* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Plant and Soil
Magnesium transporter wikipedia , lookup
Gene expression wikipedia , lookup
Genetic engineering wikipedia , lookup
Oxidative phosphorylation wikipedia , lookup
Photosynthesis wikipedia , lookup
Western blot wikipedia , lookup
Interactome wikipedia , lookup
Genetic code wikipedia , lookup
Point mutation wikipedia , lookup
Silencer (genetics) wikipedia , lookup
Protein–protein interaction wikipedia , lookup
Biochemistry wikipedia , lookup
Plant nutrition wikipedia , lookup
Two-hybrid screening wikipedia , lookup
Proteolysis wikipedia , lookup
Nitrogen dioxide poisoning wikipedia , lookup
Amino acid synthesis wikipedia , lookup
Biosynthesis wikipedia , lookup
Metalloprotein wikipedia , lookup
Evolution of metal ions in biological systems wikipedia , lookup
Plant and Soil 137: 127-134, 1991. © 1991 Kluwer Academic Publishers. Printed in the Netherlands. Control of nitrogenase in PLSO NL19 Azospirillum sp. R O B E R T H. B U R R I S , A N T O N H A R T M A N N , Y A O P I N G Z H A N G and H A I A N FU Department of Biochemistry, University of Wisconsin, Madison, WI 53706, USA Key words: ammonium, Azospirillum sp., dinitrogenase reductase, 'switch off' Abstract Many N2-fixing organisms can turn off nitrogenase activity in the presence of N H ] and turn it on again when the N H 4 is exhausted. One of the most interesting systems for accomplishing this is by covalent modification of one subunit of dinitrogenase reductase by dinitrogenase reductase ADP-ribosyltransferase ( D R A T ) . The system can be reactivated when NH~- is exhausted, by dinitrogenase reductase activating glycohydrolase ( D R A G ) which removes the inactivating group. It is fascinating that some species of the genus Azospirillum possess the D R A T and D R A G systems (A. lip@rum and A. brasilense), whereas A. amazonense in the same genus lacks D R A T and D R A G . A. amazonense responds to N H 4 but does not exhibit modification of dinitrogenase reductase characteristic of the action of D R A T . However, it has been possible to clone D R A T and D R A G and to introduce them into A. amazonense, whereupon they become functional in this organism. The D R A T and D R A G system does not appear to function in Acetobacter diazotrophicus, an organism isolated from sugar cane, that fixes N 2 at a p H as low as 3.0. A. diazotrophicus does show a rather sluggish response to N H 4 . A level of about 1 0 / z M N H 4 is required to 'switch off' the system. The response to N H 4 is influenced by the dissolved oxygen concentration ( D O C ) as has been reported for Azospirillum sp. A D O C in equilibrium with 0.1 to 0.2 kPa 0 2 seems optimal for the response in A. diazotrophicus. Introduction Nitrogen fixation is very energy demanding. Chemical fixation via the Haber-Bosch process achieves breakage of the N - N bond and reduction of N 2 to 2 N H 3 by employing high temperature and high pressure in the presence of a catalyst. Biological fixation of N 2 faces the same energy barriers but operates in a more subtle fashion at ambient temperature and a pN 2 of 0.78 atm. by invoking the combined action of dinitrogenase and dinitrogenase reductase. The action of these enzymes is supplemented by ferredoxin or flavodoxin as a reductant and M g A T P as an energy source. When one provides optimal conditions for nitrogenase activity it still requires 16 M g A T P for the reduction of 8 H ÷ + N 2 -+ 2 N H 3 + H 2. Under natural conditions, 20 to 30 M g A T P are used per N 2 reduced. Faced with this energy demand, there have been evolutionary pressures for organisms possessing nitrogenase to develop a means to turn off nitrogenase when fixed nitrogen is available to supply the organism's needs. These control systems vary among N2-fixing organisms. Our interest has centered on the system that invokes covalent modification of one unit of the dimer of dinitrogenase reductase. When N H 4 enriched in 15N became available, the 15N furnished a means for quickly establishing the initiation and cessation of N 2 fixation and the influence of fixed nitrogen on the process of biological N a fixation. When Azotobacter vinelandii was given the choice of 15N2 and N H 4 or other nitrogenous compounds, Wilson et al. (1943) found that the organisms showed a clear preference for N H 4 over nitrate, nitrite and a variety of amino acids. When N H 4 at 121 or 128 Burris et al. 315 ppm N was present, only 0.9% or 0.0%, respectively, of the cell nitrogen was derived from the I5N2. Burris and Wilson (1946) expanded on these observations and followed the time course of utilization of NH 4 and nitrate by A. vinelandii. Although switch-off of N 2 fixation of A. vinelandii by NH 4 was demonstrated may years ago, the detailed mechanism of switch-off in A. vinelandii has not been elucidated. In contrast to A. vinelandii, the control in Rhodospirillum rubrum has been worked out in considerable detail. Kamen and Gest (1949) and Gest and Kamen (1949) demonstrated that Rhodospirillum rubrum is capable of fixing N 2. Although Schneider et al. (1960) reported that they had prepared cell-free extracts from R. rubrum that fixed N2, there was difficulty in getting consistent N 2 fixation with extracts from this organism. Munson and Burris (1969) found that the time course of reductions catalyzed by R. rubrum nitrogenase preparations was nonlinear and exhibited a lag phase. The inconsistencies arose because R. rubrum has a method for quickly turning off its N 2 fixation system and then turning it on again. Ludden and Burris (1976) first reported the presence of an activating system for nitrogenase in R. rubrum chromatophores. The factor could be separated from the dinitrogenase and dinitrogenase reductase on a column of DEAE cellulose. ATP and Mg 2+ or Mg 2÷ plus Mn 2+ were necessary for the activation of dinitrogenase reductase. The activating factor if preincubated with R. rubrum preparations abolished the lag. It was apparent that the action of the activating factor was centered on dinitrogenase reductase and had no influence on dinitrogenase. Several students in the laboratory had problems obtaining active nitrogenase preparations from R. rubrurn despite the initial success of Schneider et al. (1960). Paul Ludden took up the problem and made the crucial observation that an extract from R. rubrurn chromatophores would activate dinitrogenase reductase from the organism. The responses suggested that an enzyme was activating the Fe protein of nitrogenase, and with incubation the activity increased. The agent was designated the 'activating factor' by Ludden and Burris (1976). It is interesting that Nordlund et al. (1977) in Sweden observed the factor independently. They derived the activating agent from NaC1 extracts of chromatophore membranes. They also reported that it was sensitive to oxygen, reacted specifically with the Fe protein of nitrogenase and that its activity required ATP and Mg 2+. Ludden and Burris (1979) reported that when inactive dinitrogenase reductase was activated it released an 'adenine-like molecule'. ATP and a divalent metal ion were required for activation of the protein. Nordlund and Eriksson (1979) observed that preparations were not active in H 2 production until treated with the activating factor. Ludden and Burris (1976) reported that the addition of Mn 2+ and Mg 2+ enhanced the activity of extracts from R. rubrum. Anaerobic collection of R. rubrum aided in recovering active extracts from R. rubrum. In 1980 Ludden could verify that the activating factor was oxygen labile and sensitive to trypsin. It could by stabilized, and it could be precipitated with 30% polyethylene glycol 4000. The extract could be purified further on a column of DEAE cellulose. The presence of NH 4 in the culture medium did not suppress the production of activating factor. The dinitrogenase reductase recovered from R. rubrum had its full complement of Fe and S whether it was active or inactive. It proved possible to separate the inactive Fe protein from R. rubrum into two components on polyacrylamide gels (Ludden and Burris, 1978). The inactive Fe protein carried phosphate, ribose and an 'adenine-like' unit. With an active Fe protein available it was possible to purify both dinitrogenase and dinitrogenase reductase from R. rubrum, and the properties of both were similar to those of other nitrogenase proteins. The modifying group did not appear on dinitrogenase reductase isolated from other organisms. Nordlund et al. (1977) had suggested that the 'activating factor' formed a stable complex with dinitrogenase reductase to restore its catalytic activity. Ludden (1980) in contrast presented evidence that activity was restored when the activating enzyme removed the adenine-ribosephosphate unit whose addition had inactivated the enzyme. The process required the activating enzyme, a divalent metal ion (Mg z+, Mn 2+, Control of nitrogenase in Azospirillurn Fe 2+) and ATP. He pointed out that upon activation the second band disappeared from a gel used to separate components of inactive dinitrogenase reductase. So the picture emerged that R. rubrum dinitrogenase reductase was inactivated in the dark or aerobically by adding a unit with phosphate, ribose and adenine, and that the dinitrogenase reductase could be reactivated by enzymatic removal of these blocking units. From this point the story has developed primarily with observations by Ludden and colleagues, and the observations and conclusions have been summarized by Ludden (1980) and by Ludden and Roberts (1989). They depict their scheme for activation and inactivation of dinitrogenase reductase as shown in Figure 1. We will define their terms and then develop the rationale. Starting at the left, we have the active form of dinitrogenase reductase with its shared Fe/S center and with two equivalent protein subunits. As illustrated at the top of the figure the dinitrogenase reductase is altered by the enzyme D R A T (dinitrogenase reductase ADPribosyltransferase) which with NAD and MgADP, ADP-ribosylates one of the two equivalent protein subunits and thus inactivates dinitrogenase reductase. This is accompanied by the release of nicotinamide. The inactivation oc- 129 curs by ADP-ribosylation of the arginine in position 101. Note that D R A T is coded for by draT and that the inactivation is triggered by NH2 or darkness. The inactive complex can be reactivated by D R A G (dinitrogenase reductase-activating glycohydrolase) acting with MgATP and Mn a+. ADP ribose is removed from arginine-101 of the inactive dinitrogenase reductase and is released. The D R A G is coded for by draG. Dinitrogenase reductase is restored to its normal active state in which the two subunits are equivalent. The story is well told by Ludden and Roberts (1989) of how the activation and inactivation of dinitrogenase reductase was resolved and how the various components in the system fit together. Ludden and Burris (1979) presented evidence that both adenine and ribose were covalently bound to inactive dinitrogenase reductase and that both were removed when the dinitrogenase reductase was reactivated. Pope et al. (1985a,b) purified the displaced adenine moiety and demonstrated its identity as ADP-ribose by NMR and mass spectrometric analysis. Jouanneau et al. (1988) found ADP ribosylation also functioned in the inactivation of the nitrogenase system in Rhodobacter capsulatus. The attachment of the ADP to ribose was established by Pope et al. (1986). MgADP NAD DRAT nicotinamide [ NHI [darkness I PMS - (active) ~ f ADPR IcccP (-) J DRAG MgATP MnZ* (inactive) Fig. i. Control of dinitrogenase reductase activity by ADP-ribosylation as depicted by Ludden and Roberts (1989). ADPR, ADP-ribose; PMS, phenazine methosulfate; CCCP, chlorocarbonylcyanidephenylhydrazone. 130 Burris et al. Experimental We now shift to some of the specifics of our work and the work of others on the control of nitrogenase in Azospirillum sp. and related organisms. Under proper conditions Azospirillum species fix nitrogen vigorously. Characteristically they thrive at a low partial pressure of oxygen, and certain species have their N 2 fixation systems turned off readily by low concentrations of NH 4 . Azospirillum lip@rum is the most studied species of the group. A. lipoferum was first described by Beijerinck (1925) as Spirillum lipoferum, reexamined by Becking (1963), and brought into prominence when D6bereiner and Day (1976) reported that they had found it associated with the roots of grasses that did not exhibit symptoms of nitrogen deficiency whereas surrounding grass did. They described how to culture the organism and pointed out that it was microaerobic. Since then the organism has been renamed and studied extensively. Okon et al. (1976a) reported that S. lipoferum grew vigorously on malate, lactate, succinate and pyruvate and poorly on glucose. It grows aerobically when supplied NH 4 , but when it is fixing N 2 the optimal pO 2 is a bit below 0.01 atm. (1 kPa). Okon, Houchins, Albrecht and Burris (1977) reported that on NH 4 the doubling time of S. lipoferum was 1 h, whereas on N 2 it was 5.5 to 7 h. They found that its dinitrogenase reductase was activated by the activating factor from R. rubrum. Ludden et al. (1978) separated S. lipoferum extracts into three components: dinitrogenase, dinitrogenase reductase and an activating factor comparable to that recovered from R. rubrum. The efficacy of the activating factor was increased by added Mn 2+ and Mg 2+ The activation required ATP. Hartmann and Burris (1987) reported that the nitrogenase activity of A. brasilense and A. lipoferum was completely inhibited by 2 kPa of oxygen; some activity was recovered by transferring to 0.2 kPa of oxygen. There was no covalent modification of dinitrogenase reductase comparable to that arising from switch-off by NH 4 as evidenced by Western-blotting and 32p-labeling experiments. The question was posed by Ludden et al. (1982), 'Why is the Fe protein inactive as purified from R. rubrum grown on glutamate as its nitrogen source'? Although the inactive Fe protein proved capable of binding MgATP and undergoing a conformational shift upon binding MgATP, and it could undergo oxidation-reduction, it was incapable of transferring electrons to dinitrogenase at a substantial rate. Hence, it could not function in the reduction of N 2 or acetylene. Apparently the only reported purification of the nitrogenase from the Azospirillum group is that of Song et al. (1985). Purification from A. amazonense was not unlike that for other nitrogen fixers and specific activities were comparable. They were unable to find an activating enzyme for A. amazonense so the organism differed in this respect from A. lipoferum and A. brasilense. Vignais and coworkers (1987) reviewed information on regulation of N 2 fixation in the photosynthetic bacteria. They pointed out that it is generally accepted that NH 4 itself is not the real effector of switch-of of nitrogenase, and that glutamine synthetase may be involved directly or indirectly. However, they concluded that there is no direct evidence that glutamine is the in vivo effector of switch-off. The switch-on-off process can be very rapid in N2-fixing organisms. Sweet and Burris (1981) followed nitrogenase activity by measuring H 2 production with a hydrogen electrode. This permitted continuous measurement, and rapid changes could be observed. R. rubrum grown on N 2 on a glutamate medium turned nitrogenase off quickly when NH 4 was added and turned production of H a back on in seconds after exhaustion of the NH 4. Cejudo et al. (1984) found that low levels of NH~- turned off nitrogenase immediately in Azotobacter chroococcum, and that the activity returned when the NH 4 was exhausted. Inhibitors of NH 4 assimilation, such as methionine sulfoximine, prevented NH 4 switch-off. Hartmann et al. (1986) reversibly inhibited the nitrogenase activity of A. brasilense, A. lipoferum and A. amazonense. The interesting observation was that methionine sulfoximine abolished switch-off in A. lipoferum and A. brasilense but not in A. amazonense. Obviously Control of nitrogenase in Azospirillum there was something different about the control of nitrogenase among species of the same genus. In A. amazonense, the in vitro nitrogenase activity of NH4-treated cells was not decreased, and no evidence could be found for a modified Fe protein band on electrophoretic examination. Hartmann et al. (1988) noted further differences among species and strains of the azospirilla. A. brasilense grew poorly on glutamate, aspartate, serine or histidine, and these amino acid even at concentrations of 10mM showed little inhibition of N 2 fixation. In contrast, A. lip@rum and A. amazonense grew well on these amino acids as a source of carbon and nitrogen. Azospirillum spp. thus differ substantially in their metabolism of amino acids. Herbaspirillum seropedicae, a N2-fixing organism isolated and characterized by D6bereiner (1989), is microaerobic. Fu and Burris (1989) found its optimal pO 2 is from 0.04 to 0.2 kPa for N 2 fixation. NH 4 inhibited N 2 fixation only partially even at 20 mM concentration. Treatment with NH 4 gave no detectable change in the electrophoretic pattern of the nitrogenase components, so there is no evidence that ADPribosylation plays a part in turning off nitrogenase in this organism. Control of nitrogenase by the DRAT and D R A G system is by no means universal among N2-fixing microorganisms. The question arises whether one can transfer the system to other microorganisms devoid of it. Fu, Wirt et al. (1989) demonstrated that the DRAT system could be moved to E. coli and KlebsieIla pneumoniae. When the DRAT was expressed in wildtype K. pneumoniae strains, they lost their ability to fix N 2. If the arginine 101 site on dinitrogenase reductase was eliminated, the dinitrogenase reductase could not be inactivated by ADP-ribosylation. The mechanism of NH 4 switch-on-off in A. brasilense and A. lipoferum was investigated by Fu, Hartmann et al. (1989), and they established a correlation between in vivo regulation of nitrogenase activity by NH 4 or glutamine and the reversible covalent modification of dinitrogenase reductase. Dinitrogenase reductase ADP-ribosyltransferase (DRAT) activity was found in extracts of A. brasilense with NAD serving as the donor molecule. Dinitrogenase reductase- 131 activating glycohydrolase (DRAG) activity was present in extracts of both A. lipoferum and R. rubrum. The region homologous to R. rubrum draT and draG was identified in the genomic DNA of A. brasilense as a 12-kilobase EcoRI fragment and in A. lip@rum as a 7-kilobase EcoRI fragment. The authors concluded that a posttranslational regulatory system for nitrogenase activity is present in A. brasilense and A. lipoferum, and that it operates via ADP-ribosylation of dinitrogenase reductase as it does in R. rubrum. The question of whether the regulatory system characteristic of R. rubrum could be transferred to a N2-fixing organism lacking it and whether it would function there was posed by Fu, Burris and Roberts (1990). They found that they could transfer draT and draG, the genes coding for DRAT and DRAG, from R. rubrum or A. lip@rum to K. pneumoniae (an organism normally devoid of these genes) and demonstrate that they could function. The expressed draT and drag genes allowed K. pneumoniae to respond to NH 4 with a reversible regulation of nitrogenase activity that was correlated with the reversible ADP-ribosylation of dinitrogenase reductase. One can conclude that DRAT and D R A G in K. pneumoniae are sufficient to support NH 4+-switch-off-on, " and that ADP-ribosylation can serve as a regulatory mechanism for this organism. Fu et al. (1990) cloned the draTG genes from A. lipoferum and demonstrated that they are located next to nifH (structural gene for dinitrogenase reductase) as in R. rubrum. The similarity in organization strongly suggests that these genes have been conserved in these two organisms during evolution. This is the first identification of this regional homology outside the photosynthetic bacteria. Herbaspirillum seropedicae is an organism that has an unusually low optimal pO 2 for nitrogen fixation as measured by reduction of acetylene. As indicated in Figure 2, the optimal acetylene reduction occurred at about 0.1 kPa. A. brasilense had optimal acetylene reduction at about three times this pO 2. Acetobacter diazotrophicus (formerly Acetobacter nitrocaptans) is a nitrogen-fixing organism with intriguing properties. This organism was Burris et al. 132 300 800 ! x 600 ...200 - g oJ ,..~ 400 "6 E= ~_t00 200 o 0 i i t 2 4 6 kPa Oxygen Fig. 2. Influence of pO 2 on acetylene reduction by Acetobacter diazotrophicus (-[3-) in comparison with AzospirilIurn brasilense Sp7 (-[]-) and Herbaspirillurn seropedicae Z176 (-Ii,-). Samples (4 mL) of N2-fixingcultures were transferred to an O2-electrode chamber, C2H2 was added, and samples were removed for measurement of C2H 2 reduction at different dissolved O 2 levels. The OD580 of the culture was about 1.0 and contained about 500/xg protein mL-:. isolated f r o m sugar cane and described by Boddey and D 6 b e r e i n e r (1988) and D 6 b e r e i n e r (1989). They have summarized recent studies on associative nitrogen fixation. Most free-living and associative nitrogen fixers achieve limited N 2 fixation because of the dearth of energy-yielding substrates available to them. H o w e v e r , A. diazotrophicus by growing in association with sugar cane has chosen a location rich in carbohydrate. It takes advantage of the available sucrose, grows actively in or on the cane stem and roots, and fixes an abundance of N 2. The physiology of the organism is particularly r e m a r k a b l e in two aspects: (a) it grows in high concentrations of sucrose (we grow it routinely in 10% sucrose and D 6 b e r e i n e r , 1989, reports that it grows in 30% sucrose), and (b) it fixes N 2 at a remarkably low p H (it grows and reduces acetylene below p H 3.0). T h e r e is no clear definition of the internal p H of the organism, but it grows on a nitrogen-free m e d i u m at p H 3.0 and lower. T h e r e are indications that A. diazotrophicus and other organisms that m a y be associated with it m a y fix over 200 kg N per hectare in a year. The data in Figure 3 are from an experiment in which amino acids and azaserine were added 0 0 10 20 M1nutes 30 40 50 Fig. 3. Effect of amino acids and azaserine on nitrogenase activity of Acetobacter diazotrophicus PAL-5. At 15 min, 2mM azaserine (-[3-), 5mM asparagine (-~-), 5mM glutamic acid (-<>-) or water as control (-[]-) was added. Reactions were run in an O2-electrode chamber at 0.10.2 kPa 02. to Ne-fixing cultures of A. diazotrophicus. A t 15 min 2 m M azaserine, 5 m M glutamic acid and 5 m M asparagine were added to the chamber. Glutamic acid and asparagine inhibited about 25% and azaserine about 50%. As discussed, it has been shown that N 2 fixation can be turned off by N H 4 and that the mechanism in R. rubrum, A. lipoferurn and A. brasilense is by ADP-ribosylation of arginine 101 of one of the protein subunits of dinitrogenase reductase. W h e n this occurs, the two subunits can be separated by gel electrophoresis, because one has had its properties altered by adding the A D P - r i b o s e at arginine 101. We found no evidence for such a switch-off-on process in A. diazotrophicus, so we examined the electrophoretic pattern of dinitrogenase reductase from this organism. Figure 4 indicates that A. diazotrophicus is not subject to control by covalent modification of its dinitrogenase reductase. The left lane shows the separation of two c o m p o n e n t s f r o m A. brasilense that had been treated with N H 2 ; this is an organism that undergoes switch-off by ADP-ribosylation of its dinitrogenase reductase. The normal subunit and the modified subunit from its dinitrogenase reductase are clearly separated. In contrast, only one band appears in extracts from A. diazo- Control of nitrogenase in Azospirillum 133 trophicus. Lane 2 is before and lane 3 after a 30 min treatment with 10 mM NH,Cl. Cells for lane 3 were held under 6 kPa 0, for 30 min and those for lane 5 were anaeorobic for 30min. Under no conditions was the dinitrogenase reductase altered by covalent modification as it is in A. brasilense. Figure 5 supports these conclusions and indicates that A. diazotrophicus lacks draT and draG genes. Fig. 4. Immunoblots of crude extracts from Acetobacfer diazotrophicus PAL-5 separated by SDS-PAGE with antiserum against dinitrogenase reductase of Azotobacter vinelandii. Lane 1, crude extract of Azospirillum brasilense treated with 5 mM NH,Cl for 30 min. Lanes 2, 3, 4, 5 from crude extracts of A. diazotrophicus PAL-5 before (lane 2) and after 30 min treatment with 10mM NH,Cl (lane 3). Lane 4, 30min at 6.0 kPa 0,; and lane 5, anaerobic for 30min. The crude extracts were prepared by sonication (lanes 1 and 2) or by quick extraction (lanes 3, 4, 5). Fig. 5. Southern hybridization of genomic DNA of Acetobacter diazotrophicus PAL-5 with a draTG probe from Azospirillum lipoferum. Total DNAs from A. lipoferum (lane 1) and A. diazobophicus (lane 2) were digested with EcoRI and hybridized to a “P-labeled draTG probe from A. lipoferum. (A)Ethidium bromide-stained DNA on the agar- ose gel, (B) autoradiogram. Bacteriophage A-DNA cut with Hind III was used as a size marker and control (lane 3). References Becking J H 1963 Fixation of molecular nitrogen by an aerobic Vibrio or Spirillum. Antonie van Leeuwenhoeck 29, 326. Beijerinck M W 1925 Uber ein Spirillum welches freien Stickstoff binden kann? Zenthl. Bakt. Parasitkde. 63, 353359. Boddey R M and Diibereiner J 1988 Nitrogen fixation associated with grasses and cereals: Recent results and perspectives for future research. Plant and Soil 108; 53-65. Burris R H and Wilson P W 1946 Ammonia as an intermediate in nitrogen fixation by Azotobacter. J. Bacterial. 52, 505-512. Cejudo F J, de la Torre A and Paneque A 1984 Short-term ammonium inhibition of nitrogen fixation in Azotobacter. Biochem. Biophys. Res. Commun. 123, 431-437. Dobereiner J 1989 Isolation and identification of root-associated diazotrophs. In Nitrogen Fixation with Non-Legumes. Eds. F A Skinner et al. pp 103-108. Kluwer Academic Publishers, Dordrecht, The Netherlands. Dobereiner J and Day J M 1976 Associative symbioses in tropical grasses: Characterization of microorganisms and dinitrogen-fixing sites. In Proc. 1st Internat. Symp. on Nitrogen Fixation. Eds. W E Newton and C J Nyman. pp 518-538. Washington State University Press, Pullman, WA. Fu H and Burris R H 1989 Ammonium inhibition of nitrogenase activity in Herbaspirillum seropedicae. J. Bacterial. 171, 3168-3175. Fu H, Burris R H and Roberts G P 1990 Reversible ADPribosylation is demonstrated to be a regulatory mechanism in prokaryotes by heterologous expression. Proc. Natl. Acad. Sci. USA 87, 1720-1724. Fu H, Fitzmaurice W P, Roberts G P and Burris R H 1990 Cloning and expression of draTG genes from Azospirillum lipoferum. Gene 86, 95-98. Fu H, Hartmann A, Lowery R G, Fitzmaurice W P, Roberts G P and Burris R H 1989 Posttranslational regulatory system for nitrogenase activity in Azospirillum spp. J. Bacterial. 171, 4679-4685. Fu H-A, Wirt H J, Burris R H and Roberts G P 1989 Functional expression of a Rhodospirillum rubrum gene encoding dinitrogenase reductase ADP-ribosyltransferase in enteric bacteria. Gene 85, 153-160. 134 Control o f nitrogenase in A z o s p i r i l l u m Gest H and Kamen M D 1949 Photoproduction of molecular hydrogen by Rhodospirillum rubrum. Science 109, 558559. Hartmann A and Burris R H 1987 Regulation of nitrogenase activity by oxygen in Azospirillurn brasilense and Azospirillure lipoferum. J. Bacteriol. 169, 944-948. Hartmann A, Fu H and Burris R H 1986 Regulation of nitrogenase activity by ammonium chloride in Azospirillum spp. J. Bacteriol. 165, 864-870. Hartmann A, Fu H and Burris R H 1988 Influence of amino acids on nitrogen fixation ability and growth of AzospiriIlure spp. Appl. Environ. Microbiol. 54, 87-93. Jouanneau Y, Meyer C and Vignais P M 1988 Regulation of nitrogenase activity in Rhodobacter capsulatus: ADP ribosylation of the Fe protein. In Nitrogen Fixation: Hundred Years After. Eds. H Bothe, F J deBruijn and W E Newton. p. 173. Gustav Fischer, Stuttgart. Kamen M D and Gest H 1949 Evidence for a nitrogenase system in the photosynthetic bacterium Rhodospirillum rubrum. Science 109, 560. Ludden P W 1980 Nitrogen fixation by photosynthetic bacteria: Properties and regulation of the enzyme system from Rhodospirillum rubrum. In Nitrogen Fixation. Eds. W E Newton and W H Orme-Johnson. pp 139-156. University Press, Baltimore, MD. Ludden P W and Burris R H 1976 Activating factor for the iron protein of nitrogenase from Rhodospirillum rubrum. Science 194, 424-426. Ludden P W and Burris R H 1978 Purification and properties of nitrogenase from Rhodospirillurn rubrum, and evidence for phosphate, ribose and an adenine-like unit covalently bound to the iron protein. Biochem. J. 175, 251-259. Ludden P W and Burris R H 1979 Removal of an adeninelike molecule during activation of dinitrogenase reductase from Rhodospirillum rubrum. Proc. Natl. Acad. Sci. USA 76, 6201-6205. Ludden P W, Hageman R V, Orme-Johnson W H and Burris R H 1982 Properties and activities of 'inactive' Fe protein from Rhodospirillum rubrum. Biochim. Biophys. Acta 700, 213-216. Ludden P W, Okon Y and Burris R H 1978 The nitrogenase system of Spirillum lipoferum. Biochem. J. 173, 10011003. Ludden P W and Roberts G P 1989 Regulation of nitrogenase activity by reversible ADP ribosylation. In Current Topics in Cellular Regulation 30, 23-56. Academic Press, New York. Nordlund S and Eriksson U 1979 Nitrogenase from Rhodos- pirillum rubrum. Relation between 'switch-off' effect and the membrane component. Hydrogen production and acetylene reduction with different nitrogenase component ratios. Biochim. Biophys. Acta 547, 429-437. Nordlund S, Eriksson U and Baltscheffsky H 1977 Necessity of a membrane component for nitrogenase activity in Rhodospirillum rubrum. Biochim. Biophys. Acta 462,187195. Okon Y, Albrecht S L and Burris R H 1976 Factors affecting growth and nitrogen fixation of Spirillum lipoferum. J. Bacteriol. 127, 1248-1254. Okon Y, Houchins J P, Albrecht S L and Burris R H 1977 Growth of Spirillum lipoferum at constant partial pressures of oxygen, and the properties of its nitrogenase in cell-free extracts. J. Gen. Microbiol. 98, 87-93. Pope M R, Murrell S A and Ludden P W 1985a Purification and properties of the heat-released nucleotide-modifying group from the inactive iron protein of nitrogenase from Rhodospirillum rubrum. Biochemistry 24, 2374-2380. Pope M R, Murrell S A and Ludden P W 1985b Covalent modification of the iron protein of nitrogenase from Rhodospirillum rubrum by adenosine diphosphoribosylation of a specific arginine residue. Proc. Natl. Acad. Sci. USA 82, 3173-3177. Pope M R, Saari L L and Ludden P W 1986 N-glycohydrolysis of adenosine diphosphoribosyl arginine linkages by dinitrogenase reductase activating glycohydrolase (activating enzyme) from Rhodospirillum rubrum. J. Biol. Chem. 261, 10104-10111. Schneider K C, Bradbeer C, Singh R N, Wang L C, Wilson P W and Burris R H 1960 Nitrogen fixation by cell-free preparations from microorganisms. Proc. Natl. Acad. Sci. USA 46, 726-733. Song S-D, Hartmann A and Burris R H 1985 Purification and properties of the nitrogenase of Azospirillum amazonense. J. Bacteriol. 164, 1271-1277. Sweet W J and Burris R H 1981 Inhibition of nitrogenase activity by NH 4 in Rhodospirillum rubrum. J. Bacteriol. 145, 824-831. Vignais P M, Willison J C, Allibert P, Ahombo G and Jouanneau Y 1987 Regulation of nitrogen fixation in photosynthetic bacteria. In Inorganic Nitrogen Metabolism. Eds. Q Ullrich et al. pp 154-159. Springer-Verlag, Berlin. Wilson P W, Hull J F and Burris R H 1943 Competition between free and combined nitrogen in nutrition of Azotobacter. Proc. Natl. Acad. Sci. USA 29, 289-294.