JOURNAL OF BACTERIOLOGY,0021-9193/97/$04.0010

July 1997, p. 4319–4327 Vol. 179, No. 13

Copyright © 1997, American Society for Microbiology

Transcriptional and Translational Regulation of Nitrogenase inLight-Dark- and Continuous-Light-Grown Cultures of the

Unicellular Cyanobacterium Cyanothece sp. Strain ATCC 51142MILAGROS S. COLON-LOPEZ,1† DEBRA M. SHERMAN,2 AND LOUIS A. SHERMAN1*

Department of Biological Sciences1 and Department of Botany and Plant Pathology,2

Purdue University, West Lafayette, Indiana 47907

Received 20 February 1997/Accepted 22 April 1997

Cyanothece sp. strain ATCC 51142 is a unicellular, diazotrophic cyanobacterium which demonstrated ex-tensive metabolic periodicities of photosynthesis, respiration, and nitrogen fixation when grown under N2-fixing conditions. N2 fixation and respiration peaked at 24-h intervals early in the dark or subjective-darkperiod, whereas photosynthesis was approximately 12 h out of phase and peaked toward the end of the lightor subjective-light phase. Gene regulation studies demonstrated that nitrogenase is carefully controlled at thetranscriptional and posttranslational levels. Indeed, Cyanothece sp. strain ATCC 51142 has developed anexpensive mode of regulation, such that nitrogenase was synthesized and degraded each day. These patternswere seen when cells were grown under either light-dark or continuous-light conditions. Nitrogenase mRNAwas synthesized from the nifHDK operon during the first 4 h of the dark period under light-dark conditions orduring the first 6 h of the subjective-dark period when grown in continuous light. The nitrogenase NifH andNifDK subunits reached a maximum level at 4 to 10 h in the dark or subjective-dark periods and were shownby Western blotting and electron microscopy immunocytochemistry to be thoroughly degraded toward the endof the dark periods. An exception is the NifDK protein (MoFe-protein), which appeared not to be completelydegraded under continuous-light conditions. We hypothesize that cellular O2 levels were kept low by decreasingphotosynthesis and by increasing respiration in the early dark or subjective-dark periods to permit nitrogenaseactivity. The subsequent increase in O2 levels resulted in nitrogenase damage and eventual degradation.

During nitrogen (N2) fixation, the enzyme nitrogenase cat-alyzes the reduction of atmospheric N2 to ammonia. This en-ergetically demanding process requires reducing power andapproximately 16 ATPs per fixed N2 (13, 44). Nitrogenase isnotorious for its oxygen (O2) sensitivity. Both of the nitroge-nase components, the Fe-protein and the MoFe-protein, arequickly inactivated upon exposure to O2 (8, 49), and bothsubunits are known to be inactivated in vivo and in vitro (17).Therefore, the cellular concentration and the activity of thenitrogenase enzyme are carefully monitored and strictly regu-lated at the transcriptional and/or posttranslational level, phe-nomena well documented for many nitrogen-fixing organisms(i.e., Klebsiella sp., Anabaena sp., and Azotobacter sp.) (9, 13).

A very diverse group of microorganisms, such as cyanobac-teria, photosynthetic bacteria, and symbiotic and free-livingbacteria, are capable of fixing N2 (diazotrophy) (11, 60). Cya-nobacteria are unique within this group since they are the onlydiazotrophs with the ability to conduct the two processes ofO2-evolving photosynthesis and O2-sensitive N2 fixation withinthe same cell. Therefore, these N2-fixing microorganisms havedeveloped sophisticated strategies to protect nitrogenase fromambient O2 as well as from photosynthetically generated O2.Although there is not a universal protection mechanism, cya-nobacteria combine several strategies, including avoidance,physical barriers to O2 diffusion, high rates of respiration, andspatial and temporal separation of N2 fixation and photosyn-

thesis, to protect nitrogenase from O2 inactivation (9, 11, 13,57).

Nitrogen fixation has been found in all morphologicalgroups of cyanobacteria: filamentous heterocystous (e.g.,Anabaena sp. and Nostoc sp.), filamentous nonheterocystous(e.g., Trichodesmium sp., Oscillatoria sp., and Plectonema sp.),and unicellular (e.g., Gloeothece sp. and Synechococcus sp.) (9,16, 53). Anabaena sp. fix N2 aerobically by restricting the ac-tivities of N2 fixation and photosynthesis to the heterocyst andvegetative cell, respectively (spatial separation) (16, 58). Inaddition, low intracellular levels of O2 are maintained by highrates of respiration and by using the heterocyst envelope as abarrier to O2 diffusion. Other cyanobacteria, such as Plec-tonema sp. and Synechococcus sp. strain SF1, fix N2 only undermicroaerobic conditions (51, 55).

Interestingly, some unicellular cyanobacteria are capable ofaerobic N2 fixation. The most studied strains are representedby Gloeothece sp., Synechococcus sp. strain RF1, and Synecho-coccus sp. strains Miami BG 43511 and Miami BG 43522 (18,36, 59). Temporal separation of N2 fixation and photosynthesishave been shown in these strains when grown under 12-hlight–12-h dark periods (LD conditions) (36, 39, 45). Underthese conditions, cyclic nitrogenase and photosynthetic activi-ties, which are restricted to the dark and light phases, respec-tively, were observed (19, 25). Studies with Gloeothece sp. andSynechococcus sp. strains Miami BG 43511 and Miami BG43522 indicated that the cyclic buildup and breakdown of car-bohydrate along with high rates of respiration provided reduc-tant and energy to drive N2 fixation in the dark (12, 32, 37). Inaddition, respiration may protect nitrogenase from O2 inacti-vation (10, 14).

It has been shown that the cyclic pattern of nitrogenaseactivity reported in LD-grown cultures of Gloeothece sp. andSynechococcus sp. strain RF1 was under transcriptional and

* Corresponding author. Mailing address: Department of BiologicalSciences, Purdue University, 1392 Lilly Hall, West Lafayette, IN47907-1392. Phone: (765) 494-4407. Fax: (765) 496-1495. E-mail:lsherman@bilbo.bio.purdue.edu.

† Present address: Abbott Laboratories, Hospital Products Division,Abbott Park, IL 60064-3500.

4319

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

posttranslational regulation (4, 5, 12, 20, 21). In Synechococcussp. strain RF1, Northern blot analysis resulted in the identifi-cation of nitrogenase-specific transcripts only during the darkphase of the diurnal regimen (4, 20). In addition, two distinc-tive forms of the Fe-protein were immunodetected only onsamples from the dark period. Chow and Tabita (4) suggestedthat the high-molecular-weight form corresponds to a post-translationally modified Fe-protein which is inactive and maybe targeted to be rapidly degraded. Two forms of the Fe-protein have also been identified in light-dark grown culturesof Gloeothece sp. (5), but neither the molecular signal nor thetype of modification has been described for either case.

Cyanothece sp. strain ATCC 51142 (formerly BH68) is aunicellular cyanobacterium capable of fixing N2 under aerobicconditions (42, 48). This strain was isolated and characterizedby our laboratory with the major goal of developing a suitableexperimental system with which to study the interrelationshipbetween N2 fixation and oxygenic photosynthesis in a unicel-lular cyanobacterium. The Cyanothece sp. fixes N2 cyclicallywhen grown under LD conditions (42, 48). During the diurnalregimen, rhythmic nitrogenase activity is restricted to the darkphase and peaks approximately every 24 h. Cycles of the in-tracellular concentration of carbohydrate have also been re-ported, with maxima every 24 h just before the peak of N2fixation (48). A similar pattern of cycling of these metabolicactivities is observed in continuous-light (CL)-grown cultures.(The terms subjective-light phase and subjective-dark phaseare used in CL studies to identify the periods that correspondto the light and dark phases in LD growth.) This is a uniquefeature of Cyanothece sp., since these cultures were not previ-ously synchronized by any specific light or dark pattern.

In this work, we have studied the temporal relationshipbetween the metabolic activities of N2 fixation, photosynthesis,and respiration of Cyanothece sp. cultures grown under LD andCL conditions. We also determined that the cyclic nitrogenaseactivity was modulated at the mRNA and protein levels. Ourresults demonstrated that under both light regimens (i) allthree metabolic activities cycle, with an approximately 24-hperiodicity; (ii) N2 fixation and photosynthesis are temporallyseparated, and respiration occurred at high rates during thepeak of N2 fixation; and (iii) nitrogenase activity is strictlyregulated at the transcriptional and posttranslational levels inLD- or CL-grown cultures.

MATERIALS AND METHODS

Growth conditions. Cyanothece sp. strain ATCC 51142 (Cyanothece sp. strainBH68) was cultured at 30°C in a modified minimal salt medium (ASP2) with orwithout 1.5 g of NaNO3 per liter (42). Cultures were grown in different volumes:(i) 100-ml cultures in 250-ml Erlenmeyer flasks with shaking at 120 rpm under alight intensity of 60 microeinsteins m22 s21, (ii) 500-ml cultures in 1-liter Erlen-meyer flasks with shaking at 120 rpm under a light intensity of 60 microeinsteinsm22 s21, and (iii) 12- to 15-liter cultures in 20-liter glass carboys bubbled with airat a flow rate of 18 ft3 h21 and surrounded by light bulbs emitting an average of75 to 145 microeinsteins m22 s21. Subculture was routinely done to a finalconcentration of 106 cells ml21, using a stationary culture with a cell density ofat least 2.3 3 107 cells ml21. The cell density of the cultures was determined bythe use of a Coulter Counter (Coulter Electronics Inc., Hialeah, Fla.). The100-ml cultures were used as inocula for the 500-ml cultures which were used forphysiology experiments and as inocula for the 15-liter carboys.

The experimental design described by Schneegurt et al. (48) was used tomonitor the metabolic activities of N2 fixation, O2 photoevolution, and respira-tion over a 24-h cycle for several consecutive days. The day before the experi-ment, two experimental cultures were inoculated 12 h apart at a cell density ofapproximately 106 cells ml21, using a stationary culture with a cell densityranging between 2.6 3 107 and 5.9 3 107 cells ml21. Physiology experiments weredone under two different light regimens, CL and LD. Cultures were submitted toalternating light-dark cycles by covering them with layers of dark cloth after 12 hof light.

Nitrogenase activity. Nitrogenase activity was assayed by a modified acetylenereduction method (42, 48, 54). Acetylene was generated from calcium carbide

(Aldrich Chemical Co., Milwaukee, Wis.). Assays were performed in Vacutainertubes containing 2.0 ml of the corresponding cell sample and 15% acetylene (2.2ml) in the gas phase. Each sample was incubated for 1 h under conditions of light,temperature, and shaking similar to those of the mother cultures. The reactionswere stopped by adding 0.2 ml of 5 M NaOH. A 0.2-ml gas sample was analyzedfor its ethylene concentration by gas chromatography (Hewlett-Packard [Naper-ville, Ill.] 5890 series II gas chromatograph) with a 6-ft Poropak N column and aflame ionization detector. Gas flow rates of 40 lb of nitrogen, 20 lb of hydrogen,and 35 lb of air per in2 per min were used. The temperatures of the column andinjector port were set at 100°C. Duplicate samples were analyzed for each timepoint and corrected for the small amount of ethylene present during the assay.To calculate the amount of N2 reduced, the peak area values for each peak in thechromatograph were divided by the calculated conversion factor of 13,736.77,resulting in nanomoles of N2 reduced per hour (for assay mixtures incubated for1 h) and normalized by calculating the number of cells in 2 ml of culture. Thisconversion factor is different from that used in our previous study (48); the ratesof nitrogenase activity are reported in nanomoles of N2 reduced/108 cells/hour.

Determination of oxygen evolution and respiration rates. The photosyntheticand respiratory activities were determined by measuring the O2 production andconsumption, respectively, using a Clark electrode. Samples of different volumes,depending on cell density, were withdrawn from cultures every 2 h, centrifugedat 7,500 3 g for 10 min at 4°C, and resuspended to a final volume of 2.7 ml withfresh medium which lacked NO3. The chlorophyll (Chl) concentration of thesample was estimated from spectrophotometric measurements by using the fol-lowing equation: [Chl] (mg/ml) 5 [(OD678 2 OD750) 3 14.97] 2 [(OD620 2OD750) 3 0.615] (1), where OD is optical density and subscripts indicate wave-lengths in nanometers. These activities were measured at 30°C by using a Clark-type electrode (model 125/05; Instech Laboratories Inc., Horsham, Pa.) in awater-jacketed 3-ml stirred cuvette under heat-filtered illumination. O2 evolutionor O2 consumption of each sample was measured in the presence of 10 mMNaHCO3 and at a light intensity of 1,800 microeinsteins m22 s21 or in the dark.The rate of O2 production was obtained by adding the rates of O2 evolution inthe light and O2 consumption in the dark due to photorespiration. The rates ofO2 evolution and respiration are reported in micromoles of O2/milligram ofChl/hour.

RNA isolation and Northern blot analysis. Total RNA was extracted andpurified, using phenol-chloroform extractions and CsCl gradient purification, aspreviously described (43), with several modifications to improve the efficiency ofcell breakage. RNA was isolated from 500-ml samples every 1 to 2 h fromcultures in mid-logarithmic phase (4 3 106 to 8 3 106 cells ml21). One modifi-cation consisted of the inclusion of 425- to 600-mm-diameter glass beads (SigmaChemical, St. Louis, Mo.) during lysis. Glass beads were prepared by soakingovernight in sulfuric acid, rinsing 10 times each with distilled and nanopure water(Nanopure II, Barnstead, Dubuque, Iowa), autoclaving for 45 min, drying, andbaking at 250°C overnight. RNase-free beads were distributed in 1.9-ml bakedEppendorf tubes. Optimum conditions for cell lysis were found to be heavyvortexing with chloroform for 5 min, followed by the addition of phenol and 10min of heavy vortexing in the cold room.

Northern blots. Total RNA (5 to 15 mg/lane) was fractionated by electrophore-sis on a 1.2% agarose gel with 0.6 M formaldehyde as previously described (43,46). Gels were photographed by short-wave UV transillumination (300 nm) witha red filter, using black-and-white prints (Polaroid Film 667; Polaroid Corpora-tion, Cambridge, Mass.) and with positive-negative film (Positive/Negative 4 3 5Instant Sheet Film Polaroid; Polaroid). The gels were soaked twice in 103 SSC(13 SSC is 150 mM NaCl plus 15 mM sodium citrate; Mallinckrodt Chemical,Paris, Ky.) for 10 min and transferred to nylon membranes as previously de-scribed (46). RNA was fixed to the nylon membrane by baking at 80°C for 2 h ina vacuum oven.

The blots were hybridized with a-32P-labeled DNA probes prepared by ran-dom primer labeling using either an Oligolabelling kit or a Ready-to-go kit(Pharmacia Biotech, Piscataway, N.J.). The optimum time for labeling was foundto be 2 to 3 h at 37°C. Unincorporated radioisotope was removed by using Nickor Probe Quant G-50 columns (Pharmacia Biotech) for probes labeled with theOligolabelling or Ready-to-go kit, respectively, as recommended by the manu-facturer. Radioactive probes were boiled for 5 min and immediately put on icefor 5 min before addition to the hybridization solution. The amount of probeadded to the hybridization solution was estimated to be 10 or 2 ng/ml for probeswith a specific activity of 108 or 109 cpm/mg, respectively. Generally, the specificactivity for most of the probes was around 108 cpm/mg; very few probes werelabeled at levels as high as 109 cpm/mg.

Washes of the membranes were decided empirically depending on the ex-pected homology of the radiolabeled DNA probe and the RNA message fromCyanothece sp. The standard wash protocol (with shaking) was (i) twice for 10min each time with 23 SSC–0.1% sodium dodecyl sulfate (SDS) at room tem-perature, (ii) twice for 10 min each time with 0.13 SSC–0.1% SDS at roomtemperature, and (iii) twice for 10 min each time with 0.13 SSC–0.1% SDS at67°C. The resulting autoradiograms were scanned by using Adobe Photoshop3.0. Blots were reused by removing the probe from the membranes. Filters wereplaced in 200 ml of 13 TE buffer (10 mM Tris, 1 mM EDTA [pH 7.6]) plus 1%SDS and microwaved (Kenmore microwave; Sears Roebuck, Chicago, Ill.) athigh power for 5 min. The solution was replaced with 200 ml of 13 TE buffer,and the membrane was shaken for 30 min at 65°C. To determine the efficiency of

4320 COLON-LOPEZ ET AL. J. BACTERIOL.

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

removal of the probe, the filters were checked with a Geiger counter (modelPUG1; Technical Associates, Caloga Park, Calif.) and exposed to film overnight.Generally, one treatment was enough to remove most of the probes, and theblots were prehybridized and reused up to four times.

DNA probes. The following heterologous probes were used (15): (i) a 1.8-kbfragment containing nifH and 700 bp of upstream DNA from Anabaena sp. strainPCC 7120 (plasmid pAn154 [35]), (ii) a 0.8-kb EcoRI-KpnI fragment of nifDfrom Anabaena sp. strain PCC 7120 (plasmid pAn154 [29]), and (iii) a 0.7-kbHindIII fragment of nifK from Anabaena sp. strain 7120 (plasmid pAn207.8 [34]).A homologous probe (0.4-kb EcoRV-ApaI fragment) that corresponded to the 59coding region of nifK, the nifDK intergenic region, and some of the region 39 ofnifD (plasmid pMSC1B50) from Cyanothece sp. strain ATCC 51142 was used.

Western blot analysis. Whole-cell extracts were prepared as described previ-ously (48). Briefly, cell pellets were thawed on ice and resuspended in 10 ml ofTG buffer (10 mM Tris-HCl [pH 8.0], 10% glycerol) containing 1 mM phenyl-methylsulfonyl fluoride. The cell suspension was broken by two to three passagesthrough a French pressure cell (SLM Aminco, SLM Instruments, Inc., Urbana,Ill.) at 20,000 lb/in2. Unbroken cells and debris were pelleted at 7,500 3 g for 10min. The amount of protein was determined by using the Bradford reagent (0.1mg of Coomassie blue G-250 [Bio-Rad, Hercules, Calif.] per ml, 5% ethanol,8.5% phosphoric acid [Mallinckrodt]) and bovine serum albumin (Sigma) as thestandard (2).

Protein samples (10 to 15 mg/lane) were solubilized with 83 sample buffer (10ml of 0.5 M Tris [pH 6.8], 15 ml of 70% glycerol, 8 ml of 20% lithium dodecylsulfate, 4 ml of b-mercaptoethanol, 4 ml of 0.1% bromphenol blue) at 70°C for10 min and loaded into 10 to 20% lithium dodecyl sulfate (Gallard SchlesingerChemical Mfg. Corp., Carle Place, N.Y.)-polyacrylamide gradient gels (28).Electrophoresis was carried out at 6°C, 1.5 V, and constant power for 16 to 19 hin a Sturdier vertical slab gel unit (model SE 400; Hoefer Scientific Instruments,San Francisco, Calif.). After electrophoresis, the gels were equilibrated in trans-fer buffer {10 mM CAPS (3-[cyclohexylamino]-1-propanesulfonic acid; Sigma)(pH 11.0)} for 15 min. Proteins were transferred to nitrocellulose (Protran;Schleicher & Schuell, Keene, N.H.) membranes as previously described (56).Transfer was done in a Transphor electrophoresis unit (Hoefer), using the CAPSbuffer system (33) at 90 V for 45 min in a cold room. Filters were blocked with5% bovine serum albumin for 1 h and incubated overnight with primary antibodyin a rocker at 6°C.

Antibody probes. Antibodies were diluted with 13 TBS (50 mM Tris, 150 mMNaCl [pH 8.0]) and 0.1% NaN3. The following antibodies were provided by PaulLudden (University of Wisconsin, Madison) (30, 31) and were used in theexperiments represented in Fig. 6 to 8: (i) NifH, (ii) NifDK, (iii) dinitrogenasereductase activating glycohydrolase (DRAG), and (iv) dintrogenase reductaseADP-ribosyl transferase (DRAT). Nitrogenase antibodies, provided by DennisDean (Virginia Polytechnic Institute and State University, Blacksburg), were alsoused and provided results identical to those reported.

Electron microscopy (EM). Samples were prepared for immunolocalization byusing a freeze-substitution technique that has been described previously (50). Inbrief, the cells were embedded in a thin sheet of 1.5% agarose (Sigma type VIIlow-gelling-temperature agarose) and then plunged into liquid propane cooledto 2185°C with liquid N2. The samples were transferred to vials of 100% ethanolprecooled to 280°C, and the water was removed by substitution in ethanol overa 5-day period. The cells were infiltrated with Lowicryl HM-20, and the resin waspolymerized under UV light at 4°C. Immunolocalization on ultrathin sectionswas performed with a 10-nm colloidal gold label. Thin sections were cut, stainedwith uranyl acetate and lead citrate, and examined with a Philips EM-400 elec-tron microscope at an accelerating voltage of 80 kV. This technique preservedprotein antigenicity and permitted the use of the same antibody concentrationsas were used in the Western blot experiments (1:1,000 dilutions). The level ofnitrogenase labeling was quantitated by using IP Spectrum (Signal Analytics,Vienna, Va.) on an Apple Macintosh Quadra 950 computer. The total area of thegold particles over the cells was normalized to the total cellular area and re-ported as percentage of cell area.

RESULTS

Temporal relationship between N2 fixation, photosynthesis,and respiration in LD- and CL-grown cultures. Cyclic nitro-genase activity has previously been demonstrated in Cyanoth-ece sp. cultures grown under LD or CL conditions (42, 48). Todetermine the temporal relationship between N2 fixation, O2evolution, and respiration, these metabolic activities were as-sayed simultaneously every 1 or 2 h during several consecutivedays. Rhythms of N2 fixation, photosynthesis, and respirationwere observed for LD- and CL-grown cultures, as shown inFig. 1 to 3.

Under diurnal growth, nitrogenase activity was restricted tothe dark period and peaked approximately 4 h into this phase(Fig. 1). Maximum rates of respiration were detected just be-

fore or during the peak of N2 fixation, and the maximumcapacity of the cells to evolve O2 during photosynthesis oc-curred 6 to 8 h into the light phase (L6 to L8). Therefore, thepeak of photosynthesis is approximately 8 to 16 h out of phasewith maximum nitrogenase activity, whereas the lowest capac-ity of the cells to evolve O2 occurred during maximum nitro-genase activity. This cyclic behavior of nitrogenase was re-peated approximately every 24 h for many days (48).

This cyclic pattern was highly reproducible, although therewere some minor fluctuations as to the exact time at which N2fixation and photosynthesis reached their peak activity. Whensamples were monitored every hour, nitrogenase activitypeaked at either 3, 4, or 5 h into the dark phase (D3, D4, orD5) (data not shown). Nonetheless, the activity curve wasalways very sharp and went from a minimum to a maximumand back to the minimum within 8 h. As seen in Fig. 1, thecurves for photosynthesis and respiration activities were notnearly as sharp. Photosynthesis activity usually peaked ataround L6 to L8 and began declining while cells were stillilluminated. Respiration showed a minor peak during the lightphase but reached a true maximum at the time of peak nitro-

FIG. 1. Rhythms of N2 fixation, photosynthesis, and respiration in LD-grownCyanothece sp. strain ATCC 51142. Metabolic activities were assayed every 2 hfor 3 consecutive days. For convenience, the metabolic rates are plotted on thesame scale, and photosynthesis (maximum rate 5 616 mmol of O2/mg of Chl/h)and respiration (maximum rate 5 188 mmol of O2/mg of Chl/h) were divided by5 and 4, respectively. Nitrogenase activity maximum rate 5 140 nmol of N2reduced/108 cells/h. Solid and open bars represent dark and light growth periods,respectively. ■, nitrogenase; X, photosynthesis; å, respiration. (Inset) Growthcurves of the duplicate cultures (500 ml) used during this experiment. a.u.,arbitrary units.

VOL. 179, 1997 NITROGENASE REGULATION IN UNICELLULAR CYANOBACTERIA 4321

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

genase activity. The maximum rate of respiration of close to 200mmol of O2 consumed/mg of Chl/h is extremely high for cya-nobacteria (40, 47).

Despite these slight fluctuations, the reproducibility overmany experiments was extremely high. Figure 2 represents themeans and the standard errors of the means for 27 different24-h cycles for N2 fixation and 10 each for photosynthesis andrespiration. The analysis indicated that the data were consis-tent and highly reproducible and that on average, the maxi-mum nitrogenase, photosynthesis, and respiration activities oc-curred at D4, L6, and D4, respectively (Fig. 2).

A similar periodicity of these metabolic activities was ob-served in the CL cultures, as shown in Fig. 3. The majordifference between CL conditions (Fig. 3) and the patternsfound under LD conditions (Fig. 1) was that the cycles weresomewhat broader in CL conditions, a phenomenon that weascribed to less synchronization (48). Nitrogenase activity wasrarely depressed totally, and the period from minimumthrough maximum to minimum was 16 to 20 h, although thepeak-to-peak periodicity was generally 24 h. Respiratory activ-

ity was always significant and sometimes peaked slightly beforeor after the peak of nitrogenase activity. Photosynthesis wasalways high, varied within a narrower range than under LDgrowth conditions, and was again 8 to 16 h out of phase withnitrogenase activity. In general, these rhythms were seen topersist for many days (Fig. 3). In summary, the peaks of N2fixation and photosynthesis were temporally separated in CLconditions, as in LD conditions, although O2 evolution capac-ity was always high.

Transcriptional analysis of nif genes in LD- and CL-growncultures. Northern blot analysis was used to determine theaccumulation of nifH, nifD, and nifK transcripts throughout a24-h period for samples obtained from LD- and CL-growncultures. In initial experiments, transcripts corresponding tonifH, nifD, and nifK were detected by using gene-specific het-erologous probes from Anabaena sp. strain PCC 7120 (Fig. 4).The three genes in this operon gave similar results, and mR-NAs for the genes were present only during the first 4 h of thedark period. The estimated sizes for the transcripts are shownin Fig. 4 and correspond closely to the appropriate sizes forthese highly conserved genes (15, 16). Although the size of thelargest transcript differed somewhat in different gels, we willuse 5.0 kb as our best estimate. The transcriptional pattern was

FIG. 2. Average nitrogenase activity, photosynthesis, and respiration of LD-grown Cyanothece sp. strain ATCC 51142. The calculated rates of the metabolicactivities of N2 fixation (■), photosynthesis (X), and respiration (å) were aver-aged for every 24-h period throughout the different LD experiments. Each errorbar indicates the standard error of the mean (standard deviation divided by thesquare root of the number of samples). The units of the rates are nanomoles ofN2 reduced/108 cells/hour for N2 fixation and micromoles of O2/milligram ofChl/hour for photosynthesis and respiration. Solid and open bars indicate 12 h ofdark and light, respectively.

FIG. 3. Cycles of N2 fixation, photosynthesis, and respiration in CL-grownCyanothece sp. strain ATCC 51142. Nitrogenase activity (■) was assayed every2 h for 5 consecutive days. Photosynthesis (X) and respiration (å) were assayedevery 2 h for 3 consecutive days. The relative rates of photosynthesis (maximumrate 5 750 mmol of O2/mg of Chl/h) and respiration (maximum rate 5 158 mmolof O2/mg of Chl/h) are plotted on the same scale for convenience and weredivided by 5 and 4, respectively. Nitrogenase activity maximum rate 5 92 nmolof N2 reduced/108 cells/h. The open bar indicates that cultures were grown undercontinuous light. a.u., arbitrary units.

4322 COLON-LOPEZ ET AL. J. BACTERIOL.

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

consistent with an operon in which the genes are transcribed inthe direction nifH, nifD, and nifK, which is similar to othercyanobacteria and bacteria (16). The results may also suggestthe presence of mRNA processing or RNA polymerase paus-ing which resulted in stable, gene-sized transcripts (e.g., 1.3 kbfor nifH). These transcripts were stable enough to be detected,but we are uncertain if these smaller transcripts are functionalin directing synthesis of the specific proteins.

The relative accumulation of transcript was studied in fur-ther detail by isolating RNA every hour throughout the darkphase. As shown in Fig. 4D, the level of mRNA throughout thedark phase was determined by using a homologous probewhich corresponded to the nifK gene from the Cyanothece sp.strain ATCC 51142 nifHDK operon, which we have cloned and

sequenced (4a). The mRNA levels peaked at D2, and tran-scription of the operon virtually ended by D5. These experi-ments were again highly reproducible and indicate that thelevels of accumulation of nifHDK transcripts differ substan-tially throughout the 24-h cycle.

The hypothesis that these phenomena are under the controlof a circadian oscillator gained credence with the experimentin Fig. 5. In this experiment, RNA was extracted from cellsgrown under LD conditions every 2 h over a 2-day period andwas hybridized to the homologous nifK probe. RNA sampleswere collected during an experiment in which the peak ofnitrogenase activity occurred at D4 to D6. The net level oftranscription of nifK reached a maximum at D4, then stopped,and resumed the next day at D4. Therefore, transcription fol-lows a circadian pattern, and net nifHDK transcription wasshut down almost 20 h each day.

All of the foregoing results were paralleled by those ofexperiments done with cells grown under CL conditions (Fig.5B). Once again, net accumulation of the nifHDK transcriptsoccurred almost exclusively in the subjective-dark phase (CL14to CL20), and the phenomenon repeated each subjective nightover a 2-day period (Fig. 5B). A difference was that, similar tothe physiology, the curve for this phenomenon was less sharpand net transcription was at significant levels for approximately8 h. On both days, the net level of nifHDK mRNA was low ornegligible during the subjective-light phase (CL0 to CL10).Taken together, these results indicated significant transcrip-tional modulation of the nifHDK operon under both LD andCL growth conditions.

Analysis of nitrogenase levels in LD- and CL-grown cul-tures. To determine if the levels of the nitrogenase proteins(the Fe-protein and the MoFe-protein) remained constantthroughout a 24-h period, whole-cell extracts were made fromsamples collected every 2 h and analyzed via Western blottingwith appropriate antibody. Immunoreactive bands correspond-ing to the NifH (Fe-protein) and NifDK (MoFe-protein) pro-teins were detected by using antisera from Rhodospirillumrubrum (Fig. 6). For cells grown under LD conditions (Fig.6A), the levels of both proteins were low at the beginning ofthe dark phase, increased to a maximum at approximately D6,and decreased to virtually zero by early in the light phase. Asimilar level of accumulation was obtained for NifH in both LDand CL samples (compare Fig. 6A and B). On the contrary, theaccumulation of the NifDK proteins differed for cultures

FIG. 4. Northern blot analysis of total RNA from Cyanothece sp. strainATCC 51142 grown under light-dark N2-fixing conditions. (A to C) Total RNAsamples (15 mg/lane) were extracted every 2 h over a 24-h period. HeterologousDNA probes from Anabaena sp. strain PCC 7120 corresponding to the genesnifH (A), nifD (B), and nifK (C) were used for hybridization. Arrows indicate theestimated sizes of the transcripts. (D) Similar to panels A to C except that RNAwas extracted every hour (beginning 108 h after subculturing) during the darkphase of a different growth experiment. Total RNA samples (5 mg/lane) werehybridized to a 0.4-kb fragment of nifK from Cyanothece sp. strain ATCC 51142.The uppermost band (4.5 kb) represents the nifHDK transcript, whereas theother arrows correspond to rRNA.

FIG. 5. Northern blot analysis of nifHDK transcription over two consecutivedays. Total RNAs from Cyanothece sp. strain ATCC 51142 grown under LD (A)or CL (B) conditions were extracted every 2 h and loaded onto gels. In bothexperiments, 5 mg of total RNA/lane was hybridized against a 0.4-kb nifK frag-ment from Cyanothece sp. strain ATCC 51142. Arrows indicate the estimated sizeof the nifHDK transcript.

VOL. 179, 1997 NITROGENASE REGULATION IN UNICELLULAR CYANOBACTERIA 4323

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

grown under either LD or CL conditions. In LD-grown cells,NifDK declined in a fashion similar to that of NifH. However,in extracts from CL-grown cells, NifDK never completely dis-appeared and remained at lower but significant levels through-out the 24-h cycle.

The temporal accumulation of NifH and NifDK proteinswas also analyzed by EM immunocytochemistry using freeze-substituted cells. The synthesis and degradation of nitrogenasewere graphically depicted by this technique, as indicated in Fig.7. This experiment used the NifH antibody, although virtuallyidentical results were obtained with the NifDK antibody. It wasevident that very little NifH was present at D0 and that therewas a significant increase in the level of NifH by D2. Theamount of protein increased through D4 and D6, with theenzyme present throughout the cytoplasm and around the car-bohydrate granules. Based on the EM and additional biochem-istry results (47b), we do not believe that the nitrogenase islocated in any compartment but rather that it is free in thecytoplasm. As depicted in Fig. 7, a major change took place inthe accumulation of the NifH protein between D6 and D9. ByD9, the amount of Fe-protein present was extremely low, sug-gesting rapid degradation of the protein during the D6 to D9time period. The amount of nitrogenase remained at a low,basal level throughout the dark period, as shown for L9 in Fig.7.

The quantitation for both NifH and NifDK antibodies canbe seen in Fig. 8. It is clear that either nitrogenase subunit goesfrom a very low level to a peak within the first 4 h in the darkand then declines rapidly after D6. Similar to what was seen inthe Western blots, the NifH declines more rapidly to a lowerlevel, whereas NifDK continues at some modest level through-out the diurnal cycle. These results, obtained by using bothWestern blotting and EM immunocytochemistry, representedstrong evidence for a periodic proteolysis of nitrogenase inCyanothece sp. strain ATCC 51142.

DISCUSSION

The results presented above document the role of transcrip-tion and proteolysis in the metabolic activities of Cyanothecesp. strain ATCC 51142 during diazotrophic growth under ei-ther LD or CL conditions. The nitrogenase enzyme, encodedby the nifHDK operon, is under very tight control, with net

transcriptional accumulation restricted to approximately 4 to6 h in LD cells or approximately 8 h in CL cultures. Undereither light regimen, the accumulation of nitrogenase tran-scripts peaked early in the dark or the subjective-dark periodand began to decline soon thereafter. The similarity in tran-scriptional patterns between the two growth conditions sug-gested that accumulation of nitrogenase mRNA is not directlyregulated by light.

The fate of nitrogenase after the peak of activity was mostinteresting. The quantity of enzyme declined from a peak to aminimum within a few hours in LD cultures, a conclusionverified by both Western blotting and EM immunocytochem-istry. Typically, protein degradation was less well synchronizedin CL cultures, and it took a few hours longer before all of theNifH protein was destroyed. However, our results indicatedthat the NifDK protein remained at substantial levels through-out the 24-h period in CL cultures. As indicated in Fig. 5B,there was some low level of transcription of nifK in the sub-jective-light period. However, there was a similar level of tran-scription of the nifH gene, and the NifH protein has virtuallydisappeared by CL6. Thus, it would appear that most of thedifference between NifDK and NifH is that NifDK is not de-graded completely under CL conditions.

The fact that the metabolic cycles for N2 fixation, photosyn-thesis, and respiration occurred in both LD and CL culturesstrongly suggested that the regulation is based on an underly-ing circadian oscillator. In other studies, we have shown thatthese rhythms meet all of the criteria for circadian rhythms,including temperature compensation, and we concluded thatthe metabolic oscillations represent true circadian rhythms(47a). This phenomenon was first identified in the diazotrophiccyanobacterium Synechococcus sp. strain RF1 (22, 23). Morerecently, such circadian rhythms have been demonstrated inSynechococcus sp. strain PCC 7942, a unicellular, nondiazotro-phic cyanobacterium (24). By fusing an upstream regulatoryregion of the psbA gene to a luciferase reporter gene (26), thetranscription of psbA was shown to be controlled by a circadianrhythm in Synechococcus sp. strain PCC 7942. Circadianrhythm mutants of this organism have now been isolated (27),and it was demonstrated that the circadian clock controls geneexpression, even in cells doubling more rapidly than once every24 h (24, 38).

Respiration has an important role in diazotrophy in Cyan-othece sp. strain ATCC 51142. Respiration may provide asource of ATP for N2 fixation, and there was a substantial peakof respiratory activity near the peak of nitrogenase activity.Under both LD and CL growth conditions, respiration is veryrobust, with maximal rates approximating 200 mmol of O2/mgof Chl/h, which is extremely high for cyanobacteria (40, 47).The peak of respiration correlated well with the decline in theglycogen granules, which are likely to be the stored substratefor respiration (48). It is also likely that respiration plays a rolein scavenging free O2 in the cytoplasm, thereby preventingoxidative damage to the nitrogenase proteins. Since photosyn-thetic O2 evolution declines to a minimum at about the sametime, it is logical to assume that the decrease in intracellular O2concentration is one strategy used by Cyanothece sp. to protectnitrogenase from oxidative damage. We have also shown thatmaximal nitrogenase activity can take place in the absence ofphotosynthesis (34a), thus supporting the idea that respirationcan provide the energy and reductant needed for N2 fixation.

Most of the studies of posttranslational modification con-cerning the nitrogenase enzyme have focused on the Fe-pro-tein. Degradation of the Fe-protein upon illumination hasbeen suggested in studies performed with Gloeothece sp., Syn-echococcus sp. strain RF1, and Trichodesmium sp. (3, 4, 61). In

FIG. 6. Western blot analysis of the Fe-protein and the MoFe-protein ofCyanothece sp. strain ATCC 51142 grown under LD (A) or CL (B) conditions.Samples were withdrawn at 2-h intervals for a 24-h period. Antisera against theFe-protein (anti-NifH, 1/1,000 dilution) and MoFe-protein (anti-NifDK, 1/1,000dilution) of R. rubrum were used. (A) Ten micrograms of protein/lane; (B) 15 mgof protein/lane. The arrows show the estimated sizes for the NifH (Fe-protein)and NifDK (Mo Fe-protein) proteins.

4324 COLON-LOPEZ ET AL. J. BACTERIOL.

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

FIG. 7. Immunolocalization of nitrogenase Fe-protein (NifH) on LD-grown cultures of Cyanothece sp. strain ATCC 51142. Cells were collected and prepared atdifferent times during the light and dark periods as indicated in Materials and Methods. The antibody against NifH was used at a 1:1,000 dilution for cells collectedat the indicated times.

VOL. 179, 1997 NITROGENASE REGULATION IN UNICELLULAR CYANOBACTERIA 4325

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

every case, two forms of the Fe-protein have been immunode-tected, suggesting that the high-molecular-weight form corre-sponded to a posttranslationally modified Fe-protein and thatthis modification may have a role in the degradation pathway.With our protein gel system, we have not detected differentforms of the Fe-protein in Cyanothece sp. In cyanobacteria,neither the type of modification, the signal, nor the degrada-tion pathway is known, although O2 has been suggested as asignal (17). Posttranslational modification of the Fe-proteinhas also been found in Anabaena sp. (6, 7), where it has beendemonstrated that it is not ADP-ribosylated as in R. rubrum(30, 31). Attempts to detect the enzymes responsible for ADP-ribosylation (e.g., DRAG and DRAT) with the appropriateantibodies using whole-cell extracts of LD-grown Cyanothecesp. strain ATCC 51142 have been unsuccessful (4a). Targetedproteolytic degradation of the nitrogenase subunits may bepossible, since ubiquitin and ubiquitinated Fe-protein havebeen detected in Anabaena sp. (6). Recently, it was reportedthat both the modified and unmodified forms of the Fe-proteinwere proteolytically degraded as a consequence of O2 concen-tration and ATP depletion (7). Posttranslational regulation ofthe MoFe-protein has not been studied in as much detail as theprocess for the Fe-protein.

This work has described the nature of temporal separationof N2 fixation and photosynthesis in a unicellular, diazotrophiccyanobacterium. This seems to be the strategy used by unicel-lular cyanobacteria such as Synechococcus sp. strain RF1, Syn-echococcus sp. strain Miami BG 43511, and Gloeothece sp., aswell as nonheterocystous, filamentous cyanobacteria such asPlectonema sp. and Oscillatoria sp. (36, 39, 41, 45, 52, 53). Wesuggest that high rates of respiration and decreased photosyn-thesis at D4 generate an intracellular O2 concentration lowenough for N2 fixation. After this time, a decrease in respira-tion permits the intracellular O2 concentration to rise, leadingto oxidative damage of the subunits and subsequent degrada-tion. Thus, Cyanothece sp. strain ATCC 51142 utilizes a ratherexpensive, but safe, approach to O2 protection of nitroge-nase—the enzyme is synthesized de novo and degraded eachday by an underlying mechanism that is under the control of acircadian oscillator.

ACKNOWLEDGMENTS

We gratefully acknowledge the technical help of Wendy O. Adam-owicz and Hsiao-Yuan (Vicky) Tang at various times throughout thisproject and the kindness of Paul Ludden and Dennis Dean for pro-viding antibodies. We particularly acknowledge the input of MarkSchneegurt for providing the cells used in the EM experiment. We alsoacknowledge the use of the facilities of the Electron Microscopy Cen-ter in the School of Agriculture at Purdue University for this research.

This work was supported by USDA grant 93-37306-9238 (to L.A.S.)and by fellowships from the American Society for Microbiology, fromthe National Hispanic Scholarship Foundation, and from NIH (F31GM16400-01) (to M.S.C.-L.).

REFERENCES

1. Arnon, D. I., B. D. McSwain, H. Y. Tsujimoto, and K. Wada. 1974. Photo-chemical activity and components of membrane preparations from blue-green algae. 1. Coexistence of two photosystems in relation to chlorophyll aand removal of phycocyanin. Biochim. Biophys. Acta 357:231–245.

2. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye bind-ing. Anal. Biochem. 72:248–254.

3. Capone, D. G., J. M. O’Neil, J. Zehr, and E. J. Carpenter. 1990. Basis for dielvariation in nitrogenase activity in the marine planktonic cyanobacteriumTrichodesmium thiebautii. Appl. Environ. Microbiol. 56:3532–3536.

4. Chow, T.-J., and F. R. Tabita. 1994. Reciprocal light-dark transcriptionalcontrol of nif and rbc expression and light-dependent posttranslational con-trol of nitrogenase activity in Synechococcus sp. strain RF-1. J. Bacteriol.176:6281–6285.

4a.Colon-Lopez, M. S., and L. A. Sherman. Unpublished data.5. Du, C., and J. R. Gallon. 1993. Modification of the Fe protein of the

nitrogenase of Gloeothece (Nageli) sp. ATCC 27152 during growth underalternating light and darkness. New Phytol. 125:121–129.

6. Durner, J., and P. Boger. 1995. Ubiquitin in the prokaryote Anabaena varia-bilis. J. Biol. Chem. 270:3720–3725.

7. Durner, J., I. Bohm, H. Hilz, and P. Boger. 1994. Post translational modifi-cation of nitrogenase. Differences between the purple bacterium Rhodospi-rillum rubrum and the cyanobacterium Anabaena variabilis. Eur. J. Biochem.220:125–130.

8. Eady, R. R., S. Imam, D. J. Lowe, R. W. Miller, B. E. Smith, and R. N. F.Thorneley. 1980. The molecular enzymology of nitrogenase, p. 19–35. InW. D. P. Stewart and J. R. Gallon (ed.), Nitrogen fixation. Academic PressLtd., London, England.

9. Fay, P. 1992. Oxygen relations of nitrogen fixation in cyanobacteria. Micro-biol. Rev. 56:340–373.

10. Gallon, J. R., and A. F. Hamadi. 1984. Studies on the effect of oxygen onacetylene reduction (N2-fixation) in Gloeothece sp. ATCC 27152. J. Gen.Microbiol. 130:495–503.

11. Gallon, J. R., and A. E. Chaplin. 1987. An introduction to nitrogen fixation.Cassell, London, England.

12. Gallon, J. R., S. M. Perry, T. M. A. Rajab, K. A. M. Flayeh, J. S. Jones, andA. E. Chaplin. 1988. Metabolic changes associated with the diurnal patternof N2-fixation in Gloeothece. J. Gen. Microbiol. 134:3079–3087.

13. Gallon, J. R. 1992. Tansley review no. 44. Reconciling the incompatible: N2fixation and O2. New Phytol. 122:571–609.

14. Grobbelaar, N., H. Y. Lin, and T.-C. Huang. 1987. Induction of a nitrogenaseactivity rhythm in Synechococcus and the protection of its nitrogenase againstphotosynthetic oxygen. Curr. Microbiol. 15:29–33.

15. Haselkorn, R. 1986. Organization of the genes for nitrogen fixation in pho-tosynthetic bacteria and cyanobacteria. Annu. Rev. Microbiol. 40:525–547.

16. Haselkorn, R., and W. J. Buikema. 1992. Nitrogen fixation in cyanobacteria,p. 166–190. In G. Stacey, R. H. Burris, and H. J. Evans (ed.), Biologicalnitrogen fixation. Chapman and Hall, New York, N.Y.

17. Hill, S. 1988. How is nitrogenase regulated by oxygen? FEMS Microbiol.Rev. 54:111–130.

18. Huang, T.-C., and T. J. Chow. 1986. New type of N2-fixing unicellularcyanobacterium (blue-green alga). FEMS Microbiol. Lett. 36:109–110.

19. Huang, T.-C., and T. J. Chow. 1988. Comparative studies of some nitrogen-fixing unicellular cyanobacteria isolated from rice fields. J. Gen. Microbiol.134:3087–3089.

20. Huang, T.-C., and T. J. Chow. 1990. Characterization of the rhythmic nitro-gen-fixing activity of Synechococcus sp. RF1 at the transcription level. Curr.Microbiol. 20:23–26.

21. Huang, T.-C., T. J. Chow, and I. S. Hwang. 1988. The cyclic synthesis of thenitrogenase of Synechococcus RF1 and its control at the transcriptional level.FEMS Microbiol. Lett. 50:127–130.

22. Huang, T.-C., J. Tu, T. J. Chow, and T. H. Chen. 1990. Circadian rhythm ofthe prokaryote Synechococcus sp. RF1. Plant Physiol. 92:531–533.

23. Huang, T.-C., and N. Grobbelaar. 1995. The circadian clock in the pro-karyote Synechococcus RF-1. Microbiology 141:535–540.

24. Johnson, C. H., S. S. Golden, M. Ishiura, and T. Kondo. 1996. Circadian

FIG. 8. Quantitation of the immunolocalization experiments with the NifHand NifDK antibodies for Cyanothece sp. strain ATCC 51142 grown under LDconditions. Experiments such as the one depicted in Fig. 7 were analyzed quan-titatively as described in Materials and Methods. D0, D2, D4, D6, D9, L4, and L9indicate the time at which the sample was taken from the culture.

4326 COLON-LOPEZ ET AL. J. BACTERIOL.

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

clocks in prokaryotes. Mol. Microbiol. 21:5–11.25. Khamees, H. S., J. R. Gallon, and A. E. Chaplin. 1987. The pattern of

acetylene reduction by cyanobacteria grown under alternating light anddarkness. Br. Phycol. J. 22:55–60.

26. Kondo, T., C. A. Strayer, R. D. Kulkarni, W. Taylor, M. Ishiura, S. S.Golden, and C. H. Johnson. 1993. Circadian rhythms in prokaryotes: lucif-erase as a reporter of circadian gene expression in cyanobacteria. Proc. Natl.Acad. Sci. USA 90:5672–5676.

27. Kondo, T., N. F. Tsinoremas, S. S. Golden, C. H. Johnson, S. Kutsuna, andM. Ishiura. 1994. Circadian clock mutants of cyanobacteria. Science 266:1233–1236.

28. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature 227:680–685.

29. Lammers, P. J., and R. Haselkorn. 1983. Sequence of the nifD gene codingfor the a subunit of dinitrogenase from the cyanobacterium Anabaena.Proc. Natl. Acad. Sci. USA 80:4723–4727.

30. Ludden, P. W., G. P. Roberts, R. G. Lorvery, W. P. Fitzmaurice, L. L. Saari,L. Lehman, D. Lies, D. Wochle, H. Wirt, S. A. Murrell, M. R. Pope, and R. H.Kanemato. 1988. Regulation of nitrogenase activity by reversible ADP-ribo-sylation of dinitrogenase reductase, p. 157–162. In H. Bothe, F. J. deBruijn,and W. E. Newton (ed.), Nitrogen fixation: hundred years after. Fisher,Stuttgart, Germany.

31. Ludden, P. W., and G. P. Roberts. 1989. Regulation of nitrogenase activity byreversible ADP ribosylation. Curr. Top. Cell Regul. 30:23–56.

32. Maryan, P. S., R. R. Eady, A. E. Chaplin, and J. R. Gallon. 1986. Nitrogenfixation by Gloeothece sp. PCC 6909; respiration and not photosynthesissupports nitrogenase activity in the light. J. Gen. Microbiol. 132:789–796.

33. Matsudaira, P. 1987. Sequence from picomole quantities of proteins elec-troblotted onto polyvinylidene difluoride membranes. J. Biol. Chem. 262:10035–10038.

34. Mazur, B. J., C. F. Chui, and R. Haselkorn. 1982. Sequence of the genecoding for the b-subunit of dinitrogenase from the blue-green algaAnabaena. Proc. Natl. Acad. Sci. USA 79:6782–6786.

34a.Meunier, P., and L. A. Sherman. Unpublished data.35. Mevarech, M., D. Rice, and R. Haselkorn. 1980. Nucleotide sequence of a

cyanobacterial nifH gene coding for nitrogenase reductase. Proc. Natl. Acad.Sci. USA 77:6476–6480.

36. Mitsui, A., S. Kumazawa, A. Takahashi, H. Ikemoto, S. Cao, and T. Arai.1986. Strategy by which N2-fixing unicellular cyanobacteria grow photoau-totrophically. Nature 323:720–722.

37. Mitsui, A., and S. Kumazawa. 1988. Nitrogen fixation by synchronouslygrowing unicellular aerobic nitrogen-fixing cyanobacteria. Methods Enzy-mol. 167:484–490.

38. Mori, T., B. Biner, and C. H. Johnson. 1996. Circadian gating of cell divisionin cyanobacteria growing with average doubling times of less than 24 hours.Proc. Natl. Acad. Sci. USA 93:10183–10188.

39. Mullineaux, P. M., J. R. Gallon, and A. E. Chaplin. 1981. Acetylene reduc-tion (nitrogen fixation) by cyanobacteria grown under alternating light darkcycles. FEMS Microbiol. Lett. 10:245–247.

40. Peschek, G. A., K. Villgrater, and M. Wastyn. 1991. ‘Respiratory protection’of the nitrogenase in dinitrogen-fixing cyanobacteria. Plant Soil 137:17–24.

41. Rai, A. N., M. Bozthakur, and B. Bergman. 1992. Nitrogenase derepression,its regulation and metabolic changes associated with diazotrophy in thenon-heterocystous cyanobacterium Plectonema boryanum PCC73110. J. Gen.Microbiol. 138:481–491.

42. Reddy, K. J., J. B. Haskell, D. M. Sherman, and L. A. Sherman. 1993.Unicellular, aerobic nitrogen-fixing cyanobacteria of the genus Cyanothece.

J. Bacteriol. 175:1284–1292.43. Reddy, K. J., R. Webb, and L. A. Sherman. 1990. Bacterial RNA isolation

with one hour centrifugation in a table-top ultracentrifuge. BioTechniques8:250–251.

44. Robson, R. L., and J. R. Postgate. 1980. Oxygen and hydrogen in biologicalnitrogen fixation. Annu. Rev. Microbiol. 34:183–207.

45. Rojek, R., C. Harms, M. Hebeler, and L. H. Grimme. 1994. Cyclic variationsof photosynthetic activity under nitrogen fixing conditions in SynechococcusRF-1. Arch. Microbiol. 162:80–84.

46. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

47. Scherer, S., U. Hafele, G. H. J. Kruger, and P. Boger. 1988. Respiration,cyanide-insensitive oxygen uptake and oxidative phosphorylation in cya-nobacteria. Physiol. Plant. 72:379–384.

47a.Schneegurt, M. A., M. S. Colon-Lopez, and L. A. Sherman. Unpublisheddata.

47b.Schneegurt, M. A., and L. A. Sherman. Unpublished data.48. Schneegurt, M. A., D. M. Sherman, S. Nayar, and L. A. Sherman. 1994.

Oscillating behavior of carbohydrate granule formation and dinitrogen fix-ation in the cyanobacterium Cyanothece sp. strain ATCC 51142. J. Bacteriol.176:1586–1597.

49. Shah, V. K., and W. J. Brill. 1977. Isolation of an iron-molybdenum cofactorfrom nitrogenase. Proc. Natl. Acad. Sci. USA 74:3249–3253.

50. Sherman, D. M., T. A. Troyan, and L. A. Sherman. 1994. Localization ofmembrane proteins in the cyanobacterium Synechococcus sp. PCC7942.Plant Physiol. 106:251–262.

51. Spiller, H., and K. T. Shammugan. 1987. Physiological conditions for nitro-gen fixation in a unicellular marine cyanobacterium, Synechococcus sp. strainSF1. J. Bacteriol. 169:5379–5384.

52. Stal, L. J., and W. E. Krumbein. 1985. Nitrogenase activity in the non-heterocyst cyanobacterium Oscillatoria sp. grown under alternating light-dark cycle. Microbiology 143:67–71.

53. Stal, L. J., and W. E. Krumbein. 1987. Temporal separation of nitrogenfixation and photosynthesis in the filamentous, non-heterocystous cyanobac-terium Oscillatoria sp. Arch. Microbiol. 149:76–80.

54. Stewart, W. D. P., G. P. Fitzgerald, and R. H. Burris. 1968. Acetylenereduction by nitrogen-fixing blue-green algae. Arch. Microbiol. 62:336–348.

55. Stewart, W. D. P., and M. Lex. 1970. Nitrogenase activity in the blue-greenalga Plectonema boryanum strain 594. Arch. Microbiol. 73:250–260.

56. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: procedure andsome applications. Proc. Natl. Acad. Sci. USA 76:4350–4354.

57. Tuli, R., S. Naithari, and H. S. Misra. 1996. Cyanobacterial photosynthesisand the problem of oxygen in nitrogen-fixation: a molecular genetic view. J.Sci. Ind. Res. 55:638–657.

58. Wolk, C. P., A. Ernst, and J. Elhai. 1994. Heterocyst metabolism and devel-opment, p. 769–823. In D. A. Bryant (ed.), The molecular biology of cya-nobacteria. Kluwer Academic Publishers, Dordrecht, The Netherlands.

59. Wyatt, J. T., and J. K. G. Silvey. 1969. Nitrogen fixation by Gleocapsa.Science 165:908–909.

60. Young, J. P. W. 1992. Phylogenetic classification of nitrogen-fixing organ-isms, p. 43–86. In G. Stacey, R. H. Burris, and H. J. Evans (ed.), Biologicalnitrogen fixation. Chapman and Hall, New York, N.Y.

61. Zehr, J. P., M. Wyman, V. Miller, L. Duguay, and D. G. Capone. 1993.Modification of the Fe protein of nitrogenase in natural populations ofTrichodesmium thiebautii. Appl. Environ. Microbiol. 59:669–676.

VOL. 179, 1997 NITROGENASE REGULATION IN UNICELLULAR CYANOBACTERIA 4327

on April 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from