INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

In the nitrogen-fixing (nif) enteric bacteria Klebsiella pneumoniae and Enterobacter cloacae, the nif genes are regulated by the activity of the ntrC, ntrA, nifA, and nifL genes at the transcriptional level (2, 3, 4, 24). nifL and nifA constitute an operon which is regulated by the product of ntrC or autoregulated by the product of nifA, NifA (5, 9). NifA acts as a positive regulator in conjunction with NtrA to activate nif genes (19), while the product of nifL, NifL, acts as a repressor of nif genes under oxygen or in an excess of fixed nitrogen (13).

Our previous investigations showed that the nifLA promoter is highly sensitive to oxygen and that NifA is inactivated by NifL under oxygen (5, 15, 23). Accordingly, we hypothesized that nif regulation by oxygen is mediated at two different levels. First, at the transcriptional level, the oxygen sensitivity of the nifLA promoter ensures that the nifLA promoter and, hence, all other nif promoters are repressed by oxygen; second, at the posttranslational level, NifL interacts with NifA in the presence of oxygen. As the result of a shortage of active NifA, the expression of other nif genes is blocked.

Direct interaction of NifL and NifA has been demonstrated by the two-hybrid system test (16, 23). As to the question of aerial regulation of the nifLA promoter, most investigators have claimed that neither FNR nor the oxrC gene product is involved in the expression of nifLA in response to oxygen (14). Since the activity of the K. pneumoniae nifLA promoter can be prevented by inhibition of DNA gyrase activity, they thus held that aerobic-anaerobic regulation of the nifLA promoter is mainly mediated through the level of DNA supercoiling (6, 8). Our present work demonstrated that an upstream sequence of the nifLA promoter characteristic of the FNR binding sequence, when bound by the trans-acting factor from the bacterial cells, activates anaerobic expression of the nifLA operon. We also proved that DNA gyrase is truly important for anaerobic transcription of the nifLA promoter. A mechanism for aerobic regulation of the nifLA promoter is proposed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Strains and plasmids. The bacterial strains and plasmids used in this work are listed in Table 1.

DNA manipulations. Preparation of plasmids DNA, endonuclease digestion, ligation, and transformation were carried out essentially as described by Sambrook et al. (21).

Construction of nifLA promoter mutants. Initially, the E. cloacae nifLA promoter (Fig. 1) was cloned in vector M13mp19. The resultant clone is referred to as M13mp19-nifLAp. The M13mp19-nifLAp DNA was digested with BglI and HpaI, filled in with Klenow polymerase, and ligated to yield M13mp19-nifLAp-D, where the FNR site consensus sequence upstream of the nifLA promoter was deleted. Promoter mutant clones M13mp19-nifLAp-m1 and M13mp19-nifLAp-m2, with a mutation in the FNR site consensus sequence, were constructed by oligonucleotide-mediated mutagenesis as described by Sambrook et al. using a 0.5-kb BamHI-to-HindIII fragment of the nifLA promoter region carried in M13mp19-nifLAp as a target (21). A 26-mer mutant primer (5'AACAGGCGTTAACAGGGCCCAGATCG3') complementary to bases 63 to 38 with the four most conserved bases mismatched with respect to the FNR consensus was synthesized and used to construct M13mp19-nifLAp-m1. Another 24-mer mutant primer (5'ACAGGCGTTAACATCGATCCAGAT3') complementary to bases 62 to 39 with one most conserved base mismatched was synthesized and used to construct M13mp19-nifLAp-m2. The 26-mer primer and the 24-mer primer create a unique restriction ApaI or ClaI endonuclease site in each case. It can be used to screen the desired mutants.

View larger version (29K): [in this window] [in a new window]   FIG. 1.   Nucleotide sequence of the E. cloacae nifLA promoter region. The NtrC binding site is shown as a dashed box, the consensus sequence of the FNR site is boxed, the NifA binding site is shadowed, the 54-RNA polymerase recognition site is double underlined, the transcription start site is indicated by the arrow, and the restriction endonuclease site is underlined. The conserved sequence of the FNR site is indicated by the star.

Assay of -galactosidase. Bacteria were grown aerobically for 24 h at 28°C in nitrogen-free minimal medium supplemented with 0.01% casein hydrolysate, 1 mg of vitamin B₁ per liter, 20 mg of glutamine per liter, and appropriate antibiotics (5). Cells were pelleted, washed, and resuspended in the same medium and then grown under aerobic or anaerobic (flashing with nitrogen) conditions for 8 h. -Galactosidase activity was assayed as described by Miller (17).

DNA binding assay. Binding of protein to the FNR site consensus sequence was monitored by measuring the reduction in the electrophoretic mobility of the labeled probe DNA fragments as described by Fried and Crothers (11). In a 20-µl total sample volume, 3.5 pmol of an [-³²P]dATP-labeled double-stranded probe was incubated with 14 µg of protein extracts containing 3 µg of calf DNA for 25 min at 25°C. A DNA fragment with the sequence 5'TTAATGCCGTCGAATACCGATCTGGATCAATGTTAACGCCTGTT3', followed by the nifLA promoter containing the FNR binding site consensus sequence, and a DNA fragment with the sequence 5'TTAATGCCGTCGAATACCGATCTGGGGCCCTGTTAACGCCTGTT3', followed by the nifLA promoter with the FNR binding site consensus sequence mutated, were synthesized as probes. Protein-DNA complexes were separated by 12% polyacrylamide gel electrophoresis and then autoradiographed.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

A sequence upstream of the nifLA promoter and its role in the activity of the nifLA operon. As reported previously for E. cloacae E26, there are three cis-acting elements residing in the region upstream of the nifLA operon: the ⁵⁴-RNA polymerase recognition site at 24 to 12, the NtrC binding site at 171 to 135, and the NifA binding site at positions +44 to +59 from the transcription start site (TSS) (5). Through analysis of the DNA sequence of the nifLA promoter, we found a sequence, CCGAT-N₄-ATCAA, at positions 69 to 48 from the TSS (Fig. 1) which is characteristic of the sequence of an FNR binding site (10, 22). In order to know the role of this defined sequence in the activity of the nifLA operon under anaerobic conditions, we constructed a promoter mutant with a 48-bp BglI-to-HpaI fragment including the consensus FNR binding site upstream of the TSS deleted and cloned it into plasmid pGD926 to form a nifLAp-lacZ translational fusion. After it was introduced into E. cloacae E26 or Escherichia coli YMC9, the -galactosidase activity of the fusion was measured. As shown in Table 2, deletion of the consensus FNR binding site caused a marked decrease in the activity of the nifLA promoter under anaerobic conditions. Furthermore, we made promoter mutants by site-directed mutagenesis. One mutant contains the sequence CCGAT CTGG GGCCC, where the conserved sequence ATCAA was changed to GGCCC, and another mutant contains the sequence CCGAT CTGG ATCGA, where the ATCAA sequence was changed to ATCGA. In addition, a mutant with base pairs T¹¹¹, C¹¹², and C¹²¹ outside the FNR binding site consensus sequence changed to G¹¹¹, T¹¹², and G¹²¹ was constructed and used as a control. After these mutants were cloned into pGD926 to form nifLAp-lacZ fusions, their activities were measured. Data in Table 2 show that either deletion or alteration of the sequence of the consensus FNR site produced low activity under anaerobic conditions, whereas mutation at sites outside of this sequence did not affect the anaerobic expression of the fusion (Table 2). These results substantiate the evidence that the consensus FNR site upstream of the nifLA promoter is important for regulation of the nifLA operon.

A trans-acting factor binds to the defined sequence upstream of the nifLA promoter. To test if there is a trans-acting factor bound to the defined sequence upstream of the nifLA promoter which enhances the anaerobic expression of the nifLA operon, a gel mobility shift assay was conducted. The fragments encompassing the consensus FNR site or its variants were incubated with the soluble protein extracts from E. cloacae or E. coli and assayed for DNA-protein complex formation. As shown in Fig. 2, a slower-migrating complex was detected following incubation of the probe containing the FNR site consensus sequence with the cell extracts. In contrast, no complex was detected following incubation of the probe containing the mutated consensus sequence of the FNR site with the cell extracts. We thus concluded that a trans-acting factor binding to the FNR consensus sequence is present in E. cloacae and also in E. coli.

View larger version (25K): [in this window] [in a new window]   FIG. 2.   Gel mobility shift assay of a DNA fragment containing the FNR site consensus sequence incubated with protein extracts from E. cloacae E26 (lanes 2 to 5), E. coli YMC9 (lanes 7 and 9), and E. coli JRG2865 (lane 6). Lanes: 1 and 8, labeled DNA fragment; 2, 3, 6, and 7, labeled DNA fragment incubated with protein extract from anaerobic culture; 4, labeled DNA fragment incubated with protein extract from aerobic culture; 5 and 9, labeled DNA fragment containing mutated FNR site consensus sequence incubated with protein extract from an anaerobic culture. The reaction mixture in lane 2 also contained 300 pmol of unlabeled probe DNA as a specific competitor.

Superhelical status of DNA and activity of the nifLA promoter. Before the identification of a regulatory factor responding to oxygen status, it has been reported that the K. pneumoniae nifLA promoter requires a specific degree of negative supercoiling for expression, which is only possible under anaerobic conditions (6, 8). To examine this possibility, we tested influence of gyrase activity on the transcription of the E. cloacae nifLA promoter by introducing the nifLAp-lacZ fusion into an E. coli DH5 gyrA mutant or into gyrA⁺ strains of E. coli and E. cloacae in the presence of a gyrase inhibitor. The known gyrase-dependent K. pneumoniae nifLAp-lacZ fusion (6, 8) was also run as a control. The results showed that expression of the nifLAp-lacZ fusion has been markedly halted both in the gyrA mutant and in the gyrA⁺ strains with the presence of the gyrase inhibitors under aerobic or anaerobic conditions (Tables 2 and 3). When a plasmid carrying constitutively expressed gyrA was introduced into the gyrA DH5 mutant harboring the nifLAp-lacZ fusion, the activity of the fusion was restored. However, a gyrB clone did not have the same effect (Table 2). Curiously, the nifH operon of Rhizobium meliloti, which is known to be insensitive to oxygen (unpublished data), appears to be DNA gyrase dependent too. These results confirm the earlier inference that DNA gyrase activity is crucial for transcription of the nifLA operon and possibly other nif genes.

View this table: [in this window] [in a new window]   TABLE 3.   Effects of coumermycin A1 and novobiocin on -galactosidase activities of nifLAp-lacZ translation fusions