ABSTRACT
In the case of nitrogenase-based photobiological hydrogen production systems of cyanobacteria, the inactivation of uptake hydrogenase (Hup) leads to significant increases in hydrogen production activity. However, the high-level-activity stage of the Hup mutants lasts only a few tens of hours under air, a circumstance which seems to be caused by sufficient amounts of combined nitrogen supplied by active nitrogenase. The catalytic FeMo cofactor of nitrogenase binds homocitrate, which is required for efficient nitrogen fixation. It was reported previously that the nitrogenase from the homocitrate synthase gene (nifV) disruption mutant of Klebsiella pneumoniae shows decreased nitrogen fixation activity and increased hydrogen production activity under N2. The cyanobacterium Nostoc sp. strain PCC 7120 has two homocitrate synthase genes, nifV1 and nifV2, and with the ΔhupL variant of Nostoc sp. strain PCC 7120 as the parental strain, we have constructed two single mutants, the ΔhupL ΔnifV1 strain (with the hupL and nifV1 genes disrupted) and the ΔhupL ΔnifV2 strain, and a double mutant, the ΔhupL ΔnifV1 ΔnifV2 strain. Diazotrophic growth rates of the two nifV single mutants and the double mutant were decreased moderately and severely, respectively, compared with the rates of the parent ΔhupL strain. The hydrogen production activity of the ΔhupL ΔnifV1 mutant was sustained at higher levels than the activity of the parent ΔhupL strain after about 2 days of combined-nitrogen step down, and the activity in the culture of the former became higher than that in the culture of the latter. The presence of N2 gas inhibited hydrogen production in the ΔhupL ΔnifV1 ΔnifV2 mutant less strongly than in the parent ΔhupL strain and the ΔhupL ΔnifV1 and ΔhupL ΔnifV2 mutants. The alteration of homocitrate synthase activity can be a useful strategy for improving sustained photobiological hydrogen production in cyanobacteria.
The amount of solar energy received on the earth's surface is vast, and consequently, the photobiological production of H2 by cyanobacteria and microalgae that use H2O as the electron donor is expected to become an environmentally acceptable renewable energy alternative on a large scale (9, 30, 31, 35, 36). The enzyme that catalyzes H2 production in cyanobacteria is either nitrogenase or hydrogenase (31, 42). We have chosen the nitrogenase system as a starting point for hydrogen production based upon its potential for optimization as part of long-term research and development strategies (35, 36). Notably, nitrogenase catalyzes essentially unidirectional production of H2, in contrast with reversible or bidirectional hydrogenase, which catalyzes reversible reactions of evolution and absorption of H2 in the presence of O2.
The biological challenge for the future development of cyanobacterial nitrogenase-based H2 production is the presence of hydrogenases, especially uptake hydrogenase (Hup), which reabsorbs the H2 produced and decreases the efficiency of H2 production (42). The inactivation of Hup has been proven to be an effective strategy for improving the efficiency of nitrogenase-based photobiological H2 production by several Anabaena and Nostoc spp. strains of cyanobacteria (10, 23, 26, 38, 50). When these inactivation mutants were transferred into combined-nitrogen-free medium, they developed N2 fixation and net H2 evolution activity in a few days, with the activities of the mutants under Ar about three to seven times higher than those of the respective wild-type strains from which they were derived (10, 26, 38, 50). At the maximum-activity stages, the mutant strains frequently achieved activities of about 100 μmol of H2 mg of chlorophyll a−1 h−1 or higher under Ar (10, 36, 38, 50). However, under other gas phases, such as air that contained N2, the high H2 production activities of these mutants typically decreased quickly over a few tens of hours. In the absence of N2, such as under Ar, a stage of high-level H2 production lasted much longer than under air, and the culture accumulated H2 to 29% (vol/vol), but the activity gradually decreased in about a week (50). These results seem to be closely related to the nutritional state of combined nitrogen in cyanobacterial cells: nitrogenase-based H2 production activity is high or low according to the high or low requirement for combined nitrogen, respectively. On the other hand, cells seem to require at least a modest supply of combined nitrogen for maintaining various cell activities. When a low concentration of N2 (1%) was included in the Ar mixture (8) or low concentrations of ammonium (10−4 to 5 × 10−4 M) were frequently added (every 1 to 3 days) to the culture (45), the nitrogen-fixing cyanobacterium Anabaena cylindrica showed sustained H2 production lasting for about 20 to 30 days under a continuous flow of gas.
The nitrogenase reaction is catalyzed by the association of two proteins, catalytic dinitrogenase (MoFe protein) and dinitrogenase reductase (Fe protein), which supplies electrons from reduced ferredoxin or flavodoxin to the catalytic dinitrogenase in an ATP-dependent manner. Under optimal conditions for N2 reduction, the nitrogenase reaction is expressed as follows: $$mathtex$$\[\mathrm{N}_{2}{+}8e^{{-}}{+}8\mathrm{H}^{{+}}{+}16\mathrm{ATP}{\rightarrow}\mathrm{H}_{2}{+}2\mathrm{NH}_{3}{+}16(\mathrm{ADP}{+}\mathrm{P}_{\mathrm{i}})\]$$mathtex$$(1) H2 is produced as an inevitable by-product, and the reaction is irreversible (17, 32). Equation 1 represents the optimum case for N2 fixation, and the reaction is more generally expressed as follows: $$mathtex$$\[\mathrm{N}_{2}{+}(6{+}2n)e^{{-}}{+}(6{+}2n)\mathrm{H}^{{+}}{+}2(6{+}2n)\mathrm{ATP}{\rightarrow}\ n\mathrm{H}_{2}{+}2\mathrm{NH}_{3}{+}2(6{+}2n)(\mathrm{ADP}{+}\mathrm{P}_{\mathrm{i}})\ (n{\geq}1)\]$$mathtex$$(2) In the absence of N2 (e.g., under Ar), all the electrons are allocated to H2 production: $$mathtex$$\[2e^{{-}}{+}2\mathrm{H}^{{+}}{+}4\mathrm{ATP}{\rightarrow}\mathrm{H}_{2}{+}4(\mathrm{ADP}{+}\mathrm{P}_{\mathrm{i}})\]$$mathtex$$(3) The ratio of electrons allocated to N2 fixation to the total electrons used in the nitrogenase reaction is called an electron allocation coefficient. It is 0.75 in equation 1 but generally smaller than 0.75 in equation 2 and 0 in equation 3. It is expected that by decreasing the electron allocation coefficient, a larger fraction of the electrons will be allocated to H2 production and the activity of nitrogenase will be sustained by decreasing the supply of combined nitrogen even under the gas phase containing N2.
The catalytic center of dinitrogenase normally binds the FeMo cofactor to which homocitrate is ligated (13, 14, 15, 16, 20). Homocitrate is synthesized by the nifV-encoded enzyme homocitrate synthase, which catalyzes the condensation of acetyl coenzyme A and 2-oxoglutarate (51). Nuclear magnetic resonance (22) and crystallographic (27) analyses have revealed that dinitrogenase purified from a nifV mutant of the facultative anaerobic bacterium Klebsiella pneumoniae has a modified FeMo cofactor that binds citrate instead of homocitrate. In vitro biochemical studies indicated that the modified citrate-containing dinitrogenase from K. pneumoniae nifV mutants reduces N2 poorly; however, the reduction of acetylene and proton under Ar proceeds at rates comparable to those for the wild-type enzyme (11, 12, 28). The modified enzyme has a lower electron allocation coefficient (0.54 [21] versus about 0.75 [34]) and a higher Km for N2 (approximately 0.24 atm [21] versus approximately 0.12 atm [7]) than the wild-type enzyme.
In vivo activities differ among several nifV mutants constructed from various microorganisms, as reported in the references cited below. The nifV mutants from K. pneumoniae retain about 80% of the acetylene reduction activity of the wild type but are unable to grow diazotrophically (with N2 as the sole source of nitrogen nutrition) (28). The Azotobacter chroococcum nifV mutant fails to grow diazotrophically and shows no ability to reduce acetylene and N2 (6). In contrast, the nifV mutants from Azotobacter vinelandii (24) and Rhodobacter capsulatus (25) retain about 10% of the acetylene reduction activities of the respective wild-type strains and are capable of very slow diazotrophic growth. Citrate was suggested to be present in about 30% of the purified MoFe protein from the A. vinelandii nifV mutant (29) and about 50% from the K. pneumoniae nifV mutant (27). The growth rates and the in vivo activities of N2, acetylene, and proton reduction of several nifV mutants may be differently affected among bacterial strains, probably due to subtle differences in the ratio of the occupancy of citrate in the FeMo cofactor, in the arrangement of nif-related genes downstream of nifV, and in the nitrogen nutritional conditions of the starting culture, etc.
In Anabaena (Nostoc) sp. strain PCC 7120 (hereafter referred to as Nostoc sp. strain PCC 7120), Stricker et al. (40) identified a nifV1 gene that is contiguous with the nifZ and nifT genes and noted the presence of another nifV gene in the genome that was not sequenced in the study. They generated nifV1 disruption mutants and found that the mutants exhibited about 70% of the diazotrophic growth of the wild type 10 days after nitrogen deprivation. Subsequently, whole-genome sequencing of Nostoc sp. strain PCC 7120 (19) revealed the presence of two nifV genes, designated nifV1 (alr1407) and nifV2 (alr2968) (see CyanoBase [http://bacteria.kazusa.or.jp/cyano/index.html ]). The nifV2 gene shows a high degree of sequence homology to the nifV1 gene of the same strain (74% identity at the nucleotide level and 78% identity at the amino acid level). In contrast with nifV1, which is located adjacent to nifZ and nifT, nifV2 has no known homologous nif genes nearby.
In fungi such as yeasts and in extremely thermophilic bacteria, homocitrate is used as an intermediate in lysine biosynthesis through the α-aminoadipic acid pathway (39, 47). The Kyoto Encyclopedia of Genes and Genomes metabolic pathway database (18) suggests that lysine biosynthesis in Nostoc sp. strain PCC 7120 is likely to proceed by the diaminopimelic acid pathway, which does not involve homocitrate, rather than the α-aminoadipic acid pathway, which does involve homocitrate.
The effects of the nifV mutation on net H2 production by cyanobacteria have yet to be examined. In order to study the effects of the nifV disruption on photobiological H2 production in Nostoc sp. strain PCC 7120, we have created three nifV mutants from the ΔhupL parental strain, the ΔhupL ΔnifV1 strain (with the hupL and nifV1 genes disrupted) (Table 1), the ΔhupL ΔnifV2 strain, and the ΔhupL ΔnifV1 ΔnifV2 double mutant, and determined their growth, acetylene reduction activities, and H2 production.
Bacterial strains and plasmids used in this study
MATERIALS AND METHODS
Bacterial strains and growth conditions.The strains and plasmids used in this study are described in Table 1. Cyanobacterial strains were grown photoautotrophically at 28°C under continuous illumination with white fluorescent light at a level of photosynthetically active radiation of about 100 μmol of photons m−2 s−1, unless otherwise indicated, by bubbling with air essentially as described previously (26, 50). The media used throughout the study were liquid BG11 and BG110 (BG11 lacking NaNO3) supplemented with 1 μM NiCl2 and 10 mM N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (TES)-KOH buffer (pH 8.2).
For time course measurements of the induction of nitrogenase and H2 production and heterocyst formation by the ΔhupL strain from Nostoc sp. strain PCC 7120 (26) and the derivative strains, cells were initially grown in BG11 medium by bubbling with air and cells in the exponential growth phase were washed and transferred into 1.5-liter flasks of BG110 medium at a concentration equivalent to 0.20 or 0.30 μg of chlorophyll a ml−1 essentially as described by Masukawa et al. (26). Escherichia coli strains were grown in Luria-Bertani medium at 37°C. The medium for the selection and maintenance of Nostoc and E. coli contained the appropriate antibiotics as described previously (26).
Growth rates were estimated from the chlorophyll a concentrations in the cultures as previously described (26).
Construction of nifV mutants.All molecular manipulations were carried out according to standard methods (37). Cyanobacterial genomic DNA was isolated by the glass bead method (41). The target genes for disruption were nifV1 (alr1407) and nifV2 (alr2968), as identified from the complete genome sequence information for Nostoc sp. strain PCC 7120 (19) in the CyanoBase database. The nifV1 and nifV2 regions of 4.0 and 4.4 kb, respectively (Fig. 1), were amplified by PCR from genomic DNA of Nostoc sp. strain PCC 7120 by using KOD-Plus polymerase (Toyobo Co. Ltd., Osaka, Japan) and primers for the nifV1 region (V1-f1 [5′-GGTATCGAGAATCCACCTGATTCGAC-3′] and V1-r1 [5′-TCTTACCCAACATGAGAACACAGTTAGG-3′]) and the nifV2 region (V2-f1 [5′-TGCTAGCCGAGACGAATTAATAGCTTAC-3′] and V2-r1 [5′-GGATTTTGGTACAGTGTGCTTTACGG-3′]). The amplified fragments were cloned into the EcoRV site of pBluescript II SK(+) (Stratagene, La Jolla, CA) to form pNifV1 and pNifV2. The inserts in pNifV1 and pNifV2 were sequenced. nifV1 in plasmid pNifV1 was disrupted by the insertion of a neomycin resistance (Neor) cassette from pRL648 (4) at the EcoRV site in nifV1, and nifV2 in plasmid pNifV2 was disrupted by the insertion of a chloramphenicol and erythromycin resistance (Cmr-Emr) cassette from pUC19CmEm (Table 1) at blunt-ended HindIII sites in nifV2. The 5.1-kb XhoI-SacI fragment containing the disrupted nifV1 and the 6.3-kb SpeI-HindIII fragment containing the disrupted nifV2 were ligated into the corresponding sites in pRL271 (1) and pRL271KmNm (Table 1), respectively. The resulting plasmids, designated pΔnifV1Nm and pΔnifV2CmEm, were transferred into the ΔhupL strain and Nostoc sp. strain PCC 7120 by conjugation via triparental mating (5), and the double-recombinant ΔhupL ΔnifV1 and ΔhupL ΔnifV2 strains were selected on medium containing the relevant antibiotics and 5% sucrose (2). To construct the nifV1 nifV2 double mutant, the plasmid pΔnifV2CmEm was transferred into the ΔhupL ΔnifV1 strain from the ΔhupL strain by conjugation and a double-recombinant strain, the ΔhupL ΔnifV1 ΔnifV2 mutant, was selected essentially as described above. Complete replacement of the parental nifV1 or nifV2 gene by the mutated versions in the ΔhupL ΔnifV1, ΔhupL ΔnifV2, and ΔhupL ΔnifV1 ΔnifV2 strains was confirmed by Southern blot analysis. The hybridization probes for nifV1 and nifV2 were amplified by PCR from pNifV1 and pNifV2, respectively, using primers for nifV1 (V1-f2 [5′-TTCTGCATACTTGCTAACAACTAGG-3′] and V1-r2 [5′-CTGAGATCACAGCGCGGTTCC-3′]) and nifV2 (V2-f2 [5′-TGCAATTGGTGTGCATGAAATCGAAG-3′] and V2-r2 [5′-TCGTAGGTTTGAGGATTTTGCAGTACAC-3′]).
Insertional inactivation of nifV1 (A) and nifV2 (B). The nifV1 and nifV2 coding regions were amplified by PCR from genomic DNA with the primer pairs V1-f1 and V1-r1 and V2-f1 and V2-r1, respectively. The triangles show the insertion and orientation of a Neor and a Cmr-Emr cassette in nifV1 and nifV2, respectively. Thick lines labeled probe a and probe b indicate the DNA fragments used as probes in Southern blot analyses for the verification of mutants.
Hydrogen production and acetylene reduction assays.Cells were concentrated by centrifugation and suspended in BG110 medium to a chlorophyll a concentration of about 3 to 6 μg ml−1, and 1-ml samples were transferred into 7.5-ml Fernbach flasks equipped with butyl rubber stoppers. The flasks were preilluminated at 28°C on a rotary shaker for about 10 min with photosynthetically active radiation of about 200 μmol of photons m−2 s−1 from M-26 halogen bulbs (Kondo Sylvania Ltd., Tokyo, Japan), which were used as a source of artificial sunlight (44). The gas phase was changed to Ar or N2 for the assay of H2 production activity or to 12% C2H2 (vol/vol) in Ar for the assay of nitrogenase activity, and then the flasks were illuminated further. After about 1 h, 0.2- and 0.5-ml gas samples were withdrawn for gas chromatography analysis of ethylene and H2 production, respectively, essentially as previously described (26).
RNA isolation and Northern blot analysis.Total RNA was isolated using ISOGEN according to the instructions of the manufacturer (Nippon Gene Co. Ltd., Tokyo, Japan). For Northern blot hybridization, total RNA (about 20 μg) was separated on 1% formaldehyde-agarose gels and then blotted onto Hybond-N+ membranes (GE Healthcare UK Ltd., Buckinghamshire, United Kingdom). The fixed membranes were hybridized at 68°C overnight with digoxigenin (DIG)-labeled RNA probes in DIG Easy Hyb (Roche Diagnostics GmbH, Penzberg, Germany). The DNA fragments of nifV1 and nifV2 were amplified by PCR using primers for nifV1 (nifV1-f3 [5′-TTCTGCATACTTGCTAACAACTAGG-3′] and nifV1-r3 [5′-TTTTCGCTAAGATTTCCCCAAGGTCTCT-3′]) and nifV2 (nifV2-f2 and nifV2-r2), the PCR products were cloned into the pGEM-T Easy vector (Promega, Madison, WI) and the EcoRV site of pBluescript II SK(+), and the plasmids were digested by HindIII and XbaI, respectively. The resulting digested plasmids were used as templates for the synthesis of antisense DIG-labeled RNA probes essentially as previously described (50). To prepare RNA probes for nifZT and alr1410, the DNA fragments containing nifZ-nifT and nifZ-nifT-alr1410 were amplified by PCR using primers for nifZT (nifZT-f [5′-ATGCAGCGCGATGAAAACGAGTT-3′] and nifZT-r [5′-TTGCGAGCAGGGTTGAGCTATAGA-3′]) and alr1410 (nifZT1410-f [5′-ACGAATCAGGTATTCATGCACATGG-3′] and nifZT1410-r [5′-CTGGAGAGAAAGCCAATGCGA-3′]), respectively, and the amplified products were cloned into the EcoRV site of pBluescript II SK(+) to produce plasmids pNifZT-554 and pNifZT1410-1580, respectively. The latter plasmid was digested by HincII and self-ligated to produce plasmid pNifZT1410-1440. The 3.52-kb EcoRI-digested pNifZT-554 and 3.66-kb XbaI-digested pNifZT1410-1440 were used as templates for the synthesis of the respective RNA probes. Northern blot hybridization and detection were performed according to the instructions of the manufacturer (Roche).
Construction of green fluorescent protein (GFP) fusion reporters.The promoter regions of nifV1 (PnifV1)- and nifV2 (PnifV2)-gfp transcriptional fusions were constructed as follows. A 756-bp fragment containing an upstream region of nifV1 (511 bp) and a 5′ portion of nifV1 (245 bp) was amplified by PCR from pNifV1 by using primers nifV1-f4 (5′-CCGGGAGTTTGATAATGGGCCTCTAC-3′) and nifV1-r2 and subsequently cloned into the EcoRV site of pBluescript II SK(+), producing plasmid pPnifV1-756. A 2.0-kb HincII fragment containing an upstream region of nifV2 (1.64 kb) and a 5′ portion of nifV2 (373 bp) from pNifV2 was cloned into the EcoRV site of pBluescript II SK(+), producing plasmid pPnifV2-Hc2. A 2.7-kb SmaI-PvuII fragment containing the promoterless gfp plus the omega streptomycin-spectinomycin resistance (Strr-Sptr) cassette from pRL2379 (48) was inserted into the SmaI site downstream of the translational start of nifV1 in pPnifV1-756 and nifV2 in pPnifV2-Hc2. The resulting plasmids were referred to as pPnifV1-gfp and pPnifV2-gfp, respectively. The ClaI-SpeI fragments bearing the PnifV1-gfp-omega construct (3.5-kb) from pPnifV1-gfp and the PnifV2-gfp-omega construct (4.7-kb) from pPnifV2-gfp were inserted into the NspV-NheI site of the shuttle vector pRL25C (46). The resulting plasmids were transferred into the wild-type Nostoc sp. strain PCC 7120 by conjugation as described above, and Strr-Sptr recombinants were selected.
Microscopy.Heterocyst frequency was determined visually using an Eclipse E600 microscope (Nikon, Tokyo, Japan) by counting over 2,000 cells per strain at each time point (see Fig. 3). Heterocysts were distinguished by their thick cell envelope, and dividing vegetative cells were counted as two cells.
GFP fluorescence and bright-field images were captured by a DP70 digital camera (Olympus, Tokyo, Japan) mounted on a Nikon Eclipse E600 microscope and were processed with Adobe Photoshop version 6.0 (Adobe Systems, CA). Images of GFP fluorescence were photographed with excitation and emission wavelengths of 465 to 496 nm and 515 to 555 nm, respectively.
RESULTS
Growth rates and nitrogenase and H2 production activities of the mutants.A total of three types of mutants, in which either or both nifV1 and nifV2 were inactivated, were constructed by genetically engineering the ΔhupL variant of Nostoc sp. strain PCC 7120. In BG11 medium containing 17.6 mM NO3−, the growth rates, as estimated by chlorophyll a content, of all the nifV mutants were indistinguishable from that of the parent ΔhupL strain (data not shown). In BG110 medium lacking combined nitrogen, all of the mutants grew, but the growth rates of all the nifV mutants were lower than those of the parent strain to various degrees (Fig. 2A). The growth rate of the ΔhupL ΔnifV1 ΔnifV2 mutant was extremely low, with a doubling time of about 4.5 days compared with less than about 2 days for the other two mutants. The ΔhupL ΔnifV1 mutant grew slightly better than the ΔhupL ΔnifV2 mutant.
Time courses of N2-fixing growth, the development of nitrogenase, and H2 production activities for the ΔhupL strain and the corresponding nifV mutant cultures. (A) N2-fixing growth in BG110 medium; (B) nitrogenase activity measured by acetylene reduction under Ar; (C and D) H2 production activity under Ar (C) and N2 (D); (E and F) H2 production activity of cultures under Ar (E) and N2 (F). ○, ΔhupL strain; ▴, ΔhupL ΔnifV1 strain; ▪, ΔhupL ΔnifV2 strain; •, ΔhupL ΔnifV1 ΔnifV2 strain. Data are means ± standard deviations of results from duplicate or triplicate samples.
The time courses for the induction of nitrogenase activity are shown in Fig. 2B. The nitrogenase activity of the ΔhupL strain reached the maximum at about 1.5 days after nitrogen step down and then declined to less than 50% of the maximum in less than 1 day. The nitrogenase activity of the ΔhupL ΔnifV1 mutant reached a maximum at about 2 days after nitrogen step down and thereafter declined much more slowly than that of the ΔhupL strain, resulting in higher nitrogenase activity than that of the ΔhupL strain after about 2 days. The entire time course profile of the nitrogenase activity of the ΔhupL ΔnifV2 mutant was roughly similar to that of the ΔhupL ΔnifV1 mutant, but the magnitude of the nitrogenase activity of the ΔhupL ΔnifV2 mutant was slightly lower than that of the ΔhupL ΔnifV1 mutant. The nitrogenase activity of the ΔhupL ΔnifV1 ΔnifV2 mutant reached the maximum about 2 days after nitrogen step down and was maintained at about the same low levels throughout the period of the experiment.
The ΔhupL ΔnifV1 ΔnifV2 mutant nitrogenase reduced acetylene to ethylene as shown in Fig. 2B, but further reduction to ethane was not detectable even after about 4 h of incubation (data not shown). Ethane formation should be detectable if the culture has at least 1% of the activity of its ethylene formation.
Photobiological H2 production activities of the cells sampled at the times indicated and assayed for 1 h under Ar and N2 are shown in Fig. 2C and D, respectively. The entire H2 production activity profiles of the three nifV mutants and the ΔhupL strain were roughly similar to the nitrogenase activity profiles in Fig. 2B. Upon closer inspection, it was found that the H2 production activity of the ΔhupL ΔnifV1 ΔnifV2 mutant relative to those of the other strains was significantly improved compared with that of acetylene reduction (Fig. 2B), and the improvement was more pronounced under N2 than under Ar. In two separate experiments (data not shown), the ratios of H2 production activity under N2 versus that under Ar were significantly higher for the ΔhupL ΔnifV1 ΔnifV2 mutant culture (47.2% ± 2.5% and 45.5% ± 1.7%) than for the cultures of the ΔhupL ΔnifV1 mutant (35.7% ± 1.8% and 36.2% ± 2.4%), the parent ΔhupL strain (32.3% ± 1.7% and 33.0% ± 1.5%), and the ΔhupL ΔnifV2 mutant (31.9% ± 1.4% and 30.2% ± 1.4%). It is noteworthy that the ratio for the ΔhupL ΔnifV1 mutant culture was slightly higher than that for the ΔhupL ΔnifV2 mutant culture.
The H2 production activity of cultures was calculated by multiplying the H2 production activities per milligram of chlorophyll a under Ar and N2 (Fig. 2C and D) by the chlorophyll a concentrations of the cultures (Fig. 2A) as shown in Fig. 2E and F, respectively. The H2 production activity of the ΔhupL ΔnifV1 mutant culture was lower than that of the culture of the parent ΔhupL strain until about the middle of day 2 but exceeded the latter after day 3, and the difference between the two became greater thereafter. It was also noted that after day 3 the difference between the two was greater under N2 than under Ar. The H2 production activity of the ΔhupL ΔnifV2 mutant culture exceeded that of the culture of the parent ΔhupL strain after day 7 under both Ar and N2.
Heterocyst frequency.Under conditions of combined-nitrogen starvation, some of the vegetative cells of the nifV mutants and the ΔhupL strain differentiated into heterocysts within a few days after nitrogen step down, and the time course of heterocyst formation in these mutants was studied morphologically by light microscopy (Fig. 3). Because it was difficult to morphologically discriminate heterocysts from vegetative cells at day 1 by light microscopy observation, only the results from day 2 and beyond are shown. In the parent ΔhupL strain culture, the heterocyst frequency was the highest at day 2 (9.1% of the total cells) and declined thereafter. In the ΔhupL ΔnifV1 and ΔhupL ΔnifV2 mutant cultures, the heterocyst frequency was also the highest at day 2, but the rates of the decrease were significantly lower than that in the parent culture. The heterocyst frequency in the culture of the ΔhupL ΔnifV1 ΔnifV2 mutant was as high as those in the parent culture and in the cultures of both of the single nifV mutants at day 2, but it continued to increase to about 11% until day 6 after nitrogen step down and remained at high levels.
Time courses of heterocyst frequency in the ΔhupL strain and the corresponding nifV mutants after the transfer of cells to BG110 medium. The frequencies between 0 and 2 days after nitrogen deprivation are shown by dashed lines. Representative results from one of two independent experiments are shown. Symbols are as defined in the legend to Fig. 2.
Expression of nifV1 and nifV2.The steady-state levels of expression of nifV1 and nifV2 transcripts were analyzed by Northern blot hybridization using total RNA isolated from whole cells of the wild type grown in BG11 medium containing 17.6 mM NO3− or in BG110 medium lacking combined nitrogen. The cells for RNA isolation were harvested at the day indicated in Fig. 4 after nitrogen step down by transfer from BG11 to BG110 medium. Northern blot experiments with the RNA probe of nifV1 showed hybridization to three transcripts of 1.3, 1.6, and 1.8 kb and weak hybridization to a 2.5-kb transcript after nitrogen step down. These transcripts correspond roughly in size with those of nifV1, nifV1-nifZ, nifV1-nifZ-nifT, and nifV1-nifZ-nifT-open reading frame alr1410, respectively. In the ΔhupL ΔnifV1 mutant culture, the levels of nifZ-nifT expression were clearly detectable irrespective of nitrogen nutritional status and were much higher than those in the parent ΔhupL strain, even under conditions of combined-nitrogen depletion (data not shown), probably because the Neor gene was inserted in the sense orientation in the ΔhupL ΔnifV1 mutant culture (Fig. 1). The transcript of open reading frame alr1410 was barely detectable by its probe in the wild type, although its presence is suggested in Fig. 1, but was clearly detectable in the ΔhupL ΔnifV1 mutant culture, probably due to the reason stated above (data not shown). These nifV1 transcripts appeared 1 day after the removal of combined nitrogen, and their levels peaked in 1 or 2 days (data not shown) and then decreased thereafter. The profile of nifV1 expression in the ΔhupL ΔnifV2 mutant was similar to that in the wild type (data not shown). The RNA probe for nifV2 hybridized strongly to a constitutively expressed transcript of 1.3 kb and weakly to a constitutively expressed 2.1-kb transcript. The 1.3-kb transcript was remarkably abundant in cultures containing combined nitrogen. The amount of this transcript greatly decreased 1 day after nitrogen starvation (data not shown) and gradually increased thereafter to the level in cells replete with combined nitrogen. The time course profile of nifV2 expression levels in the ΔhupL ΔnifV1 mutant was similar to that in the wild type (data not shown).
Northern blot analysis of nifV1 and nifV2 transcripts. Samples of total RNA (20 μg) isolated from the wild-type cells harvested before (+N) and on the indicated day (2 [2d], 3, or 4) after nitrogen deprivation (−N) were hybridized with RNA probes specific for either nifV1 or nifV2. (Top panels) Transcript sizes were estimated from the mobilities of the standard RNA markers. (Bottom panels) The band of 16S rRNA was stained with ethidium bromide and was used as a loading control.
Spatial pattern of nifV1 and nifV2 expression.We further examined the cellular expression patterns of nifV1 and nifV2 by using transcriptional gfp fusions on a shuttle vector in the wild type. Figure 5 shows GFP fluorescence from the wild-type cells carrying a nifV1::gfp fusion. When the cells were grown with nitrate, they had no GFP fluorescence (data not shown), but after combined-nitrogen deprivation in the culture medium, strong GFP fluorescence was observed exclusively in heterocysts of filaments. A nifV2::gfp transcriptional fusion bearing 0.87 or 1.64 kb upstream of alr2967 or nifV2, respectively, was constructed, but the GFP fluorescence was not discernible above the background fluorescence from photosynthetic pigments in vegetative cells and heterocysts, although the presence of nifV2 messages was detected in wild-type whole cells on Northern blots (Fig. 4).
Expression of the fusion between the promoter region of nifV1 and the gfp gene. Strain NsV1GFP was grown in BG110 medium lacking combined nitrogen for 2 days. GFP fluorescence (left panel) and the corresponding light micrograph (right panel) are shown. Arrowheads indicate heterocysts.
DISCUSSION
Nitrogen fixation and H2 production in the nifV mutants.The present study demonstrates that in Nostoc sp. strain PCC 7120 ΔhupL cells, both nifV1 and nifV2 gene products are required for the full development of N2 fixation activity but that either one can support N2-fixing growth to a considerable extent. The results show that there are no substantial differences in growth rate between the parent ΔhupL strain and its derivative nifV mutants in the nitrate-containing BG11 medium (data not shown) but that there are significant differences under N2-fixing conditions (Fig. 2A). These data suggest that in cultures of Nostoc sp. strain PCC 7120, the two nifV gene products function in the biosynthesis of homocitrate for the FeMo cofactor of dinitrogenase but not for lysine synthesis through the α-aminoadipic acid pathway, in harmony with the metabolic pathway analysis presented in the Kyoto Encyclopedia of Genes and Genomes database (18) for the Nostoc sp. strain PCC 7120 wild-type genome. The fact that the ΔhupL ΔnifV1 ΔnifV2 mutant cells could grow diazotrophically in the absence of combined nitrogen indicates that the nitrogenase of this strain can fix N2 in the absence of homocitrate, although at a much lower rate than that of the parent ΔhupL strain. A likely explanation for these observations is that citrate substitutes for homocitrate in this mutant, as in the K. pneumoniae nifV mutant (22, 27). It was reported previously that R. capsulatus nifV mutants have the ability to reduce acetylene not only to ethylene but also to ethane at a rate of 5 to 7% of the rate of ethylene formation (25), although we could not detect any ethane formation by the ΔhupL ΔnifV1 ΔnifV2 mutant (see Results above). The above-mentioned difference between the two mutants can be attributed to the subtle difference in nitrogenase substrate specificity, but we did not study this issue further.
The higher sustained levels of nitrogenase activities and higher heterocyst frequencies in both of the nifV single mutants derived from the parent ΔhupL strain indicate that the supply levels of fixed nitrogen for active growth were lower in the mutants than in the parent ΔhupL strain. The lower fixed-nitrogen supply levels likely also led to the slower declines of heterocyst frequency in these mutants than in the parent ΔhupL strain after day 2 or 3 (Fig. 3). The heterocyst frequency of the ΔhupL ΔnifV1 ΔnifV2 cells remained at a high level even after day 7, indicating a very poor supply of combined nitrogen in this mutant.
Expression of nifV1 and nifV2.In Northern blot experiments with a nifV1-specific probe, Stricker et al. (40) found a 1.8-kb transcript in N2-fixing cultures and no transcript in non-N2-fixing cultures. We have confirmed this finding and further identified the other transcripts in N2-fixing cultures by using a nifV1 RNA probe: the 1.3-, 1.6-, 1.8-, and 2.5-kb transcripts corresponded to the predicted lengths of transcripts encompassing nifV1, nifV1-nifZ, nifV1-nifZ-nifT, and nifV1-nifZ-nifT-alr1410, respectively. Hybridization to a 1.3-kb major transcript with a nifV2 RNA probe showed that nifV2 is expressed as a monocistronic message. An additional 2.1-kb weakly hybridizing transcript corresponds to the predicted length of the message transcribed from alr2967. The deduced gene product of alr2967 shows similarity to a universal stress protein (19).
The expression profiles of nifV1 and nifV2 (Fig. 4) and the corresponding promoter-gfp fusions (Fig. 5) indicate that the two homocitrate synthases are expressed differently both spatially and temporally according to the nitrogen nutritional status of the cells. The expression of nifV1 is markedly induced after nitrogen step down. On the other hand, nifV2 is expressed even in the presence of combined nitrogen and in the absence of nitrogenase activity (Fig. 4), but the protein does not seem to be specifically localized in heterocysts, in contrast with NifV1.
Differences between the ΔhupL ΔnifV1 and ΔhupL ΔnifV2 mutants in nitrogenase and H2 production.Some subtle differences between the ΔhupL ΔnifV1 and ΔhupL ΔnifV2 mutant cultures in growth rates and H2 production under both Ar and N2 were noted (Fig. 2). Possible explanations for these differences may be that supplies of homocitrate and citrate to dinitrogenase vary between these mutants due to spatial and temporal differences in expression between nifV1 and nifV2 (Fig. 4 and 5). The development of N2 fixation activity is a very complex process, requiring the mobilization of reserve nitrogen compounds for the synthesis of new proteins, including nitrogenase. Once active dinitrogenase containing the FeMo cofactor with bound homocitrate is formed, it meets the combined nitrogen requirement efficiently by itself. The activities of N2 fixation and H2 production may also be influenced by the levels of citrate in heterocysts. These factors may influence the subtle differences observed in growth rates and H2 production between the ΔhupL ΔnifV1 and ΔhupL ΔnifV2 mutant cultures.
From the average ratios of H2 production activity under N2 versus that under Ar during the study period, it is suggested that the ratio of electrons allocated to H2 production is slightly higher in the ΔhupL ΔnifV1 mutant culture than in the culture of the ΔhupL ΔnifV2 strain, which may be explained by a higher level of occupancy of citrate in dinitrogenase in the culture of the ΔhupL ΔnifV1 mutant than in the latter.
Concluding remarks.The present study demonstrates a promising strategy for improving H2 production activity over that of the parent ΔhupL strain derived from Nostoc sp. strain PCC 7120, as evidenced by the greater sustained H2 production and higher nitrogenase activities of the ΔhupL ΔnifV1 mutant culture grown under air. In future studies, it will be interesting to see if the alteration of the homocitrate synthase activity level by genetic engineering is effective in improving H2 production activity in the presence of N2 in other strains.
ACKNOWLEDGMENTS
We thank C. P. Wolk for providing RP4 and pRL plasmids.
This work was supported in part by the MEXT grant Venture (2001 to 2005), the MEXT grant Initiatives for Attractive Education for Graduate Schools (no. b043), grants for special research projects from Waseda University (2004A-085 and 2005B-083) to H.S., and a grant-in-aid for scientific research from the Japan Society for the Promotion of Science (16.9494) to H.M.
FOOTNOTES
- Received 23 May 2007.
- Accepted 30 September 2007.
- Copyright © 2007 American Society for Microbiology
