Effect of oxygen on acetylene reduction by photosynthetic bacteria

The effect of dissolved oxygen concentration on nitrogenase activity was studied in three species of photosynthetic bacteria. The O2 concentration in the cell suspension was measured with an O2 electrode inserted into the reaction vessel. Acetylene reduction by whole cells of Rhodopseudomonas capsulata, Rhodospirillum rubrum, and Chromatium vinosum strain D was inhibited 50% by 0.73, 0.32, and 0.26 microM O2, respectively. The inhibition of the activity by O2 in R. capsulata usually was reversed completely by reestablishing anaerobic conditions. In R. rubrum and C. vinosum the inhibition was only partially reversible. The respiration rate of R. capsulata was the highest of the three, that of R. rubrum was intermediate, and that of C. vinosum was lowest. R. capsulata and R. rubrum cells were broken after their acetylene reduction activity in vivo had been completely inhibited by O2, and nitrogenase was found to be active in vitro. A concentration of cyanide that did not affect acetylene reduction activity, but which inhibited 75 to 90% of the O2 uptake by whole cells of R. capsulata, shifted the O2 concentration causing 50% inhibition of nitrogenase activity from 0.73 microM to 2.03 microM. These results are in accordance with the assumption that within a limited range of O2 concentrations, the respiratory activity of the cells is enough to scavenge the O2 and to keep the interior of the cells essentially anaerobic. It is suggested that O2 inhibits nitrogenase activity by competing for a limited supply of electrons. When cyanide is present, respiration is slower but is adequate to keep the nitrogenase environment in the cell anaerobic. The lower respiration rate may allow a greater proportion of the electrons to be used for acetylene reduction.     

Characterization of a TOL-like plasmid from Alcaligenes eutrophus that controls expression of a chromosomally encoded p-cresol pathway.

Strains of all 18 species of the family Rhodospirillaceae (nonsulfur photosynthetic bacteria) were studied for their comparative nitrogen-fixing abilities. All species, with the exception of Rhodocyclus purpureus, were capable of growth with N2 as the sole nitrogen source under photosynthetic (anaerobic) conditions. Most rapid growth on N2 was observed in strains of Rhodopseudomonas capsulata. Within the genus Rhodopseudomonas, the species R. capsulata, R. sphaeroides, R. viridis, R. gelatinosa, and R. blastica consistently showed the highest in vivo nitrogenase rates (with the acetylene reduction technique); nitrogenase rates in other species of Rhodopseudomonas and in most species of Rhodospirillum were notably lower. Chemotrophic (dark microaerobic) nitrogen fixation occurred in all species with the exception of one strain of Rhodospirillum fulvum; oxygen requirements for dark N2 fixation varied considerably among species and even within strains of the same species. We conclude that the capacity to fix molecular nitrogen is virtually universal among members of the Rhodospirillaceae but that the efficacy of the process varies considerably among species.     

Continuous monitoring, by mass spectrometry, of H2 production and recycling in Rhodopseudomonas capsulata.

Hydrogen evolution and consumption by cell and chromatophore suspensions of the photosynthetic bacterium Rhodopseudomonas capsulata was measured with a sensitive and specific mass spectrometric technique which directly monitors dissolved gases. H2 production by nitrogenase was inhibited by acetylene and restored by carbon monoxide. An H2 evolution activity coupled with HD formation and D2 uptake (H-D exchange) was unaffected by C2H2 and CO. Cultures lacking nitrogenase activity also exhibited H-D exchange activity, which was catalyzed by a membrane-bound hydrogenase present in the chromatophores of R. capsulata. A net hydrogen uptake, mediated by hydrogenase, was observed when electron acceptors such as CO2, O2, or ferricyanide were present in the medium.     

Nitrous oxide reduction by members of the family Rhodospirillaceae and the nitrous oxide reductase of Rhodopseudomonas capsulata.

After growth in the absence of nitrogenous oxides under anaerobic phototrophic conditions, several strains of Rhodopseudomonas capsulata were shown to possess a nitrous oxide reductase activity. The enzyme responsible for this activity had a periplasmic location and resembled a nitrous oxide reductase purified from Pseudomonas perfectomarinus. Electron flow to nitrous oxide reductase was coupled to generation of a membrane potential and inhibited by rotenone but not antimycin. It is suggested that electron flow to nitrous oxide reductase branches at the level of ubiquinone from the previously characterized electron transfer components of R. capsulata. This pathway of electron transport could include cytochrome c', a component hitherto without a recognized function. R. capsulata grew under dark anaerobic conditions in the presence of malate as carbon source and nitrous oxide as electron acceptor. This confirms that nitrous oxide respiration is linked to ATP synthesis. Phototrophically and anaerobically grown cultures of nondenitrifying strains of Rhodopseudomonas sphaeroides, Rhodopseudomonas palustris, and Rhodospirillum rubrum also possessed nitrous oxide reductase activity.     

H2 metabolism in the photosynthetic bacterium Rhodopseudomonas capsulata: production and utilization of H2 by resting cells.

Photoproduction of H2 and activation of H2 for CO2 reduction (photoreduction) by Rhodopseudomonas capsulata are catalyzed by different enzyme systems. Formation of H2 from organic compounds is mediated by nitrogenase and is nto inhibited by an atmosphere of 99% H2. Cells grown photoheterotrophically on C4 dicarboxylic acids (with glutamate as N source) evolve H2 from the C4 acids and also from lactate and pyruvate; cells grown on C3 carbon sources, however, are inactive with the C4 acids, presumably because they lack inducible transport systems. Ammonia is known to inhibit N2 fixation by photosynthetic bacteria, and it also effectively prevents photoproduction of H2; these effects are due to inhibition and, in part, inactivation of nitrogenase. Biosynthesis of the latter, as measured by both H2 production and acetylene reduction assays, is markedly increased when cells are grown at high light intensity; synthesis of the photoreduction system, on the other hand, is not appreciably influenced by light intensity during photoheterotrophic growth. The photoreduction activity of cells grown on lactate + glutamate (which contain active nitrogenase) is greatly activated by NH4+, but this effect is not observed in cells grown with NH4+ as N source (nitrogenase repressed) or in a Nif- mutant that is unable to produce H2. Lactate, malate, and succinate, which are readily used as growth substrates by R. capsulata and are excellent H donors for photoproduction of H2, abolish photoreduction activity. The physiological significances of this phenomenon and of the reciprocal regulatory effects of NH4+ on H2 production and photoreduction are discussed.     

L-Methionine-SR-sulfoximine as a probe for the role of glutamine synthetase in nitrogenase switch-off by ammonia and glutamine in Rhodopseudomonas palustris

Methionine sulfoximine (MSX), an irreversible inhibitor of glutamine synthetase of Rhodopseudomonas palustris restored nitrogenase activity to cells in which nitrogenase had been completely inhibited by ammonia switch-off. After addition of MSX, there was a lag period before nitrogenase activity was fully restored. During this lag, glutamine synthetase activity progressively decreased, and near the time of its complete inhibition, nitrogenase activity resumed. Nitrogenase switch-off by ammonia thus required active glutamine synthetase. Glutamine itself caused nitrogenase inhibition whose reversal by MSX depended on the relative ratio of MSX to glutamine. Unlike ammonia, glutamine inhibited nitrogenase under conditions where glutamine synthetase activity was absent. This indicates that glutamine is the effector molecule in nitrogenase switch-off, for instance by interacting with the enzymatic system for Fe protein inactivation. The effects of glutamine and MSX were also dependent on the culture age. Possible explanation for this and for the competitive effects are a common binding site within the regulatory apparatus for nitrogenase, or, in part, within a common transport system. Some observations with MSX were extended to Rhodopseudomonas capsulata and agreed with those in R. palustris.     

Glutamine as a feedback inhibitor of the Rhodopseudomonas sphaeroides nitrogenase system.

In whole cells of Rhodopseudomonas sphaeroides, nitrogen fixation, as measured by hydrogen production and acetylene reduction, was totally inhibited by micromolar concentrations of ammonia. This inhibition could not be duplicated by glutamate or glutamine alone. The inhibition by ammonia was abolished by methionine sulfoximine, a glutamine synthetase inhibitor. Inhibition by glutamine was complete in the presence of methionine sulfone, a preferential inhibitor of glutamate synthase, presumably by permitting a rise in the glutamine pool. The results indicated that the level of the glutamine pool controlled the activity of nitrogenase. None of these effects could be duplicated with cell-free nitrogenase, indicating there is probably a mediator which responds to the glutamine pool and inhibits nitrogenase, rather than glutamine itself being a direct inhibitor.     

Alterations in the phospholipid composition of Rhodopseudomonas sphaeroides and other bacteria induced by Tris.

Alterations in the phospholipid head group composition of most strains of Rhodopseudomonas sphaeroides, as well as Rhodopseudomonas capsulata and Paracoccus denitrificans, occurred when cells were grown in medium supplemented with Tris. Growth of R. sphaeroides M29-5 in Tris-supplemented medium resulted in the accumulation of N-acylphosphatidylserine (NAPS) to as much as 40% of the total whole-cell phospholipid, whereas NAPS represented approximately 28 an 33% of the total phospholipid when R. capsulata and P. denitrificans respectively, were grown in medium containing 20 mM Tris. The accumulation of NAPS occurred primarily at the expense of phosphatidylethanolamine in both whole cells and isolated membranes of R. sphaeroides and had no detectable effect on cell growth under either chemoheterotrophic or photoheterotrophic conditions. Yeast extract (0.1%) and Casamino Acids (1.0%) were found to be antagonistic to the Tris-induced (20 mM) alteration in the phospholipid composition of R. sphaeroides. The wild-type strains R. sphaeroides 2.4.1 and RS2 showed no alteration in their phospholipid composition when they were grown in medium supplemented with Tris. In all strains of Rhodospirillaceae tested, as well as in P. denitrificans, NAPS represented between 1.0 and 2.0% of the total phospholipid when cells were grown in the absence of Tris. [32P]orthophosphoric acid entered NAPS rapidly in strains of R. sphaeroides that do (strain M29-5) and do not (strain 2.4.1) accumulate this phospholipid in response to Tris. Our data indicate that the phospholipid head group composition of many Rhodospirillaceae strains, as well as P. denitrificans, is easily manipulated; thus, these bacteria may provide good model systems for studying the effects of these modifications on membrane structure and function in a relatively unperturbed physiological system.     

Regulation of nitrogenase A and R concentrations in Rhodopseudomonas capsulata by glutamine synthetase.

Nitrogen-starved purple non-sulphur bacteria have an active unregulated form of nitrogenase (nitrogenase A); however, the nitrogenase of a glutamine synthetase-negative mutant of Rhodopseudomonas capsulata, when nitrogen-starved, was predominantly inactive and required activation by Mn2+ and activating-factor protein. This regulatory form of nitrogenase has been called nitrogenase R. Treatment of wild-type cells (containing nitrogenase A) with methionine sulphoximine, an inhibitor of glutamine synthetase, converted the enzyme into nitrogenase R. Glutamine synthetase thus appears to control the intracellular concentrations of nitrogenase A and R and in this way regulates nitrogenase activity in the photosynthetic bacterium.     

Derepression of nitrogenase activity in glutamine auxotrophs of Rhodopseudomonas capsulata.

In contrast to wild-type cells, glutamine auxotrophs of the photosynthetic bacterium Rhodopseudomonas capsulata synthesize nitrogenase, produce H2 (catalyzed by nitrogenase), and continue to reduce dinitrogen to ammonia in the presence of exogenous NH4+. The glutamine synthetase activity of such mutants is less than 2% of that observed in the wild type. It appears that glutamine synthetase plays a significant role in regulation of nitrogenase synthesis in R. capsulata.     

Respiratory control determines respiration and nitrogenase activity of Rhizobium leguminosarum bacteroids.

The relationship between the O2 input rate into a suspension of Rhizobium leguminosarum bacteroids, the cellular ATP and ADP pools, and the whole-cell nitrogenase activity during L-malate oxidation has been studied. It was observed that inhibition of nitrogenase by excess O2 coincided with an increase of the cellular ATP/ADP ratio. When under this condition the protonophore carbonyl cyanide m-chlorophenylhydrazone (CCCP) was added, the cellular ATP/ADP ratio was lowered while nitrogenase regained activity. To explain these observations, the effects of nitrogenase activity and CCCP on the O2 consumption rate of R. leguminosarum bacteroids were determined. From 100 to 5 microM O2, a decline in the O2 consumption rate was observed to 50 to 70% of the maximal O2 consumption rate. A determination of the redox state of the cytochromes during an O2 consumption experiment indicated that at O2 concentrations above 5 microM, electron transport to the cytochromes was rate-limiting oxidation and not the reaction of reduced cytochromes with oxygen. The kinetic properties of the respiratory chain were determined from the deoxygenation of oxyglobins. In intact cells the maximal deoxygenation activity was stimulated by nitrogenase activity or CCCP. In isolated cytoplasmic membranes NADH oxidation was inhibited by respiratory control. The dehydrogenase activities of the respiratory chain were rate-limiting oxidation at O2 concentrations (if >300 nM. Below 300 nM the terminal oxidase system followed Michaelis-Menten kinetics (Km of 45 +/- 8 nM). We conclude that (i) respiration in R. leguminosarum bacteroids takes place via a respiratory chain terminating at a high-affinity oxidase system, (ii) the activity of the respiratory chain is inhibited by the proton motive force, and (iii) ATP hydrolysis by nitrogenase can partly relieve the inhibition of respiration by the proton motive force and thus stimulate respiration at nanomolar concentrations of O2.     

Fermentation of pyruvate by 7 species of phototrophic purple bacteria]

The dark, anaerobic fermentation of pyruvate under growth conditions was examined with the following species of phototrophic purple bacteria: Rhodospirillum rubrum strains Ha and S1, Rhodopseudomonas gelatinosa strain 2150, Rhodopseudomonas acidophila strain 7050, Rhodopseudomonas palustris strain ATCC 17001, Rhodopseudomonas capsulata strains Kb1 and 6950, Rhodopseudomonas sphaeroides strain ATCC 17023, and Chromatium vinosum strain D. Fermentation balances were established for all experiments. Under fermentative conditions cell protein and dry weight increased only slightly, if at all. The species differed considerably in their fermentative activity; R. rubrum and R. gelatinosa exhibited the highest rates (2-8 mumoles pyruvate/mg protein-h). R. acidophila and R. capsulata showed an intermediate fermentation rate (0.4--2.0 mumoles pyruvate/mg protein-h), while the other strains tested fermented at quite low rates (0.2-0.4 mumoles pyruvate/mg protein-h). The extremes of fermentation times were from 30-380 hours. Based on the products of fermentation which were formed in addition to acetate, formate, and CO2, the species can be grouped as follows: a) R. rubrum, R. gelatinosa, and R. sphaeroides additionally form propionate. b) R. gelatinosa, R. palustris, R. capsulata, R. sphaeroides, and C. vinosum additionally form lactate. R. palustris also produces butyrate. c) R. acidophila and R. capsulata additionally form much 2,3-butanediol, acetoin, and diacetyl. Small amounts of acetoin were formed by the rest of the strains. A comparison of the fermentation of pyruvate by normal and starved cells (4 days in the light without a carbon source) of R. rubrum and R. gelatinosa shows that the latter ferment more slowly and produce less acetate and formate, but more propionate or lactate. The fermentation of pyruvate by R. rubrum was also studied in cultures in which the pH fell (7.2--6.6). Compared with the fermentation at neutral pH (7.3, 7.4), the following differences were found: a slower fermentation rate, an increased production of dry weight, an increased formation of propionate, but a reduced formation of acetate and a very low production of formate.     

Differential sensitivity of guanylyl cyclase and mitochondrial respiration to nitric oxide measured using clamped concentrations

Nitric oxide (NO) signal transduction may involve at least two targets: the guanylyl cyclase-coupled NO receptor (NO(GC)R), which catalyzes cGMP formation, and cytochrome c oxidase, which is responsible for mitochondrial O(2) consumption and which is inhibited by NO in competition with O(2). Current evidence indicates that the two targets may be similarly sensitive to NO, but quantitative comparison has been difficult because of an inability to administer NO in known, constant concentrations. We addressed this deficiency and found that purified NO(GC)R was about 100-fold more sensitive to NO than reported previously, 50% of maximal activity requiring only 4 nm NO. Conversely, at physiological O(2) concentrations (20-30 microM), mitochondrial respiration was 2-10-fold less sensitive to NO than estimated beforehand. The two concentration-response curves showed minimal overlap. Accordingly, an NO concentration maximally active on the NO(GC)R (20 nm) inhibited respiration only when the O(2) concentration was pathologically low (50% inhibition at 5 microM O(2)). Studies on brain slices under conditions of maximal stimulation of endogenous NO synthesis suggested that the local NO concentration did not rise above 4 nm. It is concluded that under physiological conditions, at least in brain, NO is constrained to target the NO(GC)R without inhibiting mitochondrial respiration.     

Influence of pH, O2, and temperature on the absorption properties of the secondary light-harvesting antenna in members of the family Rhodospirillaceae

In some Rhodospirillaceae, the primary light-harvesting (LH I) antenna absorbs near-infrared light around 870 nm, whereas LH II (holochrome B800-860) has a major absorption band between 850 and 860 nm (B860) and a minor absorbancy around 800 nm (B800). Results show that, unlike LH I, holochrome B800-860 (LH II) exhibits unstable light absorption properties in whole cells. This was observed in Rhodopseudomonas capsulata grown anaerobically in light in weakly buffered carbohydrate medium; cultures lost both carotenoid-dependent brown-yellow pigmentation and LH II absorbancy. The whole cell spectrophotometric changes were attributed to mild acid conditions generated during sugar metabolism. LH II absorbancy was also destroyed in both R. capsulata and Rhodopseudomonas gelatinosa when cultures growing at neutral pH were acidified to a pH value around 5.0 with HCl. In contrast, during the same time period of exposure to pH 5.0, only a 50% decrease in Rhodopseudomonas sphaeroides LH II B800 absorbancy was measured. At neutral pH, LH II absorbancy in suspensions of nongrowing Rhodopseudomonas spp. was also sensitive to O2 exposure and to incubation at 30 to 40 degrees C. During treatment with O2, the rate of LH II B800 absorption decrease in R. gelatinosa and R. sphaeroides was 60 and 40% per h, respectively, compared with their absorbancy maximum around 860 nm. Both 860-nm absorbancy and the total bacteriochlorophyll content of the cells remained unchanged. On the other hand, no significant decrease in B800 if LH II in R. capsulata occurred during O2 exposure, but a 20% absorption decay rate per h of B800 was observed in cells incubated anaerobically at 40 degrees C. These B800 LH II spectral changes Rhodopseudomonas spp. were prevented by maintaining cells at neutral pH and at 10 degrees C. The near-infrared absorption spectrum of Rhodospirillum rubrum, which does not form LH II, was not significantly influenced by these different pH, aerobic, or temperature conditions.     

D-erythrose supports nitrogenase activity in isolated Anabaena sp. strain 7120 heterocysts.

Among organic compounds tested for their ability to support nitrogenase activity in isolated heterocysts of Anabaena sp. strain 7120 under argon, D-erythrose (5 mM) was unique in supporting acetylene reduction at 10 times the control rates. Higher concentrations of D-erythrose exhibited substrate inhibition. At 50 kPa of H2, all concentrations of D-erythrose inhibited H2-supported acetylene reduction. The effects of D-erythrose on nitrogenase activity were explored. Erythrose enhanced 15N2 incorporation by heterocysts, but NADP+ did not enhance erythrose-supported acetylene reduction. H2 protected nitrogenase from O2 inactivation, but erythrose did not; erythrose did not counter protection by H2. Tests with inhibitors of electron transport showed that erythrose-supported acetylene reduction requires electron flow through ferredoxin, a b-type cytochrome, and a 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone-sensitive transfer agent whose electron flow is not mediated through the plastoquinone and Rieske iron protein.     

Nitrogen fixation and ammonia switch-off in the photosynthetic bacterium Rhodopseudomonas viridis.

Rhodopseudomonas viridis ATCC 19567 grows by means of nitrogen fixation in yeast extract-N2 or nitrogen-free medium when sparged with 5% CO2 and 95% N2 in the light at 30 degrees C. Acetylene reduction assays for nitrogenase activity revealed an initially high level of activity during early-logarithmic growth phase, a lower plateau during mid- to late-logarithmic phase, and a dramatic reduction of activity at the beginning of the stationary phase. When viewed by electron microscopy, nitrogen-fixing R. viridis cells appeared to be morphologically and ultrastructurally similar to cells grown on nitrogen-rich media. Whole cells prepared under reducing conditions in the dark for electron spin resonance spectroscopy yielded g4.26 and g3.66 signals characteristic of the molybdenum-iron protein of nitrogenase. During growth on N2 in the absence of fixed-nitrogen sources, the nitrogenase activity of R. viridis measured by acetylene reduction stopped rapidly in response to the addition of NH4Cl as has been observed in other Rhodospirillaceae. However, unlike the nitrogenase of Rhodopseudomonas palustris or Rhodospirillum rubrum, which recover from this treatment within 40 min, the nitrogenase activity of R. viridis was not detectable for nearly 4 h.     

Photoproduction of h(2) from cellulose by an anaerobic bacterial coculture

Cellulomonas sp. strain ATCC 21399 is a facultatively anaerobic, cellulose-degrading microorganism that does not evolve hydrogen but produces organic acids during cellulose fermentation. Rhodopseudomonas capsulata cannot utilize cellulose, but grows photoheterotrophically under anaerobic conditions on organic acids or sugars. This report describes an anaerobic coculture of the Cellulomonas strain with wild-type R. capsulata or a mutant strain lacking uptake hydrogenase, which photoevolves molecular hydrogen by the nitrogenase system of R. capsulata with cellulose as the sole carbon source. In coculture, the hydrogenase-negative mutant produced 4.6 to 6.2 mol of H(2) per mol of glucose equivalent, compared with 1.2 to 4.3 mol for the wild type.     

Immunochemical relationship of the major outer membrane protein of Rhodopseudomonas sphaeroides 2.4.1 to proteins of other photosynthetic bacteria.

Immunoblots of sodium dodecyl sulfate-polyacrylamide gels derived from outer membrane preparations of various strains of Rhodopseudomonas sphaeroides revealed polypeptides which cross-reacted with antibody directed against the major outer membrane protein of R. sphaeroides 2.4.1. Immunochemical quantitation of the major outer membrane protein of strain 2.4.1 showed approximately 5.5 x 10(4) molecules per cell whether cells were grown chemoheterotrophically or photoheterotrophically. Rhodospirillum rubrum outer membranes contained a cross-reactive protein, whereas the outer membranes derived from Rhodopseudomonas capsulata and Paracoccus denitrificans showed no cross-reaction with the antibody prepared against the major outer membrane protein from R. sphaeroides 2.4.1.     

Physiological sources of reductant for nitrogen fixation activity in Nostoc sp. strain UCD 7801 in symbiotic association with Anthoceros punctatus.

Pure cultures of the symbiotic cyanobacterium-bryophyte association with Anthoceros punctatus were reconstituted by using Nostoc sp. strain UCD 7801 or its 3-(3,4-dichlorophenol)-1,1-dimethylurea (DCMU)-resistant mutant strain, UCD 218. The cultures were grown under high light intensity with CO2 as the sole carbon source and then incubated in the dark to deplete endogenous reductant pools before measurements of nitrogenase activities (acetylene reduction). High rates of light-dependent acetylene reduction were obtained both before starvation in the dark and after recovery from starvation, regardless of which of the two Nostoc strains was reconstituted in the association. Rates of acetylene reduction by symbiotic tissue with the wild-type Nostoc strain decreased 99 and 96% after 28 h of incubation in the dark and after reexposure to light in the presence of 5 microM DCMU, respectively. Supplementation of the medium with glucose restored nitrogenase activity in the dark to a rate that was 64% of the illuminated rate. In the light and in the presence of 5 microM DCMU, acetylene reduction could be restored to 91% of the uninhibited rate by the exogenous presence of various carbohydrates. The rate of acetylene reduction in the presence of DCMU was 34% of the uninhibited rate of tissue in association with the DCMU-resistant strain UCD 218. This result implies that photosynthates produced immediately by the cyanobacterium can supply at least one-third of the reductant required for nitrogenase activity on a short-term basis in the symbiotic association. However, high steady-state rates of nitrogenase activity by symbiotic Nostoc strains appear to depend on endogenous carbohydrate reserves, which are presumably supplied as photosynthate from both A. punctatus tissue and the Nostoc strain.     

Genetic and physical map of the structural genes (nifH,D,K) coding for the nitrogenase complex of Rhodopseudomonas capsulata.

Functional genes coding for the structural components of the nitrogenase complex (nifH,D,K) have been cloned on an 11.8-kilobase-pair HindIII fragment of DNA from the photosynthetic bacterium Rhodopseudomonas capsulata. The genes were physically mapped by hybridization of individual cloned nif genes from Klebsiella pneumoniae and Anabaena sp. strain 7120 to Southern blots of HindIII digests of the cloned R. capsulata fragment, after introduction of HindIII sites into the latter at specified locations by insertion of Tn5. Plasmids with the 11.8-kilobase-pair HindIII fragment containing the Tn5 insertions were also used for complementation tests with chromosomal Nif- mutations and for the generation of subfragments to locate those mutations by marker rescue. The R. capsulata nifH,D,K genes comprise a single unit of expression, with the same organization and polarity as found in K. pneumoniae. However, the R. capsulata nifH,D,K fragment did not complement Nif- point mutations in the corresponding Klebsiella genes, and the Klebsiella nif genes did not function in R. capsulata.