JOURNAL OF BACTERIOLOGY, Nov. 1982, p. 706-7130021-9193/82/110706-08$02.00/0

Vol. 152, No. 2

Control of Nitrogenase in a Photosynthetic AutotrophicBacterium, Ectothiorhodospira sp.

A. BOGNAR, L. DESROSIERS, M. LIBMAN, AND E. B. NEWMAN*

Department ofBiological Sciences, Concordia University, Montreal H3G IM8, Canada

Received 11 December 1981/Accepted 23 July 1982

An Ectothiorhodospira species fixed nitrogen when grown as an autotroph incompletely inorganic medium by using a variety of electron donors. The organismalso used organic carbon sources; however, this required induction of synthesis ofvarious enzymes, whereas the enzymes needed for autotrophic growth were

synthesized constitutively. Nitrogenase induction and function were inhibited byammonium chloride. Nitrogenase activity was dependent on light and inhibited byoxygen.

Although nitrogen fixation is a property ofmany procaryotes, most environmental researchhas concentrated on symbiotic Rhizobium spe-cies, which provide the bulk of the nitrogenfixed in agricultural soils (21). The contributionof free-living organisms, particularly the photo-synthetic cyanobacteria (25) and Rhodospiril-lales (11), has become of interest more recently.Many of the Rhodospirillales fix elemental

nitrogen (25). They do so primarily anaerobical-ly in the light (24), under which conditions theygrow quite rapidly (7, 24). Such conditions canbe found in soils in association with oxygen-utilizing heterotrophic bacteria (11). Evidence ofnitrogen fixation in the dark has been obtainedfor Rhodopseudomonas acidophila and R. cap-sulata (24), in both cases with lactate as carbonand energy source. This has also been reportedfor a Thiocapsa sp. and Ectothiorhodospira sha-posnikovii (29).The use of organic carbon sources by the

Rhodospirillales is well known. Many of themcan also synthesize their cell material from car-bon dioxide if provided with an inorganic elec-tron donor. Most of the Chromatiaceae havethis ability (27). However, most of the Rhodo-spirillaceae do not use inorganic electron donorsother than hydrogen (4, 15, 16). Three species dogrow with thiosulfate and bicarbonate, R. palus-tris, (22, 28), R. sulfidophila, and R. sulfoviridis(27).The organism studied here, a newly isolated

strain of Ectothiorhodospira, can fix nitrogenwhile growing autotrophically or heterotrophi-cally, deriving its reducing power from inorganicsulfur compounds or from organic compounds.It is also quite aerotolerant. These metabolicproperties suggest that the organism may con-tribute ecologically significant amounts of fixednitrogen to its environment. Here we describe

aspects of nitrogenase regulation in this orga-nism.

MATERIALS AND METHODSOrganism. The red photosynthetic organism studied

here is a new isolate of the genus Ectothiorhodospira,on deposit with the American Type Culture Collec-tion, Rockville, Md., as no. 31751. It was isolated aspart of a single-cell protein project and is known forindustrial purposes as E. goldameirae. The organismgrows as short rods 1.0 by 2.0 ,um only under anaero-bic and microaerophilic conditions. It grows well withsulfide, sulfur, or thiosulfate as photosynthetic elec-tron donor. Several organic compounds can also beused as hydrogen donors. During growth on sulfide,elementary sulfur is deposited outside the cells in themedium and disappears during further growth. This,and the large stacks of intracellular lamellar mem-branes, classify it among the Ectothiorhodospira (3, 26).Its physiology will be reported in detail elsewhere.Media. The medium of Baalsrud and Baalsrud (1)

was used with minor modifications. The main sub-strates for growth were added at the following concen-trations (per liter): sodium thiosulfate, 16 g; sodiumsulfide, 2 g; elemental sulfur, 0.1 to 10 g; bicarbonate,10 g; and acetate and succinate, 10 g. Cultures wereincubated under nitrogen or argon at 30°C with 5,380to 10,760 lx (500 to 1,000 foot candles) of light provid-ed by 40-W incandescent light bulbs. Anaerobic condi-tions were established either by filling containers withmedium or by bubbling nitrogen through the mediumand closing the containers with rubber stoppers.Growth experiments. To assess the ability of cells to

adapt from one medium to another, exponential-phasecultures were chilled in ice water, centrifuged at 4°Cunder argon, washed, and inoculated into mediumincluding 10-4 M glutathione through which nitrogenhad been bubbled immediately before use. Cells werewashed with minimal medium which contained 10-4 Mglutathione but no electron donor. Turbidity was fol-lowed by using flasks fitted with sidearms for directreading of turbidities in a Klett-Summerson colorim-eter with a 640-nm filter.Dark microaerophilic growth. Deep agar shake tubes

706

NITROGENASE IN AN ECOTHIORHODOSPIRA sP.

were prepared according to the method of Kampf andPfennig (9). Media contained the minimal salts de-scribed above, including 1 g of ammonium chlorideand 5 g of bicarbonate per liter. Depending on theexperimental conditions, the medium was supplement-ed with 2 g of thiosulfate, 5 g of acetate, or 0.2 g ofsulfide per liter. Some tubes were inoculated, whereasothers were left uninoculated. Tubes were incubatedfor 3 weeks in the dark at 30°C.

Preparation of resting cells. Cells harvested andwashed as above were suspended in the washingmedium at 1 to 2 mg of protein per ml and incubatedunder nitrogen for 24 to 48 h at 30°C with the usualillumination. Such preparations could then be storedfor several days in the refrigerator with very littleeffect on their nitrogenase activity. For nitrogenaseassays, these cells were diluted 1:2 with washingsolution supplemented with the appropriate sub-strates.

Nitrogenase assays. Samples, usually 5 ml, of eitherexponential-phase cells or resting cells were removedwith a syringe through the stopper of the culture flaskunder a flow of argon and injected into assay flasks,usually 30 ml in total volume, through which argonwas bubbled for 5 min. After 2 ml of acetylene wasinjected, the cultures were incubated with illuminationand shaking at 30°C. Rates were measured by measur-ing ethylene content at four intervals of a 20 to 25 minincubation by injecting 10- to 100-,Jl samples of the gasphase into a Perkin-Elmer Fl gas chromatographwith a stainless steel column (6 ft. by 0.125 in. i.d. [ca.1.83 m by 0.42 cm]) containing 5% phenylisocyanateon Porasil C.

In vivo control of nitrogenase. For studies of effec-tors on nitrogenase, an initial rate of nitrogen fixationwas determined in the absence of the effector; inhibi-tors in solutions previously treated with argon wereinjected into the assay flasks, and a second rate wasdetermined.

fkbemical determinations. Protein was measured bythe method of Lowry et al. (14) on trichloroacetic acidprecipitates redissolved in NaOH.

RESULTSGrowth of Ectothiorhodospira sp. strain ATCC

31751 in various Media. Ectothiorhodospira sp.strain 31751 ATCC is a particularly efficientphotosynthetic autotroph. It grew anaerobicallyin the light with a doubling time of 4.5 h at 30°Cin a completely inorganic medium, using bicar-bonate as carbon source, ammonium chloride asnitrogen source, and thiosulfate as electron do-nor. An organic compound such as acetate orsuccinate could be used as carbon source or ascarbon source and electron donor, but this in-creased the growth rate only slightly.

Ectothiorhodospira sp. strain ATCC 31751grew with a variety of inorganic electron donorsincluding thiosulfate, sulfide, and elemental sul-fur. Although sulfide was toxic at high levels, itwas used as an electron donor at 0.8% or less.

Ectothiorhodospira sp. strain ATCC 31751derived its nitrogen from inorganic ammoniumsalts, some amino acids, urea, and dinitrogen,

but not from nitrate. In a completely inorganicmedium with bicarbonate, thiosulfate, and dini-trogen, it grew with a doubling time of 7.5 to 8 hat 30°C, i.e., at about half the rate of a culturegiven an ammonium salt.

Ectothiorhodospira sp. strain ATCC 31751 didnot grow under fully aerobic conditions on anymedium tested, organic or inorganic. It did,however, grow in the dark in microaerophilicconditions as provided by Pfennig deep agarshake cultures (9) in medium supplemented withammonium chloride and (i) acetate alone, (ii)acetate and either thiosulfate or sulfide, or (iii)bicarbonate and either thiosulfate or sulfide.Ectothiorhodospira sp. strain ATCC 31751 alsofixed nitrogen when grown in Pfennig tubes withacetate.

Constitutive synthesis of enzymes needed forautotrophic growth. As indicated above, Ectoth-iorhodospira sp. strain ATCC 31751 could growin many diverse conditions. The enzymes need-ed for autotrophic growth seemed to be formedin all growth conditions. Cells grown in hetero-trophic conditions could transfer to autotrophicconditions without a lag. The reverse shift re-quired a long period of adaptation. Thus, Ectoth-iorhodospira sp. strain ATCC 31751 grown inthiosulfate, bicarbonate, and ammonium chlo-ride transferred to the same medium without alag (Fig. 1A, curve 1) but had a long lag whentransferred to medium with acetate, bicarbon-ate, and ammonium chloride (Fig. 1A, curve 3).If a small amount of thiosulfate (0.04%) wasprovided with the acetate, the lag was abolished(Fig. IA, curve 2). The thiosulfate presumablyprovided reducing power during the transition touse of acetate, since it did not suffice for anysignificant growth in the absence of acetate (Fig.1A, curve 4). Thus, the transfer ofEcotothiorho-dospira sp. strain ATCC 31751 from thiosulfateto acetate is formally similar to the transfer ofEscherichia coli from glucose to lactose (17).

In the reciprocal experiment (Fig. 1B), anacetate-grown culture was transferred to medi-um with acetate (curve 1) and with thiosulfate(curve 3), in both cases without a lag. Synthesisof enzymes for the use of thiosulfate and bicar-bonate thus appeared to be constitutive in thisstrain. It differs in this respect from R. palustris,which also grows as an autotroph with inorganicelectron donors, but does so only after adapta-tion (22).Use of electron donors for acetylene reduction.

The ability of the cells to use electron donorswas tested, using the whole cell assay of nitroge-nase by acetylene reduction. Cells grown withthiosulfate, bicarbonate, and dinitrogen con-tained a great deal of nitrogenase. Resting cellsprepared from cells grown in this medium wereunable to reduce acetylene unless an exogenous

VOL. 152, 1982 707

708 BOGNAR ET AL.

'F100

20

0 4 8 0 4 8 12

T IMFE [hours]

FIG. 1. The growth of Ectothiorhodospira sp.strain ATCC 31751 with different electron donors.Cells from exponential phase of cultures grown in twodifferent media, (A) 0.8% thiosulfate, 0.5% bicarbon-ate, and (B) 0.5% acetate, O.S% bicarbonate, werechilled, washed, and subcultured into media with thefollowing electron donors. (A) curve 1, 0.8% thiosul-fate (c); curve 2, 0.04% thiosulfate and 0.5% acetate(x); curve 3, 0.5% acetate (O); curve 4, 0.04% thiosul-fate (A). (B) curve 1, 0.5% acetate (O); curve 2, 0.05%acetate and 0.8% thiosulfate (x); curve 3, 0.8%o thio-sulfate (O); curve 4, 0.05% acetate (O).

electron donor was added, presumably becausetheir endogenous stores of reducing power weredepleted. The activity measur'ed then by using avariety of electron donors would be a function ofthe resting cells ability to use the particularcompound.

Cells grown photoautotrophically (thiosulfate,bicarbonate, dinitrogen) used thiosulfate to re-duce acetylene at a rate of 28 nmol/min per mg ofprotein (Table 1). However, they also usedsuccinate and acetate at significant rates, thoughthey were unable to use a-ketoglutarate or ma-late.

Cells grown with acetate as carbon and elec-tron source were able to use organic compounds(acetate and succinate), and inorganic com-pounds (thiosulfate, sulfide, and sulfur) equallywell (Table 2). Their rate with thiosulfate (19,umol/min per mg of protein) was gimiiar to thatof thiosulfate-grown cells, and they were able touse sulfide at a much higher rate. This may bedue to the increase in redox potential in thepresence of sulfide.

Physiological effects of oxygen on growth. Ec-tothiorhodospira sp. strain ATCC 31751 grewphotosynthetically only under anaerobic condi-

TABLE 1. Nitrogenase activities in thiosulfate-grown cells assayed with various electron donorsa

Nitrogenase activityElectron donors (nmol acetylene reduced/

min per mg of protein)

None .............. 5.2Acetate ............ 14.6Thiosulfate .......... 28.0Succinate ........... 31.3Malate ............. 0a-Ketoglutarate ....... 5.1

a Resting cell suspensions of cells grown with thio-sulfate, bicarbonate, and dinitrogen were diluted 1:2 inmedium containing the compounds indicated, incubat-ed under argon, and assayed for nitrogenase. Theresults given are averages of several experiments,except for those for malate and oa-ketoglutarate whichare the results of single experiments.

tions and stopped growing when oxygen wasintroduced. To test whether the effect of oxygenwas reversible, air was bubbled for 1 h into ananaerobic photosynthesizing culture growingwith bicarbonate, thiosulfate, and ammoniumchloride. The culture immediately ceased toincrease in turbidity; however, soon after oxy-gen was removed by bubbling nitrogen, growthresumed.The same experiment was repeated with a

culture growing in the same medium with dini-trogen rather than ammonium chloride (Fig. 2),air being passed through the culture for 15 min inozke case and for 1 h in another. In both cases,growth ceased soon after air was bubbled intothe culture and resumed at close to the preced-ing rate 2 h after the air was replaced by nitro-gen. No significant difference was seen with thelonger period of aeration. It is clear then that theeffect of oxygen was drastic but reversible.

Regulation of nitrogenase synthesis in vivo.Nitrogenase activity was not seen in cells grown

TABLE 2. Nitrogenase activities in acetate-growncells assayed with various electron donorsa

Nitrogenase activityElectron donors (nmol acetylene reduced/

min per mg protein)

None .............. 2.1Acetate ............ 25.7Thiosulfate .......... 19.0Succinate ........... 31.2Sulfur ............. 11.5Sulfide ............. 61.4

a Resting cell suspensions of cells grown with ace-tate and dinitrogen were diluted 1:2 in medium con-taining the compounds indicated, incubated underargon, and assayed for nitrogenase. Results are aver-ages of two or more experiments, except for succinatewhich is the result of a single experiment.

J. BACTERIOL.

NITROGENASE IN AN ECOTHIORHODOSPIRA sp.

150

LJ

6a60

4 8

TIME [hours]FIG. 2. Effect of aeration on the growth of Ectoth-

iorhodospira sp. strain ATCC 31751 with dinitrogen asnitrogen source. At the time indicated by the arrow,air was bubbled through exponential-phase cultures at30°C in the light for periods of0 (0), 15 (0), and 60 (x)min. After the aeration period, nitrogen was passedthrough the culture to remove oxygen, and turbidityfollowed thereafter.

with ammonium chloride. When cells grownwith ammonium chloride were washed and sus-pended in medium under dinitrogen as solenitrogen source, nitrogenase activity could beassayed after 3 h of incubation and reached itsmaximum specific activity by 6 h (Fig. 3). Ifchloramphenicol was added to the suspendedcells, no nitrogenase activity was seen. Thisseems to indicate that cells grown with ammoni-um chloride do not contain nitrogenase and thatprotein synthesis is needed for its appearance.To investigate the regulation of the induction

of nitrogenase activity, cells grown with thiosul-fate, bicarbonate, and ammonium chloride wereincubated under argon for 24 h with variouspossible effectors of nitrogen metabolism andthen assayed for nitrogenase thereafter (Table3). Cultures incubated with dinitrogen as solenitrogen source, under argon alone, or with oneof several amino acids all showed nitrogenaseactivity. Indeed, nitrogenase activity at 24 h wasconsiderably higher than steady-state nitroge-nase activity, which may represent an overshootin synthesis. However, cells incubated with am-monium chloride or urea showed no activity.

Regulation of nitrogenase activity. (i) Inhibitionof nitrogenase activity by ammonia. As can beseen in Fig. 4A, nitrogenase activity in Ectoth-iorhodospira sp. strain ATCC 31751 as in manyother photosynthetic bacteria (10, 25) was inhib-

ited by ammonium chloride. The effect ofammo-nium ion was very dependent on the concentra-tion of the cell suspension tested. Nitrogenaseactivity of a suspension of cells containing 0.2mg of protein per ml was inhibited transiently byammonium ion at 10-5 M, for about 3 h at 10'M, and for longer at 10-3 M. The same cells inmore dilute suspension (0.04 mg of protein perml) were inhibited strongly even by l0-5 MNH4Cl. The nitrogenase assay is a whole cellassay performed under conditions in which thecells can grow and concomitantly assimilateNH4Cl. In more dilute suspension, the cellsassimilate less NH4Cl per unit time, and there-fore may be inhibited for longer by lower con-centrations. A 1i-0 M concentration is thus anoverestimate of the minimum concentration ofNH4Cl needed to inhibit nitrogenase.

(il) Effect of oxygen on nitrogenase. The nitro-genase of Ectothiorhodospira sp. strain ATCC31751 was inhibited by oxygen added to the gasphase in concentrations as low as 2% (Fig. 4B).The extent to which different concentrationsinhibited depended on the concentration of thecell suspension tested, just as in the case ofNH4Cl, more dilute cultures being inhibited bylower amounts of oxygen. This seems to implythat the cells have a considerable ability toremove oxygen metabolically from the medium,thus allowing nitrogenase to function. This couldbe a factor in the relatively long lag between the

12

CE0

E'I

E

I I0 2 4 6

TIME [hours]

FIG. 3. Induction of nitrogenase. Cells from expo-nential-phase cultures grown on medium containing0.1% ammonium chloride were centrifuged and sus-pended in the same medium without ammonium chlo-ride. One culture also contained 20 ,ug of chloram-phenicol per ml (0); the other did not (A). At the timesindicated, samples were removed from each flask andassayed for nitrogenase activity in the presence ofchloramphenicol.

VOL. 152, 1982 709

710 BOGNAR ET AL.

TABLE 3. Effect of various compounds onsynthesis of nitrogenasea

Nitrogenase activitySubstance added (nmol acetylene reduced/

min per mg of protein)

None ................. 57Nitrogen ............... 79Alanine ............... 76Arginine ............... 52Glutamate .............. 50Ammonium chloride ....... 0Urea ...... ........... 0

a Cells were grown with thiosulfate, bicarbonate,and ammonium chloride, harvested, and incubated at30°C under argon in the same medium from whichammonium chloride was omitted. The substances not-ed were added at 6 mg/ml, except for nitrogen. Whennitrogen was added it served as the atmosphere inplace of argon. Nitrogenase activity was assayed after24 h.

addition of oxygen and the onset of inhibition, amatter of some 20 min.The inhibitory effect of oxygen was readily

reversed by the removal of oxygen. When cellswhose activity had been totally inhibited by 5%oxygen were gassed with argon and reassayed, arapid recovery of 100% of the original activitywas observed (Fig. 5). This recovery was seeneven in the presence of chloramphenicol, thoughonly to the extent of 54%. We have often noticeda decrease in nitrogenase activity in cells incu-bated with chloramphenicol, perhaps because anitrogen-containing intermediate accumulatedand inhibited the enzyme. This may account forthe lower level of activity seen with chloram-

phenicol here. It seems clear that nitrogenase isinhibited in the presence of oxygen, but it is notpermanently inactivated-at least not in theshort periods of time (2 to 3 h) examined here.

(lii) Dependence of nitrogenase activity on light.Nitrogenase activity is dependent on a supply ofreducing power and energy, the latter providedin this organism by light. When light was re-moved, nitrogenase activity became unmeasur-able within 15 min (data not shown). The en-zyme was not inactivated, since restoration oflight was followed about 15 min later by arestoration of nitrogenase activity. These de-layed effects probably reflect the depletion ofATP and its resynthesis. Since photosynthesisoccurs exclusively under anaerobic conditions,this may also explain the delayed inhibition byoxygen.

DISCUSSIONWe have described here an Ectothiorhodo-

spira species, a photosynthetic bacterium whichgrew rapidly in totally inorganic media by usingbicarbonate as carbon source, thiosulfate aselectron donor, and ammonium chloride as ni-trogen source. The organism also used organicsources of carbon and electrons such as acetateand succinate. However, cells grown in anymedium tested could transfer to growth in inor-ganic medium without a lag, whereas growthwith organic sources required adaptation andwas only slightly more rapid than growth withinorganic compounds.

Ectothiorhodospira sp. strain ATCC 31751,like many other photosynthetic bacteria, can fixdinitrogen. It can use organic or inorganicsources of reducing power; these support similar

0.570c

006 t os°°-m %l0E ~~~~~~~2-10-0M 20/

io2 / NH4CI

±2 02

0 60 120 180 0 60 120

TIME [min.]

FIG. 4. Inhibitors of nitrogenase activity. Exponential-phase cells grown with dinitrogen as nitrogen sourcewere assayed for nitrogenase as described in the text. Initial activity was determined, inhibitors added at the timeindicated by the arrow, and the assay continued. The inhibitors tested were (A) ammonium chloride and (B)oxygen, stated as percent ambient gas.

J. BACTERIOL.

NITROGENASE IN AN ECOTHIORHODOSPIRA sP.

c

0

o3~/

E

E

0 20 40 0 20 40 60

TIME [min]

FIG. 5. Reversibility of nitrogenase inhibition byoxygen. (A) Exponential-phase cells grown with dini-trogen as nitrogen source were assayed for nitrogenaseactivity. At the time indicated by the first arrow,oxygen was injected to 5% of the ambient gas. (B) Atthe time indicated by the second arrow, the gas phasewas flushed with argon, acetylene was reinjected intothe flasks (at time 0 in [B]), and nitrogenase was againassayed. The experiment was carried out in the pres-ence (0) and absence (0) of chloramphenicol (20 ,g/ml).

rates of nitrogen fixation and growth. Whereasthe use of organic compounds for growth re-quired adaptation, their use as electron donorswas constitutive. It seems likely then that theorganism makes the enzymes for oxidizing thesecompounds constitutively, but does not make allthe enzymes needed to assimilate carbon fromthem.

Nitrogen fixation in the genus Ectothiorhodo-spira has been reported for both E. shaposniko-vii (5, 29, 30) and E. mobilis (2). These organ-isms use a similar range of nitrogen sources forgrowth except that E. shaposnikovii uses nitrate(13, 20) and the organism we used does not. Wethink that E. shaposnikovii is very similar to ourstrain, but differs in aspects of its nitrogenmetabolism. E. shaposnikovii has only beenshown to fix nitrogen when grown on organiccarbon sources (5, 29, 30). Under these condi-tions it will use thiosulfate as an electron donorto nitrogenase when arginine is supplied as nitro-gen source, but not with dinitrogen (5). More-over, it will not use sulfide as an electron donor(29) nor did it grow with sulfide and dinitrogenunder autotrophic conditions (30). The Ectoth-

iorhodospira sp. strain ATCC 31751 studiedhere used inorganic sulfur compounds very well,whether grown autotrophically or with organiccarbon sources. It appears that E. shaposnikoviidoes not fix nitrogen in a totally inorganic medi-um, but the authors did not state whether theytested this.The fixation of nitrogen is intensive in the use

of energy and reducing power (18, 25), particu-larly because the use of electron donors bynitrogenase is inherently inefficient. Much of thereducing power may be diverted to hydrogenformation. Nitrogen-fixing cultures of Ectothio-rhodospira sp. strain ATCC 31751 are easilyidentified visually by the copious production ofhydrogen bubbles on slight agitation.Because of the physiological demands of the

process, the capacity for nitrogen fixation isextremely sensitive to the energy state of thecell. It can be inhibited not only by inhibitors ofthe enzyme itself, but also by conditions whichdecrease the energy stores of the cell (e.g., lackof light) or the reducing power (e.g., less effec-tive electron donor). The electrons of thiosulfateare at too high a potential to act directly asreductants for nitrogenase. Cells grown withthiosulfate and bicarbonate use energy in theform of light-generated ATP to reduce the elec-trons further (6, 12).

Since the supply of both energy and reducingpower is dependent on light, it is not surprisingthat nitrogen fixation is regulated by light, atleast in light-grown cultures. The 10- to 20-minlag found in response to both the shutoff and thereturn of light suggests that the cell has anexhaustible store of a reduced intermediate,perhaps a ferredoxin (25).

Nitrogen fixation does not depend on lightwhen other sources of energy are available.Ectothiorhodospira sp. strain ATCC 31751grows in the dark with either acetate or sulfideas energy source. In either case it requiresenough oxygen to serve as terminal electronacceptor, but it is inhibited by higher levels. Therequired microaerophilic conditions are easilyprovided by Pfennig deep agar shake cultures(9). These cultures can fix nitrogen, by usingthiosulfate, sulfide, or acetate, in the dark. Wehave been unable to show this in liquid culture.

Nitrogen fixation is inhibited by oxygen, alsowith a considerable lag. Oxygen inhibits at leasttwo functions of photosynthetic bacteria: nitro-genase activity and photosynthesis (10, 19, 25).Cells in exponential growth with ammoniumchloride as nitrogen source stop growing as soonas air is bubbled into the medium and resumegrowth as soon as the air is replaced withnitrogen. It would appear that their energy sup-ply is cut off by the addition of oxygen but thatno irreversible inactivation occurs.

VOL. 152, 1982 711

712 BOGNAR ET AL.

Nitrogen-fixing cells treated in the same wayshow a longer, but still comparatively short lag.This suggests that the primary effect of oxygenin the nitrogenase assay must be on availabilityof energy and not directly on nitrogenase. This isconsistent with the fact that the effect of oxygenis reversible, at least after short periods ofexposure. It may be that some of the nitrogenaseis inactivated, since it took 2 h before an in-crease in turbidity was seen in our experiment(Fig. 2). However if the nitrogenase enzymewere extensively inactivated, the cell would bestarved for nitrogen and could not recover in theshort time it actually took to resume growth.The delay in the onset of inhibition by oxygen

may also indicate that the organism can protectitself by metabolizing oxygen, as is done by awide variety of photosynthetic organisms (8, 15,16, 23). This delay is seen in cells oxidizing bothorganic and inorganic compounds. The ability oforganisms to grow under microaerophilic condi-tions shows that the cells have oxygen-handlingfacilities at least under those conditions. Theextent to which these are available in anaerobi-cally grown cells is not known.The properties we have described for nitroge-

nase regulation in Ectothiorhodospira sp. strainATCC 31751 are very similar to those of otherphotosynthetic bacteria. Ammonia represses,and oxygen both inhibits and represses nitroge-nase in all organisms tested (10). In photosyn-thetic bacteria, light also is a regulator of nitro-genase (25). Differences arise from the metabolicprocesses used by various organisms to obtainenergy, reducing power, and protection fromoxygen (4). The organism described in this workis of particular interest because its metabolicversatility and rapid growth in autotrophic con-ditions with light and inorganic substrates maymake it useful as a commercial source of bio-mass and derivatives thereof.

ACKNOWLEDGMENTS

This work was supported by a strategic grant of the NationalScience and Engineering Research Council of Canada.The authors wish to recognize the work of several people

who have been involved in establishing conditions for growingthis organism. Particular thanks are due to James Whittaker,who was the first person to grow it in a dependable manner,and to Ian Krantz, whose devoted work over three summersadded a great deal to this project.

LITERATURE CITED1. Baalsrud, K., and K. S. Baalsrud. 1954. Studies on Thio-

bacillus denitrificans. Arch. Mikrobiol. 20:34-62.2. Bast, E. 1977. Utilization of nitrogen compounds and

ammonia assimilation by Chromatiaceae. Arch. Micro-biol. 113:91-94.

3. Cherni, N. E., Z. V. Soloviev, V. D. Fedorov, andE. N. Kondratieva. 1969. Ultrastructure of cells of twospecies of purple sulfur bacteria. Mikrobiologiya 38:479-484.

4. Dalton, H. 1974. Fixation of dinitrogen by free-living

microorganisms. Crit. Rev. Microbiol. 3:183-220.5. Gogotov, I. N., and V. P. Glinskil. 1973. A comparative

study of nitrogen fixation in the purple bacteria. Mikrobio-logiya 42:983-986.

6. Gest, H. 1972. Energy conversion and generation ofreducing power in bacterial photosynthesis. Adv. Microb.Physiology 7:243-282.

7. Herbert, R. A. A., E. Siefert, and N. Pfennig. 1978.Nitrogen assimilation in Rhodopseudomonas acidophila.Arch. Microbiol. 119:1-5.

8. Hochman, A., and R. H. Burris. 1981. Effect of oxygen onacetylene reduction by photosynthetic bacteria. J. Bacte-riol. 147:492-499.

9. Kampf, C., and N. Pfennig. 1980. Capacity of Chromatia-ceae for chemotrophic growth. Specific respiration ratesof Thiocystis violacea and Chromatium vinosum. Arch.Mikrobiol. 127:125-135.

10. Keister, D. L., and D. E. Fleischman. 1973. Nitrogenfixation in photosynthetic bacteria. Photophysiology8:157-183.

11. Kobayashi, M., and M. Z. Haque. 1971. A contribution tonitrogen fixation and soil fertility by photosynthetic bacte-ria, p. 443456. In T. A. Lie and E. G. Mulder (ed.),Biological nitrogen fixation in natural and agriculturalhabitats. Plant and soil special volume. Nijhoff, TheHague.

12. Kondratieva, E. N. 1979. Interrelation between modes ofcarbon assimilation and energy production in phototro-phic purple and green bacteria, p. 117-175. InJ. R. Quayle (ed.), Microbial biochemistry, vol. 21. Uni-versity Park Press, Baltimore.

13. Krasilnikova, E. N., L. M. Zakharchuk, and L. M. Lin-nik. 1980. Growth of E. mobilis in the dark. Mikrobiolo-giya 49:244-248.

14. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folin phenolreagent. J. Biol. Chem. 193:265-275.

15. Meyer, J., B. C. Kelley, and P. M. Vignais. 1978. Nitrogenfixation and hydrogen metabolism in photosynthetic bac-teria. Biochimie 60:245-260.

16. Meyer, J., B. C. Kelley, and P. M. Vignais. 1978. Effect oflight on nitrogenase function and synthesis in Rhodopseu-domonas capsulata. J. Bacteriol. 136:201-208.

17. Monod, J. 1942. Recherches sur la croissance des culturesbacteriennes. Hermann Cie, Paris.

18. Mortenson, L. E., and R. N. F. Thorneley. 1979. Struc-ture and function of nitrogenase. Annu. Rev. Biochem.48:387-418.

19. Pfennig, N. 1967. Photosynthetic bacteria. Annu. Rev.Microbiol. 21:285-324.

20. Pfennig, N., and H. G. Truper. 1974. The phototrophicbacteria p. 24-64. In R. E. Buchanan and N. E. Gibson(ed.), Bergey's manual of determinative bacteriology, 8thed. The Williams & Wilkins Co., Baltimore.

21. Postgate, J. R. 1974. New advances and future potential inbiological nitrogen fixation. J. Appl. Bacteriol. 37:185-202.

22. Rolls, J. P., and E. S. Lindstrom. 1967. Effect of thiosul-fate on the photosynthetic growth of Rhodopseudomonaspalustris. J. Bacteriol. 94:860-866.

23. Siefert, E. 1976. Nitrogen fixation in facultative aerobicRhodosapirillacease with photosynthetic or respiratoryenergy generation p. 149-151. In G. A. Codd andW. D. P. Stewart (ed.), 2nd International Symposium onPhotosynthetic Prokaryotes.

24. Siefert, E., and N. Pfennig. 1980. Diazotrophic growth ofRhodopseudomonas acidophila and Rhodopseudomonascapsulata in the dark. Arch. Microbiol. 125:73-77.

25. Stewart, W. D. P. 1973. Nitrogen fixation by photosyn-thetic microorganisms. Annu. Rev. Microbiol. 27:283-316.

26. Truper, H. G. 1968. Ectothiorhodospira mobilis Pelsh, aphotosynthetic sulfur bacterium depositing sulfur outsidethe cells. J. Bacteriol. 95:1910-1920.

27. Truper, H. G., and N. Pfennig. 1978. Taxonomy of the

J. BACTERIOL.

VOL. 152, 1982 NITROGENASE IN AN ECOTHIORHODOSPIRA Sp. 713

Rhodospirillales, p. 19-27. In R. K. Clayton and tion of molecular nitrogen by photosynthesizing bacteriaW. R. Sistrom (ed.), The photosynthetic bacteria. Plenum in relation to the presence of light and ATP and thePublishing Corp., New York. character of exogenous substrates. Dokl. Akad. Nauk.

28. Van NIel, C. B. 1944. The culture, general physiology, SSR 196:698-700.morphology, and classification of the non-sulfur purple 30. Zakvateeva, N. V., I. V. Malofeeva, and E. N. Kondra-and brown bacteria. Bacteriol. Rev. 8:1-118. tieva. 1970. Fixation capacity of photosynthesizing bacte-

29. Zhakvateeva, N. V., and E. N. Kondatieva. 1970. Fixa- ria. Mikrobiologiya 39:761-66.