry Reitzer, Boris Magasanik, Benjamin Mifflin, Howard Goodman, Gloria Coruzzi, and Tom Adams for the communication of results prior to pub- lication. Technical assistance was provided by Tony Bleeker and Hans Kende for the HPLC analysis; Mike Shimei and Jan Watson for oligonucle- otide synthesis and purification; Bob Haselkorn, Lee McIntosh, and Rob McClung for plasmids; and Ron Davis for sequence analysis. I acknow— ledge the DOE Plant Research Laboratory, the Biochemistry Department, and Michigan State University for financial support of my graduate studies. Finally, and most importantly, I extend thanks to the members of my family, especially my father and my wife, for their constant support and encouragement. TABLE OF CONTENTS List of Tables List of Figures CHAPTER 1 Introduction CHAPTER 2 Characterization of the Gene Encoding Glutamine Synthetase I (glnA) from Bradyrhizobium japonicum Introduction Materials and Methods Bacterial Strains and Med Materials B. japonicum Genomic Library Construction Hybridization Methods Plasmid Complementation and GS Assays Nodule Bacteria Isolation Sl Nuclease Protection Analysis Results B. japonicum Library Screening Localization of the glnA homology Complementation Analysis ia Measurements of GS activity Promoter Sequence and Expression Analysis Discussion PAGE vi CHAPTER 3 Apparent Eucaryotic Origin of Glutamine Synthetase II from the Bacterium Bradyrhizobium japonicum Introduction Materials and Methods GSII Purification and Protein Methods Oligonucleotide and DNA Methods Complementation Analysis Resluts and Discussion CHAPTER 4 Transcriptional Control of glnII, the Gene Encoding Glutamine Synthetase II of Bradyrhizobium japonicum Introduction Materials and Methods Promoter Mapping Bacterial Cultures RNA Purification Results Promoter Mapping Sequence Analysis 91311 Promoter Characterization Discussion CHAPTER 5 Summary and Conclusions LIST OF REFERENCES PAGE 35 36 36 37 38 38 81 TABLE LIST OF TABLES PAGE Complementation of B. coli ET8051 913A deficiency by 24 plasmids carrying the B. japonicum glflA gene GSII Purificaton 39 The Statistical Significance of Amino Acid Sequence 49 Homologies Between B. japonicum G511 and Other Glutamine Synthetases FIGURE 1 10 11 LIST OF FIGURES Restriction Endonuclease Map of the B. japonicum glnA Region Localization of glnA by Heterologous Hybridization Identification of GSI by Sedimentation Centrifugation The Sequence of the 5’ End of the B. japonicum glnA Gene SI Nuclease Protection Analysis of B. japonicum glnA B. japonicum glnA Promoter Sequence Comparison GSII Purification: SDS-Polyacrylamide Gel Electrophoresis Analysis GSII Amino Terminus Sequence and Mixed Oligonucleotide Probe Identification of GSII by Sedimentation Centrifugation DNA and Amino Acid Sequence of glnII and GSII Homology Matrix Comparing GSII With Phaseolus vulgaris GS Homology Matrix Comparing GSII With Anabaena GS 51 Nuclease Protection Mapping of the g1fl11 Promoter DNA Sequence Upstream of glfl11 B. japonicum Promoter Sequence Comparison B1 japonicum Minimal Media Growth Curves Comparison of glnA and glnII Transcription vi , azmeiEffilii§§§E§5r PAGE 20 22 26 27 3o 32 41 42 44 45 47 48 60 61 62 66 68 CHAPTER 1 Introduction Bacteria of the genera Rhizobium and Bradyrhizobium are unique in their ability to fix nitrogen in symbiotic association with leguminous plants. During the development of this symbiosis, the bacteria undergo metabolic and morphologic transformations to become bacteroids, the intracellular "organelles“ where nitrogen fixation occurs. Concurrent with this bacterial cell differentiation, root cells at the site of infection develop to form nodules, the multicellular organs which provide the specialized environment required for rhizobial nitrogen fixation. The legume-rhizobium symbiosis is species specific and is presumably coordinated by a series of complex interactions between the procaryotic and eucaryotic symbionts. The formation of effective nodules, therefore, can be viewed as a developmental pathway requiring the regulation of many genes and enzymes in both the bacterium and the host plant. A complete understanding of nodule development is necessary in order to develop fully the tremendous potential of the rhizobium- legume symbiosis. Because of the bacteroid’s role in nitrogen fixation, the regulation of plant and bacterial nitrogen metabolism is of special interest. In the bacteroid, atmospheric nitrogen is reduced to ammonia by nitrogenase, a complex enzyme system consisting of component I, which contains the active site of dinitrogen reduction, component II, which is involved in electron transport, and a set of nitrogenase specific cofactors. The coordinated expression of at least 18 genes (designated as Elf genes) is required for the synthesis of nitrogenase and its associated cofactors in Klebsiella pneumoniae (Roberts & Brill, 1981) and presumably also in Rhizobium and Bradyrhizobium species. All known nitrogenases are oxygen labile. Bacteria which fix nitrogen in aerobic environments have developed various mechanisms for protecting their nitrogenase from 02 inactivation. Rhizobia depend on the plant encoded, oxygen binding protein leghemoglobin, which maintains a very low concen— tration of free oxygen in the nodule, thus protecting bacteroid nitro— genase from 02 inactivation. Breaking the triple bond of dinitrogen is a very energy expensive process, requiring approximately 12 molecules of ATP and six electons for each N2 fixed (Mortenson & Thorneley, 1979). The plant provides a respirable energy source for the nitrogen fixing bacteroids in the form of dicarboxylic acids (Ronson gt _1., 1981). The oxygen required for bacteroid respiration is transported through the nodule by facilitated diffusion using leghemoglobin as a carrier protein (Bergersen, 1980). Thus, leghemoglobin serves the dual function of protecting the oxygen labile nitrogenase and transporting oxygen for bacteroid respiration. In order for rhizobium to function as an effective nitrogen fixing symbiont, it is necessary for the bacteroid to export fixed nitrogen for utilization by the host plant. Isotopic tracing studies indicate that as much as 94% of the nitrogen fixed by bacteroids in legume root nodules is exported to the plant (O’Gara & Shanmugam, 1976). The primary pathway of ammonia assimilation is by the coordinated activity of glutamine synthetase (GS) and glutamate synthase (GOGAT; Tyler, 1978). GS catalyzes the production of glutamine from glutamate and ammonia. GOGAT produces two molecules of glutamate from glutamine and Z-ketoglutarate. During the development of nitrogen fixing bacteroids, GOGAT activity remains relatively constant but GS activity is repressed in concert with the derepression on nitrogenase (Brown & Dilworth, 1975; Robertson gt .g1., 1975; Upchurch & Elkan, 1978). As a result, the ammonia formed by the nitrogenase complex is exported from the bacteroid to the plant cell where it is assimilated by a very abundant, nodule t al., specific glutamine synthetase (Brown & Dilworth, 1975; Cullimore 1983; Evans & Crist, 1984; Lara gt _1., 1983; McParland gt g1., 1976; O’Gara & Shanmugam, 1976; Upchurch & Elkan, 1978). Because GS is the first enzyme of the ammonia assimilation pathway, its regulation is critical in the control of overall nitrogen metabolism. Some bacteria use glutamate dehydrogenase (GDH), which catalyzes the formation of glutamate from 2-ketoglutarate and ammonia, to assimilate ammonia present at high concentrations (Tyler, 1978). In many strains of rhizobia, however, GDH is only used for glutamate catabolism and GS/GOGAT is the only pathway of ammonia assimilation (Howitt & Gresshoff, 1985; Ludwig, 1976; 0’Gara & Shanmugam, 1976; Vairinhos gt g1., 1983). The regulation of GS in rhizobia is unusual when compared to non-symbiotic nitrogen fixing bacteria, such as Klebsiella ggggmggigg, which activate both GS and nitrogenase during nitrogen starvation. It has been suggested that the ability to simultaneously activate ammonia production (nitrogenase) and repress ammonia assimilation (GS) is an adaptation of rhizobial nitrogen metabolism necessary for symbiotic nitrogen fixation (Ludwig & Signer, 1977; O’Gara & Shanmugam, 1976). Although the export of ammonia by bacteroids is logical in terms of their role as symbiotic nitrogen fixing "organelles,“ it is not immed- iately apparent how rhizobia could have evolved this apparently altru- istic behavior. Kahn gt Q1. (1985) have suggested that bacteroid nitrogen fixation is not altruistic but rather is necessary for self feeding. According to their model, nitrogen in the nodule is used as a carrier of carbon into the bacteroids, entering as amino acids and excreted as ammonia. The purpose of nitrogen fixation is to replace the nitrogen that has been utilized by the plant to ensure a continual flow of carbon into the bacteroid. Thus, nitrogen fixation is a response to carbon rather than nitrogen starvation. The excretion of nitrogen is not unusual among diazotrophic bacteria. Some species of Bradyrhizobium are able to fix nitrogen gx glgfltg, but the reduced nitrogen is excreted into the culture medium and not utilized for growth (Evans & Crist, 1984; Ludwig, 1980a; O’Gara & Shanmugam, 1976; Upchurch & Elkan, 1978). Some actinomycetes can fix nitrogen in a symbiotic association with non-leguminous plants similar to the rhizobium—legume symbiosis (Akkermans & Roelofsen, 1980). The filimentous blue—green algae Anabaena fix nitrogen in terminally differ- entiated cells termed heterocysts (Haselkorn, 1978). The heterocysts "sacrifice" themselves in order to fix nitrogen for the benefit of other cells in the culture. The development of specialized nitrogen fixing cells is characteristic of nitrogen exporting diazotrophs. In contrast to rhizobial bacteroids which repress GS and export ammonia, Anabaena heterocysts produce large amounts of GS and export fixed nitrogen in the form of glutamine (Thomas gt g1., 1977). This observation supports the conclusion that the regulation of GS is an important component of bacteroid metabolism. The study of GS regulation in rhizobia is complicated by the presence of two species of GS, designated G81 and GSII (Darrow & Knotts, 1977). G51, in all respects, is typical of procaryotic glutamine synthetases. It has 12 subunits of 59,000 daltons and is post transla- tionally modified by a regulatory system similar to that of B. ggli GS in which any or all of the 12 subunits can be repressed by the adenylyl- ation of a specific tyrosine residue (Darrow, 1980; Darrow & Knotts, 1977; Ludwig, 1980b). Furthermore, GSI will cross-react with antibodies raised against B. coli GS (Tronick et 1., 1973). Although the synthe— tic activity of GSI is regulated by adenylylation, the enzyme is pro— duced at a relatively constant level (Ludwig, 1980b). The gene encoding GSI, termed glnA, has been isolated from Rhizobium meliloti and is homologous to the glnA gene of B. coli (Somerville & Kahn, 1983). Whereas GSI is similar to glutamine synthetases from all bacteria, GSII is found only in the Rhizobiaceae family, which includes the genera Rhizobium, Bradyrhizobium, Aqrobacterium, and Phyllobacterium (Fuchs & Keister, 1980). GSII is encoded by a gene distinct from the gene encoding GSI (Darrow, 1980; Somerville & Kahn, 1983). The structure of GSII (8 subunits of 36,000 daltons) is characteristic of eucaryotic glutamine synthetases (DeVries gt g1., 1983). DeVries gt g1. (1983) demonstrated that GSII of thtgtjgm lequminosarum will cross-react with antibodies raised against Pisum satuvum GS and suggested that the gene encoding GSII may be of eucaryotic origin. However, this hypothesis was not confirmed by Cullimore and Miflin (1984) who found that Rhizobium phaseoli GSII would not cross-react with antibodies raised against Phaseolus vulgaris GS and that a partial cDNA clone of the gene encoding B. vulgaris GS had no detectable homology to R. phaseoli genomic DNA. Thus the origin of GSII remains a mystery. G81 and GSII are differentially regulated in response to a number of different stimuli, including nitrogen source, carbon source, oxygen concentration, and symbiotic development (Cullimore gt g1., 1983; Darrow gt g1., 1981, Fuchs & Keister, 1980; Ludwig, 1980b; Rao gt g1., 1978). G51 activity is regulated by the adenylylation cascade system. In contrast, there is no known regulation of GSII by post translational modification. Therefore, changes in total cellular GSII activity are presumably due to the regulation of transcription of the gene encoding GSII or translation of its mRNA. It is not yet clear what function GSII serves in rhizobial nitrogen metabolism. A B. meliloti 913A deletion strain, which has no GSI, grows normally, indicating that GSII alone is able to satisfy all of the ammonia assimilation needs of the cell (Somerville & Kahn, 1983). This is in contrast to a report by Ludwig (1980b) who found that GSII did not function in ammonia assimilation. Darrow gt _1. (1981) have suggested that GSII provides extra ammonia assimilatory capacity during nitrogen limited growth. This conclusion is based on the observation that GSII is induced in B. jgpgfliggm when grown on poor nitrogen sources, such as amino acids, or good carbon sources, such as succinate and gluconate. The concentration of oxygen in the culture medium also seems to affect the regulation of GSII. Rao gt g1 (1978) found that, at 02 concentra— tions below 0.40%, GSII levels decline dramatically in concert with the _m;-/ adenylylation of GSI. The repression of GSII at low oxygen concentra— tions may be especially significant since similar conditions exist in root nodules. In all cases that have been examined, GSII repression is concurrent with an increase in GSI adenylylation. The simultaneous repression of GSI and GSII suggests the presence of a general nitrogen regulatory system which can control both gene transcription and protein modification. The regulation of bacterial nitrogen metabolism has been most extensively studied in the enteric bacteria Klebsiella pneumoniae and Escherichia ggli (Alvarez-Morales gt _1., 1984; Magasanik, 1982). In these bacteria, GS and other enzymes, responsible for the catabolism of various nitrogen sources, are under the control of a complex regulation system, termed Ntr. Ntr regulated genes include gtgl (arginine uptake), hisJQHP (histidine uptake), hgtBB (histidine utilization) and nif (nitrogen fixation). This group of coordinately regulated, unlinked genes is sometimes referred to as the Ntr regulon. All genes activated by the Ntr system require the Ntr specific sigma factor, 060, which substitutes for the normal sigma factor and directs RNA polymerase binding and transcriptional initiation at Ntr promoters (Hirschman gt g1., 1985; Hunt & Magasanik, 1985). 060 is the product of the rggN gene (formerly designated glflB and flttA). All Ntr activated genes have DNA sequence homology in the region 10 to 30 base pairs (bp) upstream of the RNA start site. The consensus sequence (—27 CTGGCAC—N -TTGCA —IO) 5 defines the Ntr promoter (Dixon, 1984b; 0w _t _1., 1983). RNA polymerase and 060 alone are insufficient for transcription from Ntr promoters. An additional nitrogen regulatory protein, NRI, is required. NR1 is the product of the ntrC gene (also designated glnG; i 1 r |———. _. -4,§;v§f--r!_.-r 4,. , ' :aha—sc"—~/ A l Reitzer & Magasanik, 1983). NRI is a dimeric protein which binds double AAAA TTTT 1985). The mechanism by which NRI activates Ntr promoters is not stranded DNA at the sequence 5’-TGCACC TGGTGCA-3’ (Ames & Nikaido, completely understood. It has been suggested that the partial homology between the NRI binding sequence and the Ntr consensus promoter (TGCA) is responsible for the direct interaction of NRI with the DNA at the RNA polymerase binding site (Hunt & Magasanik, 1985). In addition, Reitzer and Magasanik (1986) have shown that NRI binding sites upstream of glnA are required for Ntr promoter activation. The function of these up- stream activating sequences is independent of their exact position and orientation relative to the RNA polymerase binding site. Position independence is unusual for cis—acting regulatory sequences in pro— caryotes but is typical of eucaryotic enhancer elements. The key to the control of the Ntr system seems to be regulating the amount and activity of NRI (Bueno gt g1., 1985). The gene encoding NRI, nttt, is in the complex glgA-gttBB operon, located downstream from glgA (the gene encoding GS), and flttB (also termed 913B) which encodes another nitrogen regulatory protein, NRII. This operon is transcribed from three differentially regulated promoters (Alvarez-Morales gt g1., 1984). Upstream of glnA is an NRI dependent Ntr promoter and a standard promoter which is regulated by catabolite repression (Dixon, 1984b; Reitzer & Magasanik, 1985). In the intergenic region between glflA and ELLE is another standard promoter which can be repressed by the DNA binding activity of NRI (Hawkes gt ._1., 1985; Reitzer & Magasanik, 1983). Thus NR1 autoregulates its own production by both promoter activation and repression. NRI has also been implicated in gene regula- tion by antitermination (Ames & Nikaido, 1985). The conversion of NRI from an activating to a repressing form is central to the control of the entire Ntr system (Bueno gt _1., 1985). This conversion is carried out by NRII, perhaps by a covalent modification of NRI. NRII, in turn, is regulated by two other proteins, a uridylyltransferase and the P11 protein, encoded by the glflB and glflB genes respectively. The activity of these proteins is controlled by the intracellular ratio of glutamine to 2—ketoglutarate which is an indicator of the cell’s nitrogen nutri— tional status (Bueno gt g1., 1985). The reversible uridylylation of P11 also regulates the adenylyltransferase responsible for the modification of GS (Stadtman & Ginsburg, 1974). Thus the Ntr system is able to regulate nitrogen metabolism at both the transcriptional and post translational levels. One of the operons induced by the K. pneumoniae Ntr system is gifBA. The products of the gift and nifA genes are regulatory proteins with homology to NRII and NRI respectively, and are responsible for the regulation of the Bit genes (Buikema et l., 1985; Dixon, 1984a; Drummond et 1., 1986). nit promoters, like Ntr promoters, require the Ntr specific sigma factor and have the consensus Ntr promoter sequence (Benyon t l., 1983; Ow gt g1., 1983). nit promoters, however, require the product of the nifA gene instead of NRI as an additional activating factor (Ow & Ausubel, 1983). All git genes also have a distinct up— stream activating element (5’-TGT-N10—ACA—3’; Alvarez—Morales gt g1., l., 1986) Thus, there are two classes of 060 dependent 1986; Buck gt __ 60 promoters, Ntr and git. Both classes share the a RNA polymerase binding site. Differences in the Ntr and git upstream activating sequences presumably account for their differential regulation. The molecular mechanism of nitrogen regulation in rhizobia is not as well understood as the Ntr system of the enteric bacteria. The git genes from both Rhizobiuni meliloti and Bradyrhizobium japonicum have been isolated and were found to have promoter structures characteristic of Klebsiella pneumoniae git promoters, including 060 binding sites and upstream activating sequences (Adams & Chelm, 1984; Alvarez-Morales, gt g1., 1986; Ausubel, 1984; Buck gt g1., 1986; Fuhrmann & Hennecke, 1984). These observations indicate that rhizobia have a git regulatory system homologous to K. pneumoniae. A number of rhizobial mutations that block git expression also alter the regulation of GSI and GSII, suggesting that the regulation of nitrogen fixation is tied in with general nitrogen metabolism (Donald & Ludwig, 1984; Kondorosi gt gi,, 1977; Ludwig, 1980a; Morett gt gi., 1985). To date, no nitrogen regulated promoters, other than git, have been characterized from rhizobia. In this dissertation, I describe the isolation and characterization of the genes encoding GSI and GSII (gigg and gig11 respectively) from Bradyrhizobium japonicum, the soybean symbiont. These genes were chosen because the regulation of GS is central in the control of nitrogen metabolism. Furthermore, the presence of two differentially regulated enzymes which catalyze identical reactions despite extensive structural differences is unprecedented. Thus a comparison of the gigA and gig11 genes and their regulatory elements will be of interest. Finally, these genes will serve as molecular probes for the characterization of GS expression. Differences between GS regulation in B. japonicum and other bacteria may reflect adaptations of nitrogen metabolism necessary for symbiotic nitrogen fixation. , “T :’—‘—‘:i—-/ CHAPTER 2 Characterization of the Gene Encoding Glutamine Synthetase I (glnA) from Bradyrhizobium jagonicum Introduction In free-living Bradyrhizobium iaponicum, ammonia is assimilated primarily by the coordinate activity of glutamine synthetase (GS) and \ glutamate synthase (Brown & Dilworth, 1975; Vairinhos gt g1., 1983). However, in bacteroids, the differentiated symbiotic form of these bacteria, GS activity is repressed in concert with the derepression of nitrogenase activity (Brown & Dilworth, 1975; Upchurch & Elkan, I978). The ammonia formed by the nitrogenase enzyme complex is then exported to the plant cell cytoplasm, where it is incorporated by plant assimilatory enzymes (Brown & Dilworth, 1975; Evans & Crist, 1984; Lara gt g1., I983; O’Gara & Shanmugam, 1976; Upchurch & Elkan, 1978). Therefore, the regulation of the rhizobial genes involved in nitrogen metabolism is an important part of the bacterial developmental process which leads to symbiotic nitrogen fixation. The study of GS regulation in rhizobia is complicated by the presence of two GS species, designated GSI and GSII (Darrow & Knotts, 1977). GSI is very similar to the single GS enzyme found in most other 11 i1, Gram negative bacteria. It is a polymeric enzyme consisting of 12 identical subunits of 59,000 daltons, is relatively heat stable, and is regulated by a reversible adenylylation cascade system (Darrow, 1980; Darrow & Knotts, 1977). In contrast, GSII has eight subunits of 36,000 daltons, is heat labile, and is not known to be modified after trans— lation (Darrow, 1980; Darrow & Knotts, 1977). These proteins are products of different genes (Darrow, 1980; Somerville & Kahn, 1983) and are differentially regulated in response to changes in nitrogen source (Ludwig, 1980b), carbon source (Darrow gt g1., 1981), and oxygen concen- tration (Darrow gt gi., 1981; Rao gt gi., 1978). The mechanisms by which rhizobia regulate GSI and GSII activities are not well understood. In a variety of rhizobial species, glutamine auxotrophs have been isolated which have very low GS activity and are ineffective in symbiotic nitrogen fixation (Donald & Ludwig, 1984; Kondorosi gt g1., 1977; Ludwig, l980a). The complex pleiotrophic phenotypes of these strains suggest that they have mutations in a general nitrogen regulation system. Such regulatory systems have been described for other bacteria (Magasanik, 1982). To study directly the mechanism of GS regulation during nodule development, I isolated the genes encoding GSI and GSII from B. japonicum, the soybean symbiont. With cloned genes it will be possible to characterize those factors involved in the regulation of GS expression. In this chapter, I de- scribe the characterization of the gene encoding GSI, designated gigg. I determined that the rhizobial gigg gene is highly homologous to the B. coli glnA gene and is constitutively transcribed from a single promoter. The sequence of the glnA promoter has weak homology to the B. coli consensus promoter and is distinct from the differentially regulated nif 13 promoters of the same organism. The isolation and characterization of this gene has been reported (Carlson gt_ gi., 1983; Carlson gt ,gi., 1985). Materials and Methods Bacterial Strains and Media. The E. coli K-12 strain ED8654 (galK galT tr rpRm etB Lstks upE sup F) was used for routine plasmid construc- tion and maintenance. The phage lambda recombinant library was con- structed and amplified on B. coli K802 (lacY metB galT Lstk galK sup E) Plaque hybridization, complementation, and GS assay studies used the bs Nalr, which glutamine auxotroph ,B. coli ET8051¢o(rha—glnA ) LutCk contains a deletion of the entire gigA region (Fisher rgt gi., 1981). B. ggii strains were grown at 37 0C in either LB medium (Davis gt gi., 1980) or M9 medium (Miller, 1972) with 0.4% glycerol. B. japonicum strain USDA 110 was grown at 300C in minimal salts arabinose medium (MA) (Ludwig & Signer, 1977) with trace elements (O’Gara & Shanmugam, 1976); modified Bergersen’s medium (Bergersen, 1961) with 0.2% xylose and 10 mM of the designated nitrogen source (MBX); formate medium (Manian & O’Gara, 1982) with 2.5 mM (NH4)ZSO4; or anaerobic yeast extract xylose medium (YEX; Adams gt gi., 1984) with 10 mM KNO3. Materials. Genomic DNA from B. japonicum was purified by phenol extraction (Mar armur & Doty, 1962). Plasmid DNA was isolated from B. coli \ by CsCl ethidium bromide equilibrium centrifugation (Clewell & Helinski, 1972). Phage particles were prepared by two rounds of CsCl block m—a» 4.: _ 59.x.- .7 ___.. 71,—: . B 14 density gradient sedimentation (Davis gt _i., 1980), and phage DNA was then extracted by formamide treatment (Davis & Thomas, 1974). DNA restriction endonuclease fragments to be used for cloning, DNA sequenc- ing, SI nuclease protection analyses, and hybridization probes were isolated by separation on and elution from polyacrylamide gels by the method of Maxam and Gilbert (1980). Unless otherwise stated, recombi— nant plasmids are derivatives of pBR322 (Bolivar et l., 1977). B. japonicum Genomic Library Construction. B. jagonicum DNA, partially digested with the restriction endonuclease MBgI, was ligated to the ngHI sites of the cloning vector XBFIOI (Maniatis gt g1., 1982). The vector DNA had been digested with BgiI in addition to ngH1 to cleave the stuffer fragment and thus lower the background of intact \ vector in the library. The ligated mixture was packaged into lambda 5 phage particles (Hohn, 1979) and plated on E. coli K802; 1.4x10 plaques were obtained. Assuming an estimated genome size of 10,000 kilobase pairs (kbp) for B. japonicum USDA 110 and a minimum insert size of 6 kbp, a complete representation of the genome (P=0.99) would be expected in less than 7,700 recombinant phage. Hybridization Methods. Lambda library plaques were prepared by standard procedures (Maniatis t l., 1982) with E. coli ET8051 grown in LB plus 3 mM glutamine, 10 mM M9804 and 0.2% maltose. Phage DNA from plaques was transferred to cellulose nitrate sheets as described by Benton & Davis (1977). Hybridization with nick-translated DNA probes were carried out for 24 h at 45°C in 40% formamide, 2x Denhardt solution (Denhardt, 1966), 5x SSPE (1x SSPE is 180 mM NaCl, 10 mM NaPO4 [pH 7.7], 15 1.0 mM EDTA), and 0.2 mg of sheared and denatured salmon sperm DNA per ml. The filters were then washed at room temperature twice for 15 min in 2x SSPE, 0.1% sodium dodecyl sulfate (SDS) followed by two 15 min washes in 0.25x SSPE, 0.1% SDS. Southern transfers of restriction endonuclease fragments separated by agarose gel electrophoresis (Southern, 1975), were hybridized and treated in a similar fashion. Hybridization signals were detected by autoradiography at -70°C. Plasmid Complementation ggg _§ Agggys. Plasmid complementation of glutamine auxotrophy was tested in B. ggii ET8051 with various plasmids constructed from the expression vector pUC8 (Vieira & Messing, 1982). The cells were streaked on M9 glycerol defined medium agar plates with 0.5 mM isopropylthio-B-D-galactoside (IPTG), which is included as a synthetic inducer of the 1gg promoter (Vieira & Messing, 1982). This medium contains NH4+ as the sole nitrogen source. GS activity was determined on cell—free extracts of B. japonicum USDA 110 grown to the late log phase in MA medium or B. coli grown to the stationary phase in LB broth plus 1 mM IPTG, 80 ug of ampicillin per ml, and 3 mM glutamine. Cell-free extracts were prepared by the procedure of Tronick gt g1. (1973), with the following minor modifications. Cells were suspended in 5 volumes of grinding buffer (10 mM imidazole-hydrochloride [pH 7.0], 1.0 mM MnCl and disrupted by two passes through a French pressure cell 2) at 12,000 lb/inz. The streptomycin sulfate precipitation was omitted. To separate GSI and GSII, cell-free extracts were loaded over a contin- uous gradient of 5% to 20% (wt/vol) sucrose prepared in grinding buffer and centrifuged for 4 h at 45,000 rpm in a Beckman SW 50.1 rotor at 4°C (Darrow, 1980). The gradient fractions were assayed for total glutamine synthetase activity by the Y—glutamyl transferase assay as described by Shapiro & Stadtman (1970), with Mn2+ as the activating cation to measure GSII and both adenylylated and non—adenylylated forms of GSI (Darrow & Knotts, 1977). Heat inactivation was carried out by incubation of the crude extract at 58°C for 30 min. One unit of enzyme produces 1 umol of Y—glutamylhydroxamate in 1 min at 37°C. Protein concentrations were determined by a modification of the Lowry procedure (Markwell t al., 1978) with bovine serum albumin as a standard. nggig Bacteria Isolation. The total bacterial population from frozen soybean nodules was prepared as described previously (Adams & Chelm, 1984). This preparation can be separated into three developmental fractions by centrifugation through a discontinuous sucrose gradient (Ching _t gi., 1977). This procedure has been adapted in our laboratory for use with a zonal ultracentrifuge rotor as follows. The total bacterial fraction from 50 g of nodules was loaded on top of a discontinous sucrose gradient and centrifuged at 40,000 rpm for 4 11 at 4°C in a Beckman 14 Ti zonal ultracentrifuge rotor. The gradient had been prepared by sequentially loading 90, 140, 160, and 275 ml of 45%, 50%, 52%, and 57% sucrose (wt/vol in 50 mM KPO [pH 7.5]), respectively, from the outer edge of the zonal 4? rotor. The fractions containing each of the bacterial forms from three separate gradients were combined, and each of the three combined fractions was rerun through a second similar gradient. The peak fractions from these second gradients constituted the doubly purified nodule bacteria, transforming bacteria, and bacteroids. S1 Nuclease Protection Analysis. Transcriptional initiation sites were mapped by a modification of the SI nuclease protection technique of Berk and Sharp (1977). The 360 basepair (bp) BgiI frag- ment, subcloned in pBJ93 was purified and 5’ end-labeled with T4-poly- nucleotide kinase (Maniatis gt gi., 1982). The two labeled strands were separated by boiling in 80% formamide for 5 min followed by electro— phoresis for 2 days at 4°C on an 8% polyacrylamide gel (Maniatis gt gi., 1982). The slower—migrating fragment on this gel was determined to be the coding strand by DNA sequence analysis. The 5’ end—labeled coding strand DNA was precipitated with 20 ug of RNA, suspended in 10 ul of hybridization buffer (Adams & Chelm, 1984), boiled for 10 min, and allowed to hybridize for 3 h at 58°C. The sample was then added to 0.3 ml of SI digestion buffer (Adams & Chelm, 1984) containing 40 units of SI nuclease (PL Biochemicals) and incubated for 30 min at 37°C. The digestion was stopped with 75 ul of 2.5 M ammonium acetate, 50 mM EDTA and precipitated with 0.8 ml of ethanol. The pellet was washed in 80% ethanol and suspended in electrophoresis sample dye (Maxam & Gilbert, 1980). The resultant DNA fragments were separated on denaturing 6% polyacrylamide gels (Maxam & Gilbert, 1980). Results B. japonicum Library Screening. A genomic library, constructed in the B. coli phage lambda vector XBFIOI, was screened for recombinant phage carrying B. japonicum GS genes by plaque hybridization with an B. coli GS probe. A 600 bp BgmHI—ngRI restriction endonuclease fragment __‘ from the middle of the B. coli glnA gene was subcloned from p811 (Fisher t al., 1981) to yield pBJ6 and used to probe cellulose nitrate trans- fers of lambda library plaques formed on B. coli ET8051. This strain carries a deletion of the glnA gene and thus eliminates background hybridization of the B. coli probe to the B. coli chromosomal DNA within the plaques (Fisher gt _i., 1981). Several copies of the library were screened; two different phage, N6IA and N6IG, were isolated. Restric— tion endonuclease mapping of these two phage indicated that they contain a 6.5 kbp overlapping region of rhizobial DNA with the cloned fragments in opposite orientation with respect to the lambda vehicle, indicating that, at least in the region of overlap, they contain a contiguous portion of the B. japonicum genome. The restriction map of the com- bined 11.8 kbp region of the cloned DNA is shown in Figure 1. Localization gt the glnA Homology. To localize the region of x6IA and x6IG that is homologous to the B. coli glnA probe, DNA from the recombinant phage k61A was purified and analyzed by Southern blot hybridization. Three restriction endonuclease fragments of the B. coli gigA gene, representing the 5’ end, middle, and 3’ end of the gene hybridized with approximately equal efficiency to X6IA DNA. Each probe, however, hybridized to different regions of the cloned DNA (Figure 2). The location and direction of transcription of the B. japonicum GS gene as shown in Figure 2 is inferred from these results by analogy to the structure of the E. coli glnA gene (Backman gt gi., 1981). Complementation Analysis. To confirm the presence of a GS gene in these recombinant phages, several subclones that contained the entire 19 .3occm nmcopm; mmoco ago an umpmuwnce m? wcmm «dam mgp mo cowpmmoq mgh .meoeoca wMH ms“ Eocc cowpqwcumcmcp mo covpumcwu mgp mcwpmuwucw mzogcm mg“ cpwz .wusa couum> cowmmmcaxw mgp oycw coco—093m mam; gasp mcowmwc mzp mpmomucw mcsmwm mcp we soppon mgp pm mean mgh .mex can I—_ l pBJ53 A pBJ55 —_> _> I pB56 21 .mmnoca HHdw .m moon“ mco op xmo_oEo; mo mcowmmc mcp >3 umpuwumca mw awe cowpuwcpmmc amwfldmmMfi .m mzp zonn cmcmwpm mm Azoccm cacopmgv mcmm a 8.558 e: 482 1% Ha £828 535888 :23 E... can .233 95% czogm wcmm «dam HHdw .m ms“ wo vcw .m ucm .mFuqu .vcw .m mcp Eocw manoca wgp 0p wN? -chnz; was“ mpcmsmmcm co_powcpmmc <29 amwflddQMH .m on“ mmpmuwucw mczm_w ago mo cowpcoa Logor one .C, T+C, and C indicate lanes with Maxam and Gilbert sequence reactions on the labeled coding strand of the 232 bp EI—Hian fragment (Figure 4). SI nuclease protection experiments used the 390 bp El fragment with th+e following RNA+typesz MBX medium with glutamate (U), MBXglutamate—NH4 (UN), MBX-NH (N), formate-NH (F), anaerobic YEX (AA), total bacterial populat on from nodules lTN), nodule bacteria (NB), transforming bacteria (TB), bacteroids (B), and B. coli ET8051 (E). GACTCC UUNN FAATNNBTBB E 533?“:- . 31 A comparison of the DNA sequence surrounding the B. japonicum glnA promoter to promoter sequences from B. japonicum and B. coli is shown in Figure 6. Weak homology between the B. jagonicum glnA promoter and the B. coli consensus promoter (Hawley & McClure, 1983) can be detected. In contrast, no apparent homology is observed between the glnA promoter and the B. japonicum git promoters (Adams and Chelm, 1984). Discussion The regulation of GS is of special interest in the Rhizobium-legume symbiosis, where repression of GS activity in the differentiated bacter— oid allows for newly fixed nitrogen to be exported to the plant cyto— plasm in the form of ammonia, where it is assimilated by a plant—en— coded, nodule—specific GS (Lara gt gi., 1983). In order to initiate the study of rhizobial GS regulation on the molecular level, I have isolated a GS gene from B. japonicum USDA 110. I conclude that this gene, designated as gigg, encodes the GSI protein of B. japonicum based on several lines of evidence. B. japonicum gigA is homologous to B. ggii gigA throughout a region of at least 1.35 kbp, a region larger than that necessary for a gene encoding the 36,000 dalton subunit of GSII, but in good agreement with the amount of DNA necessary to encode the 59,000 dalton subunit of GSI. The enzyme encoded by the rhizobial gene, when expressed in B. coli, cosediments with native B. japonicum GSI and exhibits the same relative s-ability to heat, a property specific to GSI (Darrow & Knotts, 1977). We do not know whether the rhizobial GSI can 32 B. coli consensus a ------ tcTTGACat--t -------- t-tg-TAtAaT ------ cat * B. japonicum glnA AatccgggCTgGACgctaTctgagccgggTGcTAchcgcattttgc * B. japonicum nifH TaagGTGchgthAGACCtTGGCAchCtGTTGCtgAtaachGca * B. japonicum nifDK TttaGTGth-aTgAGACCcTGGCAtGcCgGTTGCaaAgtcttGGat Figure 6. B. japonicum glnA Promoter Sequence Comparison. The upper portion is a comparison between the B. coli consensus promoter (Hawley & McClure, 1983) and the B. japonicum glnA promoter. Uppercase letters in the B. coli sequence indicate highly conserved bases among B. ggii promoters. Uppercase letters in the glnA sequence indicate homol— ogies to the B. coli consensus promoter. The lower portion is a compar— ison between two B. jagonicum git promoters (Adams & Chelm, 1984). Uppercase letters in the git sequences indicate homologies between the nifH and nifDK promoters. The asterisks indicate points of transcrip— tional initiation. 33 be adenylylated in B. coli as it can be in B. japonicum (Darrow & Knotts, 1977). The $1 nuclease protection data show that, in all cases which we have examined, gigg is transcribed from a single promoter. However, RNA from cells grown under a variety of conditions yield different amounts of protected fragment (Figure 5). Under the hybridization conditions used here, the amount of protected fragment is proportional to the amount of total RNA in the hybridization reaction, indicating that the gigA probe is in excess over its transcript. Therefore, the amount of protected fragment is proportional to the abundance of gigg message in each of the RNA types. Assuming that the degradation rate of the 5’ end of the mRNA does not vary significantly in the different cell types, it is possible to use these results to estimate the relative activity of the gigA promoter in cells grown under various conditions. 0f partic— ular interest is the activity of the gigg promoter in the three bacterial forms found in nodules, with the bacteroid RNA having less gigA message than the RNA from either transforming bacteria or nodule bacteria (Figure 5). These data suggest that the low bacterial GS activity in nodules is at least partially due to lowered amounts of gigA transcript in the nitrogen fixing bacteroids. The mechanism of this response is unknown, although it may be due to specific repression of the gigA promoter in the latter stages of differentiation. This tran- scriptional control would be similar to the response of the P11 promoter of Anabaena gigg, which is also repressed during nitrogen fixation (Tumer gt g1.,1983). In B. japonicum, total biosynthetic GS activity is dependent on the amount of GSI and GSII as well as on the adenylylation state of GSI 34 (Darrow, I980; Darrow gt _1., 1981). I find that total GS activity is not necessarily coordinated with the regulation of gigA transcription. For example, cells grown on glutamate and NH4+ have a high concentration of gigA message (Figure 5), but total GS activity is low due to the adenylylation of GSI and the repression of GSII (Ludwig, 1980b). I conclude that the constitutive gigg promoter is responsible for provid- ing a relative constant level of GSI and contributes little to the regulation of total GS activity in free living cells. This pattern of GS regulation can be compared to that found in other procaryotes. In B. coli, glnA is transcribed from two promoters. One provides a basal level of activity, and the other is induced during nitrogen starvation (Reitzer & Magasanik, 1985). In Anabaena, gigg is transcribed from several promoters in ammonia-grown cells, but a single git-like promoter is specifically induced when cells are derepressed for nitrogenase (Tumer et l., 1983). A situation similar to this exists in Klebsiella pneumoniae (Dixon, 1984). Thus in B. coli, Anabaena and K. pneumoniae, a single GS gene is transcribed from two or more differentially regu- lated promoters. In B. jagonicum, however, GS regulation has been divided between two genes. Since I detect little transcriptional control of gigg, I conclude that GSI activity is primarily regulated by adenylylation. Although GSII activity, like GSI, is dramatically regulated in response to oxygen concentration and nitrogen source t al., 1981; Ludwig, 1980b; Rao gt gi.,1978), there is no known (Darrow regulation of GSII by post—translational modification. It is therefore likely that GSII activity is controlled at the transcriptional level. A comparison of the promoter sequences of the two GS genes of B. japonicum will be of interest. CHAPTER 3 Apparent Eucaryotic Origin of Glutamine Synthetase II from the Bacterium Bradyrhizobium japonicum Introduction The Rhizobiaceae family of bacteria is characterized by their ability to form cortical hypertrophies on plants, and by the fact that bacteria can be reisolated from these galls or nodules (Jordan, 1984). This family includes the genera Rhizobium, Bradyrhizobium, Aqrobacterium and Phyllobacterium. Another unique feature of these bacteria is that they contain two forms of the enzyme glutamine synthetase, termed GSI and GSII (Darrow & Knotts, 1977; Fuchs & Keister, 1980). GSI is typical of procaryotic glutamine synthetases with respect to enzyme structure, the modulation of activity by post—translational modification, immuno— logical cross—reactivity, and amino acid sequence (Darrow & Knotts, 1977; Tronick _t _i., 1973; Carlson gt al., 1985). By contrast, GSII is distinct from all other known procaryotic glutamine synthetases in structure and immunological reactivity, and is not known to undergo post—translational modification. In these respects GSII is similar to eucaryotic glutamine synthetases (DeVries gt g1., 1983). I have isolated and characterized the gene encoding GSII, which 35 36 I term glnII, from Bradyrhizobium japonicum, the soybean symbiont. In this chapter, I show that the amino acid sequence of GSII, as inferred from the gene sequence, is highly homologous to plant glutamine synthe- tases, suggesting that this bacterial gene is of eucaryotic origin. Materials and Methods B§11 Purification ggg Protein Methods. Bradyrhizobium japonicum USDA 110 was grown to mid-log phase in 2 l of MBX medium (Chapter 2) with 10 mM glutamate. Cell extracts were prepared by the method of Tronick gt g1. (1973) using GS Buffer (GSB is 10 mM imidazole—HCl [pH 7.15] and 1.0 mM MnClz). GS activity is expressed in units of umoles of Y-glutamylhydroxamate per min as measured by the reverse transferase assay (Bender gt _i., 1977). Heat inactivation of GSII was carried out on undiluted extracts by incubation at 580C for 30 min. In the extract used for purification, GSII accounted for 88% of the total GS activity. All purification steps were done at 40C. The extract was diluted with GSB to a protein concentration of 1 mg per ml and loaded on a 10 ml Affi—Gel Blue (Bio-Rad) column at a flow rate of 0.5 ml per min. The column was washed with 50 ml of 088 and eluted with a 50 ml linear gradient of 0 to 5 mM ATP in GSB. A broad peak of GS activity, including both GSI and GSII, eluted immediately upon the initiation of the ATP gradient. The fractions with the highest specific GS activity were combined and loaded onto a Bio-Rad Bio-Gel TSK DEAE-5-PW ion exchange HPLC column, (75 x 7.5 mm) equilibrated with GSB, at a flow rate of 1 ml per min. The column was eluted with a 40 ml 37 linear gradient of 0 to 500 mM KCl in GSB. The first and second peaks of GS activity corresponded to GSII and GSI respectively. One ml fractions for sequence determination were desalted on a 10 ml Bio-Rad P—6 column and lyophylized. Total protein was assayed by the method of Lowry gt g1. (1951). SDS polyacrylamide electrophoresis gels (Laemmli, 1970) were stained with Coomassie Brilliant Blue. Protein sequencing was performed on an Applied Biosystems Model 470A gas phase sequencer at the University of Michigan protein sequencing facility, Ann Arbor, Michigan. Oligonucleotide ggg Bfl_ Methods. Oligonucleotide probes were synthesized by the phosporamidite method on an Applied Biosystems Model 380A DNA synthesizer. Oligonucleotides were purified by denaturing 12% polyacrylamide gel electrophoresis (Maxmn & Gilbert, 1980). DNA was sequenced by the dideoxy nucleotide chain termination method (Sanger gt g1., 1977). Random fragments for sequencing were generated by soni- cation (Deininger, 1983) and subcloned into the ngI site of M13—mp19 (Norrander gt gi., 1983). All regions were sequenced either on both strands or from three different fragments of the same strand. An ordered B. japonicum genomic library was constructed by cloning B. jagonicum DNA, partially digested with BthI, into the cosmid cloning vector pLAFRI (Adams gt al., 1984; Friedman gt 31., 1982). Cultures containing individual cosmids were replicated onto cellulose nitrate sheets layered on LB plates. The resulting colonies were screened by hybridization (Grunstein & Hogness, 1975) with the 5’ end radiolabelled mixed Oligonucleotide probe (Whitehead gt gi., 1983; Wood t l., 1985). 38 Complementation Analysis. Plasmids were tested for their ability to complement glutamine auxotrophy in Escherichia coli ET8051 Lo(tgg- gigg) ggttk rt; Nalr] (Fisher gt g1., 1981) on M9 defined medium agar plates (Miller, 1972) with 0.2% glucose and 1 mM thiamine. GSI and GSII were separated by sucrose density gradient centrifugation as described previously (Carlson et l., 1985) except that the centrifugation was carried out at 50,000 rpm for 3 h. Results and Discussion GSII of B. japonicum was purified as described in the materials and methods section. The results of the purification procedure are summar- ized in Table 2. This preparation was greater than 95% pure as indi— cated by SDS—polyacrylamide gel electrophoresis (Figure 7). The amino acid sequence of the amino terminus of GSII was determined by sequential Edman degradation and was used to design a mixed Oligonucleotide probe with homology to the DNA which encodes the first six amino acids of GSII (Figure 8). An ambiguity at position 16 of the Oligonucleotide was not included because two of the leucine codons, UUA and UUG, are infrequent— ly utilized in B. japonicum. Two cosmids, pchos7—20 and pchos13-79, were isolated from a genomic library by hybridization to the mixed Oligonucleotide probe. The region of hybridization in pchos7-20 was localized to 5.0 kbp ngRI, 1.0 kbp B9111, and 2.1 kbp Bgil fragments. These fragment sizes agree with those detected by hybridization to Southern transfers of total genomic DNA restriction digests. Table 2. GSII Purification 39 Fraction GSII Total Specific Purification Percent Activity Protein Activity Factor Yield (units) (mg) (units mg ) Extract 149 48 3.10 (1) (100) Affi-Gel Blue 39.1 0.89 43.9 14 26 DEAE HPLC 9.92 0.12 82.7 27 6.7 fraction 17 40 Figure 7. GSII Purification: SDS—Polyacrylamide Gel Electro— phoresis Analysis. Samples from the GSII purification procedure as described in the methods section were separated on a 10% gel. The major band in DEAE HPLC fractions 17—21 represents GSII. The large and small molecular weight markers are bovine serum albumin and ovalbumin, respec— tively. GSI eluted in fractions 22 through 25 and produced a single band on a similar polyacrylamide gel. DEAE HPLC fractions Extract Affi-Gel Blue 16 17 18 19 20 21 22 23 42 5 10 15 Met Thr Lys Tyr Lys Leu Glu Tyr Ile Trp Leu Asp Gly Tyr Thr 3,TAC TGG TTC ATG TTC GA 5, A A T T C 20 * 25 * 30 * Pro Thr Pro Asn Leu Arg Gly Lys Val Gln Ile Lys Glu His Ala Glu Figure 8. GSII Amino Terminus Sequence and Mixed Oligonucleotide Probe. The first 31 amino acid residues of B. japonicum GSII, as determined by protein sequencing, are shown. The first 5 amino acids of the enzyme were used to design the structure of the mixed Oligonucleo— tide probe as indicated. Asterisks indicate positions of discrepancies with the GSII amino acid sequence as determined from the DNA sequence. 43 The 2.1 kbp Bgil fragment with homology to the Oligonucleotide probe was subcloned in both orientations in pBR322. The resulting plasmids, pBJI96A and pBJ196B, were tested for their ability to comple— ment glutamine auxotrophy in B. ggii ET8051. ET8051(pBJlQ6A) grew well on defined medium plates with ammonia as the sole nitrogen source and yielded extracts with 0.82 units of GS activity per mg of protein. This GS activity was completely heat labile, a property specific to GSII, and cosedimented with the GSII activity of B. japonicum in sucrose density gradient centrifugation (Figure 9). These data indicate that the gene encoding B. japonicum GSII, gig11, is located entirely on the 2.1 kbp _g11 fragment insert of pBJlg6A. ET8051(pBJ196B), ET8051(pchos7-20), and the control strain, ET8051(pBR322), gave no detectable growth on defined medium plates and produced no detectable glutamine synthetase activity (less than 0.001 units per mg protein). Apparently the comple— mentation observed with ET8051(pBJI96A) is due to gig11 expression from the tgt promoter of the vector, in agreement with the gene orientation as indicated by sequence determination. In order to determine the precise location of the gig11 gene, I sequenced the 2.1 kbp BgiI fragment cloned in pBJI96A. The DNA sequence contains a single long open reading frame which encodes GSII (Figure 10). The subunit molecular weight of GSII predicted by the complete inferred amino acid sequence is 36,904 daltons, in good agreement with 36,000 daltons as determined by SDS—polyacrylamide gel electrophoresis (Darrow, 1980). The amino terminus of GSII matches the protein sequenc- ing data in 28 of 3] positions (Figure 8). The discrepancies at posit— ions 24, 29, and 31 are presumably due to inaccuracies in the final cycles of the protein sequencing. 44 OD500 o 5 10 15 2o 25 Fraction Number Figure 9. Identification of GSII by Sedimentation Centrifugation. The dashed line shows the separation of GSI and GSII of B. jagonicum. The larger GSI (fraction 10) sediments further than the smaller GSII (fraction 18). The solid line shows that GS from B. coli 45 U1 * * * * * ATG ACC AAG TAT AAG CTC GAG TAC ATC TGG CTC GAC GGA TAT ACG CCG ACT CCG AAC M T K Y L E I W L G Y T P T P N *2. ”5 9% XE 45* X- )1- TI V > a x§4 ME “3 >§. we «a e .06) >1- *5 C") re mg 32%!- 5') v5 ('5 (‘3 1% og* 0% _4 e .4 08H 0&3 TTGGGGC LWG e ('3 0% r—i U7 Mn _‘ 3(- * 225 C GTC TTC CCG GAC GCC GCG CGC ACC AAC GGC GTG V F P D A A R N G TGC TG C V >5 mi mfi Igi we _4 “2 eg. (7) ('3 *3 Kg ('5 we (I) <3 gx- Z> * * * CTC GTG ATG TGC GAA GTC ATG ATG CCC GAT GGC L V M C E V M M P D G ‘k * 3% C AAC AAG CGC GCC ACC AlT CTC GAC N K R A T I L D ('3 C GCC A x§4 we AC T n “Om IE. * ‘k * TTCGAGCAGGAATACTTCTTCTACAAGGACGGCCGT F E Q E Y F F Y K D G R * * 375 C lTC CCG CTC GGC TTC CCG ACC P L G F P T F >R" =§ GAC D >9 CGGC G c>§ 4 ‘It * 'k 450 TCC GGT TAT CCG GCG CCG CAG GGC CCG TAC TAC ACC GGC GTC GGC lTC TCG AAC GTC GGC GAC GTC GCC CGC AAG S G P Q P Y Y T G V G F S N V G D V A R K * * 525 11' ATCGTCGAAGAGCATCTCGACCTCTGCCTCGCGGCCGGCATCAACCATGAAGGCATCAACGCGGAAGTCGCGAAG I HLDLCLAAGINHEGINAEVAK ~k * TG TGG ATG GCC CGC TAC CTG M W M L 0% gag 4 so we a: fili- as e a mgi as on a»; a2 as >5. GCTGACGAA A D E > * * 675 GCTTGGCGACACGGAC GGAAC L G D T D W N 750 3(- X- ‘- SFV Z G) .< * 990 GTTCGCAGATCCTGA V R R S - Figure 10. DNA and Amino Acid Sequence of glnII and GSII. The DNA sequence of glnII is shown starting with the ATG initiation codon and ending with the TGA termination codon. The predicted amino acid sequence of GSII is shown with single letter amino acid codes. 46 The comparisons of the B. japonicum GSII amino acid sequence with Phaseolus vulgaris root GS (Gebhardt gt 1., 1986) and Anabaena 7120 GS (Tumer gt gi., 1983) are shown as homology matrixes in Figure 11 and Figure 12. These matrixes were generated using the analysis program of Pustell and Kafatos (1984) with parameters set so that each letter within the matrix represents a match of 47% or greater over a span of 23 amino acids. It can be seen that GSII has only limited homology to the bacterial GS but extensive homology to the plant GS. GSII was compared to a variety of glutamine synthetases using the method of Lipman and Pearson (1985) which optimizes the alignment between amino acid sequences and quantitates the significance of the similarity (Table 3). Despite extensive homology among most procaryotic glutamine synthetases, as indicated by the ability of their genes to cross hybridize, the homology between GSII and a typical bacterial glutamine synthetase of Anabaena is only marginally significant. In contrast, GSII has extensive homology with all eucaryotic glutamine synthetases that we have examined. Because all suggested cases of convergent evolution result in similarities of enzyme function without extensive sequence homology (Bannister & Parker, 1985), and because GSII is found only in the Rhizobiaceae (Fuchs & Keister, 1980), we conclude that the B. japonicum g1g11 gene is the result of a eucaryote to pro- caryote gene transfer event. We suggest that a plant served as the source of the progenitor gig11 gene because of the plant pathogenic nature of the Rhizobiaceae. The presence of the gig11 gene in the Rhizobiaceae is the first evidence of gene transfer to symbiotic bacteria from the eucaryotic host. Another suggested example of eucaryote to procaryote gene 47 so 100 150 200 250 300 x 50 ....... Yo; ................... : ................................... s z 100 ............... i ................................................. 5 wx 150 ....................... th ....................................... K . u : sQR 200 ................................... PX... ......................... w: Va 250 ............................................ Vv .................. vY . a2 . W110 300 .......................................: ............... 0 ........ . RP . . 5' ° 2 'z : Y : PN : : v 350 ................................................................. Figure 11. Homology Matrix Comparing GSII With Phaseolus vulgaris Root GS. The B. japonicum GSII amino acid sequence is plotted along the x axis and is compared to the amino acid sequence of Phaseolus vulgaris root GS plotted along the y axis. The homology matrix was plotted using the following parameters: range=11, scale factor=0.75, minimum value plotted=47%, compression=5. Each letter represents a 2% range of homology for that region of the matrix (A=100%—99%, B=98%—97%,... Z=50%-49%, a=48%—47%). 50 100 150 200 250 300 50 ........................................................ a I ..... x la a 100 ................................................................. Iaa 150 ................................................................. a 200 ................................................................. a 250 ................................................................. SS 2 300 ................................................................. v E 350 .......................................................... Sa ..... 400 ................................................................. 450 ................................................................. Figure 12. Homology Matrix Comparing GSII With Anabaena GS. The B. japonicum GSII (x axis) is compared to Anabaena GS (y axis). The homology matrix was plotted as described in Figure 11. 49 Table 3 The Statistical Significance of Amino Acid Sequence Homologies Between B. japonicum GSII and other Glutamine Synthetases Glutamine Synthetase Source % Identity Similarity z Valueb to GSII Score Bradyrhizobium japonicum 100 1719 208.4 Phaseolus vulgarisC 43.8 663 77.3 Biggm sativumd 46.6 642 73.9 Medicago staivae 43.8 674 78.9 Nicotiana giggbaqinifolad 42.6 667 74.0 Anabaena 7120f 24.4 114 8.83 All scores are from optimized alignments calculated with ktup = 1 and 1000 randomized sequences using the method of Lipman and Pearson (1985). C' The 2 value equals the similarity score minus the mean of the similarity scores of the randomized sequences divided by the standard deviation of the scores of the randomized sequences. A similarity score is considered significant if its 2 value is greater than 10. Z values less than 3 are not significant. C Gebhardt gt al., 1986 d Coruzzi at gi., 1986 e Tischer _t al., 1986 f Tumer gt g1., 1983 50 transfer, involving bacteriocuprein of Photobacterium leiognathi (Bannister & Parker, 1985), has recently come under dispute (Dunlap & Steinman, 1986). Although sequence homology suggests the origin of the g1g11 gene, questions pertaining to the mechanism, frequency, or function of gene transfer between symbionts or pathogens and hosts cannot be directly addressed. Since plant GS genes are known to contain introns (Tischer t al., 1986), B. japonicum gig11 had to evolve further by the loss of the introns, or the gene transfer must have occurred prior to the development of introns in the plant genes. In addition, there is no obvious reason why the aquisition of a eucaryotic gene would confer a selective advantage on a plant symbiotic bacterium. GSII is not known to serve an essential function, acting only to provide extra ammonia assimilatory capacity during growth under nitrogen limited conditions (Darrow gt gi., 1981). Nevertheless, it appears that the ability to acquire genetic information from eucaryotes can be a source of genetic diversity in the evolution of bacteria. CHAPTER 4 Transcriptional Control of glnII, the Gene Encoding Glutamine Synthetase II of Bradyrhizobium japonicum Intoduction In bacteria, nitrogen is generally assimilated in the form of ammonia by the coordinated activity of the enzymes glutamine synthetase (GS) and glutamate synthase. When nitrogen becomes limiting, the cell’s ammonia assimilatory capacity is enhanced by an increased synthesis and activity of GS. Nitrogen starvation also causes the induction of a number of other enzymes responsible for the production of ammonia from alternative nitrogen sources such as nitrate and amino acids. In the case of severe nitrogen starvation, diazotrophs have the ability to produce ammonia through the reduction of dinitrogen by the activity of the enzyme nitrogenase. The large number of genes and enzymes involved in nitrogen metabo- lism are carefully regulated by a complex nitrogen regulatory (Ntr) system (Magasanik, 1982). Two genera of nitrogen fixing bacteria, Rhizobium and Bradyrhizobium are unique in that their ability to fix nitrogen is limited, for the most part, to the highly differentiated symbiotic form termed bacteroids which are found within the root nodules 51 52 of leguminous plants. These bacteria are also unusual because as much as 94% of the fixed nitrogen is exported from the bacteroid and is assimilated by the plant symbiont (O’Gara & Shanmugam, I976). The purpose of this study is to investigate the mechanisms by which B. jagonicum regulates the unusual physiology of symbiotic nitrogen fixation. Nitrogen regulation has been best characterized in the enteric bacteria B. coli and K. pneumoniae (Magasanik, 1982; Alvarez—Morales gt gi., 1984). Genes that are regulated by the Ntr system have the unusual promoter sequence -27 CTGGCAC-NS-TTGCA -10 (Dixon, 1984b; 0w gt gi., 1983). Transcriptional activation of Ntr promoters requires the RNA polymerase core enzyme, an Ntr-specific sigma factor (060), and the regulatory protein NRI, encoded by the gtgt (gigB) gene (Hirschman gt g1., 1985; Hunt & Magasanik, 1985; Merrick & Gibbins, 1985). NRI also has DNA binding activity specific for the sequence AAAA TTTT tion of NRI bas been implicated in transcriptional activation (Reitzer & 5’-TGCACC TGGTGCA-3’ (Ames & Nikaido, I985). The DNA binding func— Magasanik, 1986), repression (Reitzer & Magasanik, I983; Hawkes gt gi., 1985) and antitermination (Ames & Nidaido, 1985). The conversion of NRI from its DNA binding form to its Ntr promoter activating form seems to _1., 1985). be central to the control of the Ntr system (Bueno gt One of the Ntr activated operons of K. pneumoniae is nifLA. The nifA gene product is a regulatory protein with homology to NRI (Buikema t al., 1985; Drummond et al., 1986) and is required for the activation (D of the genes encoding the nitrog-nase _nzyme complex (git genes; Dixon, 1984a). git promoters, like Ntr promoters, require the Ntr—specific sigma factor and have the U60 consensus promoter sequence (0w t g1., 53 1983). In addition, nif promoters have the upstream regulatory sequence 5’—TGT—N10—ACA-3’ which is required for efficient gitA dependent tran- scriptional initiation and is presumably responsible for the differen- tial regulation of git and Ntr promoters (Alvarez-Morales gt gi., 1986; Buck gt _1., 1986). The complex interactions among the Ntr promoters and regulatory proteins result in the effective regulation of bacterial nitrogen metabolism. There is evidence that B. japonicum has a nitrogen regulatory system which is homologous to the Ntr system of the enteric bacteria. GSI of B. jagonicum is regulated by an adenylylation cascade system homologous to the Ntr regulated adenylylation of GS in B. ggii. The best evidence for the existence of an Ntr system in B. japonicum is from the characterization of the git promoters. B. japonicum git promoters are very similar in structure to the nif promoters of B. pneumoniae. Both have the 060 recognition sequence (Adams & Chelm, 1984; Fuhrmann & Hennecke, 1984) and the upstream git activation sequence (Alvarez- Morales gt gi., 1986; Buck gt a1., 1986). Furthermore, the gitA gene product of B. gnuemoniae is able to activate transcription from B. japonicum git promoters (Alvarez—Morales & Hennecke, 1985). The conservation of git promoter sequences indicates extensive homology between the Ntr systems of B. japonicum and B. pnuemoniae. However, there may be differences in these systems which allow B. jagonicum to express the unusual metabolism of a symbiotic nitrogen fixing bacterium. A number of studies report the characterization of mutations in Rhizobium and Bradyrhizobium that affect both the Ntr dependent GS regulation and the effectiveness of symbiotic nitrogen fixation (Donald & Ludwig, 1984; Konderosi et l., 1977; Ludwig, 1980). 54 The complex phenotypes of these strains suggests that they carry muta- tions in the Ntr system. Recently, several genes have been isolated from B. japonicum with homology to the K. gnuemoniae nifA gene (Adams & Chelm, 1986; Fischer gt _1., 1986). Mutations in some of these genes have similar pleiotrophic phenotypes. Bacteria of the Rhizobiaceae family are distinct from all other bacteria by having a second form of GS, designated GSII (Darrow & Knotts, 1977), which seems to have been acquired from plants by a eucaryote to procaryote gene transfer event (Carlson & Chelm, 1986). The level of GSII activity is regulated in response to nitrogen and carbon source (Darrow gt gi., 1981; Ludwig, 1980b) and oxygen concen- tration (Rao gt g1., 1981). Because GSII is not known to be post translationally regulated, the gene encoding GSII, g1g11, is presumably regulated by an Ntr promoter. One possible role of the B. japonicum Ntr system is the repression of GS activity in nitrogen fixing bacteroids (Brown & Dilworth, 1975; Cullimore gt _1., 1983; DeVries _t _1., 1983). This repression is presumably responsible for the observed export of ammonia from bacteroids during symbiotic nitrogen fixation (O’Gara & Shanmugam, 1976). To date, no Ntr promoters have been characterized from B. japonicum. In Chapter 3 I described the isolation of the gene encoding GSII (gig11) of B. japonicum. In this chapter, I describe the characterization of the structure and expression of the Ntr regulated glnII promoter and discuss its potential role in the regulation of nitrogen metabolism in B. japonicum. 55 Materials and Methods Promoter Mapping. Transcriptional start sites were mapped by the SI nuclease protection method of Berk and Sharp (1977). Single stranded 5’ end—labeled probes were generated by the primer extension method (Adams & Chelm, 1986). 80 ng of Oligonucleotide were labeled with 10 units of T4-polynucleotide kinase (Bethesda Research Laboratory) at 37°C for 1 h in 100 mM tris(hydroxymethyl)aminomethane (Tris) -HCl [pH 7 5], 10 mM MgCl 6 mM dithioerythritol and 100 uCi of [Y—32P]ATP (3000 Ci 2, per mmole). In experiments with mixed probes, the specific activities of the 5’ end-labeled oligomers were adjusted to 250 cpm per pg with unlabeled Oligonucleotides prior to the elongation reaction. Labeled Oligonucleotides were combined with 20 ug of single stranded M13 recom- binant phage DNA, ethanol precipitated, suspended in 50 ul of Klenow buffer (10 mM Tris-HCl [pH 8.5], 10 mM MgClZ), heated to 900C for 10 min and allowed to hybridize for 1 h at 370C. The Oligonucleotide primer was extended with 3 units of the large fragment of B. coli DNA poly- merase (Klenow fragment, Bethesda Research Laboratory) for 1 h at 370C in 100 ul of Klenow buffer with 0.6 mM each of ATP, CTP, GTP and TTP. The partially double stranded DNA was ethanol precipitated, suspended in Bill reaction buffer (Bethesda Research Laboratory), digested with 10 units of Bill (Bethesda Research Laboratory) for 3 h at 370C, ethanol precipated, and suspended in formamide dye buffer. The elongated primer was purified by denaturing polyacrylamide gel electrophoresis (Maxam & Gilbert, 1980), detected by autoradiography, and eluted by the crush and soak method (Maxam and Gilbert, 1980). 56 SI nuclease protection experiments were carried out as described in Chapter 2, using 9000 cpm of probe hybridized with 10 ug of total cell RNA. Protected fragments were separated by denaturing polyacrylamide gel electrophoresis. Gels were dried onto filter paper by aspiration under vacuum prior to autoradiography. Bands of radioactivity from dried gels were cut out and the amount of radioactivity was determined by scintillation counting. Bacterial Cultures. Batch cultures of B. japonicum USDA 110 were grown at 28°C in 14 l fermenters (New Brunswick) with an agitation rate of 200 RPM and an aeration rate of 500 ml per min. The culture medium consisted of 50 mM 3-(N-morpholino)-propanesulphonate (MOPS), 0.03% 4, 0.03% NaZHPO4, 4 2 minerals and vitamins to the following concentrations (per liter): 10 (w/v) KHZPO 0.012% MgSO '7 H 0, 0.45 mM CaCl2 and trace mg H3B0 1.0 mg ZnSO ’7 H O, 0.5 mg CuSO '5 H 0, 0.5 mg MnCl2‘4 H O, 4, 0.1 mg Na 4 2 4 2 2 MoO4‘2 H20, 1.0 mg FeCl3, 0.2 mg riboflavin, 0.12 mg biotin, 2 0.8 mg thiamine HCl, 0.48 mg inositol, 0.08 mg p-aminobenzoic acid, 0.5 mg nicotinic acid, 0.8 mg pantothenic acid, and 0.001 mg cyanocobal— amine. The pH of the medium was adjusted to 6.8 by the addition of NaOH prior to autoclaving. The CaClz, vitamins, carbon source and nitrogen source were filter sterilized and added after autoclaving. Cultures were inoculated with cells grown to stationary phase in YEX medium (Chapter 2) with 5 mM NH4Cl. Samples from the batch cultures were taken every 12 h. Culture density was measured by optical density at 420 nm. Protein concentration was measured on whole cell lysates by the method of Lowry gt .gi. (1951) with the following modifications. Protein samples were prepared by mixing 0.5 ml of the bacterial culture 57 with an equal volume of trichloroacetic acid. The protein was precipi- tated in a microcentrifuge and suspended in 0.1 ml of 0.3 M NaOH. Bfl_ Purification. Cells for RNA purifications were harvested from 1 l samples by centrifugation, washed in 20 ml of H20, recentrifuged, suspended in 8.6 ml of H20 and stored at -70°C. To 8.6 mls of cell suspension was added 0.4 mls of 10% (w/v) N—lauroyl sarkosine and 1 ml of 1 M Tris-HCl [pH8.0]. The cells were lysed by two passes through a French pressure cell at 12,000 psi and immediately added to 10 gm of CsCl. The resulting mixture was layered over a 7 ml cushion of 5.6 M CsCl. The tube was filled with 0.1 M Tris-HCl [pH8.0], 0.4% (w/V) N-lauroyl sarkosine and centrifuged in a Beckman SW28 rotor at 22,500 RPM at 20°C for 36 h. RNA pellets were suspended in autoclaved H20, extracted twice with H20-saturated phenol, extracted four times with diethylether, precipitated with ethanol, suspended in autoclaved H20 and stored at -7o°c. Results Promoter Mapping. The site of transcriptional initiation upstream of g1g11 was mapped by the SI nuclease protection method. Single stranded 5’ end-labeled probe was generated using the recombinant phage M13glnII, which has the 2.1 kbp Bgil fragment of pBJI96 (Chapter 3) cloned in M13mp18 (Yanisch-Perron gt 1., 1985) oriented so that the non—coding strand is produced in the single stranded phage. Second strand synthesis was primed with the Oligonucleotide TCII 58 (5’-CGACGCGAATTCCTTGA—3’) which hybridizes to the DNA encoding amino acids 26 through 31 of GSII (Figure 10). The resulting probe, which includes about 700 bases upstream of the initiation codon of glnII, was hybridized to RNA prepared from B. japonicum USDA 110 and treated with $1 nuclease. A single protected fragment of about 170 bases in length was observed. The precise location of transcriptional initiation upstream of gig11 can be determined by comparing the length of the protected fragment to a sequencing ladder prepared with M13glnII and TCII. The sequence of the gig11 promoter can be read directly from the gel (Figure 13). Sequence Analysis. The DNA sequence upstream of gig11 was deter- mined from pBJI96 as described in Chapter 3 and has some interesting features (Figure 14). 40 bp upstream of the g1g11 initiation codon is the center of an imperfect inverted repeat. This sequence should produce a "hairpin" secondary structure in the 5’ untranslated region of the g1g11 mRNA. Upstream of the promoter is an unusual run of 13 or 14 consecutive G—C base pairs. The exact length of this structure is uncertain due to difficulties in sequencing regions of high G-C content. The promoter region of gig11 is shown in Figure 15. As expected, there is little homology between the promoters of gig11 and gigA (Figure 6). This lack of promoter homology is presumably responsible for the differential regulation of the two glutamine synthetases of B. japonicum. However, Figure 15 shows that there is extensive homology between the gigii promoter and several B. japonicum git promoters. This suggests that gig11 and git genes are regulated, at least in part, by a common mechanism. However, the differential expression of glnII and git 59 Figure 13. SI Nuclease Protection Mapping of the gigII Promoter. Lanes ACGT show DNA sequence reactions (Sanger gt gi., 1977) for the region upstream of glnII. SI nuclease protection reactions used the following RNA types: E, B. coli ET8051; B, B. japonicum USDA 110 grown to mid-log phase in MOPS buffered medium (Methods) plus 0.2% xylose and 10 mM glutamate. EB ACGT 1 61 -280 CATCGGCAAC CTCCGGCACC GGCCGGCCTG CTCCAGAATC AGGCGTTTTC CGCGGGATTT -220 [- - - -NRI— - — -] TTGCAGTGCA GCTTCACGCA AAGGTGCGCC GCTATGACGC Acoccoccoo GGACGACCGG -160 CCGCTCGCGG GGGGGGGGGG GGLGGCCGCG GTGGAAAACC TCCCGCAATG CGGCCTTTTG —100 GCACGCTAAA TGCTTGTAAA CGGTCGGCCG ATGGTGGCCG GGTACAAACG TGGGGGGCCC M T K Y K L CCGGGCCCCC AACCTTTCGC ATCATCTACA GAGAGGCTCA ATG ACC AAG TAT AAG CTC Figure 14. DNA Sequence Upstream of gigII. The DNA sequence was determined as described in Chapter 3. The asterisk marks the point of transcriptional initiation as indicated in Figure 13. The lines over the sequence from -55 to —24 mark the inverted repeat in the 5’ untrans— lated region of the glnII mRNA. The region with a potential NR1 binding site is marked in brackets. 62 B. l nifH taaggthcgggttaGaCCtTGGCACGgctGTTGCTgATaagCGGZa B. j nifDK ttTtagtgctcatGaGaCCcTGGCAtGchGTTGCaaAgtcttGGat B. i. nifB caTtcgcgtcatcctchCacGGCAtGCaAGTTGCTaAchthtgaE B. l fixA gcggtccCaagchchggaTGGtACaagAcTTGCTgtTctcttcc: B. i. g1g11 ccchchaatchgcctttTGGCACGCtAaaTGCTthaaaCGGtZ B. i. consensus --T-—-—C-—-—-G-G—CC—TGGCACGC—AGTTGCT—AT---CGG*** Ntr consensus ------------------- CTGGCAC ----- TTGCA —————————— * Figure 15. B. japonicum Promoter Sequence Comparison. The B. japonicum glnII promoter and a variety of git promoters are aligned to maximize homology. The B. japonicum consensus sequence was determined by matches in 3 out of 5 sequences. Homologies between the B. japonicum promoters and the B. japonicum consensus are shown as uppercase letters. Transcriptional initiation sites are shown with asterisks. The git promoters are from Adams and Chelm (1984) and Adams gt g1i (1986). The tiyg promoter is from Fuhrmann gt g1. (1985). The consensus sequence for Ntr promoters from enteric bacteria is from Dixon (1984b). 63 genes suggests that other sequences in these promoters are involved in transcriptional regulation. In B. coli, the transcription of glnA is partially controlled by NRI through binding at a number of NRI binding sites (see Introduction). A search of the 600 bp of DNA sequence upstream of the B. japonicum gigii promoter reveals only one region with better than 65% homology to the consensus NRI binding sequence of enteric bacteria (Figure 14). gig11 Promoter Characterization. The control of the gig11 promoter was characterized in a series of experiments which measure changes in the amount of cellular gig11 mRNA during the transition of B. japonicum cultures from log to stationary phase. In each culture, either the carbon or the nitrogen source is added at a low concentration so that the transition to stationary phase is concurrent with the depletion of a known nutrient. Therefore, any changes in gig11 transcription at this point in the growth curve can be attributed to specific physiological changes. The activity of the gig11 promoter was determined using quantita- tive S1 nuclease protection analysis so that the amount of protected fragment is proportional to the amount of g1g11 mRNA (Chapter 2). RNA for these experiments was purified from B. japonicum culture samples taken at six points in the growth curve, three before and three after the transition to stationary phase. The transcription of gigA was also monitored for comparison. The gigA SI nuclease probe was prepared using the primer extension method described above with the Oligonucleotide TCA (5’—CCCCTTTTGGCTCTGCG—3’) and the template M13glnA, which has the 391 bp BgiI fragment of pBJ93 cloned in M13mp18. This gives a 5’ end—labeled 64 single stranded probe 331 bases long and produces a 125 base SI nuclease protected fragment. The first culture was grown on 0.2% xylose and 1 mM NH4Cl. These conditions were chosen so that the culture would deplete the nitrogen source before growing to an 00420 of 1.0. The growth curve of the low ammonia xylose culture is shown in Figure I6N. The results of the quantitative SI protection assay (Figure 17N) show that there is no detectable gig11 transcription in B. japonicum when growing on xylose and ammonia, but when the nitrogen source is depleted, the gig11 pro- moter is activated. In contrast, gigg transcription is relatively constant throughout the growth curve. There is about a 3-fold decrease in gigA transcription during stationary phase. This type of differen- tial regulation between the gigg promoter and GSII was observed in Chapter 2. The continued increase in the optical density of the culture following the cessation of protein accumulation (Figure 16N) is due to the production of extracellular polysaccharide (EPS) after the depletion of nitrogen in stationary phase. Both the activation of gig11 and the production of EPS seem to be the result of the nitrogen limited and carbon excess condition following the depletion of ammonia. This conclusion is supported by the results of an analysis using a similar culture grown with a low concentration of xylose (0.02%) and excess nitrogen (5 mM NH4Cl; Figure 16C). Just as with the first culture, there is no detectable g1g11 transcription during growth on xylose and ammonia. However, when stationary phase is brought on by carbon limitation, the culture does not induce lnII transcription or EPS production (Figure 17C). tn— ':'.-_ I 'I . ._ fl“: 1 flfiii 43:. '1 "hit-L: - 65 .AFuaxz ze o.m ucm mpmcFuozm x¢.ov mczszu choEEm mchFouzm 30F 85F FOF m>czo gpgocm mLF mF m Foams .FFuvzz :5 o. m :%czm mmoFxx xmo. ov wcsFon «FCOEEm mmoFxx 30F wcF LOF m>cso LFZOLm 85F mF 0 chma 225 o. F van owe—xx xm. ov mczstu mmoFxx choEEm 30F 65F LOF m>c=u Lpzocm 85F mF z :quchm .m>onm umnFLUmmn mmFaEmm me mLF cF 3 umpompocq FFF>FFumonac Fo chogm mcF mchoLm an mpmu coFFuwFoca mmmwFusc Fm mLF mmNFLmE53m chmq Loam Fo FFm; cmzoF mLF .coFchmamca cmc mam; mFFwo cog: chFoq wEFF mcp 305m mzoccw 85F .coFchpcwucoo :Fmposa FmFOF use E: owe Fm oo mcp >5 umcsmmms mm mczsto cmFsuFchq m LOF m>czu ngocm 85F mzocm chma comm Fo :oFFcoq swag: wcF .muocpmz cam mFmFFmez cF cmnFcommn mm UquEwm new czocm mew; OFF ;=u guzocw mFumz FmEFch amwdmmde .m .oF wcszd (V) "mud lul/fin’ c: an a a, N I 66 l l I 4" 500 ~ 400 ~ 300 e zoo — 100 ~ ['1 ll uni ‘(01 vuio uldo Days Days 67 .vwFumev mm; mpoca deflm cpmcmF FF3F oz .mnoca «dam cpmcmF FF2F mF camp Fmoecmqaz mgF .umpmoFucF mom monoca deflm vcm «QHm umpomuoca zFFwFFLmQ mcF Fo mcoFFFmoa one .oF wcszd :F vamoncF mm m use .u .z mmczszu coan mcu Fo sumo EocF :wxmp mquEmm me EocF vwcmamca mm; FFmFFchzc an nmczmmme was FFmFmL mcF .conqFFumcmcF Mamam cam mafia Fo :ochaQEou .FF mesmFm omwmmw m mmumwr O U U U... <:_m II. . =5 mmemmF 69 A third culture was designed to study the effect of a preferred carbon source on GSII production. This culture is similar to the low xylose ammonia culture except that the xylose was replaced with succi— nate (0.04%). Succinate proved to be a very effective carbon source, supporting a rapid growth rate, even at this low concentration (Figure 16S). Succinate also had a dramatic effect on g1g11 transcription. Figure 17S shows that the gig11 promoter is active during growth on succinate and ammonia but is repressed after carbon source depletion. In the succinate culture, as in the low ammonia xylose culture, EPS production parallels gig11 expression. EPS accumulates during the growth phase and disappears after carbon source depletion. The pattern of‘ gigA expression in the succinate culture is very similar to the pattern observed in the first culture (compare Figure 17N and 175). These data are in agreement with earlier observations (Chapter 2) that the gigA promoter is not under general nitrogen control. The gig11 promoter, on the other hand, seems to be induced specifically during conditions of nitrogen limitation. In all conditions that I have tested, gig11 is transcribed from a single promoter. There is no other partial or full length protection of the gig11 probe. However, several RNA samples gave full length protec- tion of the gigA probe (Figure 17) suggesting the presence of an up- stream promoter. Earlier SI nuclease protection studies using strand separated probes did not detect any transcription upstream of the primary gigA promoter (Chapter 2). This discrepancy may be due to differences in the methods of probe s1nthesis although the possibility of a second glnA promoter cannot be excluded. 70 DISCUSSION I have previously demonstrated that the gigA gene of B. japonicum is transcribed from a single promoter which is not regulated by nitrogen source (Chapter 2). In contrast, GSII levels fluctuate greatly in response to the cell’s nitrogen nutrition (Ludwig, I980b; Darrow gt g1. 1981). These data suggested that gig11, may be transcribed from a nitrogen regulated promoter similar to the nitrogen regulated promoter of the gigA genes in other bacteria (Dixon, 1984b, Reitzer & Magasanik, 1985; Tumer gt _1., 1983). In this chapter I have shown that gig11 is transcriptionally regulated in response to changes in nitrogen metabo- lism and has a promoter distinct from the gigA promoter but very similar to B. japonicum git promoters. These data indicate that B. japonicum has a nitrogen regulation (Ntr) system homologous to the well character- ized Ntr systems of B. coli and K. pnuemoniae. gig11 is the first Ntr regulated gene to be characterized from B. japonicum. The Ntr regulation of gig11 is best illustrated by the results of the nutrient limited batch culture experiments, which show that gig11 is induced in conditions of nitrogen limitation. This response occurs in the low ammonia xylose culture (Figures 16N and 17N) when the cells go from a carbon limited to a nitrogen limited condition. B. japonicum growing on xylose and ammonia is presumably carbon limited. Xylose in not a very good carbon source since substituting even a low concentra- tion of succinate results in a much more rapid growth rate (compare Figures 16N and 165). Furthermore, ammonia is an excellent nitrogen source since even at low concentrations it will repress GS activity in both B. coli (Magasanik, 1982) and B. japonicum (Ludwig, 1978). 71 However, when ammonia is depleted and the cells become limited for nitrogen, B. japonicum responds by inducing gig11 transcription. This response is not due to general starvation since a similar low xylose ammonia culture does not induce gig11 in stationary phase (Figure 17C). The amount of gigii transcription during exponential growth is also controlled by nitrogen limitation. However, the extent of nitrogen limitation is not controlled by the nitrogen source but rather by the relative qualities of the carbon and nitrogen sources. For example, growth on xylose and ammonia is carbon limited as previously explained. However, by substituting the relatively poor carbon source xylose with a good carbon source such as succinate, the cells become nitrogen limited and gig11 is transcribed during the growth phase (Figure 175). When the succinate is depleted, the cells are no longer nitrogen limited and g1g11 transcription stops. This interpretation is supported by the observation that EPS accumulates during growth on succinate and ammonia. In most bacteria, EPS is produced during nitrogen limited growth (Harder & Dijkhuizen, 1983). The disappearance of EPS in stationary phase indicates that, unlike most bacteria, B. japonicum is able to utilize its own EPS as a carbon source (Dudman, 1977; Patel & Gerson, 1974). Apparently succinate is a sufficiently good carbon source to cause growth to be nitrogen limited even with ammonia as the nitrogen source. Similarly, carbon limited growth on xylose and ammonia can be made carbon rich by substituting ammonia with a relatively poor nitrogen source such as glutamate. Thus xylose glutamate grown cells have high levels of g1g11 transcription (Figure 13). Because gig11 is subject to nitrogen regulation, a comparison of the DNA sequence of the glnII promoter with well characterized Ntr 72 promoters of other bacteria may show that homologous mechanisms of gene regulation are functioning in B. japonicum. A number of different DNA sequences have been shown to be important in Ntr promoters, including the consensus Ntr promoter, the NRI binding sequence and the upstream git activating sequence. An analysis of the DNA sequence upstream of gig11 reveals some similarities and some differences to these Ntr promoter features. A comparison of Figure 6 and Figure 15 shows that there is little homology between the B. japonicum gigA and gig11 pro- moters but a large block of homology among gig11 and B. japonicum git promoters. This region of homology is similar in structure to the sequence of other bacterial Ntr promoters responsible for transcrip— tional initiation by RNA polymerase and the Ntr-specific sigma factor, encoded by the gene tpgfl. It is therefore reasonable to conclude that the conserved sequence upstream of gig11 is the RNA polymerase binding site and that B. japonicum has a gene analogous to gppfl which encodes an Ntr-specific sigma factor. Although Ntr and git genes share an Ntr specific promoter sequence, only git genes have the upstream regulatory sequence. This sequence has been found upstream of git promoters in many nitrogen fixing bacteria (Buck gt _1., 1986) and may account for the differential regulation of git genes and other Ntr activated promoters. In accordance with this model, I find that the git activating sequence does not occur upstream of the glnII promoter. Another feature of Ntr promoters is the NRI binding site. NRI is a DNA binding protein encoded by the gene gttt (gigB) and is required for Ntr promoter activation. NRI may also bind to DNA to function as a transcriptional activator, repressor, or antiterminator. I find only 73 one potential NRI binding site upstream of glnII (Figure 14). It is not known if this sequence is involved in glnII regulation by NRI binding, or even if B. japonicum has a regulatory protein analogous to NRI. In contrast, the Ntr regulated glnA promoter of ,B. coli has five NRI binding sites involved in gigA transcriptional regulation (Reitzer & Magasanik, 1986). More experiments are necessary in order to determine if a similar mechanism of regulation functions in B. japonicum. Despite extensive differences between the structures of B. gpii and B. japonicum GS genes, the overall regulation of GS seems quite similar. In B. coli a single GS gene, gl A, is transcribed from two differen- tially regulated promoters (Reitzer and Magasanik, 1985). One promoter is homologous to the consensus B. coli promoter and provides a constant low level of GS during growth on carbon limited media. This function is served in B. japonicum by the gigA gene which is transcribed in all conditions from a promoter independent of the cell’s nitrogen nutrition. The other B. ggii gigA promoter has multiple NRI binding sites and is activated by the Ntr system to provide high levels of GS during growth on nitrogen limited media. In B. japonicum, nitrogen starvation causes the production of high levels of GSII by means of a single inducible nitrogen regulated promoter. Although the glnII promoter is responding to nitrogen limitation, there is evidence that other factors are involved in gig11 regulation. Adams and Chelm (1986) have shown that g1g11 is transcribed at signifi- cant levels in bacteroids and in free living cultures grown at low oxygen concentrations, and that this transcription originates from the same promoter as the aerobic expression. gigii transcription in these conditions requires the presence of a gene, och (also termed nifA in 74 Fischer gt _1., 1986), which is homologous to the regulatory genes nifA and gttt of K. pneumoniae. gggB, however, is not responsible for aerobic expression of g1g11 since pggB deletion strains show normal aerobic regulation of gig11. The microaerobic specific activation of gig11 is unknown in the well characterized Ntr systems of the enteric bacteria and contrasts with previous studies which reported the repres— sion of B. japonicum GSII in bacteroids (Cullimore gt l., 1983) and microaerobic cultures (Rao gt l., 1978). It is not yet known if the ppr dependent transcription of gig11 results in the production of functional GSII, although it is possible that GSII is more abundant in bacteroids than previously thought. pggB mutants do not fully differ- entiate to form nitrogen fixing bacteroids. It will be of interest to characterize the role of och and other nitrogen regulatory factors in the regulation of nitrogen metabolism in symbiotic nitrogen fixing bacteria. The glnII promoter, along with git promoters, will serve as important tools in the characterization of the B. japonicum Ntr system. CHAPTER 5 Summary and Conclusions Glutamine synthetase (GS) catalyzes the first step in the primary pathway of bacterial ammonia assimilation. Thus the regulation of GS is an important element in the control of overall nitrogen metabolism. The enteric bacteria regulate GS yig a complex nitrogen regulatory system (Ntr) which has been the subject of extensive genetic and biochemical characterization (Magasanik, 1982). In contrast, the mechanisms controlling rhizobial nitrogen metabolism are not well understood. Rhizobial genes encoding the enzyme nitrogenase have been isolated and characterized. The concurrent activation of nitrogenase and repression of GS is apparently responsible for the export of fixed nitrogen from symbiotic bacteroids. In this dissertation, I have demonstrated the isolation and characterization of the genes encoding GSI and GSII of Bradyrhizobium japonicum, the soybean symbiont. This work will ulti- mately lead to a more complete understanding of the specific adaptations necessary for rhizobium to function as a symbiotic nitrogen fixing bacteria. In Chapter 1, I discussed the isolation and characterization of the gene encoding GSI. Unlike GSII, GSI is homologous in structure and sequence to the single GS of B. coli. Furthermore, GSI and B. coli GS 75 76 are both regulated by an adenylylation cascade system. Therefore, I chose the B. coli nomenclature glnA for the gene encoding B. japonicum GSI. The most significant finding of Chapter 1 is that glnA is consti- tutively transcribed from a single promoter. This is in contrast to other bacteria (B. coli, B. pnuemoniae, and Anabaena) where GS is transcribed from multiple promoters. Therefore, GSI is primarily regulated post-translationally by adenylylation. The constitutive production of a reversibly modified GS presumably allows for a very rapid adjustment of total GS activity as determined by the changing needs of the cell. The gigA promoter is the first "normal" promoter to be characterized from B. japonicum. Although the gigA promoter has some homology to the B. coli consensus promoter, it appears to have little or no activity in B. ggii. A similar result was observed with the Bgiip- gigg meliloti gigA promoter. Because it is always active, the B. japonicum gigA promoter may be useful for the expression of foreign genes in rhizobia. GSII is not known to be regulated by a post-translational modifi- cation. Therefore, the gene encoding GSII is probably transcriptionally regulated from an Ntr-like promoter. For this reason, the isolation and characterization of the GSII gene (gig11) was of particular interest. Numerous screenings of B. japonicum genomic libraries using heterologous hybridization probes resulted in the identification of only the gigA gene. gig11 was not detected because of the limited homology between the eucaryote-like gig11 sequence and the procaryotic probes. Attempts to isolate glnII by complementation of an B. coli glutamine auxotroph were also unsuccessful. Ultimately, glnII was identified in a cosmid clone bank using a mixed Oligonucleotide hybridization probe based on 77 a partial amino acid sequence determined from purified GSII (Chapter 3). This is the first report of the isolation of a gig11 gene. Amino acid sequence comparisons revealed extensive homology between GSII and eucaryotic glutamine synthetases but limited homology between GSII and procaryotic glutamine synthetases. Based on these data, I suggest that GSII evolved as the result of a eucaryote to procaryote gene transfer event. This conclusion is supported by the observation that the GSII isozyme is found in a single family of bacteria, the Rhizobiaceae. Because bacteria of this family are characterized by their ability to form cortical hypertrophies on plants, I suggest that a plant provided the progenitor glnII gene. Little is known of the role of horizontal gene transfer in evolution. I know of no selective pressure for the maintenance of a plant GS gene in plant pathogenic bacteria. GSII is not known to serve any special role in the develop- ment of rhizobial symbioses. It will now be possible to construct a g1g11' rhizobium using site directed gene replacement mutagenesis. It will be of interest to analyze the symbiotic phenotype of a glnII mutant rhizobium. Glutamine auxotrophs have been utilized in the study of rhizobial physiology. Because there are two GS genes in rhizobia, glutamine auxotrophs generated by traditional mutagenesis techniques will generally carry mutations in regulatory genes, resulting in com- plex, pleiotropic phenotypes. I have shown that GSII is transcription- ally regulated to provide high levels of total GS activity during nitrogen limited growth. I predict that a gig11' rhizobia would not be a glutamine auxotroph (due to the presence of a functional gigg gene) and would not grow well on poor nitrogen sources (due to the lack of a transcriptionally inducible GS). 78 The site of transcriptional initiation upstream of the gig11 coding region was mapped by $1 nuclease protection analysis. Upstream of the RNA start site is a region of homology to B. japonicum git promoters. This conserved region is the same sequence responsible for 660 dependent RNA polymerase binding in K. pneumoniae Ntr promoters. Furthermore, the gig11 promoter has no significant homology to the glnA promoter. Based on these observations, 1 suggest that gig11 and git genes of B. jgpgg- iggg require an alternate sigma factor for transcriptional initiation, analogous to the Ntr system of enteric bacteria. This model, however, does not account for the differential expression of gig11 and git genes. Presumably, sequences other than the RNA polymerase binding sites are involved in gig11 and git regulation. In K. pneumoniae, upstream activation sequences and the protein NRI are involved in Ntr promoter regulation. A similar system may function in B. japonicum. The gig11 promoter will serve as a useful tool in the characterization of possible NRI-like proteins. Successive deletions upstream of gig11 can be used to identify regions necessary for promoter activity. Promoter activity can be monitored ig vivo GS assays or ig vitro with a reconsti— tuted RNA polymerase transcription system. Similar experiments have identified possible upstream activation sequences in B. japonicum git promoters (Alvarez-Morales, _t _1., 1986). I have shown that, in free-living aerobic cultures, the gig11 promoter is activated by either nitrogen starvation or nitrogen limited growth. Adams and Chelm (personal communication) have identified a possible gttB-like gene which is required for induction of gig11 in nitrogen limited cultures. Although these data are still preliminary, it appears that the B. japonicum "ntrC" gene is not required for git 79 gene activation as it is in B. pneumoniae. Therefore, nif regulation is not linked to general nitrogen regulation. It is not clear what factors affect the expression of the gig11 promoter in microaerobic cultures and bacteroids. Rao, gt ._1. (1978) have shown that GSII activity (and presumably gig11 transcription) in a nitrogen limited culture decreases dramatically when the 02 concentration drops below 0.40%. In contrast, Adams and Chelm (1986) have demonstrated significant gig11 transcription in nitrogen rich microaerobic cultures. The microaerobic induction of glnII is dependent on another regulatory gene, och, which is required for bacteroid development and git activation. These data can be ex- plained as follows. glnII transcription in nitrogen limited aerobic cultures is "ntrC" dependent. When the 02 concentration drops below a threshold level, the cells become oxygen (or energy) rather than nitro- gen limited, and "gtgt" dependent gig11 transcription stops. These same conditions (but not "gtgt") cause the induction of ngB dependent promoters, including git genes. Because of the homology between the och and "ntrC" gene products, och stimulates transcription of glnII at a lowered level. Thus, gig11 regulation is "normal," responding primar- ily to nitrogen metabolism. Alternately, it is possible that micro- aerobic cultures are still nitrogen limited and that the gig11 promoter is inactivated by a microaerobic-specific regulatory system, possibly mediated by och. It would be of interest to characterize the micro— aerobic response of gig11 in pggB- mutants grown on a poor nitrogen source. The preliminary results of Adams and Chelm, which suggest that git transcription is not activated by general nitrogen regulation, are highly significant since this model implies that the regulation of git, and not glnII, has been modified so that B. japonicum can export fixed 80 nitrogen to its host plant. However, the induction of git gene tran— scription gt piggtg does not necessarily result in the assembly of a functional nitrogenase enzyme. Further investigations are needed to determine what additional regulatory factors are involved in the devel— opment of the rhizobial—legume symbiosis. LIST OF REFERENCES LIST OF REFERENCES Adams, T. H. & Chelm, B. K. (1984) B. M91. Appl. Genet. 2, 392-405 Adams, T. H. & Chelm, B. K. (1986) Manuscript in Preparation Adams, T. H., McClung, C. R. & Chelm, B. K. (1984) B. Bacteriol. 159, 857—862 Akkermans, A. D. L. (Stewart, W. D. P. London & Roelofsen, W. (1980) in Nitrogen Fixation & Gallon, J. R., eds.) pp. 279—299, Academic Press, Alvarez-Morales, A., Betancourt—Alvarez, M, Kaluza, K. & Hennecke, H. (1986) Nucleic Acids Egg. 14, 4207-4227 Alvarez-Morales, A., Dixon, R. & Merrick, M. (1984) E BO B. 3, 501-507 Alvarez—Morales, A. & Hennecke, H. (1985) M_1. ng. Genet. 199, 306-314 Ames, G. F.—L. & Nikaido, K. (1985) E BO B. 4, 539—547 Ausubel, F. M. (1984) Cell 37, 5—6 Backman, K., Chen, Y —M. & Magasanik, B. (1981) Proc. Natl. Acad. Bgi. B. B. A. 78, 3743—3747 21> Bannister, J. V. & Parker, M. W. (1985) Proc. Natl. B. B. A. 82, 149-152 cad. Bgi. Bender, R. A., Janssen, K. A., Resnick, A. K., Blumenberg, M., Foor, F. & Magasanik, B. (1977) B. Bacteriol. 129, 1001-1009 Benton, W. D. & Davis, R. W. (1977) Science 196, 180—182 Benyon, J., Cannon, M., Buchanan-Wollaston, V. & Cannon, F. (1983) Cell 34, 665-671 Bergersen, F. J. (1961) Aust. B. Biol. Sci. 14, 349—360 Bergersen, F. J. (1980) in Nitrogen Fixation (Stewart, W. D. P. & Gallon, J. R., eds.) pp. 139-160, Academic Press, London 82 Berk, A. J. & Sharp, P. A. (1977) Cell 12, 721-732 Bolivar, F., Rodriguez, R. L., Greene, P. J. Betlach, M. C. Heynecker, H. L. & Boyer, H. W. (1977) Gene 2, 29-113 Brown, C. M. & Dilworth, M. J. (1975) Q. ng. Microbiol. 86, 39-48 Buck, M., Miller, S., Drummond, M. & Dixon, R. (1986) Nature 320, 374-378 Bueno, R., Pahel, G. & Magasanik, B. (1985) 9. Bacteriol. 164, 816-822 Buikema, W. J., Szeto, W. W., Lemley P. V. Orme—Johnson, W. H. & Ausubel, F. M. (1985) Nucleic Acids ng. 13, 4539—4555 Carlson, T. A. & Chelm, B. K. (1986) Nature in press Carlson, T. A, Guerinot, M. L. & Chelm, B. K. (1983) in Plant Molecular Biology (Goldberg, R. B., ed.) pp. 291—302, Alan R. Liss, New York Carlson, T. A, Guerinot, M. L. & Chelm, B. K. (1985) 9. Bacteriol. 162, 698-703 Chelm, B. K., Carlson, T. A. & Adams, T. H. (1985) in Nitrogen Fixation Research Progress (Evans, H. J., Bottomley, P. J. & Newton, W. E., eds.) p. 217, Martinus Nijhoff Publishers, Dordrecht Ching, T. M., Hedtke, S. & Newcomb, W. (1977) Plant Physiol. 60, 771-774 Clewell, D. B. & Helinski, D. R. (1972) B. Bacteriol. 110, 1135—1146 Coruzzi, G., Tingey, S. & Walker, B. (1986) Personal Communication Covarrubias, A. A. & Bastarrachea, F. (1983) M91. ng. Genet. 190, 171—175 Cullimore, J. V., Gebhardt, C., Saarelainen, R., Miflin, B. J., Idler, K. B. & Barker, R. F. (1984) 9. M91. Appl. Genet. 2, 589—599 Cullimore, J. V., Laura, M., Lea, P. J. & Miflin, B. J. (1983) Planta 157, 245—253 Cullimore, J. V. & Miflin, B. J. (1984) 9. B19. B9t. 35, 581—587 Darrow, R. A. (1980) in Glutamine Synthetase: Metabolism, Enzymology gg9 Regulation (Mora, J. & Palacios, R., eds.) pp. 139—166, Academic Press, Inc., New York 83 Darrow, R. A., Crist, D., Evans, W. R., Jones, B. L., Keister, D. L. & Knotts, R. R. (1981) in Current Perspectives ig Nitroqen Fixation (Gibson, A. H. & Newton, W. E., eds.) pp. 182—185, Australian Academy of Science, Canberra Darrow, R. A. & Knotts, R. R. (1977) Biochem. Biophys. Egg. Commun. 78, 554-559 Davis, R. W., Botstein, D. & Roth, J. R. (1980) Advanced Bacterial Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Davis, R. W. & Thomas, M. (1974) 9. M91. Biol. 91, 315-328 Deininger, P. L. (1983) Anal. Biochem. 129, 216—223 Denhardt, D. T. (1966) Biochem. Biophys. Egg. Commun. 23, 641-646 DeVries, G. E., Oosterwijk, E. & Kinjne, J. W. (1983) Plant Bgi. Letts. 32, 333-341 Dixon, R. (1984a) J. ng. Microbiol. 130, 2745-2755 Dixon, R. (1984b) Nucleic Acids Egg. 12, 7811—7830 Donald, R. G. K. & Ludwig, R. A. (1984) 9. Bacteriol. 158, 1144—1151 Drummond, M., Whitty, P. & Wootton, J. (1986) MBO 9. 5, 441—447 Dudman, W. F. (1977) in Surface Carbohydrates 9t th Prokaryotic Cell (Sutherland, 1., ed.) pp. 357—414, Academic Press, London Dunlap, P. V. & Steinman, H. M. (1986) B. Bacteriol. 165, 393-398 Evans, W. R. & Crist, D. K. (1984) Arch. Microbiol. 138, 26-30 Fischer, H.—M., Alvarez-Morales, A. & Hennecke, H. (1986) E BO 9. 5, 1165-1173 _a. o Fisher, R., Tuli R. & Haselkorn, R. (1981) Proc. Natl. Acad. _g_ B. B. A. 78, 3393-3397 Friedman, A. M., Long, S. R., Brown, S. E., Buikema, W. J. & Ausubel, F. M. (1982) Gene 18, 289-296 Fuchs, R. L. & Keister, D. L. (1980) B. Bacteriol. 144, 641—648 Fuhrmann, M. & Hennecke, H. (1984) 9. Bacteriol. 158, 1005—1011 Fuhrmann, M. Fischer, H.-M. & Hennecke, H. (1985) M91. ng. Genet. 199, 315-322 Gebhardt, C., Oliver, J., Forde, B. G., Saarelainen, R. & Miflin, B. (1986) EMBO J 5, 1429—1435 84 Grunstein, M. & Hogness, D. (1975) Proc. Natl. Acad. Sci. 9. B. A. 72, 3961-3965 Harder, W. & Dijkhuizen L. (1983) Ann. Rev. Microbiol. 37, 1-23 Haselkorn, R. (1978) Ann. Rev. Plant Physiol. 29, 319-344 Hawkes, T., Merrick, M. & Dixon, R. (1985) Mol. Gen. Genet. 201, 492-498 Hawley, D. K. & McClure, W. R. (1983) Nucleic Acids Res. 11, 2237-2255 Hirschman, J., Wong, P.—K., Sei, K., Keener, J. & Kustu, S. (1985) Proc. Natl. Acad. Sci. 9. B. A. 82, 7525-7529 Hohn, B. (1979) Methods Enzymol. 68, 299-309 Howitt, S. M. & Gresshoff, P. M. (1985) 9. Gen. Microbiol. 131, 1433-1440 Hunt, 2. P. & Magasanik, B. (1985) Proc. Natl. Acad. Sci. 9. B. E. 82, 8453-8457 Jordan, D. C. (1984) in Berqey’s Manual 9t Systematic Bacteriology (Drieg, N. R., ed.) Vol. 1, pp. 234—242, Williams & Wilkins, Baltimore Kahn, M. L., Kraus, J. & Somerville, J. E. (1985) in Nitrogen Fixation Research Progress (Evans, H. J., Bottomley, P. J. & Newton, W. E., eds.) pp. 193-199, Martinus Nifhoff Publishers, Dordrecht Kondorosi, A., Svab, Z. Kiss, G. B. & Dixon, R. A. (1977) M91. Gen. Genet. 151, 221-226 Laemmli, U. K. (1970) Nature 227, 680-685 Lara, M., Cullimore, J. V., Lea, P. J., Miflin, B. J., Johnston, A. W. B. & Lamb, J. W. (1983) Planta 157, 254-258 Lipman, D. J. & Pearson, W. R. (1985) Science 227, 1435—1441 Lowry, O. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) 9. Biol. Chem. 193, 265-275 Ludwig, R. A. & Signer, E. R. (1977) Nature 267, 245-248 Ludwig, R. A. (1978) 9. Bacteriol. 135, 114—123 Ludwig, R. A. (1980a) Proc. Natl. Acad. Sci. 9. B. E. 77, 5817-5821 Ludwig, R. A. (1980b) 9. Bacteriol. 141, 1209-1216 Magasanik, B. (1982) Annu. Egy. Genet. 16, 135-168 85 Manian, S. S. & O’Gara, F. (1982) Arch. Microbiol. 131, 51—54 Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Markwell, M. A. K., Haas, S. M., Bieber, L. L. & Tolbert, N. E. (1978) Anal. Biochem. 87, 206-210 Marmur, J. & Doty, P. (1962) 9. M91. Biol. 5, 109-118 Maxam, A. M. & Gilbert, W. (1980) Methods Enzymol. 65, 499-580 McParland, R. H., Guevara, J. G. Becker, R. R. & Evans, H. J. (1976) Biochem. 9. 153, 597-606 Merrick, M. J. & Gibbins, J. R. (1985) Nuclic Acids Egg. 13, 7607—7619 Miller, J. H. (1972) Experiments ig Molecular Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Morett, E., Moreno, S. & Epsin, G. (1985) M91. Egg. Genet. 200, 229-234 Mortenson, L. E. & Thorneley, R. N. F. (1979) Egg. Egy. Biochem. 48, 387-418 Norrander, J., Kemp, T. & Messing, J. (1983) Gene 26, 101—106 O’Gara, F. & Shanmugam, K. T. (1976) Biochem. Biophys. Acta 437, 313—321 Ow, D. W. & Ausubel, F. M. (1983) Nature 301, 307-313 Ow, D. W., Sundaresan, V., Rothstein, D. M., Brown, S. E. & Ausubel, F. M. (1983) Proc. Natl. Acad. Bgi. B. B. A. 80, 2524-2528 Patel, J. J. & Gerson, T. A. (1974) Arch. Microbiol. 101, 211-220 Pustell, J. & Kafatos, F. C. (1984) Nucleic Acids Egg. 12, 643-655 Rao, V. R., Darrow, R. A. & Keister, D. L. (1978) Biochem. Biophys. Egg. Commun. 81, 224-231 > o n) o. m n k: m |:1> I Reitzer, L. J. & Magasanik, B. (1983) Proc. Natl. 80, 5554-5558 Reitzer, L. J. & Magasanik, B. (1985) Proc. Natl. 82, 1979-1983 > n D) o. m o _1 IC m |> (D _I Reitzer, L. J. & Magasanik, B. (1986) C l 45, 785—792 Roberts, G. P. & Brill, W. J. (1981) Egg. Egy. Microbiol. 35, 207—235 86 Robertson, J. G., Farnden, K. J. F., Warbuton, M. P. & Banks, J.-A. M. (1975) Aust. 9. Plant Physiol. 2, 265—272 Ronson, C. W., Lyttleton, P. & Robertson, J. G. (1981) Proc. Natl. Acad. B91. 9. B. E. 78, 4284-4288 Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. B91. 9. B. E. 74, 5463-5467 Shapiro, B. M. & Stadtman, E. R. (1970) Methods Enzymol 17A, 910-922 Somerville, J. E. & Kahn, M. L. (1983) 9. Bacteriol. 156, 168-176 Southern, E. M. (1975) 9. M91. Biol. 98, 503—517 Stadtman, E. R. & Ginsburg, A. (1974) in 1gg Enzymes (Boyer, P. 0., ed.) Vol. 10, pp. 755—807, Academic Press, New York Thomas, J., Meeks, J. C., Wolk, C. P., Shaffer, P. W., Austin, S. M. & Chien, W.—S. (1977) 9. Bacteriol. 129, 1545-1555 Tischer, E., DasSarma, S. & Goodman, H. M. (1986) E 80 J. in press Tronick, S. R., Ciardi, J. E. & Stadtman, E. R. (1973) 145, 858-868 IL; Bacteriol. Tumer, N. E., Robinson, S. J. & Haselkorn R. (1983) Nature 306, 337—342 Tyler, B. (1978) Egg. Egy. Biochem. 47. 1127—1162 Upchurch, R. G. & Elkan, H. G. (1978) 9. Egg. Microbiol. 104, 219—225 Vairinhos, F., Bhandari, B. & Nicholas, D. J. D. (1983) Planta 159, 207—215 Vieira, J. & Messing, J. (1982) Gene 19, 259—268 Whitehead, A. S., Goldberger, G., Woods, D. E., Markham, E. F. & Colton, H. R. (1982) Proc. Natl. Acad. Sci. 9. B. E. 80, 5387-5391 hier, J., Lasky, L. A. & Lawn, R. M. (1985) Proc. 9. B. E. 82, 1585—1588 Yanisch-Perron, C., Vieira, J. & Messing, J. (1985) Gene 33, 103-119 II IIIIIIIIIIIIIIIIIIIIIIIII III 111121111311111111111171111113111511