Previous Article | Next Article ![]()
Journal of Bacteriology, April 2004, p. 2107-2114, Vol. 186, No. 7
0021-9193/04/$08.00+0 DOI: 10.1128/JB.186.7.2107-2114.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Laboratory of Molecular Plant Physiology, Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya 464-8601, Japan
Received 27 August 2003/ Accepted 11 December 2003
|
|
|---|
|
|
|---|
In accordance with the large variety of organisms that perform nitrate assimilation, there are different types of nitrate transporters, NR enzymes, and NiR enzymes in nature. Nitrate is reduced in dissimilatory and respiratory processes as well, which adds more variety to the NR enzymes. Although all NRs contain a molybdenum cofactor as the prosthetic group, the eukaryotic NRs, assimilatory enzymes which use NAD(P)H as the electron donor, share little sequence similarity with the prokaryotic NRs (5, 21). The prokaryotic enzymes are classified into three major classes, i.e., assimilatory, respiratory, and periplasmic NRs (21, 34). Although the three types of NR have different subunit compositions and physiological roles, their catalytic subunits have sequence similarities. The prokaryotic assimilatory NRs are further classified into two classes: the NAD(P)H-dependent enzyme of heterotrophic bacteria and the ferredoxin-dependent enzyme of cyanobacteria. While NR from heterotrophic bacteria consists of a catalytic subunit and a flavin adenine dinucleotide-binding subunit involved in electron transfer from NAD(P)H to the catalytic subunit (20, 29), the cyanobacterial NR consists only of the catalytic subunit (28).
Among cyanobacteria, the NR-related genes are best characterized in Synechococcus elongatus strain PCC7942, where three genetic loci, narA, narB, and narC, have been identified and characterized as essential for the nitrate reduction (17, 18). The narB locus has been identified as the NR structural gene (narB) (4), which is clustered with the NiR gene (nirA) (23, 40) and the nitrate and nitrite transporter genes (nrtABCD) (30-32) to form the nirA-nrtABCD-narB operon (40). The narA locus contains five genes, which encode homologs of the Moa proteins and the MoeA protein of Escherichia coli (36), and the narC locus carries a gene encoding the homolog of the E. coli MobA protein (35). As in E. coli, the Moa, Moe, and Mob proteins are involved in biosynthesis of molybdopterin guanine dinucleotide (MGD), which is the molybdenum cofactor of prokaryotic NRs (33). Although its function is unclear, the homolog of the moeB gene of E. coli has been also identified in S. elongatus strain PCC7942 (35). Another gene known to be related to MGD biosynthesis in E. coli is mog, which is required for the efficient incorporation of molybdate into molybdopterin (14). The mog gene product is homologous to MoaB protein, and open reading frames (ORFs) that would encode a protein homologous to Mog or MoaB are found in several cyanobacteria, except for Synechocystis sp. strain PCC6803 and Thermosynechococcus elongatus BP-1. It thus seems that most of the NR-related genes have been identified and characterized in cyanobacteria. In this paper, however, we report the identification of a novel gene (narM) in S. elongatus strain PCC7942 which is unique to cyanobacteria and required for the expression of NR activity. Possible role(s) of the narM gene product are discussed.
|
|
|---|
Random insertional mutagenesis and isolation of mutants. Cells of the NA3 mutant of S. elongatus strain PCC7942, lacking the nitrate and nitrite transporter genes nrtABCD (27), were transformed with a kanamycin resistance (Kmr) marker-tagged library by double-sided homologous recombination, and transformants were selected on plates containing ammonium and kanamycin as described previously (25). Transformants were cultivated for 7 days in liquid medium containing ammonium. The cells were then transferred to nitrate-containing medium and cultivated for 16 h. After this period, cells were cultivated in the presence of ampicillin for 5 days under otherwise the same conditions as before, with daily supplementation of ampicillin amounting to 500 µg per ml of medium. After the ampicillin enrichment step, cells were harvested by centrifugation, washed three times by resuspension and recentrifugation, and plated on solid medium containing ammonium. Putative mutant colonies were isolated and screened on replicate plates containing ammonium, nitrite, and nitrate, respectively.
Retrieval and analysis of the tagged genomic DNA fragments. Genomic DNA isolated from the selected mutants was digested with SacI and fractionated by electrophoresis on a 0.7% agarose gel. DNA fragments of 3 to 10 kbp were eluted from the gel and ligated into the SacI site of pUC19, and the resulting plasmids were used for transformation of E. coli JM109. Kanamycin-resistant E. coli transformants were isolated and shown to contain a plasmid carrying a Synechococcus genomic DNA fragment, which was tagged with the Kmr marker. Nucleotide sequences of the DNA regions flanking the Kmr marker were discontiguous, indicating that a PstI-PstI genomic fragment had been replaced by the Kmr gene cassette during construction of the Kmr-tagged genomic DNA library (Fig. 1). The missing 0.5-kbp PstI-PstI genomic region was obtained by PCR amplification with genomic DNA from the wild-type strain as a template. The genomic DNA fragment and the PCR product were used for determination of the genomic DNA sequence.
![]() View larger version (13K): [in a new window] |
FIG. 1. Map of the genomic region of narM in the cyanobacterium S. elongatus strain PCC7942. Open arrows show the locations and directions of orf170, orf353, narM (orf161), and lpxB (orf364). The bar with a closed box below the open arrows represents the DNA fragment retrieved from the NR-less mutant obtained by random insertional mutagenesis. Open triangles show the sites of insertion of the Kmr gene in the insertional mutants of orf353 (NM1 and NM11) and orf161 (NM2 and NM21). Closed triangles show the locations and directions of the primers p1 to p5, which were used for PCR amplification of the genomic sequences (Fig. 2) and RT-PCR analysis of RNA (Fig. 6). Restriction endonuclease sites are abbreviated as follows: M, MscI; P, PstI; S, SacI; X, XhoI; (X), engineered XhoI.
|
Isolation and analysis of DNA and RNA. Chromosomal DNA was extracted and purified from S. elongatus strain PCC7942 cells as described by Williams (45). Manipulations and analyses of DNA were performed according to standard protocols. Total RNA was extracted and purified from the Synechococcus cells by the method of Aiba et al. (1). RNA samples were treated with RNase-free DNase I (TaKaRa) for elimination of any remaining DNA. For synthesis of the cDNAs to be used for analysis of the transcripts of narM, narB, moaC, moeA, and mobA, 0.5 µg of total RNA was mixed with 20 pmol of the oligonucleotides p5 (5'-CTCAGTCAGCCTCAAGGG-3', complementary to nucleotides +470 to +487 with respect to the translation start of the narM gene), p6 (5'-GGCAAAGACTGTAAGCCG-3', complementary to nucleotides +503 to +520 with respect to the translation start of the narB gene), p7 (5'-TGTTCCATGATCTGCGCCTTTC-3', complementary to nucleotides +927 to +948 with respect to the translation start of the moaC gene), p8 (5'-CAACTGCTCAATCAAAAAAGGCTG-3', complementary to nucleotides +625 to +648 with respect to the translation start of the moeA gene), and p9 (5'-AATCAGTCGGGGTGTTGCAGTT-3', complementary to nucleotides +541 to +562 with respect to the translation start of the mobA gene), respectively. cDNAs were synthesized with the SuperScript III first-strand synthesis system (Invitrogen). The PCR consisted of 30 cycles of template denaturation (95°C for 30 s), annealing with oligonucleotide primers (60°C for 30 s), and product extension (72°C for 40 s). The primers used were p5 and p3 for narM, p6 and p10 (5'-GGAATGTTCGATCTCTCG-3', corresponding to nucleotides -3 to +15 with respect to the translation start of the narB gene) for narB, p7 and p11 (5'-GATGTCGGTGATAAGGCGGTC-3', corresponding to nucleotides +7 to +27 with respect to the translation start of the moaC gene) for moaC, p8 and p12 (5'-GATCTTGTCCTATTCGGCAGCG-3', corresponding to nucleotides +3 to +24 with respect to the translation start of the moeA gene) for moeA, and p9 and p13 (5'-CGCGATGAATTTTGCTGCCTTG-3', corresponding to nucleotides -4 to +18 with respect to the translation start of the mobA gene) for mobA. The PCR products were resolved by electrophoresis in 1.5% agarose gels with 0.5 µg of ethidium bromide/ml.
Other methods. NR and NiR activities were determined at 30°C by using toluene-permeabilized cells with dithionite-reduced methyl viologen as the electron donor (11, 12). Nitrite was determined with a flow injection analyzer (NOX 1000; Tokyo Chemical Industry Co., Ltd.). A derivative of the plasmid pYK5, carrying a transcriptional fusion of the promoter region of nirA (nucleotides -275 to -15 with respect to the translation start site) and the coding sequences of luxA and luxB, was used to transfer the promoter-reporter fusion (PnirA::luxAB) into NM11 by double-sided homologous recombination at the cmpB-cmpC region as previously described (26). In vivo bioluminescence was measured with a luminometer (ARGUS-50; Hamamatsu Photonics) as previously described (26). Chlorophyll was determined according to the method of Mackinney (24).
|
|
|---|
Retrieval and analysis of tagged genomic DNA fragments from the mutants. To exclude the mutants in which the Kmr gene cassette was inserted in known genes involved in nitrate assimilation, DNA samples extracted from the mutants were analyzed by PCR, with the primers amplifying the Kmr marker from the two ends and the third primer amplifying nirA, narB, moeA, or moaA from the 3' end of the respective gene. In each of the three mutants defective in nitrite reduction, a 1.1-kbp product was amplified with the nirA-specific primer, indicating that the Kmr gene was inserted in the nirA gene. Fourteen of the 16 NR-less mutants were shown to have the Kmr gene inserted in either moeA or moaA. In the remaining two NR-less mutants, no DNA fragment was amplified with the primers specific to known genes required for nitrate reduction. The Kmr marker and its flanking regions of these 2 mutants were retrieved into the SacI site of pUC19 (Fig. 1). Nucleotide sequence analysis showed that the two mutants are identical with respect to the DNA sequences flanking the Kmr gene. One of the flanking sequences encoded an amino acid sequence homologous to the C-terminal part of Sll1455 of the Synechocystis gene product while the other encoded an amino acid sequence homologous to the internal part of the Synechocystis Sll1634 gene product, suggesting that the flanking sequences were discontiguous, due to loss of a PstI-PstI genomic fragment during construction of the Kmr-tagged genomic library. This assumption was confirmed by PCR amplification of a 0.5-kb fragment of wild-type genomic DNA that bridges the flanking sequences of the Kmr gene cassette. The nucleotide sequence of the 3.3-kb SacI-SacI fragment of genomic DNA, including the 0.5-kb region that was replaced by the Kmr gene in the NR-less mutant, was determined. The region contained two partial ORFs (orf170 and orf364) and two entire ORFs (orf353 and orf161), and the NR-less mutants were found to lack the 5'-terminal region of orf353 and most of orf161 (Fig. 1). The deduced amino acid sequences of ORF353 and ORF161 were 51 and 49% identical, respectively, to Sll1634 and Sll1455 of the Synechocystis gene products of unknown function (15). The deduced amino acid sequence of the C-terminal part of ORF364 was found to be 48% identical to lipid A disaccharide synthase, the product of the lpxB gene, of E. coli (6) and hence identified as a part of the lpxB gene of S. elongatus strain PCC7942. The other incomplete ORF, orf170, encoded an amino acid sequence 53% identical to the C-terminal part of Slr1262, a Synechocystis gene product of unknown function (15).
Characterization of insertional mutants.
To determine which of orf353 and orf161 is required for nitrate assimilation, defined mutants of these genes were constructed from the wild-type strain and the NA3 mutant (
nrtABCD) of S. elongatus. Homozygous mutants, each inactivated in the relevant wild-type genes by inserting the Kmr marker, were obtained both from the wild-type strain and NA3 (Fig. 2). The orf353::Kmr mutants constructed from the wild-type strain and NA3 were designated NM1 and NM11, respectively (lanes 2 and 3); the orf161::Kmr mutants constructed from the wild-type strain and NA3 were designated NM2 and NM21, respectively (lanes 9 and 10). Figure 3 compares the growth curves of the parental strains and the mutants. The wild-type strain grew well in all the media tested. The NA3 mutant lacking the nitrate and nitrite transporter grew as fast as the wild-type strain in ammonium- or nitrite-containing medium (Fig. 3A and B) but grew slowly in nitrate (60 mM)-containing medium (Fig. 3C). All of the newly constructed mutants grew as fast as the wild-type strain in ammonium- or nitrite-containing medium (Fig. 3A and B); however, both orf161 mutants, i.e., NM2 and NM21, failed to grow in the nitrate-containing medium (Fig. 3C). In contrast, the orf353 mutants NM1 and NM11 grew as fast as the respective parental strains in the nitrate-containing medium (Fig. 3C). These results indicated that orf161, but not orf353, is required specifically for nitrate assimilation. Unlike the NA3 strain that grows in the presence of 60 mM nitrate, the orf161 mutants were totally defective in growth on nitrate (Fig. 3C), confirming that the nitrate reduction step but not the nitrate transport step was impaired in the mutants. Thus, we have named orf161 narM.
![]() View larger version (78K): [in a new window] |
FIG. 2. PCR analysis of the genomic DNA of the insertional mutants. DNA from the wild-type (WT) strain (lanes 1 and 6), NM1 (orf353::Kmr) (lanes 2 and 7), NM11 ( nrtABCD orf353::Kmr) (lanes 3 and 8), NM2 (orf161::Kmr) (lanes 4 and 9), and NM21 ( nrtABCD orf161::Kmr) (lanes 5 and 10) was PCR amplified with the gene-specific primer combinations p1 plus p2 for orf353 (lanes 1 to 5) and p3 plus p4 for orf161 (lanes 6 to 10). The PCR products were electrophoresed on a 1% agarose gel. The location of each primer on the genomic genes is illustrated in Fig. 1. The size of each of the amplified DNA fragments is shown. Lane M, 1 kb plus DNA ladder (Invitrogen) used as the size marker.
|
![]() View larger version (12K): [in a new window] |
FIG. 3. Growth curves of the wild-type strain and NA3 and of the orf353 and orf161 mutants derived therefrom. Cells were grown in the presence of ammonium and transferred at time zero to media containing 5 mM ammonium (A), 5 mM nitrite (B), and 60 mM nitrate (C). White circles, wild-type strain; white squares, NA3 ( nrtABCD); gray circles, NM1 (orf353::Kmr); gray squares, NM11 ( nrtABCD orf353::Kmr); black circles, NM2 (orf161::Kmr); black squares, NM21 ( nrtABCD orf161::Kmr); OD730, optical density at 730 nm.
|
|
View this table: [in a new window] |
TABLE 1. NR and NiR activities in the wild-type and mutant strainsa
|
![]() View larger version (61K): [in a new window] |
FIG. 4. Alignment of the deduced NarM protein and its homologs in various strains of cyanobacteria. S7942, NarM of S. elongatus strain PCC7942; BP-1, Tlr1356 of T. elongatus BP-1; A7120, Alr0614 of Anabaena sp. strain PCC7120; N.P., a hypothetical protein of N. punctiforme strain ATCC 29133; S6803, Sll1455 of Synechocystis sp. strain PCC6803; IMS101, a hypothetical protein of T. erythraeum IMS101; S8102, a hypothetical protein of Synechococcus sp. strain WH8102. The three cysteine residues conserved in all but one (the hypothetical protein of Synechococcus sp. strain WH8102) of these proteins are indicated by closed triangles. Asterisks indicate amino acids common to all sequences at a particular position. Dashes indicate gaps introduced to enhance the alignment.
|
![]() View larger version (18K): [in a new window] |
FIG. 5. Comparison map of the genomic regions of Synechocystis sp. strain PCC6803, Anabaena sp. strain PCC7120, T. elongatus BP-1, and N. punctiforme strain ATCC 29133, which contain a cluster of genes involved in nitrate assimilation. Genes encoding the homologs of NarM and ORF353 are indicated by black and gray bars, respectively.
|
![]() View larger version (55K): [in a new window] |
FIG. 6. RT-PCR analysis of RNA extracted from cells of the wild-type strain (lanes 1 to 3, 7 to 9, and 13 to 16) and the narM insertional mutant NM2 (lanes 4 to 6, 10 to 12, and 16 to 18) with the primer sets specific to narM (panel A, lanes 1 to 6), narB (panel A, lanes 7 to 12), moaC (panel B, lanes 1 to 6), moeA (panel B, lanes 7 to 12), and mobA (panel B, lanes 13 to 18). Cells were grown with ammonium as the nitrogen source and then transferred to ammonium-containing medium (lanes 1, 4, 7, 10, 13, and 16), nitrate-containing medium (lanes 2, 5, 8, 11, 14, and 17), or nitrogen-free medium (lanes 3, 6, 9, 12, 15, and 18). RNA was extracted from the cells 30 min after the transfer and subjected to the RT-PCR analysis. Samples were resolved by electrophoresis on 1.5% agarose gels with 0.5 µg of ethidium bromide/ml. Lane M, 1 kb plus DNA ladder (Invitrogen) used as the size marker.
|
Analysis of nirA operon expression in the orf353 mutant.
Although the mutant of orf353 expressed normal levels of nitrate assimilation enzymes and showed no defect in nitrate utilization (Table 1; Fig. 3), homologs of orf353 were found in every cyanobacterial strain that performed nitrate assimilation, located in most cases within the nitrate assimilation gene cluster (Fig. 5). This suggested that orf353 has some role in nitrate assimilation or its regulation, and indeed, Frías et al. very recently reported that a mutant of the orf353 homolog (cnaT) is defective in transcriptional activation of the nitrate assimilation operon in Anabaena sp. strain PCC7120 (8). To quantitatively examine the expression levels of the nitrate assimilation operon in the orf353 mutant of S. elongatus strain PCC7942 we introduced into NM11 a promoter-reporter fusion consisting of the promoter of the nitrate assimilation operon (PnirA) and the coding sequences of luxAB (26) and compared the bioluminescence from the resulting strain (designated YKA1c; PnirA::luxAB
nrtABCD orf353::Kmr) with that from the reference strains YKA1 (PnirA::luxAB
nrtABCD) and YKA1b (PnirA::luxAB
nrtABCD
ntcB::Kmr) (26) under various nitrogen conditions (Fig. 7). When nitrogen assimilation was inhibited by DON (an inhibitor of glutamine amidotransferase) treatment, the luciferase activity of all strains was increased by about 100-fold, representing transcriptional activation by NtcA (22, 42, 44). The addition of further nitrite increased the bioluminescence levels of the YKA1 and YKA1c strains by 21- and 17-fold while decreasing that of YKA1b by 15%. These results showed that the NtcB-mediated nitrite-dependent activation of nirA operon transcription (2, 16) is normally operating in the orf353 mutant. The results indicated that orf353 has nothing to do with the NtcA- and NtcB-mediated activation mechanism of the nitrate assimilation operon of S. elongatus strain PCC7942.
![]() View larger version (17K): [in a new window] |
FIG. 7. Effects of DON (an inhibitor of glutamine amidotransferase) and nitrite on luciferase activities of Synechococcus reporter strains YKA1 (PnirA::luxAB nrtABCD), YKA1b (PnirA::luxAB nrtABCD ntcB::Kmr), and YKA1c (PnirA::luxAB nrtABCD orf353::Kmr). Bioluminescence was measured before (control, white bars) and 12 h after the addition of DON alone (gray bars) and of DON plus nitrite (black bars) to ammonium-grown cultures of the Synechococcus reporter strains. The bioluminescence data are the means of five measurements, with standard deviations indicated.
|
|
|
|---|
The NiR of cyanobacteria is a ferredoxin-dependent enzyme consisting of an 80-kDa polypeptide (the narB gene product), MGD cofactor, and a [4Fe-4S] cluster. The NarB polypeptide is homologous to the catalytic subunit of the NAD(P)H-dependent assimilatory NR and periplasmic NR of bacteria, but unlike the latter two types of NR, it does not require a second polypeptide mediating electron transfer to the catalytic subunit (34). NarM is hence clearly not a protein component of the cyanobacterial NR. Since the narM mutant accumulated the transcripts of the NR structural gene (narB) and the genes required for MGD biosynthesis (the moa operon, moeA, and mobA) to levels comparable to those in the wild-type strain (Fig. 6), the defect of the narM mutant is not at the step of transcription of these genes. Possible roles for NarM in expression of functional NR include (i) stabilization of the molybdenum cofactor and/or the NarB polypeptide and (ii) assembly of the cofactors (MGD and the iron-sulfur cluster) with the NarB protein to yield functional NR. It should be noted that the bacteria containing other types of prokaryotic NR do not have NarM homologs, although the prokaryotic NR enzymes have a similar catalytic subunit that is homologous to NarB. The role of NarM thus seems to be specific to the NarB-type NR of cyanobacteria. The only exception to this might be the case of Burkholderia fungorum, which has a gene encoding a protein 27% identical to NarM of S. elongatus strain PCC7942. The gene is located upstream of fdhD, a gene supposed to be required for expression of functional formate dehydrogenase (http://www.tigr.org/tdb/mdb/mdbinprogress.html). Since formate dehydrogenase is a molybdoenzyme containing an iron-sulfur cluster and its amino acid sequence is about 30% identical to the NR sequences, the NarM homolog of this organism may be involved in biosynthesis of functional formate dehydrogenase. It should also be noted that expression of narB in E. coli leads to expression of functional ferredoxin-dependent NR even though E. coli does not have narM (37). This suggests that a gene product(s) of E. coli plays the role of NarM in E. coli. Several gene products involved in MGD biosynthesis have been identified and characterized in S. elongatus strain PCC7942 (35, 36). The only MGD-related gene of E. coli that is not found in S. elongatus stain PCC7942 is mog, which is required for the efficient incorporation of molybdate into molybdopterin (14). Introduction of the mog gene of E. coli into the NM2 mutant by shuttle expression vector, however, did not complement narM mutation in S. elongatus (S. Maeda, unpublished data), and the function of NarM remains unclear. Since the deduced NarM protein has no significant similarity to known proteins, biochemical studies are needed for elucidation of the function of NarM.
While mutation of orf353 in S. elongatus strain PCC7942 had little effect on expression of the nitrate assimilation activities (Table 1), the orf353 homolog (cnaT) of Anabaena sp. strain PCC7120 was recently reported to be required for normal activation of the nitrate assimilation operon and for expression of the nitrate assimilation enzymes (8). Regarding the regulation of the nirA operon, different responses of different cyanobacterial strains have been reported for the mutation of NtcB, the nitrite-responsive LysR-type protein that enhances nirA operon expression when transcription is promoted by NtcA (2). While NtcA alone can maintain the basal level transcription of the nirA operon in S. elongatus strain PCC7942, allowing normal growth of the cells under nitrate-replete conditions (39), mutation of ntcB severely impairs expression of the nitrate assimilation operon and growth on nitrate of Synechocystis sp. strain PCC6803 (3) and Anabaena sp. strain PCC7120 (7), indicating that NtcB is required not only for nitrite enhancement of transcription but also for basal level transcription from the nirA operon transcription in these strains. Although this suggested a possibility that orf353 might have a role related to the function of NtcB, the results obtained from the luxAB reporter strains clearly demonstrated that both NtcA- and NtcB-mediated regulation of nirA operon transcription are normally operating in the orf353 mutant of S. elongatus strain PCC7942 (Fig. 7), excluding the possible relation of ORF353 and NtcB. Therefore, the reason for the difference in phenotype of the orf353 mutant of S. elongatus strain PCC7942 and the cnaT mutant of Anabaena sp. strain PCC7120 remains unknown. As ORF353 has similarity to anthranilate phosphoribosyltransferase (about 20% identity), it may modify other proteins and regulate their activity. Further characterization is needed for elucidation of the function of ORF353.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»