Functional expression of a Rhodospirillum rubrum gene encoding dinitrogenase reductase ADP-ribosyltransferase in enteric bacteria

Gene, 85 (1989) 153-160

153

Elsevier GENE 03309

Functional expression of a Rhodospirillum rubrum gene encoding diiitrogenase ribosyltransferase in enteric bacteria (Recombinant

DNA; nitrogen

fixation;

reductase ADP-

fat promoter)

ADP-ribosylation;

H.-A. Fun, Heidi J. Wirta, R.H. Burris” and G.P. Robertsb Departments of a Biochemist?y and b Bacteriology, and Center for the Study of Nitrogen Fixation, College of Agricultural and Lfe Sciences, Universityof Wisconsin-Madiron, Madkon. WI 53706 (U.S.A.)

Received by C.R. Hutchinson: 30 June 1989 Revised: 26 August 1989 Accepted: 29 August 1989

SUMMARY The function of the cloned draT gene of Rhodospirillum rubrum was studied by placing it under the control of the tat promoter in the vector, pKK223-3. After induction with isopropyl-B-D-thiogalactopyranoside,

dinitrogenase reductase ADP-ribosyltransferase (DRAT) activity was detected in crude extracts of the heterologous hosts Escherichia coli and Klebsiella pneumoniae. In addition, the expression of draT produced a Nif- phenotype in the otherwise wild-type K. pneumoniae strains, the result of the ADP-ribosylation of accumulated dinitrogenase reductase (DR). DR from a nifF_ background was also susceptible to ADP-ribosylation, indicating that the oxidized form of DR will serve as a substrate for DRAT in vivo. A mutation that changes the Arg-101 residue of DR, the ADP-ribose attaching site, eliminates the ADP-ribosylation of DR in vivo, confirming the necessity of this residue for modification.

INTRODUCTION

to the high energy cost of reduction, nitrogen-fixing microorganisms have evolved efficient mechanisms

Biological nitrogen fixation is catalyzed by the nitrogenase complex which consists of two proteins: dinitrogenase (nz$DK product) and DR (nigh product) (see Ludden and Bun-is, 1986, for review). Due

to regulate nitrogenase synthesis and activity. In the purple nonsulfur photosynthetic bacterium R. rubrum, nitrogenase activity is apparently regulated by the covalent modification of DR with an ADP-ribose

Correspondenceto: Dr. G.P. Roberts, Department of BacteriolUniversity of Wisconsin-Madison, Madison, WI 53706 (U.S.A.) Tel. (608)262-3567; Fax (608)262-9865.

gene encoding DRAT; IPTG, isopropyl-8-D-thiogalactopyranoside; kb, 1000 bp; Km, kanamycin; Kp2, dinitrogenase reductase of K. pneumoniae;Mops, 4-morpholinepropanesulfonic acid; nt, nucleotide(s); PAGE, polyacrylamide-gel electrophoresis; RBS, ribosome-binding site; SDS, sodium dodecyl sulfate; lac, hp-lac hybrid promoter; TCA, trichloroacetic acid; wt, wild type; [ 1, denotes plasmid-carrier state.

ogy,

Abbreviations: Ap, ampicillin; atm, atmosphere; bp, base pair(s); Cb, carbenicillin; cpm, counts/min; DR. dinitrogenase reductase; DRAG, DR activating glycohydrolase; draG, gene encoding DRAG; DRAT, DR ADP-ribosyltransferase; draT, 0378-l 119/89/$03.50

0 1989 Elsevier

Science Publishers

B.V. (Biomedical

Division)

154

group (Kanemoto and Ludden, 1984; Pope et al., 1985; Ludden et al., 1988). The modification reaction is carried out by DRAT with NAD as donor (Lowery et al., 1986; Lowery and Ludden, 1988). The ADP-ribose group is attached to one of the two identical subunits of DR at the Arg-101 residue (Pope et al., 1985). This modification can regulate nitrogen fixation, because the ADP-ribosylated DR is unable to donate electrons to dinitrogenase (Lowery et al., 1989). Experimentally, the modified subunit can be separated from the unmodified subunit by SDS-PAGE, because of the difference in their mobility (Kanemoto and Ludden, 1984). The reversal of the modification is effected by DRAG, which removes the ADP-ribose group from the inactivated DR (Ludden and Burris, 1976; Saari et al., 1984; 1986). Substantial evidence has been provided that this ADP-ribosylation system is also operating in other phototrophs, such as Rhodobacter capsulatus (Jouanneau et al., 1983) and Chromatium vinosum (Gotto and Yoch, 1985), as well as in the nonphotosynthetic bacteria, Azospirillum brasilense and A. lipoferum (Ludden et al., 1978; Hartmann et al., 1986; Fu et al., 1989). Whereas the R. rubrum genes coding for DRAT (draT) and DRAG (draG) have been cloned and sequenced, difficulties in transposon mutagenesis of the dru region have delayed the functional characterization of these cloned genes (Fitzmaurice et al., 1989). Here, we report (i) the functional expression of the draT gene of R. rubrum in E. coli and in the diazotroph K. pneumoniae; and (ii) the substrate requirement for the draT gene product in vivo in K. pneumoniae.

RESULTS AND DISCUSSION

(a) Bacterial strains and plasmids

Bacterial strains and plasmids used in this work are listed in Table I. (b) Construction of the druT expression vector The draT expression vector was based on the pt,

expression

vector

pKK223-3

(Pharmacia

Ltd.;

Amann et al., 1983). The vector, pKK223-3, is a derivative of pBR322 and carries the tat promoter followed by the M13mp8 polylinker and a strong transcription terminator from the E. coli rrnB operon. A 1.7-kb SmaI-Hind111 fragment, containing the draT coding sequence and a portion of draG, was isolated from pWPF102 (Fitzmaurice et al., 1989). This SmaI-Hind111 fragment contains the predicted RBS and the start codon, TTG, and was cloned into SmaI-Hind111 sites in the polylinker of pKK223-3, to yield pHAF305 (Fig. 1). Plasmid pHAF305 was subsequently transformed into E. coli strain CAG2041 and K. pneumoniae strain UN5350, both of which carry F’lacI Q. The expression of draT was therefore controlled by repression of pt,, and could be induced by the addition of IPTG. (c) Detection of DRAT activity synthesized in Escherichia coli and Klebsiella pneumoniae

Although parental strains of E. coli and K. pneumoniae had no detectable DRAT activity, the introduction of pHAF305 into CAG2041 and UN5350 caused DRAT activity to appear in these strains (Table II). For the assay, the purified DR of K. pneumoniae was used, because it is an efficient substrate for R. rubrum DRAT in vitro (Lowery et al., 1988). The crude extracts of E. coli strains were prepared from cultures grown in LB medium and induced with IPTG (1 mM) for 4 h. Table II shows that the extracts of E. coli UQ741 catalyzed the incorporation of 32P from [ c+~~P]NAD into the TCA-precipitable fractions in the presence of MgCl, and ADP. The DRAT assay is dependent on the presence of DR. In the absence of added DR, only background labeling was recorded. In these strains, the DRAT activity is also dependent on the presence of pHAF305. Fig. 2 demonstrates that the 32P-labeling in the TCA-precipitable fraction was specifically incorporated into the DR present in the reaction mixture. When proteins in the reaction sample containing the extract of UQ741 (Table II) were separated on SDS-PAGE, only one band was radiolabeled (lane 3). This band corresponded to the position of DR from K. pneumoniue as verified by immunoblotting. Consistent with the filter assay results described above, the data in Fig. 2 show that the 32P-labeling

155 TABLE I Bacterial strains and plasmids Strains a

Genotype b

Source or reference

ara dlac-pro thi supC’” [F’lacJQpro + IacZ: :TnJ]

CAG2041 [pKK223-31 CAG2041[pHAF305]

c. Gross N. Franklin, this laboratory This study

;4001 nifH4384 nip4806 UN729[F’IacZopro+ lacZ::TnS] UN5350[pKK223-31 UN5350[pHAF305] UN[pHAF305] UN1041[pHAF305] UN1837[pHAF305]

MacNeil et al. (1978) MacNeil et al. (1978) MacNeil et al. (1978) MacNeil et al. (1978) N. Franklin, this laboratory N. Franklin, this laboratory This study This study This study This study

E. coli

CAG204 1 UQ477 UQ741 K. pneumoniae

UN UN729 UN1041 UN1837 UN5350 UN5366 UN5371 UN5375 UN5376 UN5374

vectors ptac, rrnBT, T,, ApR, a derivative of pBR322;

pKK223-3 pHAF305 pWPFlO2

draT placed after ptac of pKK223-3 (Fig. 1) draTG, ApR, a derivative of pBR322

a Strains of E. coli were grown with aeration in LB medium (1% tryptone/0.5% yeast extract/l%

Pharmacia, Ltd. This study Fitxmamice et al. (1989) NaCl pH 7.5) at 37°C. Strains of

K. pneumoniae were grown in LC medium (1% tryptone/0.5% yeast extract/0.5% NaCl) at 30°C. The expression of draT was induced

with 1 mM IPTG for 4 h or as otherwise indicated. Antibiotics were used in selective media at the following concentrations: 100 &ml for E. co& Cb, 800 &ml, and Ap, 60 &ml for K. pneumoniae; Km, 50 pg/ml for both strains. b ts, thermosensitive; TI and 22, transcriptional terminators in rrnB.

draT

PIxC

L

s

draG’

XrmI PVUI

'%T 1%

SphI SElI

Ecg11 +

pKK223-3

!_I

XmnI

pBR3 2 ori PVUII

pHAF305

Ap,

Fig. 1. Construction of the draT expression vector, pHAF305. The 1.7-kb SmaI-Hind111 fragment of pWPFlO2 containing the draT coding region and the predicted RBS was inserted into polylinker of pKK223-3 immediately after the tat promoter. The RBS and the start codon of draT are underlined. Plasmid DNA was isolated by the alkaline extraction procedure (Bimboim, 1983). DNA fragments were puritied from the agarose gel with a GeneClean Kit (Bio 101, La Jolla, CA) according to the manufacturer’s specifications. Restriction enzyme digestions, ligations and agarose gel electrophoresis of DNA samples were performed by standard techniques (Maniatis et al., 1982). Competent cells were prepared and DNA transformations were performed according to the method of Ham&an (1983). Restriction enzymes and other enzymes used in cloning are from Bethesda Research Laboratories (Gaithersburg, MD), New England BioLabs (Beverly, MA)or Pharmacia Ltd. (Piscataway, NJ); other chemicals used in this work were purchased from BioRad, Boehringer-Mannheim (Indianapolis, IN), or Sigma (St. Louis, MO).

156

TABLE II DRAT activity in crude extracts of enteric bacteria Sources of crude extracts a

Strains”

DRAT activity b Expression’ vector

CPM”

nmoi ADP-ribose

+ Kp2

- Kp2

mg proteinC x min

E, eoli

UQ477 UQ741

pKK223-3 pHAF305

202 14582

130 139

0.0007 0.0560

K. pnenmoniae UN5350 UN5371 (LC medium) UN5371 (NF medium)

pHAF305 pHAF305

372 6013 6482

321 145 155

0.0002 0.0210 0.0220

202

ND

0.0007

No extract

a For DRAT assays, E. coli or K. pneumoniuecultures were grown aerobically (500 ml each) in a rich medium and were collected by centrifugation at 6000 x g for 10 min (15 mm for K. pneumoniue),resuspended in 1.5 ml Mops buffer (100 mM Mops pH 7.5/l mM ADP/SO PM EDTA/l mM dithiothreitol) and lysed in the presence of lysozyme (1 mg), DNase (1.5 mg), and RNase (1.5 mg) at 4°C for 30 min. The cells were disrupted by sonication as previously described (Hartmann et al., 1986). After centrifugation at 125000 x g for 2 h at 4”C, the supernatant was used as the crude extract. To obtain the nitrogen fixing cells of 1y.p~umoni~e, an LC-grown culture was inoculated (1: 100) into minimal medium with ammonium acetate as the nitrogen source (Nieva-Gomez et al., 1980) and was grown aerobically for 24 h. The cells were then collected (4000 x g, 5 min, 4°C) and resuspended in four volumes of minimal medium (without ammonium ions) in the presence of serine (0.015%; NF medium) and incubated anaerobically at 30°C for 4-5 h without shaking. The crude extracts of nitrogen fixing K. pneumoniaewere prepared as described above. b DRAT activity was assayed under anaerobic conditions by dete~ation of [a-32P]ADP-ribose incorporation into DR with NAD as donor molecule (Lowery et al., 1986). The reaction mixture contained 0.25 mM NAD, 0.6 @i [a-32P]NAD, 0.25 mM ADP, 5 mM MgCl,, 100 @gpurified active DR from K. pneumoniue(when present), and the crude extracts in Mops buffer (100 mM pH 7.0), to a total volume of 50 ~1. The reactions were started by adding active DR, and the mixture then was incubated at 30°C for 30 min. At this time, 10 ~1 of the reaction mixture was precipitated in 5% TCA and resuspended in SDS-PAGE buffer for gel assay (see Fig. 2). The remaining mixture was precipitated in 5% TCA and collected on a ~tr~~ulose filter. The radioactivity incorporated was measured by liquid scintillation counting. The specific incorporation of 32P into DR was confirmed by immunoblotting and autoradiography. The specific activity was expressed as nmol ADP-ribose incorporated into DR per min per mg of crude extract protein. c see Table I. d CPM, cpm per reaction mixture collected on filter; + Kp2, 100 pg of Kp2 was used in the assay; ND, not determined. e Protein con~ntrations were determined with the micro-biuret method (Goa, 1953) with bovine serum albumin as the standard.

in the reaction mixture is dependent on both the druT expression vector and DR. In the absence of the draT insertion in the plasmid (lanes 1 and 2) or in the absence of DR in the reaction mixture (lanes 2 and 4), no detectable labeling was observed. These experiments suggest that the DRAT activity synthesized in E. c5& shows the same specificity for DR as does the enzyme synthesized in the endogenous host, R. rubrum.

DRAT

activity

was also detected in the Kpneumoniae strain carrying pHAF305 when grown in LC medium or in a nitrogen-free medium under nif-derepression conditions (Table II). For

cells grown in LC medium, IPTG was added at 0.7 and the cells were incubated for another 3 h before h~esting. When the cells were grown in minimal medium, IPTG was added when the culture was transferred to the serine derepression medium, and the cells were incubated for 4 h before harvesting. There was no significant difference in the accumulated DRAT activity between the two growth conditions. The level of DRAT activity expressed in the heterologous hosts, E. coIi and Kpneumoniae, is comparable to that seen in the endogenous host, R. rubrum; the DRAT specific activity in crude extracts of A 6oo =

157

-

123456

2

3

4

Fig. 2. Autoradiogram of DR exposed to DRAT in vitro. DRAT reaction mixtures (10 ~1) were precipitated in 5% TCA and resuspended in SDS-PAGE buffer (lOOnI). After boiling for 2 min, 50 ~1 of the samples were loaded onto SDS-PAGE. The SDS-PAGE system of Laemmli (1970) as modified by Kanemoto and Ludden (1984) was used to resolve the modified and the unmodified subunits of DR on a 10% polyacrylamide gel. The enzyme-linked immunoblotting procedure was followed essentially as described earlier (Hartmann et al., 1986). Following SDS-PAGE, proteins were electrophoretically transferred to a nitrocellulose membrane (Towbin et al., 1979), incubated with antiserum raised against DR ofAzoto6acter vinelandii, and crossreacting material was visualized by using horseradish peroxidase conjugated to goat anti-IgG (BioRad, Richmond, CA). The autoradiograph was prepared from the immunoblot. The reaction mixture contained crude extracts of UQ477 (draT_) with added Kp2 (lane 1), or without Kp2 (lane 2); or crude extracts of UQ741 (draT+ ) with added Kp2 (lane 3), or without Kp2 (lane 4). The positions of modified (upper) and unmodified (lower) subunits of DR are indicated by the arrows. The signs ( - , + ) indicate the direction of the electric field.

Fig. 3. Immunoblot of DR from K. pneumoniae strains carrying pHAF305. To monitor the modification state of DR following draT expression, 5-10 ml of nif-derepressed K. pneumoniae cultures were quickly collected on glass microfiber filters (Whatman GF/A) under N, gas, and the filters were frozen immediately in liquid nitrogen (Kanemoto and Ludden, 1984; Hartmann et al., 1986). Cells were disrupted by grinding the frozen filters with carborundum powder in an anaerobic buffer. Following centrifugation at 6700 x g for 30 set, the supernatants were used as samples for SDS-PAGE and immunoblotting. ABer boiling for 2 mitt, 20 fig of the ‘quick extract’ protein was loaded on SDS-PAGE (10% T, 0.59% C). %T = {[acrylamide (g) + Bis (g)]/volume (ml)} x 100. %C = {[Bis (g)]/[acrylamide (g) + Bis (g)]} x 100. The immunoblot was obtained as described in the legend for Fig. 2. DR from the following strains was examined: UN5371 (nif’ draT’, lane 2); UN5374 (nifF_ draT’, lane 3); UN5376 (nigh_ draT+, lane 4); UN5375 (nif’ draT+, lane 5); and UN (nry , lane 6). Lane 1 is the in vivo “P-labeled ADPribosylated subunit of DR from UN5371. For 32P-labeling experiments, an LC-grown culture was inoculated (0.4 : 100) into low phosphate minimal medium (0.6 mM NaHPO; ) supplemented with 200 mM Mops pH 7.5 with ammonium acetate, and was grown aerobically for 8 h. This log-phase culture then was inoculated (3.3 : 100) into the low phosphate derepression minimal medium with 0.06% serine, but without ammonia and incubated under argon at 30°C. H332P04 was added to the culture (5 &i/ml) at the beginning of the derepression procedure, about 20 h before sampling. The positions of modified (upper) and unmodified (lower) subunits of DR are indicated by the arrows. The normal ‘doublet’ often seen for DR is visible in both lanes 5 and 6 as the fainter band immediately above the major band in each lane (Roberts et al., 1978). The nature of the antigenic bands seen above the position of the ADP-ribosylated subunit (lanes 2-5) and below the unmodified subunit (lanes 2-6) is unknown.

‘,‘

1.58

R. rubrum is 0.01-0.04 units (Lowery and Ludden,

(e) Expression of druT in mutant strains of Klebsief-

1986). The low production of DRAT activity in emetic bacteria, despite the strong promoter, may be because of the presence of the unusual start codon, TTG, in the draT gene and the poor translation of the draT message, resulting from suboptim~ codon usage (Fitzmaurice et al., 1989). DRAT activity was detectable in crude extracts prepared from E. coli cultures without IPTG induction. This result is consistent with the observation of others (Amann et al,, 1983; Kleiner et al., 1988) that a single copy of l&o on F’ is insufficient to repress expression from multiple copies of the tat promoter.

la pneumoniae

(d) Effect of druT expression on nitrogenase activity in vivo The functional expression of the draT gene in the enter& nitrogen-fucer K. pneumoni~e makes it possible to test the effect of DRAT activity on nitrogenase activity in vivo. Nitrogenase activity measurement was based on the acetylene reduction rate (Burris, 1972). Two ml of derepressed cultures were transferred to anaerobic vials (25 ml) and the reaction was started by injecting acetylene (0.1 atm). The ethylene produced was measured with a gas chromatography unit equipped with a flame ion~ation detector (Shimadzu GC-8A). The nitrogenase activity of whole cells is expressed as nmol ethylene formed per ml of cell culture per unit time when the absorbance at 600 nmol was normalized to 1.0. For the assay, nitrogen-fixing cultures were induced for DRAT synthesis for 4 h with IPTG. It was demonstrated that UN5366 (draT_) had a linear increase in acetylene reduction activity (6.0 nmol C,H,/ml x min), whereas UN5371 (druT+) had only a trace amount of activity (0.05 nmol C&H&l x min). The inability of UN5371 to reduce acetylene is not due to the absence of nitrogenase components, but rather to the covalent modification of DR (lane 2 of Fig. 3). Thus, it is demons~ated clearly that DRAT protein synthesized in K. ~~urnoni~~ strain UN5371 is capable of catalyzing the ADP-ribosylation of DR in vitro with NAD as the donor molecule, and also in vivo under physiological conditions. The ADPribosylation of DR in vivo by DRAT is independent of the DRAG activity.

Because numerous mutants affected in the activity of nif gene products are available in K. pneumon~e (banes et al., 1978), the expression of druT in this organism allowed us to test the substrate requirement for ADP-ribosylation of DR in vivo. It has been proposed that the redox state of DR might play a role in its regulation in vivo (Ludden et al., 1988). It already has been shown that the reduced form of DR can serve as a substrate for both DRAT and DRAG, because the DR in a nifL)- mut~t of R. rubrum can be ADP-ribosylated (Ludden et al., 1988). The et& cacy of the oxidized form of DR as a DRAT substrate in vivo is unknown. The physiological electron transport pathway to nitrogenase is well defined in K. pneumoniue (NievaGomez et al., 1980; Shah et al., 1983). The nifJ gene product (p~vate: flavodoxin o~dor~u~tase) oxidizes pyruvate and transfers two electrons to two molecules of the n&F gene product, a flavodoxin. The reduced flavodoxin supplies electrons to DR and sequentially to dinitrogenase where the N, is reduced. The results in Fig. 3 demonstrate that the DR from a NifF - strain of K. pneumoniae can be ADPribosylated, as evidenced by the slower migrating band on SDS-PAGE (lane 3 of Fig. 3). Since the dinitrogenase was intact in the NifF- strain and the defective flavodoxin could no longer reduce DR, the DR would be expected to accumulate in the oxidized form. Thus, the oxidized form of DR, as well as the reduced form, will serve as a substrate for ADPribosylation by DRAT in vivo. K.p~eumon~ strain UN1041 has a nigh mutation in which Arg-101, the site of ADP ribosylation, has been replaced by a His residue (Chang et al., 1988). This strain displays a DR band of slower mobility than the wild type, unmodified subunit and faster than the modified subunit (Chang et al., 1988). As shown in lane 4 of Fig. 3, strain UN5376 (a derivative of UN 104 1) showed the same single band of DR with unusual mobility, but no modified subunit was detectable in this strain when the druT expression vector was introduced, That the uppermost band in lane 4 is not the modified form was demonstrated by a lack of 32P at that position following

159

in vivo labeling (not shown). This observation is consistent with the in vitro results of others that purified DR from UN 104 1 could not be ADP-ribosylated by DRAT with NAD as a donor molecule (R.G. Lowery, personal communication), and indicates that the Arg-101 residue is necessary for ADP ribosylation by DRAT. As controls, wt K.pneumoniue strain, UN, without pHAF305, showed only the unmodified form of DR (lane 6), and both UN5371 or UN5375 with pHAF305 showed the ADP-ribosylated subunit of DR in addition to the unmodified form (lanes 2 and 5). The in vivo 32P-labeling experiment demonstrated that this slower migrating subunit of DR from UN5371 was labeled with 32P (lane l), further suggesting that this is the ADP-ribosylated form of the protein. (f) Conclusions The draT gene of R. rubrum is functionally exin the heterologous hosts, E. co/i and K. pneumoniae, as determined by in vitro and in vivo assays. The expression of the draT gene produces a Nif- phenotype in a nif” K. pneumoniae strain, and the analysis of mutant strains of K. pneumoniue pro-

pressed

vides evidence that the oxidized (nz#F- background) DR will serve as substrate for the DRAT reaction.

ACKNOWLEDGEMENTS

We thank Dr. P.W. Ludden for generous support, M.M. Gosink and Dr. W.P. Fitzmaurice for enlightening discussions, Dr. V.K. Shah for the derepression method, and N.M. Franklin for use of unpublished strains. The work was supported by the College of Agricultural and Life Sciences, University of WisconsinMadison, and by Department of Energy grant DEFG02-87ER13707 and USDA grant 87-CRCR-l2561.

REFERENCES Amann, E., Brosius, J. and Ptashne, M.: Vectors bearing a hybrid trp-lac promoter useful for regulated expression of cloned genes in Escherichic co&.Gene 25 (1983) 167-178.

Bimboim, H.C.: A rapid alkaline extraction method for the isolation of plasmid DNA. Methods Enzymol. 100 (1983) 263-255. Burris, R.H.: Nitrogen ftxation-assay methods and techniques. Methods Enzymol. 24B (1972) 415-431. Chang, C., Davis, L.C., Rider, M. and Takemoto, D.J.: Characterization of n#J mutations of Klebsiella pneumoniae. J. Bacterial. 170 (1988) 4015-4022. Fitzmaurice, W.P., Saari, L.L., Lowery, R.G., Ludden, P.W. and Roberts, G.P.: Genes coding for the reversible ADP-ribosylation system of dinitrogenase reductase from Rhodospirirum rubrum. Mol. Gen. Genet. 218 (1989) 340-347. Fu, H.-A., Hartmann, A., Lowery, R.G., Fitzmaurice, W.P., Roberts, G.P. and Burris, R.H.: Posttranslational regulatory system for nitrogenase activity in Azospirillum spp. J. Bacteriol. 171 (1989) 4679-4685. Goa, J.: A micro-biuret method for protein determination. Determination oftotal protein in cerebrospinal fluid. Stand. J. Clin. Lab. Invest. 5 (1953) 218-222. Gotto, J.W. and Yoch, DC.: Regulation of nitrogenase activity by covalent modification in Chromatium vinosum. Arch. Microbiol. 141 (1985) 40-43. Hanahan, D.: Studies on transformation of Escherichia cob with plasmids. J. Mol. Biol. 166 (1983) 557-580. Hartmann, A., Fu, H.-A. and Burris, R.H.: Regulation of nitrogenase activity by ammonium chloride in Azospirillum spp. J. Bacterial. 165 (1986) 864-870. Jouanneau, Y., Meyer, C.M. and Vignais, P.M.: Regulation of nitrogenase activity through iron protein interconversion into an active and an inactive form in Rhodopseudomonas capsulata. Biochim. Biophys. Acta 749 (1983) 318-328. Kanemoto, R.H. and Ludden, P.W.: Effect of ammonia, darkness, and phenazine methosulfate on whole-cell nitrogenase activity and Fe protein modification in Rhodospirillum rubrum. J. Bacterial. 158 (1984) 713-720. Kleiner, D., Paul, W. and Merrick, M.J.: Construction of multicopy expression vector for regulated over-production of proteins in Klebsiella pneumoniae and other enteric bacteria. J. Gen. Microbial. 134 (1988) 1779-1784. Laemmli, U.K.: Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature 227 (1970) 680-685. Lowery, R.G. and Ludden, P.W.: Purification and properties of dinitrogenase reductase ADP-ribosyltransferase from the photosynthetic bacterium Rhodospirillum rubrum. J. Biol. Chem. 263 (1988) 16714-16719. Lowery, R.G., Saari, L.L. and Ludden, P.W.: Reversible regulation of the nitrogenase iron protein from Rhodospirillum rubrum by ADP-ribosylation in vitro. J. Bacterial. 166 (1986) 513-518. Lowery, R.G., Chang, C.L., Davis, L.C., Stevens, P.W., McKenna, M.C. and Ludden, P.W.: Substitution of hi&tine for arginme 101 of dmitrogenase reductase disrupts electron transfer to dinitrogenase. Biochemistry 28 (1989) 1206-1212. Ludden, P.W. and Burris, R.H.: Activating factor for the iron protein of nitrogenase from Rhodospbillum rubrum. Science 194 (1976) 424-426. Ludden, P.W. and Burris, R.H.: Nitrogenase: properties and

160 regulation. In Neyra, CA. (Ed), Biochemical Basis of Plant Breeding, Vol. II. CRC Press, Boca Raton, FL, 1986, pp. 41-58. Ludden, P.W., Okon, Y. and Burris, R.H.: The nitrogenase system of Spriiffum I@oferum. Biochem. J. 173 (1978) 1001-1003. Ludden, P.W., Roberts, G.P., Lowery, R.G., Fitzmaurice, W.P., Saari, L.L., Lehman, L., Lies, D., Woehle, D., Wirt, H., Murrell, S.A., Pope, M.R. and Kanemoto, R.H.: Regulation of nitrogenase activity by reversible ADP-ribosylation of dinitrogenase reductase. In Bothe, H., de Bruijn, F.J. and Newton, W.E. (Eds.), Nitrogen Fixation: Hundred Years After. Gustav Fisher, Stuttgart, 1988, pp. 157-162. MacNeil, T., MacNeil, D., Roberts, G.P., Supiano, M.A. and Brill, W.J.: Fine-structure mapping and complementation analysis of nif (nitrogen fixation) genes in Klebsiellu pneumoniae. J. Bacterial. 136 (1978) 253-266. Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982. Nieva-Gomez, D., Roberts, G.P., Klevickis, S. and Brill, W.J.: Electron transport to nitrogenase in Klebsiella pneumoniae. Proc. Natl. Acad. Sci. USA 77 (1980) 2555-2558.

Pope, M.R., Murrell, S.A. and Ludden, P.W.: Covalent modification of the iron protein of nitrogenase from Rhodospirilum rubrum by adenosine diphosphoribosylation of a specific arginyl residue. Proc. Natl. Acad. Sci. USA 82 (1985) 3173-3177. Roberts, G.P., MacNeil, T., MacNeil, D. and Brill, W.J.: Regulation and characterization of protein products coded by the nif (nitrogen fixation) genes of Klebsiella pneumoniae. J. Bacteriol. 136 (1978) 267-279. Saari, L.L., Triplett, E.W. and Ludden, P.W.: Purification and properties of the activating enzyme for iron protein of nitrogenase from the photosynthetic bacterium Rhodospirillum rubrum. J. Biol. Chem. 259 (1984) 15502-15508. Saari, L.L., Pope,M.R., Murrell, S.A. and Ludden, P.W.: Studies on the activating enzyme for iron protein of nitrogenase from Rhodospirihm rubrum. J. Biol. Chem. 261 (1986) 4973-4977. Shah, V.K., Stacey, G. and Brig, W.J.: Electron transport to nitrogenase. J. Biol. Cl-rem. 258 (1983) 12064-12068. Towbin, H., Staehelin, T. and Gordon, J.: Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76 (1979) 4350-4354.