The nifH promoter region of Rhizobium leguminosarum: nucleotide sequence and promoter elements controlling activation by NifA protein

The nifH promoter region of Rhizobium leguminosarum: nucleotide sequence and promoter elements controlling activation by NifA protein

Gone, 87 (1990) 31-36 Elsevier 31 GENE 03427 The nifH promoter region of Rhizobium leguminosmm: nueleotide sequence and promoter elemeuts controll...

695KB Sizes 0 Downloads 2 Views

Gone, 87 (1990) 31-36



GENE 03427

The nifH promoter region of Rhizobium leguminosmm: nueleotide sequence and promoter elemeuts controlling activation by NifA protein (Recombinant D NA; consensus sequence; multicopy inhibition; pseudo promoter; niuogen fixation; pea)

Peter W. Roelvink*, Miehiel Harmsen, Ab van Kammen and Rommert C. van den Bos

Depanm¢;~lof MolecularBiology, WageningenAgriadmral Uniwrs~y, Wageningen(The Netherland.v) Received by H.M. Krisch: 15 April 1989 Revised: 18 July 1989 Accepted: 21 July 1989


The nucleotide (nt) sequence of the Rhizobium leguminosarum nifH promoter region contains a consensus promoter, a consensus upstream activator sequence (UAS), a pseudo (~) promoter and a ql UAS. We mapped the transcription start point for the consensus promoter sequence by primer extension. This promoter differs from the consensus in one of the four supposedly invariant nt and can be activated by the Klebsiella pneumoniae nifA product in Escherichia coll. Under these conditions the ~ promoter and ~ UAS do not function. A low-copy-number plasmid construct containing the q~UAS as well as the consensus UAS delayed the onset of symbiotic nitrogen fixation in nodules induced on Pisum sadmm. Studies of high-copy-number n/.fH promoter constructs showed that partial deletion of the consensus UAS does not alter the ability to inhibit nitrogen fixation by titration of NifA suggesting that NifA can also complex with RNA polymerase containing the alternative c.factor RpoN.


The soil bacterium R. lefuminosarum induces root nodules on the pea (P. satimm L.). During endosymbiosis Correspondence to: Dr. R.C. van den Bos, Department of Molecular

Biology, Wageningen Agricultural University, Dreijenlaan 3, 6703 HA Wageningen (The Netherlands) Tel. (08370)-83263; Fax (08370)-84801. * Present address: Department of Virology, Wageningen Agricultural University, P.O. Box 8045, 6700 EM Wageningen (The Netherlands) Tel. (08370)-83095; Fax (08370)-84254. Abbreviations: Ap, ampiciHin; ~3al, p-galactosidase; bp, base pai~s); Cm, chloramphenicol; hen, high copy number; IPTG, isopropyllY-e-thiogalactopyranoside;Km, kanamycin; lcn, low copy number; n~ genes encoding Nif proteins; Nif, nitrogen fixation; nt, nucleotide(s); ORF, open reading frame; P., Pisum; PA, polyacrylamide; PE, primer extension; Poilk, Klenow (large) fragment of£. coil DNA polymerase I; ~, pseudo; R, resistant; R., Rhizoblum; RBS, ribosome.binding site; Rif, rifampicin; RNAP, RNA polymerase; RpoN, sigma factor specific for n0e genes; Sm, streptomycin; Tc, tetracycline; up, transcription start point(s); UAS, upstream activator sequence; wt, wild type; XGal, 5.bromo-4-chloro-3-indolyi-p-v-galactopyranoside; ::, novel joint (fusion). 0378-1! 19/90/503.50O 1990ElsevierSciencePublishersB.V.(BiomedicalDivision)

the bacteroid, a specialized form of the bacterium, reduces nitrogen to ammonia, which is excreted and made available to the host plant. Two specialized groups of genes, the n0e and the fix genes, are involved in nitrogen fixation. The n~f genes have a homologue in the free living nitrogen-luring K. pneumoniae; .fix genes do not. Most of the n~f and ]ix genes are activated by the regulatory nifA gone (Gussin etal., 1986) acting in concert with RNAP and RpoN, encoded by the rpoN gene (Ronson et al., 1987). The promoters ofn/f and ]ix genes have a characteristic consensus sequence CTGGYRYRN4TrGCA (Y, pyrimidine; R, pm'ine; Ausubel, 1984; Gussin etal., 1986). An UAS with the consensus sequence TGT-Nm-ACA is present in most cases. The product (NifA) of the nifA gone has a DNAbinding domain at its C terminus and, as suggested by Buck et al. (1986; 1987a), binds to the UAS (Morett et al., 1988). Then the intervening DNA loops out and NifA makes contact with the RpoN-RNAP complex leading to transcription activation from the n/.fand ]ix promoters. The aim of the present study was to determine the sequence of the

32 ,/fH promoter region ofR. ieguminosarum and to study the mechanism of transcription activation from this promoter. The functionality of the different regulatory elements of the n/fH promoter region was determined by measuring expression from this promoter in different lcn and hcn plasmid constructs in both E. coli and R. leguminosarum.


(a) Cloning and sequencing of the Rhizobium leguminosaturn nifH promoter region To determine the sequence of the R. leguminosarum strain PRE n/fH promoter region, a restriction map was made of pGB5 (Table I), a recombinant plasmid that contains the ,/fH and n/fD genes (Schetgens et ed., 1984). Preliminary sequencing of n/ft-/indicated that the 5'-termined coding region contains an Accl site. Therefore, the sequence of a 524-bp Hpal-Accl fragment, probably




containing the nifH promoter region, was determined (Fig. 1). A purine-rich stretch, with an RBS at nt + 56 through + 62 and an AUG codon at nt + 70 form a probable translation initiation site for the nOTl ORF. The sequence at nt -28 through -12 differs in 3 nt from the consensus CTGGYRYRN4TTGCA (Ausubel, 1984; Gussin et ed., 1986) for RpoN-dependent promoters. One difference, at -13, concerns an nt (C in the consensus) supposed to be invariant in such promoters. A primerextension experiment (Fig. 2) indicated the G residue at + 1 as the major tsp, confnming the functionality of the identified promoter. At nt -127 through -112, a sequence consistent with the consensus ,_"or a UAS (TGT-NIo-ACA, Buck et al., 1986) is present. At nt -206 through -191 another sequence resembling an RpoN-dependent ,/f promoter is present, but the spacer between the two promoter elements (in the consensus the two promoter elements are CTGG and TTGCA, the spacer between them being 8 bp) is 1 nt shorter than required. Buck (1986) showed that the K. pneumoniae niftt promoter with a 7- instead of 8-nt spacer is not functioned. For this reason we consider this


(3 A
















÷5 5


+ 115






+ 161

Fig. 1. Nucleotide sequence of a 524-bp HpaI.Accl fragment from pGB5 containing the nifH promoter region ofR. leguminosarum PRE. UAS (nt -127/-112) and promoter sequences (nt -28/-12) are indicated by the consensus below the respective sequences. A 0 promoter (nt -206/- 191) and a 0 UAS (nt -326/-310) are underlined. The asterisks mark the first nt of the remaining sequence in the deletion mutants (A clones) generated by BAL 31 digestion. The tsp is indicated by an upward arrow. Recombinant DNA technologies were as described (Maniatis et al., 1982). Growth media for E. coli and R. leguminosarum were as described before (Schetgens et al., 1984). Antibiotics were used in the following concentrations (in pg/ml): Ap 50 (200 for pTZ derivatives), Cm 25, Km 12.5 and Tc 10. All enzymes were purchased from Boehringer (Mannheim, F.R.G.) and used according to the manufacturer's instructions. The nt sequence was determined by dideoxy-sequencing fragments cloned in M 13mpl8 (Sanger et al., 1977). Deletion endpoints were determined by sequencing fragments cloned in pTZI8R and pTZI9R vectors. Sequence data were analyzed on a VAX computer using Dbase and Analyseq programs (Staden, 1980; 1984).

glb Fig. 2. Autoradiogram of a nil//' primer extension experiment. A heptadecamer complementary to the sequence spanning the nt 73 through 57 (see Fig. 1) was extended using as a template RNA isolated from bacteroids of 21-day-old nodules. Bacteroids were isolated from root nodules as described by Katinakis et al. (1988). Isolation of total RNA from bacteroids was as described (Krol et al., 1980). The reaction products were analyzed on a 670 PA gel next to the sequence ladder of a nil// promoter clone sequenced using the same heptadecamer as a primer. The sequence shows the nt -8 through + 18. The major extension product comigrates with the nt complementary to the G-residue marked tsp ( + 1) in Fig. 1. Primer extension (P.E.) experiments were performed as described (Th0ny et al., 1987) with one modification: actinomycin D was added to a final concentration of 30 pg/ml to inhibit DNA-dependent DNA polymerase activity of the reverse transcriptase.

33 scriptionally fused to lacZ (Fig. 3). This resulted in fusions with deletion endpoints as indicated in Figs. I and 4. The nifH: :lacZ fusions in the Icn vector pMP220 were transformed into E. coil KMBL 1164 containing the recombinant plasmid pWKI31 (POlder et al., 1983) that constitutively expresses K. pneumoniae nifA and/~Gal activity due to NffA activation of the n/fH promoter was measured (Fig. 4). Deletion of the ~ UAS and the ~ promoter (pMPdl.7L through pMPd3.2L) does not decrease the/~Gal activity indicating that these elements do not contribute to activation of the promoter. Deletions including the first nt of the UAS as in pMPd3.7L, or 6 nt, as in pMPd3.6L, abolish activation by K. pneumoniae NifA. When the constructs were tested in E. coli TH 1 that has a deletion of the ntrA gene encoding a ¢ factor that also specifically recognises n/f and fix promoters, no activation of the R. leguminosarum n/fH promoter was observed (results not shown). This indicates that the/~Gal activity measured is dependent upon the

sequence as a ~ promoter. Primer extension experiments did not reveal any transcription initiation near this ~ promoter (not shown). Upstream from this ~ promoter a second UAS-like sequence is present between nt positions -326 and -310, differing from the consensus (Buck et al., 1986) in that the spacing is 11 instead of 10 nt. We therefore consider this sequence as a ~ UAS.

(b) The effect of deletions in the nifH promoter upstream region on activation by Klebsiella pneumoniae NIfA in Escherichia coU To determine the possible functions of the n/fH regulatory elements, BAL 31 nuclease deletions were generated in the region upstream from the n/fH promoter (Fig. 3). The 524-bp Hpal-Accl fragment from pGB5 was cloned in pTZI9R digested with Sinai + Accl resulting in pMHI. This plasmid was used to generate deletions by BAL 31 digestion. The fragments carrying deletions were tranRI








u P.

,xi Rx i



~////////////////j.,,,,,/-- P

52&bp I





Pst[ ligate in pTZIOR (Pstl÷Smal) R! Sin*

., x, K, .x,


nifH :: l a c Z fusions

(high copy humbert.15) A]

: Ace[


= Hind TIT

HI : HpaI K P

: Kpnl : Pst!




BIT : ggl Tr



nifH :: l a c Z fusions (low copy numbera:§)

p : promoter RZ : EcoR| S|

= Salt

Sm : Sinai Sm~-- Sinai-blunt ligation X

• Xba!

P'/JTJ,L vector DNA (pTZ19R) Al~: .... (pTZIBR) deletion upstream of promoter

Fig. 3. Cloning strategy of DNA fragments containing various deletions in the region upstream of the R. leguminosarum n~fll promoter. Recombinant plasmid pMH I was constructed by cloning a 524-bp Hpal-Accl fragment from pGB5 into pTZ 19R digested with Smal+ Accl. This results in destruction of the Smal site (Sm*). To generate deletions, pMHI was digested with EcoRl and the linearized DNA was subjected to BAL 31 digestion for I, 2 or 3 min. ARer inactivation of the BAL 3 I, the DNA was digested with Pstl and separated on gel. The resulting diffuse band o~'fragments containing various deletions was excised from the gel and the isolated DNA was recloned in pTZI8R digested with Smal + Pstl. The various deletion endpoints were determined by sequencing and are indicated in Fig. I. Transcriptional n/fH::lacZ fusions were constructed by cloning Ec~RI-Pstl fragments from the pTZ 18 plasmids containing the deletion fragments (A-clones) into the Icn vector pMP220, or by cloning EcoRI-Hindlll fragments into th~ ~cn vet',for pMPI90 (Spaink et al., 1986).


pMH l~constructs -360



.L-I '









*1 I...I ' p




nit H

plumids pMP220

- 3&3 t - 320

i t

- 296


- 278






- 127



I t__a.



pMPd 1.6 L

67 1277



1.7 L
























3.6 L


20 bp

Fig. 4. Activation of the R. leguminosarum ni./H promoter in E. coli by K. pneumoniae NifA. The n/fH promoter region is represented graphically. The positions of the q/UAS, ~ promoter, UAS, promoter, tsp ( + 1), RBS and start codon ATG are indicated. The endpoints ofthe BAL 3 l-generated deletions are indicated (see also Fig. 1) followed by the corresponding Icn plasmid codes and the assayed ~Gal activities (in Miller units; Miller, 1972).~Gal assays were routinely performed in triplicate on bacteria grown overnight in nitrogen-free derepression medium at 30°C (Cannon, 1980).

presence of a nif and :/ix specific ¢ factor. These experi~nents indicate that the ~ UA3 and the ~ promoter are not functional in n/fH activation. On the other hand, the UAS at nt -127/-112 and the promoter at nt -28/-12 are functional elements.

(c) Functional analysis of the nifH promoter region in Rhizobium leguminosarum The experiments described above show that the heterologous K. pneumoniae NifA can activate the n07/promoter of R. leguminosarumin E. coilcells and that a UAS is essential for promoter activity. In ~iazotrophs like RMzobium and K. pneum,,niae the presence of multiple copies of plasmids carrying the UAS will result in capturing of NifA. This limits the amount of activator protein available for activation of the nil and fix genes, which may result in a decrease of nitrogen fixation. We used this phenomenon known as multicopy inhibition (see, e.g., Morett et al., 1988) as a means to evaluate the functionality of the different R. leguminosarum niflt promoter elements. Both the different lcn and hen transcriptional niflt::lacZ fusions were conjugated into R. leguminosarumstrain PI3, a Rif R mutant of R. leguminosarum PRE (Selbitschka and Lotz, 1984), and the resulting strains were inoculated on pea plants. At days 21 and 24 after inoculation acetylene reduction assays were performed (Table II). The presence of the various lcn plasmid constructs in R. ieguminosarum had no influence on nodule morphology. The nodules were reddish and appeared on the main and lateral roots in the same numbers as after inoculation with a wt strain harbouring the vector pMP220. Most of these constructs, containing the nifH promoter region, do not appear to influence the level of acetylene reduction when compared to the control strain

containing the pMP220 vector. However, the construct pMPd 1.6L, which has both the 0 UAS and the consensus UA$, caused a fivefold decrease of acetylene reduction activity at 21 days. This decrease did not occur at 24 days after inoculation in three separate experiments. This reduction is most likely explained by the cumulative capturing of NifA by the 0 and the consensus UA5 present in this "construct, which is only apparent during the onset of nitrogen fixation when the amount of NifA is limited. Capture of NifA by the UA$ then limits the amount available for the activation of nil and fix genes, the expression of which is thus delayed. The suggestion that a ~ UAS can still bind NifA is in agreement with results of Buck et al. (1987b), who showed that a multicopy plasmid carrying the K. pneumoniae niftt promoter region with a consensus UAS inhibits acetylene reduction to 0.4% of the wt level; a construct with a ~ UAS having an l l-bp (instead of 10-bp) spacer limits the decrease to 30%. R. leguminosarum strain P13 harbouring the hen vector pMPl90 (copy number 15, H. Spaink, pers. commun.) behaved like the wt. Nodules induced on the main root by R. leguminosarum PI3 strains harbouring the hen plasmids pMPdl.6H, pMPd3.2H, pMPd3.6H or pMPd3.7H were smaller than wt, were greenish white in appearance and had little or no leghaemoglobin. In fact they resembled nodules induced by R. leguminosarum nifA :: Tn5 mutants (Schetgens et ai., 1985) and senesced early. The presence of approx. 15 or more plasmid-borne copies ofthe n/fH promoter region and the UAS in the pMP190 derived constructs lead to a complete absence of nitrogen fixation. Interestingly, hen constructs containing the nifH promoter and a partially deleted UAS, as in pMPd3.6H and pMPd3.TH, still cause multicopy inhibition. This is in sharp contrast with results

35 TABLE I Bacterial strains and vectors Strains or plasmids

Genotype or phenotype

Reference or source

wt strain

Schetgens et al. (1984) Selbitschka and Lotz (1984)


R. leguminosarum PRE P13 E. coil KMBL1164 DHSoeF' THI

Plasmids pGB$ pTZIgR/pTZI9R pMP220 pMPl90 pRK2013 pWKI31 pMHI pMHld constructs pMPdl.6L constructs pMPdl.6H constructs

RifR derivative of PRE ~l(lac-proAB ) Host for pTZIgR and pTZ 19R cloning and sequencing vectors A/acU169; ~g/nF

ApR; contains nt[llD from R. leguminosarum stria PRE Apn; cloning and sequencing vectors Ten; IncPI; lcn promoter probe vector CmR, StaR; IncQ; hcn promoter probe vector KmR; contains tra-genes for mobilisation during conjugation Cmt~; constitutively expresses K. pneumoniae nU'A from cat-promoter in E. coli Ap'X; 524-bp Hpal-Accl nifiFl promoter fragment from R. leguminosarum PRE in pTZI9R ApR; series of n/fl4 promoter-region deletions in pTZI8R TcR; pMHld constructs fused to lacZ in pMP220 (Ion) Cma; pMHld constructs fused to lacZ in pMPI90 (hen)

TABLE I1 Relative acetylene reducing capacity of P. satimm inoculated with R. leguminosarum P13 carrying various niftl::lacZ fusions Ion plasmids a

Relative acetylene reducing capacity (units/mg nodules/h) u

(see section e)

day 21

day 24

pMP220 pMPdl.6L pMPdl.TL pMPd2.6L pMPd3.2L

100 21 97 105 130

100 95 99 122 139

100 3 3 3 3

100 1 3 ND ND

hen plasmids pMPl90 pMPdl.6H pMPd3.2H pMPd3.6H pMPd3.TH

a See Table I and Fig. 1. b Acetylene reduction assays were performed in triplicate by incubating three whole plants in 10~o(v/v) acetylene for I h; ethylene concentrations were measured by gas chromatography. At~er the assays nodules were picked and weighed. Activities are given in units (-- nmol of ethylene produced/mg nodule/h) as average % of wt controls (vectors pMP220 or pMPl90, for lcn and hcn constructs, respectively). ND, not done.

Van de PuUe (Leiden) Bethesda Research Laboratories (Gaithersburg, MD) Ow and Ausubel (1983)

Schetgens et al. (1984) Pharmacia (Uppsala) Spaink et el. (1986) Spalnk et 8.1.(1986) Figurski and Helinski (1979) POlder et al. (1983) This work This work This work This work

obtained for K. pneumoniae nifH, where deletion of the UAS results in a complete relief of multicopy inhibition (Buck et al., 1986). As it seems unlikely that the partly deleted UASs ofpMPd3.6H and pMPd3.TH can still bind NifA, an explanation may be complex formation between RpoN-RNAP and NifA in the absence of a UAS. This suggestion is supported by the finding that an R. meliloti nifH gene driven by a promoter without a UAS after recombination into the symbiotic plasmid can be activated to wt levels in planta (Better et al., 1985). (d) Conclusions The R. leguminosarum PRE nifH promoter region contains a promoter, TI'GGYRYRN4TTGAG, which differs from the consensus, CTGGYRYRN4TTGCA (Ausubel, 1984), in 3 nt. Four nt were suggested to be invariant in the consensus promoter, i.e., GG at -26/-25 and GC at -14/-13 (Gussin et al., 1980. We found on the other hand an A at position -13 for the R. leguminosarum nell promoter while the same was found in both the R. mfolii (Watson and Schofield, 1985) and the R. phaseoli nifH promoters (Quinto et al., 1985). The identified promoter further differs from the consensus at nt positions -12 and -28; alterations at these positions in other strains do not influence promoter activation (Alvarez-Morales and Hennecke, 1985; Gussin et al., 1980. We therefore propose to redefme the consensus for Rhizobium nif and fix pro-


rooters as NYGGYRYRN4TI'GACN. Analysis of n/fH promoter activation in bacteroids by measuring levels of jgGal from the different niftl::lacZ constructs failed; all constructs tested showed only little activity over background (results not shown). This may be due to some constraint at the level of the plasmid-like superco'ding (Gubler and Hennecke, 1988). Attempts to integrate the constructs into the sym-plasmid by homologous recombination failed due to entry exclusion of the chasing plasmid used.


We thank Connie van Ours and Thijs Broos for help with some of these experiments and Pier Madern for artwork. We are indebted to Dr. Bill Broughton for valuable advice on the manuscript.

REFERENCES Alvarez-Morales, A. and Heanecke, H.: Expression of Rhizobiumjaponicum nO~! and n~DK operons can be activated by the Klebsiella pneumoniae n~A protein but not by the product of ntrC. Mol. Gun. Genet. 199 (1985) 306-314. Ausubel, F.M.: Regulation ofnitrogen fixation genes. Cell 37 (1984) 5-6. Better, M., Ditta, G. and Helinski, D.R.: Deletion analysis of Rhizobium meliloti symbiotic promoters. EMBO J. 4 (1985) 2419-2424. Buck, M.: Deletion analysis of the KlebsleUapneumoniae nitrogenase promoter: importance of spacing between conserved sequences around positions =12 and -24 for activation by the n~'A and ntrC (glnO) products. J. Bacterioi. 166 (1986) 545-551. Buck, M., Miller, S., Drummond, M. and Dixon, R.: Upstream activator sequences are present in the promoters of nitrogen fixation genes. Nature 320 (1986) 374-378. Buck, M., Cannon, W. and Woodcock, J.: Transcriptional activation of the Klel~teilapneumonlae nitrogenase promoter may involve DNA loop formation. Mol. Microbiol. I (1987a) 243-249. Buck, M., Cannon, W. and Woodcock, J.: Mutational analysis of upstream sequences required for transcriptional activation of the Klebslellapneumonlae nifH promoter. Nucleic Acids Res. 15 (1987b) 9945-9956. Cannon, F.C.: Genetic studies with diazotrophs. In Bergersen, FJ. (Ed.), Methods for Evaluating Biological Nitrogen Fixation. Wiley, Chichester, 1980, pp. 367-413. Figurski, D.H. and Helinski, D.R.: Replication of an origin containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad. Sci. USA 76 (1979) 1648-1652. Gubler, M. and Hennecke, H.: Regulation ofthe flxA gene and the J'tr.BC operon in Bradyrhizobium japonicum. J. Bacteriol. 170 (1988) 1205-1214. Gussin, G.N., Ronson, C.W. and Ausubel, F.M.: Regulatioa of nitrogen fixation genes. Annu. Rev. Genet. 20 (1986) 567-591. Katinakis, P., Klein-Lankhorst, R., Louwerse, J., Van Kammen, A. and Van den Bos, R.C.: Bacteroid-encoded proteins are secreted into the

peribacteroid space by Rhizobiumleguminosarum. Plant. Mol. Biol. I 1 (1988) 183-190. Krol, AJ.M., Hontelez, J.GJ., Van den Bos, R.C. and Van Kammen, A.: Expression of large plasmids in the endosymbiotic form of Rhizobium leguminosarum. Nucleic Acids Res. 8 (1980) 4337-4347. Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982. Miller,J.: Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1972. Morett, E., Cannon, W. and Buck, M.: The DNA binding domain of the transcriptional activator protein NifA resides in its carboxy terminus, recognises the upstream activator sequences ofn/f promoters and can be separated from the positive control function of NifA. Nucleic Acids Res. 16 (1988) 11469-11488. Ow, D.W. and Ausubel, F.M.: Regulation of nitrogen metabolism by nifA gene product in Klebsiellapneumoniae. Nature 301 (1983) 307-313. P0hler, A., Klipp, W. and Weber, G.: Mapping and regulation of the structural genes n/fK, n/fD and n/fH ofR. melilo~. In POhler, A. (Ed.), Proc. 1st Int. Symp. Mol. Genet. Bacteria-Plant Interaction (Bielefeld). Springer-Verlag, Heidelberg, 1983, pp. 69-78. Quinto, C., de la Vega, H., Flores, M., Leemans, J., Cevallos, M.A., Pardo, M.A., Azpiroz, R., de Lourdes Girard, M., Calva, E. and Palacios, R.: Nitrogenase reductase: a functional multigene family in Rhizobiumphaseoli. Proc. Natl. Acad. Sci. USA 82 (1985) ! 170-1174. Ronson, C.W., Nixon, B.T., Albright, L.M. and Ausubel, F.M.: Rhizobium meliloti ntrA (rpoN) gene is required for diverse metabolic functions. J. Bacteriol. 169 (1987) 2424-2431. Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chainterminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 5463-5467. Schetgens, T.M.P., Bakkeren, G., Van Dun, C., Hontelez, J.G.J., Van den Bos, R.C. and Van Kammen, A.: Molecular cloning and functional characterization of R. leguminosarum structural nOr genes by site directed transposon mutagenesis and expression in £scherichla coil minicells. J. Mol. Appl. Genet. 2 (1984) 406-421. Schetgens, T.M.P., Hontelez, J.G.J., Van den Bos, R.C. and Van Kammen, A.: Identification and phenotypical characterization of a cluster of fix genes, including a nOr regulatory gene from R. leguminosarum PRE. Mol. Gun. Genet. 200 (1985) 368-374. Selbitschka, W. and Lotz, W.: RNA-polymerase and symbiotic properties of spontaneous rifampicin.resistant mutants of Rhizobium leguminosarum PRE and 300. In Ghai, B.S. (Ed.), Symbiotic Nitrogen Fixation. USG Publishers and Distributors, Lodhiana, India, 1984, pp. 9-18. Spaink, H., Okker, RJ.H., Wijfelman, C.A., Pees, E. and Lagtenberg, B.: Promoters in the nodulation region of the Rhizobium leguminosarum gym plasmid pRLIJI. Plant. Mol. Biol. 9 (1986) 27-39. Staden, R.: A new computer method for the storage and manipulation of DNA gel reading data. Nucleic Acids Res. 8 (1980) 3673-3694. Staden, R.: Graphic methods to determine the function of nucleic acid sequences. Nucleic Acids Res. 12 (1984) 521-567. Th0ny, B., Fisher, H.-M., Anthamatten, D., Bruderer, T. and Hennecke, H.: The symbiotic nitrogen fixation regulatory operon (./'vcRn/fA)of Bradyrhizobium japonicum is expressed aerobically and is subject to a novel n/fA-dependent type of activation. Nucleic Acids Res. 15 (1987) 8479-8499. Watson, J.M. and Schofield, P.R.: Species-specific, symbiotic plasmid located repeated DNA sequences in Rhizobium trifolU. Mol. Gun. Genet. 199 (1985) 279-289.