Alteration ofthe Fe Protein Nitrogenase Oxygen the ... · atmospheric dinitrogen. The other component, the Fe pro-tein(componentII, dinitrogenasereductase), consistsoftwo identical

Vol. 169, No. 6JOURNAL OF BACTERIOLOGY, June 1987, p. 2537-25420021-9193/87/062537-06$02.00/0Copyright © 1987, American Society for Microbiology

Alteration of the Fe Protein of Nitrogenase by Oxygen in theCyanobacterium Anabaena sp. Strain CA

RUSSELL L. SMITH,'t CHASE VAN BAALEN,1t AND F. ROBERT TABITA2*University of Texas Marine Science Institute, Port Aransas, Texas 78373,1 and The Department of Microbiology and

Center for Applied Microbiology, The University of Texas at Austin, Austin, Texas 78712-10952Received 15 December 1986/Accepted 19 March 1987

Changes in protein composition were noted when heterocysts ofAnabaena sp. strain CA were isolated fromfilaments grown in 1% CO-99% N2 and subsequently exposed to oxygen. Immunospecific Western blotanalysis showed that the Fe protein of nitrogenase is altered. In cells grown under microaerobic conditions, theFe protein was found in a form with an apparent molecular weight of 30,000. Exposure to oxygen caused a shiftin the migration of this polypeptide to a position corresponding to an apparent molecular weight of 31,500. Thismodification was reversible upon removal of oxygen from the culture. Chloramphenicol did not inhibit thealteration in either direction. Suppression by ammonium nitrate of the recovery of mntrogenase activity from theeffects of oxygen did not prevent the alteration of the protein. Other inhibitors of nitrogenase activity,(metronidazole, carbonyl cyanide nm-chlorophenylhydrazone, and phenazine methosulfate) were tested for theireffect on Fe protein modification. Alteration of the Fe protein may relate to the protection of nitrogenase fromthe deleterious effects of oxygen.

Nitrogenase is a multicomponent enzyme complex whichcatalyzes the biological reduction of N2 to NH4' (18, 33).The enzyme exists as a complex of two component proteinswhich vary only slightly from organism to organism. Thelarger of the two components, the Mo-Fe protein (compo-nent I, dinitrogenase), is arranged in an a2132 configurationwith a molecular weight around 220,000 (Mra = 56,000, MrI3= 54,000) and contains the actual site for the reduction ofatmospheric dinitrogen. The other component, the Fe pro-tein (component II, dinitrogenase reductase), consists oftwoidentical subunits with a molecular weight of about 65,000(subunit Mr = 32,500) (4). It is the Fe protein which supplieselectrons in a stepwise fashion to the Mo-Fe protein for N2reduction (13).

Regardless of the source, whether it be from anaerobicClostridium species or from aerobic organisms such asAnabaena species, the enzymes of the nitrogenase complexare exceedingly oxygen labile (25). While both the Mo-Feprotein and the Fe protein are damaged by exposure tooxygen, it is the Fe protein which is the more labile of thetwo. In Klebsiella pneumoniae and Azotobacterchroococcum, for example, the half-life for the donation ofelectrons to the Mo-Fe protein in air is on the order of 45 s(25). Inactivation of the Fe protein is thought to be due to theinability to accept electrons for subsequent donation to theMo-Fe protein (27) and may be dependent on the redox stateof the Fe-S clusters within the protein (4).

In the purple nonsulfur photosynthetic bacteria, the Feprotein also serves as the primary regulator of nitrogenaseactivity at the enzymatic level as it is rapidly, but reversibly,modified in response to various inhibitors of nitrogenaseactivity, including oxygen, NH4', and darkness (3, 5, 6, 8, 9,16). It has been shown that this rapid switch off of activity isthe result of covalent ADP-ribosylation of one of the

* Corresponding author.t Present address: D6partement de Recherche Fondamentale-

BM, Centre d'ttudes Nucleaires de Grenoble, 85-X, 38041,Grenoble Cedex, France.

t Deceased 20 January 1986.

subunits of the Fe protein (22). The reactivation of themodified protein requires an activating enzyme (6, 26), andthe presence of an inactivating enzyme has also recentlybeen proposed (15).Although nitrogenase activity in the cyanobacteria re-

sponds to many of the same nitrogenase inhibitors (2, 23) asthose of the photosynthetic bacteria, the effect of thesecompounds on the cyanobacterial system differs markedly.For example, no rapid switch off of nitrogenase is seen uponthe addition of fixed nitrogen sources to diazotrophicallygrown cultures of Anabaena species (23, 32), and it is onlyafter a relatively long period in the dark that significant levelsof nitrogenase activity are lost (24). These observationsimply that control of cyanobacterial nitrogenase activity isnot by the same mechanism as seen in the purple nonsulfurphotosynthetic bacteria (32).

This work reports that alteration of the Fe protein in thecyanobacterium Anabaena sp. strain CA is apparent, al-though this modification does not appear to lead to rapidinactivation of nitrogenase activity in vivo. The possiblefunction of the observed modification is discussed in termsof the physiological response of this organism to variouscompounds affecting nitrogenase activity, with special em-phasis on oxygen.

MATERIALS AND METHODSOrganism. The organism used for these studies was

Anabaena sp. strain CA (ATCC 33047), a filamentous,heterocyst-forming cyanobacterium (29).Growth conditions. Cultures were routinely grown in

Pyrex culture tubes (22 by 175 mm) in 20 ml of mediumASP-2 (31) with an NaCl content of 5 g/liter. When required,NH4NO3 was added to a final concentration of 10 mM. Thecultures were grown in either 1% CO2 in air or 1% C02-99%N2 at 39°C. For growth of cultures at low oxygen tensions,culture tubes were sealed with a rubber stopper throughwhich a cotton-plugged, 15-cm Pasteur pipette and ventila-tion tube had been inserted. Cultures were bubbled with 1%C02-99% N2 at a rate of 20 to 25 ml/min. The growth bathwas illuminated with four F36T12/D/HO flourescent lamps

2537

on March 20, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

2538 SMITH ET AL.

(Westinghouse, Bloomfield, N.J.) positioned on either side,12 cm from the center, of the growth tube, at an averageintensity of 300 microeinsteins perm2 per s.

Heterocyst isolation. Heterocysts were isolated essentiallyas described before (28). A suspension of 20 ml of cells at adensity of 0.09 to 0.11 mg (dry weight) per ml was washedonce with medium ASP-2 with the total concentrations ofKCI raised to 30 mM and NaCl raised to 0.37 M (assaymedium). The cells were resuspended in 5 ml of assaymedium containing 1.0 mg of lysozyme per ml, sparged for 5min with 1% CO02-99% N2 in vented 15-ml Corex centrifugetubes stoppered with a rubber plug, and placed in a39°Cshaking bath in the dark. After 20 min, the suspension wascentrifuged for 5 min at 3,000 rpm, and the cells wereresuspended in 5 ml of pregassed assay medium and subse-quently sonicated. Sonication was performed under a streamof 1% CO0-99% N2 or 1% CO2 in air (for cultures exposed tohigh levels of oxygen) for three 10-s intervals on a modelW-10 sonicator equipped with a standard 4.5-inch (11.4-cm)titanium probe (Heat Systems Ultrasonics, Inc., Plainview,N.Y.) set at full power. The sonicated suspension wascentrifuged for 5 min at 2,000 rpm, resuspended in 5 ml ofassay medium, and centrifuged again for an additional 5 min.All centrifugations were done at room temperature in anInternational Equipment Co. model UV Universal centrifuge(Curtin Scientific, Boston, Mass.) equipped with a swingingbucket rotor. This procedure yielded metabolically activeheterocysts (11, 28).

Electrophoretic analysis of proteins from isolatedheterocysts. Heterocysts were suspended in 200,ul of diges-tion mix (62.5 mM Tris hydrochloride [pH 6.8], 2% sodiumdodecyl sulfate [SDS], 10% glycerol, 5%-mercaptoethanol,0.001% bromphenol blue) (17). Samples were boiled for 1min, centrifuged at 3,000 rpm for 5 min in conical centrifugetubes, and then loaded onto gels. SDS- polyacrylamide gelelectrophoresis (PAGE) was done by the method ofLugtenberg et al. (17). Proteins were visualized by stainingovernight with 0.4% Coomassie blue R-250 in 25% isopropylalcohol-1o acetic acid, followed by destaining in 10%methanol-10% acetic acid.

Itumunoblot (Western) identffication of proteins. Proteinextracts were separated by SDS-PAGE. Fractionated pro-teins were then transferred electrophoretically to nitrocellu-lose filters (Millipore Corp., Bedford, Mass.) in 25 mM Tris(pH 8.3)-192 mM glycine-20% methanol (30) with a HoefferTransphor cell (Hoeffer Instruments Inc., San Francisco,Calif.). Proteins were identified by the method of Blake et al.(1) with goat anti-rabbit antibody-alkaline phosphatase con-jugate. Primary antibody was a 1:200 dilution of a crudeserum preparation raised against purified Rhodospirillumrubrum Fe protein (kindly provided by P. W, Ludden,University of Wisconsin-Madison).

Inhibitors of nitrogenase activity. Various inhibitors ofnitrogenase activity were added to cultures to ascertain theeffect of enzyme inactivation on the protein composition ofthe isolated heterocysts. Phenazine methosulphate (PMS),carbonyl cyanide m-chlorophenylhydrazone (CCCP) (9), andmetronidazole (MET) (7, 10) were added directly to culturesgrown in 1% C02-99% N2 to final concentrations of 0.1, 30,and 2 mM, respectively.

Chemicals and gases. Biochemicals were purchased fromSigma Chemical Co. (St. Louis, Mo.) and were of reagentgrade quality or better. Acrylamide was purchased fromServa Fine Biochemicals, Inc. (Garden City Park, N.Y.),and ammonium persulfate and N,N,N',N'-tetramethyl-ethylenediamine (TEMED) were purchased from Bio-Rad

1 2 3. .. , _ ,~~~WI..

FIG. 1. Electrophoretic analysis of crude extracts of heterocystsisolated from cultures of Anabaena sp. strain CA grown in 1%C02-999o N2. After treatment of cultures as indicated below,heterocysts were isolated and prepared for electrophoretic analysisas described in Materials-and Methods. Extracts were obtained fromcultures grown in 1% C02-99% N2 with no further treatment (lane1); cultures switched to 1% C02-99%0 02 for 3 h (lane 2); andcultures switched to 1% CO0-99% 02 for 3 h with chloramphenicoladded just before oxygen treatment (lane 3). The arrows indicate theareas of modification.

Laboratories (Richmond, Calif.). Gases were obtained fromBig Three Industries (Houston, Tex.).

RESULTSAnalysis of proteins from isolated heterocysts ofcultures

exposed to high oxygen tensions. The effect of high oxygentensions on the composition of the proteins from isolatedheterocysts is shown in Fig. 1. To maximize the effect ofoxygen on nitrogenase, cells were grown in 1% COT-99% N2before oxygen treatment and heterocyst isolation. An obvi-ous change was the increase in the levels of a protein with anapparent molecular size of 31.5 kilodaltons (kDa) and thesimultaneous decrease of a protein at 30 kDa after oxygentreatment. Addition of chloramphenicol (20,ug/ml) concom-itant with exposure to oxygen did not prevent the observedchange. Other changes in protein composition may also benoted.

Identification of altered protein as the Fe protein of nitro-genase. Since the predominant and most obvious change inthe protein profile correlated with the migration position ofthe Fe protein of nitrogenase, subsequent studies weredirected at this protein. Using antibodies raised against R.rubrum Fe protein, crude extracts of heterocysts isolatedfrom cultures exposed to high oxygen tensions were exam-ined by Western immunoblot analysis (Fig. 2). The resultsclearly indicated that there was a cross-reacting protein thatmigrated at 31.5 kDa in extracts of heterocysts preparedfrom filaments exposed to oxygen. This cross-reactive pro-tein is normally found as a 30-kDa protein in heterocystextracts of microaerobically grown filaments. Very littlenonspecific cross-reactivity with other proteins on the nitro-cellulose filter was noted. Owing to the conservation in thestructural proteins of the nitrogenase complex (18), it wasassumed that the Fe protein of the Anabaena sp. strain CAnitrogenase complex was modified or altered when filamentswere exposed to oxygen.

J.BACTERIOL.

on March 20, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ALTERATION OF NITROGENASE Fe PROTEIN 2539

1 2 3 4

4- Fe

FIG. 2. Comparison of stained proteins and Western im-munoblot analysis of crude extracts of heterocysts. Lanes 1 to 3show a Western blot of crude extracts of heterocysts isolated fromcultures grown in 1% C02-99% N2 with no further treatment (lane 1)or exposed to 1% C02-99% 02 for 1 h (lanes 2 and 3). Lane 4 showsthe staining pattern of a crude extract obtained from heterocysts ofcultures grown in 1% CO_2-99% N2 after exposure to 1% C02-99%02 for 1 h. Fe refers to the iron protein of nitrogenase.

Reversibility of Fe protein alteration. To determine moreprecisely the relationship of oxygen to the modification ofthe Fe protein of nitrogenase, cells were grown in 1% CO2 inair and then switched to microaerobic conditions (1%C02-99% N2). Heterocysts were isolated from these cul-tures, and the proteins were examined by SDS-PAGE. Uponremoval of oxygen from the culture, modification of the Feprotein resulted in the appearance of the 30-kDa form of theprotein (Fig. 3). Similar to the switch from the lower-molecular-weight form to the higher-molecular-weight format high levels of oxygen (Fig. 1 and 2), this alteration was notsensitive to the addition of chloramphenicol. It is alsoapparent that the high-molecular-weight formn of the ironprotein is present in cells grown at air levels of oxygen.

Effect of fixed nitroge'n on proteins of isolated heterocysts.Recovery of nitrogenase activity from oxygen inactivationwas prevented by the addition of NH4NO3 (Table 1) (R. L.Smith and F. R. Tabita, -manuscript in preparation). If thealteration of the Fe protein played a role in the protection ofnitrogenase from the effects of oxygen, perhaps the inhibi-tion of recovery from oxygen inactivation'by NH4NO3 mightbe reflected in a lack of modification of the Fe protein. Thiswas not the case (Fig. 4 and 5). The addition of NH4NO3 todiazotrophically growing cultures caused a very slow declinein nitrogenase activity. Over 'a period of 3 h, about 50% ofthe original activity remained owing to the growth of cellsthat do not synthesize nitrogenase (Table 1). This did notresult in a shift or disappearance of the protein migrating at30 kDa (Fig. 4). It was only upon treatment of cultures withoxygen (Fig. 4 and 5) that the transition occurred. Moreover,the addition of NH4NO3 concomitant with oxygen did notprevent the modification of the protein (Fig. 5). Thus, thelack of recovery of nitrogenase activity from oxygen inacti-

vation in the presence of NH4NO3 does not seem to be dueto the inability of the Fe protein to be modified.

It was also noted that the observed modification of the Feprotein was detectable within a very short time after expo-sure of cultures to oxygen (Fig. 4). In this case, an alteredprotein pattern was detected after only 30 min of exposure tooxygen.Table 1 summarizes the various conditions under which

the altered Fe protein was observed.During this work we noted that a protein of 31.5 kDa was

always observed even when the organism was grown onNH4NO3 to repress nitrogenase synthesis. Since cross-reactivity could only be detected on Western blots whenfilaments were exposed to oxygen, this endogenous31.5-kDa protein may be some other comigrating protein.

Inhibitors of nitrogenase activity and their effect on proteincomposition of isolated heterocysts. Since the Fe protein ofother nitrogen-fixing organisms is modified as a result of theeffect of inhibitors other than oxygen (9), we determined theability of these inhibitors to cause similar changes on theheterocyst proteins. PMS, a mid-range electron acceptor (Eh= +0.080 V), inhibited acetylene reduction by 88%, whileCCCP, an effective uncoupler of phosphorylation, and MET,a low-range electron acceptor (Eh = -0.325 V), each inhib-ited nitrogenase activity completely within 5 min after theiraddition. Neither CCCP- nor PMS-treated cultures exhibitedthe characteristic transition of the Fe protein (Fig. 6).MET-treated cultures, on the other hand, showed the shift inthe 30-kDa polypeptide. However, a general deterioration inthe resolution of the proteins in the gel was always apparentwith extracts from MET-treated cultures (perhaps owing tofree-anion generation), resulting in a general oxidation ofcellular components (14).

DISCUSSIONThis is the first demonstration of the modification or

alteration of the Fe protein of nitrogenase in cyanobacteria.

FIG. 3. Electrophoretic analysis of crude extracts of heterocystsfrom cultures grown in 1% CO2 in air. Extracts were obtained fromcultures originally grown in 1% CO2 in air and then swwitched to 1%C02-99% N2 for 1 h with the addition of chloramphenicol just beforetransition to microaerobic conditions (lane 1); cultures switched to1% C02-99% N2 for 1 h (lane 2); or cultures not further treated (lane3). The arrows indicate the areas of modification.

VOL. 1699 1987

on March 20, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

2540 SMITH ET AL.

TABLE 1. Summary of conditions influencing the transition of the Fe protein

Growth conditiona Transitionb Nitrogenase % Activity Mr (103) ofactivityc remaining Fe proteind1% C0z-99,% N2 None 1.9 100 30.0

1% C02-99%002 (0.5 h) 0.1 5.2 31.51% C02-99%/ 02 (3.0 h) 1.3 68.4 31.51% C02-99% 02 + CAM (3.0 h) 0.1 5.2 31.51% CO2 in air (0.5 h) 1.2 63.2 31.51% C02 in air (3.0 h) 1.6 84.2 31.5NH4NO3 (0.5 h) 1.7 89.5 30.0NH4NO3 (3.0 h) 1.0 52.6 30.0NH4NO3 + 1% C02-99%0 02 (3.0 h) 0.1 5.2 31.5

1% CO2 in air None 1.8 100 31.51% C02-99% N2 (0.5 h) 1.8 100 30.01% C02-99% N2 (3.0 h) 1.8 100 30.01% C02-99% N2 + CAM NDe 30.0

a Cultures were grown in either 1% C02-990o N2 or 1% CO2 in air as described in Materials and Methods.b Cultures were switched from the original growth conditions to those specified (as described in the text). CAM, chloramphenicol.c Nitrogenase activity is expressed as micromoles of C2H4 produced per milligram (dry weight) per hour. Ethylene production by whole filaments was measured

as described previously (28). Cultures under microaerophilic conditions during transition were assayed in 10o C2H2-90% N2. Others were assayed in 10% C2H2 inair.

d The apparent molecular weight of the Fe protein as observed by SDS-PAGE.eND, Not determined.

One of the major factors which led to the finding that the Feprotein was altered derived from the ability to isolate meta-bolically competent heterocysts from cultures of Anabaenasp. strain CA. Recently, Kumer et al. (11, 12) and Smith etal. (28) have addressed several aspects of the physiology ofthese cells including pigment composition, carbon fixationcapacity, hydrogen production, nitrogen fixation, and oxy-gen metabolism. The ability to follow changes in the proteincomposition of crude extracts from isolated heterocysts bySDS-PAGE, coupled with the knowledge obtained earlier,

FIG 4. Electrophoretic analysis of crude extracts of heterocystsisolated from cultures grown in 1% C02-99% N2. Extracts wereobtained from cultures grown in 1% CO2-99%o N2 with no furthertreatment (lane 2); cultures supplemented with 10 mM NH4NO3 for30 min (lane 3); cultures switched to 1% C02-99%o 02 for 30 min(lane 4); cultures supplemented with 10 mM NH4NO3 for 3 h (lane5); and cultures switched to 1% C02-99%o 02 for 3 h (lane 6). Lanes1 and 7 show standard molecular size markers of bovine serumalbumin (66.2 kDa), DNase (31.0 kDa), and cytochrome c (12.3kDa).

allowed certain conclusions to be drawn concerning theeffects of oxygen on Fe protein modification in particular,and on the activity of the nitrogenase complex in general.Under conditions of low oxygen tension, almost all the Feprotein was in the 30-kDa form. After exposure of thecultures to hyperbaric levels of oxygen (21), or even whencells, were grown in 1% CO2 in air, a change in the proteinpattern at 30 and 31.5 kDa was apparent. Our results clearlyindicate that the observed modification was the result of ashift in the migration of the polypeptide, rather than thedisappearance of the 30-kDa protein. The argument that theobserved change in the protein composition of crude ex-

SS S

FIG. 5. Electrophoretic analysis of crude extracts of heterocystsisolated from cultures grown in 1% C02--99% N2. Extracts are fromcultures grown in 1% C02.-99% N2 and: supplemented withNH4NO3 in the presence of 1% C02-999o 02 for 3 h (lane 1); bubbledwith 1% CO02-99%o N2 and supplemented with NH4NO3 for 3 h (lane2); bubbled with 1% C02-99% N2 and supplemented with NH4NO3for 30 min (lane 3); or bubbled with 1% C02--99%o N2 with no furthertreatment (lane 4). Arrowheads show areas of modification.

J. BACTERIOL.

on March 20, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ALTERATION OF NITROGENASE Fe PROTEIN 2541

FIG. 6. Electrophoretic analysis of crude extracts of heterocystsfrom cultures grown in 1% CO,-99% N2 and treated with inhibitors.Extracts were obtained from cultures grown in 1% C02-99% N2with no further treatment (lane 1); cultures switched to 1% CO,-99%02 for 3 h (lane 2); cultures treated with 30 ,uM CCCP for 1 h (lane3); cultures treated with 0.1 mM PMS for 2 h (lane 4); and culturestreated with 2.0 mM MET for 1 h (lane 5). Arrowheads show areas

of modification.

tracts on SDS gels did not involve the destruction of one

polypeptide followed by the de novo synthesis of anotherwas also supported by at least two other observations. First,the appearance of the band at 31.5 kDa was very rapid, beingdetectable within a few minutes after exposure to oxygen. Itis highly unlikely that a newly synthesized protein of 31.5kDa could be expressed in such a short period of time.Second, the addition of chloramphenicol did not prevent theappearance of the high-molecular-weight form.

Like the Fe protein of the purple photosynthetic bacteria,modification of the Fe protein in this cyanobacterium was

freely reversible upon removal of the substance whichoriginally caused the alteration, in this case, oxygen. How-ever, unlike the system found in the photosynthetic bacteria,alteration of the Fe protein did not cause inactivation of thenitrogenase system. In cultures which are routinely grown inair and express normal levels of nitrogenase activity, it is themodified form of the protein which is consistently found.Even more convincing is the fact that the protein is found inthe modified 31.5-kDa form after cells recover from inacti-vation of nitrogenase activity upon exposure to hyperbariclevels of 02. Covalent modification of the Fe protein in R.rubrum results in the ADP-ribosylation of only one of thesubunits of the Fe protein (22). This results in inactivation ofnitrogenase and an equal distribution of modified andunmodified forms of the protein on polyacrylamide gels.Clearly, as shown in Table 1 and Fig. 2, such was not thecase with this cyanobacterium.

Exactly what the nature of the modification or alteration ofthe Fe protein might be is the subject of current studies.However, experiments designed to determine the cause ofthe change in the apparent molecular weight of this proteinhave been inconclusive thus far (including attempts to labelthe Fe protein with 32P during the transition). It is possiblethat the observed alteration of the Fe protein is a

nonenzymatic process. That is, the transitions seen on gelsare simply a consequence of a change in the overall netcharge of the Fe protein which could alter its migration, evenin SDS gels (19). This idea is supported by the proteinpattern observed in MET-treated cell suspensions, in whichhighly oxidizing conditions presumably exist (14) which

could conceivably lead to an alteration of the protein oractivation of an Fe protein-modifying system. On the otherhand, if the transition was caused by some enzyme-catalyzed modification (similar to that of the activating-inactivating system of the photosynthetic bacteria), then themodifying enzyme(s) must be present at levels sufficient tocatalyze the observed modification before the addition orremoval of oxygen, since de novo protein synthesis was notrequired. If such a system exists, perhaps the modifyingenzyme(s) is synthesized concomitant with nitrogenasederepression. Nitrogenase activity, however, did not re-cover from oxygen inactivation in the presence of NH4NO3(Table 1). Since NH4NO3 controls the synthesis of nitroge-nase (but not its inactivation [Table 1]) and since Fe proteinmodification does occur in the presence of 02 plus NH4NO3(Fig. 5), then it would appear that whatever causes themodification of the Fe protein is regulated separately fromthe nitrogenase structural genes.

In summary, these results point to the conclusion that thealteration of the Fe protein observed here is in some waynecessary for the function of nitrogenase in the presence ofoxygen. If indeed it is, then for an organism which isconstantly exposed to its own photosynthetically generatedoxygen, and may also experience supersaturating oxygentensions during diazotrophic growth (20), a system for theprotection and retention of nitrogenase activity which is notregulated in response to nitrogen availability, but rather tofluctuating levels of oxygen, would be most attractive.

ACKNOWLEDGMENTSThis work was supported in part by grant 83-CRCR-1-1286 from

the Department of Agriculture. R.L.S. was supported by the E. J.Lund Memorial Fellowship for Graduate Research from the Univer-sity of Texas at Austin.We thank Paul W. Ludden for the antiserum to R. rubrum Fe

protein.

LITERATURE CITED1. Blake, M. S., K. H. Johnston, G. S. Russel-Jones, and E. C.

Gotschlic. 1984. A rapid sensitive method for detection ofalkaline-phosphatase conjugated anti-antibody on Westernblots. Anal. Biochem. 136:175-179.

2. Bottomley, P. J., J. F. Grilo, C. Van Baalen, and F. R. Tabita.1979. Synthesis of nitrogenase and heterocysts by Anabaena sp.CA in the presence of high levels of ammonia. J. Bacteriol.140:938-943.

3. Carithers, R. P., D. C. Yoch, and D. I. Arnon. 1979. Two formsof nitrogenase from the photosynthetic bacterium Rhodospiril-lum rubrum. J. Bacteriol. 137:779-789.

4. Eady, R. R., S. Iman, D. J. Lowe, R. W. Miller, B. E. Smith, andR. N. F. Thornely. 1980. The molecular enzymology of nitroge-nase, p. 10-35. In W. D. P. Stewart and J. R. Gallon (ed.),Nitrogen fixation. Academic Press, Inc., New York.

5. Gest, H., M. D. Kamen, and H. M. Bregoff. 1950. Studies on themetabolism of photosynthetic bacteria. V. Photoproduction ofhydrogen and nitrogen fixation by Rhodospirillum rubrum. J.Biol. Chem. 182:153-170.

6. Gotto, J. W., and D. C. Yoch. 1982. Regulation of Rhodospiril-lum rubrum nitrogenase activity. Properties and interconver-sions of active and inactive Fe-protein. J. Biol. Chem.257:2868-2873.

7. Hallenback, P. C., and P. M. Vignais. 1981. The effect ofelectron transport inhibitors on nitrogenase activity in thephotosynthetic bacterium, Rhodopseudomonas capsulata.FEMS Microbiol. Lett. 12:15-18.

8. Jouanneau, Y., C. M. Meyer, and P. M. Vignais. 1983. Regula-tion of nitrogenase activity through iron protein interconversioninto an inactive form in Rhodopseudomonas capsulata.Biochem. Biophys. Acta 749:318-328.

VOL. 169, 1987

on March 20, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

2542 SMITH ET AL.

9. Kanemoto, R. H., and P. W. Ludden. 1984. Effect of ammonia,darkness, and phenazine methosulfate on whole-cell nitroge-nase activity and Fe protein modification in Rhodospirillumrubrum. J. Bacteriol. 158:713-720.

10. Kelley, B. C., and D. J. D. Nicholas. 1981. Inhibition of nitroge-nase activity by metronidazole in Rhodopseudomonascapsulata. Arch. Microbiol. 129:344-348.

11. Kumar, A., F. R. Tabita, and C. Van Baalen. 1982. Isolation andcharacterization of heterocytes isolated from Anabaena sp.strain CA. Arch. Microbiol. 133:103-109.

12. Kumar, A., F. R. Tabita, and C. Van Baalen. 1983. Highendogenous nitrogenase activity in isolated heterocysts ofAnabaena sp. strain CA after nitrogen starvation. J. Bacteriol.155:493-497.

13. [jones, T., and R. H. Burms. 1978. Evidence for one electrontransfer by the iron protein of nitrogenase. Biochem. Biophys.Res. Commun. 80:22-25.

14. Lloyd, D., and J. Z. Pedersen. 1985. Metronidazole radical aniongeneration in vivo in Trichomonas vaginalis: oxygen quenchingis enhanced in a drug-resistant strain. J. Gen. Microbiol.131:87-92.

15. Lowery, R. G., L. L. Saari, and P. W. Ludden. 1986. Reversibleregulation of nitrogenase iron protein from Rhodospirillumrubrum by ADP-ribosylation in vitro. J. Bacteriol. 166:513-518.

16. Ludden, P. W., and R. H. Burns. 1978. Purification and prop-erties of nitrogenase from Rhodospirillum rubrum and evidencefor phosphate, ribose and an adenine-like unit covalently boundto the iron protein. Biochem. J. 175:251-259.

17. Lugtenberg, B., J. MeUers, R. Peters, P. van der Hoeck, and L.van Alpen. 1975. Electrtophoretic resolution of the "major outermembrane protein" of Escherichia coli K12 into four bands.FEBS Lett. 58:254-258.

18. Mortenson, L. E., and R. N. F. Thornely. 1979. Structure andfunction of nitrogenase. Annu. Rev. Biochem. 48:387-418.

19. Noel, D., K. Nikaido, and G. F.-L. Ames. 1979. A single aminoacid substitution in a histidine-transport protein drasticallyalters its mobility in sodium dodecyl sulfate-polyacrylamide gelelectrophoresis. Biochemistry 18:4159-4165.

20. Pearle, H. W., and P. E. Keller. 1979. Nitrogen-fixingAnabaena: physiological adaptations instrumental in maintain-ing surface blooms. Science 204:620-622.

21. Pienkos, P. T., S. Bodmer, and F. R. Tabita. 1983. Oxygeninactivation and recovery of nitrogenase activity incyanobacteria. J. Bacteriol. 153:182-190.

22. Pope, M. R., S. A. Murrell, and P. W. Ludden. 1985. Covalentmodification of the iron protein of nitrogenase from Rhodospiril-lum rubrum by adenosine diphosphoribosylation of a specificarginine residue. Proc. Natl. Acad. Sci. USA 82:3173-3177.

23. Ramos, J. L., and M. G. Guerrero. 1983. Involvement ofammonium metabolism in the nitrate inhibition of nitrogenfixation in Anabaena sp. strain ATCC 33047. Arch. Microbiol.136:81-83.

24. Ramos, J. L., F. Madeno, and M. G. Guerrero. 1985. Regulationof nitrogenase levels in Anabaena sp. ATCC 33047 and otherfilamentous cyanobacteria. Arch. Microbiol. 141: 105-111.

25. Robson, R. L., and J. R. Postgate. 1980. Oxygen and hydrogenin biological nitrogen fixation. Annu. Rev. Microbiol.34:183-207.

26. Saari, L. L., E. W. Triplett, and P. W. Ludden. 1984. Purifica-tion and properties of the activating enzyme for iron protein ofnitrogenase from the photosynthetic bacterium Rhodospirillumrubrum. J. Biol. Chem. 259:15502-15508.

27. Smith, B. E., and G. Lang. 1974. Mossbauer spectroscopy of thenitrogenase proteins from Klebsiella pneumoniae. Biochem. J.137: 169-180.

28. Smith, R. L., D. Kumar, Z. Xiankong, F. R. Tabita, and C. VanBaalen. 1985. H2, N2, and 02 metabolism by isolated heterocystsfrom Anabaena sp. strain CA. J. Bacteriol. 162:565-570.

29. Stacey, G., C. Van Baalen, and F. R. Tabita. 1977. Isolation andcharacterization of a marine Anabaena sp. capable of rapidgrowth on molecular nitrogen. Arch. Microbiol. 114:197-201.

30. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretictransfer of proteins from polyacrylamide gels to nitrocellulosesheets: procedure and some applications. Proc. Natl. Acad. Sci.USA 76:4350-4354.

31. Van Baalen, C. 1962. Studies on marine blue-green algae. Bot.Mar. 4:129-139.

32. Yoch, D. C., and J. W. Gotto. 1982. Effect of light intensity andinhibitors of nitrogen assimilation on NH4' inhibition of nitro-genase activity in Rhodospirillum rubrum and Anabaena sp. J.Bacteriol. 151:800-806.

33. Zumft, W. G., and L. E. Mortenson. 1975. The nitrogen fixingcomplex of bacteria. Biochim. Biophys. Acta 416:1-52.

J. BACTERIOL.

on March 20, 2020 by guest

http://jb.asm.org/

Dow

nloaded from