Glutamine synthetase and GlnR regulate nitrogen metabolism in Paenibacillus polymyxa WLY78

ABSTRACT

Paenibacillus polymyxa WLY78, a N₂-fixing bacterium, has great potential use as a biofertilizer in agriculture. Recently, we have revealed that GlnR positively and negatively regulates the transcription of the nif (nitrogen fixation) operon (nifBHDKENXhesAnifV) in P. polymyxa WLY78 by binding to two loci of the nif promoter according to nitrogen availability. However, the regulatory mechanisms of nitrogen metabolism mediated by GlnR in the Paenibacillus genus remain unclear. In this study, we have revealed that glutamine synthetase (GS) and GlnR in P. polymyxa WLY78 play a key role in the regulation of nitrogen metabolism. P. polymyxa GS (encoded by glnA within glnRA) and GS1 (encoded by glnA1) belong to distinct groups: GSI-α and GSI-β. Both GS and GS1 have the enzyme activity to convert NH₄⁺ and glutamate into glutamine, but only GS is involved in the repression by GlnR. GlnR represses transcription of glnRA under excess nitrogen, while it activates the expression of glnA1 under nitrogen limitation. GlnR simultaneously activates and represses the expression of amtBglnK and gcvH in response to nitrogen availability. Also, GlnR regulates the expression of nasA, nasD1D2, nasT, glnQHMP, and glnS.

IMPORTANCE

In this study, we have revealed that Paenibacillus polymyxa GlnR uses multiple mechanisms to regulate nitrogen metabolism. GlnR activates or represses or simultaneously activates and inhibits the transcription of nitrogen metabolism genes in response to nitrogen availability. The multiple regulation mechanisms employed by P. polymyxa GlnR are very different from Bacillus subtilis GlnR which represses nitrogen metabolism under excess nitrogen. Both GS encoded by glnA within the glnRA operon and GS1 encoded by glnA1 in P. polymyxa WLY78 are involved in ammonium assimilation, but only GS is required for regulating GlnR activity. The work not only provides significant insight into understanding the interplay of GlnR and GS in nitrogen metabolism but also provides guidance for improving nitrogen fixation efficiency by modulating nitrogen metabolism.

INTRODUCTION

Nitrogen (N) is an essential element of most macromolecules (e.g., proteins, nucleic acids, and cell wall components) in bacterial cells. Of a variety of N sources, ammonium (NH₄⁺) serves as the preferred N source for supporting bacterial growth (1). Ammonium is usually taken up by ammonium transporter (AmtB) from environments. However, at high external ammonium concentrations, the ammonia gas can freely diffuse into the cell without the need for AmtB. In addition, a small portion of bacteria and archaea (called diazotrophs) utilize nitrogenase to convert atmospheric N₂ to ammonium in a process known as biological nitrogen fixation. The fixed nitrogen (ammonia) is readily assimilated into amino acids (glutamate and glutamine) rather than released into the environment (2).

Glutamine synthetase (GS) which is ubiquitous in bacteria is the major enzyme responsible for the assimilation of ammonium into organic compounds. GS, with a high-affinity Km (0.1 mM) for NH₄⁺, catalyzes the synthesis of glutamine from glutamate and ammonium, and then glutamate synthase (GOGAT) converts glutamine into two molecules of glutamate by transferring the amide group from glutamine to 2-oxoglutarate (3 – 5). GS-GOGAT pathway consumes one ATP. Glutamate and glutamine are the major amino group donors for all nitrogen-containing compounds, for example, other amino acids, purines, pyrimidines, and vitamins (6). Also, the other major pathway for ammonium assimilation is through the glutamate dehydrogenase (GDH) pathway or alanine dehydrogenase (ADH) pathway that are redox-dependent reactions as compared to the ATP-dependent mechanism used by the GS system (2). GDH, with a high Km (1 mM) for NH₄⁺, catalyzes the formation of glutamate from ammonium and 2-oxoglutarate under excess nitrogen. ADH, a high Km value (≧1 mM) for NH₄⁺, catalyzes the synthesis of alanine from pyruvate and ammonium under high levels of nitrogen (7).

GS is divided into three types, GSI, GSII, and GSIII, depending on differences in sequence and structure (6). GSI is found in bacteria and archaea, while GSII is present in eukaryotes and some bacterial species (8 – 10). Members in the Rhizobiaceae, Frankiaceae, and Streptomyces have both GSI and GSII. The type GSIII is identified in anaerobic bacteria (e.g., Bacteroides fragilis, Ruminococcus albus 8) and cyanobacteria and it operates under conditions where ammonia is seldom limiting (11 – 15). Some members of anaerobic bacteria and cyanobacteria have both GSI and GSIII. Generally, GSI is divided into two groups: GSI-α and GSI-β. GSI-α enzymes are found in low G + C gram-positive bacteria (Bacillus sp., some thermophilic bacteria, and archaea), while GSI-β enzymes are present in enterobacteria (e.g., Escherichia coli) and the high G + C gram-positives (e.g., Streptomyces sp.) (16). Although GSI-α and GSI-β proteins have 35–41% of sequence identity, their regulation mechanisms are different. GSI-β enzymes are regulated by adenylylation/deadenylylation (17), whereas GSI-α activity is feedback inhibited by its product, glutamine (18).

GlnR, a nitrogen regulator, exists widely in gram-positive bacteria, such as Bacillus, Lactobacillus, Lactococcus, Streptomyces, Streptococcus, and Mycobacterium (5, 19 – 24). The GlnR proteins of Firmicutes and Actinobacteria are fundamentally different. GlnR of Actinobacteria belongs to the OmpR-type GlnR, while GlnR of Firmicutes belongs to MerR-type GlnR. GlnR of Firmicutes is a key transcriptional regulator of nitrogen metabolism in B. subtilis. In addition to GlnR, TnrA is another transcriptional regulator of nitrogen metabolism in B. subtilis (25 – 27). GlnR and TnrA have a high sequence similarity at their N terminal domains that bind to DNA sequences (GlnR/TnrA-binding sites) with a common consensus sequence (5′-TGTNAN7TNACA-3′) (5, 28). However, the C terminal domains of GlnR and TnrA are very different (29 – 32). TnrA functions primarily as an activator, while GlnR acts as a repressor (5). GlnR represses the transcription of the glnRA operon (negative auto-regulation) (19), tnrA (33), and the urease gene cluster (ureABC) under excess nitrogen condition (34). TnrA activates the expression of amtBglnK (also called nrgAB, encoding ammonia transporter AmtB and regulatory protein GlnK) (35), nasDEF (encoding proteins related to nitrite reduction) (28, 36), glnQHMP (glutamine uptake) (5), and its gene (tnrA) (25) under limited nitrogen condition. Meanwhile, TnrA also represses the expression of glnRA (28, 33) and gltAB (encoding glutamate synthase) (37).

The activity of B. subtilis GS is feedback inhibited by its product glutamine (38). The feedback-inhibited GS (FBI-GS) regulates the activities of GlnR and TnrA. FBI-GS acts as a chaperone to stabilize the DNA-binding activity of GlnR, which inhibits the transcription of nitrogen assimilation genes under excess nitrogen (29, 39). TnrA binds to its DNA-binding site and then turns on the transcription of nitrogen assimilation genes during nitrogen limitation, while under nitrogen excess condition, the FBI-GS forms a stable complex with TnrA and then inhibits its DNA-binding activity (40). Thus, GS, an enzyme of ammonium assimilation, directly interacts with and regulates the DNA-binding activity of GlnR and TnrA.

Paenibacillus is a genus of gram-positive, facultative anaerobic, and endospore-forming bacteria, originally included within the genus Bacillus and then reclassified as a separate genus in 1993 (41). The genus Paenibacillus comprises N₂-fixing and non-N₂-fixing members. N₂-fixing Paenibacillus strains possess a minimal nitrogen-fixation (nif) gene cluster containing 9–10 genes [nifBnifHnifDnifKnifEnifNnifX(orf1)hesAnifV] (42, 43). Unlike B. subtilis, N₂-fixing Paenibacillus polymyxa WLY78 and Paenibacillus sabinae T27 have a glnR gene but lack the tnrA gene (43, 44). These strains have two glnA genes, one of which is linked together with glnR as glnRA operon and the other (glnA1) of which is seated separately (42 – 44). P. polymyxa GS (encoded by glnA within glnRA operon) has high similarity with B. subtilis GS (GSI-α) and P. polymyxa GS1 (encoded by glnA1) exhibits high similarity with E. coli GS (GSI-β). P. polymyxa GlnR simultaneously acts as an activator and a repressor to regulate the transcription of the nif operon (nifBHDKENXhesAnifV) by binding to two GlnR-binding sites in the nif promoter region in response to nitrogen availability, and GS (encoded by glnA within glnRA operon) is essentially required for repressing nif gene expression by promoting the binding ability of GlnR to DNA site (45). The FBI-GS is also involved in repressing the nif gene expression of P. sabinae T27 under nitrogen excess condition (46) and GS stabilizes the binding of GlnR to nitrogen fixation gene operators in Paenibacillus riograndensis SBR5 (47). Recently, we have revealed that GlnR is a global regulator that regulates the transcription of 14 nitrogen metabolism genes/operons (glnRA, amtBglnK, glnA1, glnK1, glnQHMP, nasA, nasD1, nasD2EF, gcvH, ansZ, pucR, oppABC, appABCDF, and dppABC) using chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR), electrophoretic mobility shift assay (EMSA), and genome-wide transcription analyses (48). However, the GlnR-mediated regulation mechanism in nitrogen metabolism in Paenibacillus is not known and the functions of both GS proteins encoded by glnA within the glnRA operon and glnA1 are not clear.

In this study, we have revealed that GlnR of P. polymyxa WLY78 directly binds to the promoter regions of glnRA, glnA1, amtBglnK, and gcvH using EMSA and surface plasmon resonance (SPR). Mutation on GlnR-binding site(s) and qRT-PCR analyses have shown that GlnR positively or negatively or simultaneously positively and negatively regulates the transcription of nitrogen metabolism genes/operons (glnRA, glnA1, amtBglnK, gcvH, glnQHMP, glnS, nasA, nasD1, nasD2, and nasT). P. polymyxa GS and GS1 belong to GSI-α and GSI-β, respectively. Both GS and GS1 of P. polymyxa WLY78 possess enzyme activity, but only GS is involved in regulating the transcription of its gene (glnA), glnR, amtB, and other genes. Although P. polymyxa GS1 (encoded by glnA1) has high similarity with E. coli GS (GSI-β), P. polymyxa GS1 is not regulated by adenylation.

RESULTS AND DISCUSSION

Paenibacillus polymyxa GS and GS1 belong to distinct groups

P. polymyxa WLY78 has two glnA genes encoding GS, one of which is organized with glnR as the glnRA operon, and the other (glnA1) is located separately (Fig. S1A). P. polymyxa GS encoded by glnA within glnRA operon and GS1 encoded by glnA1 showed 39% identity, but they exhibited 39% and 76% identities with B. subtilis GS, respectively, and they showed 54% and 41% identities with E. coli GS, respectively (Fig. S1B). The data suggest that P. polymyxa GS1 has a higher identity to E. coli GS (GSI-β), while P. polymyxa GS has a higher identity to B. subtilis GS (GSI-α).

Sequence alignments of GS enzymes from P. polymyxa WLY78, B. subtilis and E. coli are shown in Fig. 1A. B. subtilis GS has a specific residue Arginine (R) at position 62 that plays a central role in glutamine binding and glutamine feedback inhibition (31). P. polymyxa GS has the specific residue R at position 61, but P. polymyxa GS1 does not have the R residue at position 61 or 62 (Fig. 1A). Like E. coli GSI-β, P. polymyxa GS1 sequence has a unique 25–27-aa insertion, but P. polymyxa GS and B. subtilis GSI-α lack the insertion (Fig. 1A) (16, 49). The enzyme activity of GSI-β in E. coli is regulated by reversible adenylylation and deadenylylation (50) and the tyrosine (Y) residue 401 located in the C-terminus of E. coli GS is the site of adenylylation/deadenylylation (51). Like E. coli GS, B. subtilis GS and P. polymyxa GS and GS1 have tyrosine (Y) residue located in their C-terminus, consistent with the report that this tyrosine is conserved in most of the GSI-type proteins (16). However, the enzyme activity of B. subtilis GS is not regulated by adenylylation (49). Phylogenetic analysis revealed that P. polymyxa GS1 is clustered in a large group with E. coli GS, while P. polymyxa GS is clustered together with B. subtilis GS (Fig. 1B), consistent with the above blast results. These data indicate that GS and GS1 of P. polymyxa WLY78 belong to GSI-α and GSI-β, respectively.

Fig 1.

Sequence alignments and phylogeny analysis of GS proteins. (A) Alignments of GS proteins from P. polymyxa WLY78, E. coli, and B. subtilis. The red arrow indicates the specific residue arginine (R) that is the site for glutamine feedback inhibition. The green box indicates a 25–27-aa insertion in the E. coli GS sequence. The blue arrow indicates the tyrosine (Y) residue which is the site of adenylylation/deadenylylation for E. coli GS. (B) The phylogenetic tree was constructed using MEGA (6.0).

Phenotype of the ΔglnA and ΔglnA1 mutants

Phenotypic characteristics of the wild-type strain and ΔglnA, ΔglnA1, and ΔglnAglnA1 mutants were analyzed on plates containing glucose minimal medium with glutamate or glutamine as the only nitrogen source (Fig. S2A). The ΔglnA and ΔglnA1 mutants grew well on glutamate or glutamine as the wild-type strain did. The double deletion mutant ΔglnAglnA1 only grew well on glutamine, but it did not grow on glutamate, suggesting that both glnA and glnA1 are essential for ammonium assimilation in P. polymyxa WLY78. Our data are consistent with the reports that the glnA deletion mutant in B. subtilis (52) and Haloferax mediterranei (53) was glutamine auxotrophic.

The wild-type, ΔglnA, and ΔglnA1 strains were cultivated in a broth medium containing 2 mM glutamate without NH₄Cl or with 100 mM NH₄Cl (Fig. S2B). When grown in a broth medium without NH₄Cl, the wild-type strain and both ΔglnA and ΔglnA1 strains grow at similar rates with 0.2 optical density at 600 nm after 24 h of culture. When cultivated in a broth medium containing 100 mM NH₄Cl, the wild-type strain grew more quickly than both ΔglnA and ΔglnA1 mutants did. Both wild-type strain and ΔglnA1 mutant reached the growth peak at 14 h of cultivation with 1.4 optical density at 600 nm and 1.1 optical density at 600 nm, respectively, whereas ΔglnA mutant reached the growth peak at 24 h of cultivation with 0.8 optical density at 600 nm. The data that inhibition of high concentration of NH₄⁺ on the growth of both ΔglnA and ΔglnA1 mutants suggest that both glnA and glnA1 genes, especially glnA, are required for ammonium assimilation.

Furthermore, wild-type, ΔglnA, ΔglnA1, and ΔglnAglnA1 strains were cultivated in a broth medium supplemented with 5 mM sodium nitrate as the only nitrogen source (Fig. S2C). The wild-type strain started to grow rapidly after incubation for 10 h. Compared to the wild-type strain, both ΔglnA and ΔglnA1 mutants grew much more slowly and the double mutant strain ΔglnAglnA1 did not grow at all. The data suggest that both glnA and glnA1 are required for nitrate assimilation, consistent with the report that the regulation of the nitrate assimilation system is impaired in the glnA-disrupted mutant of cyanobacterium Synechocystis sp. strain PCC 6803 (54). Thus, we infer that P. polymyxa WLY78 can also use nitrate as the only nitrogen source to maintain cell growth when glutamate or glutamine is missing in the environment. It is generally recognized that nitrate is converted into nitrite by nitrate reductase, then nitrite is converted into NH₄⁺ by nitrite reductase, and finally NH₄⁺ is converted into glutamine through GS.

The intracellular concentrations of NH₄⁺, glutamate and glutamine

The intracellular concentrations of NH₄⁺, glutamate, and glutamine of wild-type, △glnA1, and △glnA strains were determined when grown in a broth medium containing 2 mM glutamate without NH₄Cl or with 100 mM NH₄Cl at different times (Fig. S3). Our results showed that the intracellular concentrations of NH₄⁺ of wild-type, △glnA1, and △glnA strains grown in medium without NH₄Cl or with 100 mM NH₄Cl for 16 h were much higher than those of these strains grown in medium without NH₄Cl or with 100 mM for 0 h (Fig. S3A), suggesting that high concentration of the intracellular NH₄⁺ was produced during the period of the bacterial rapid growth. The data are consistent with the findings that all wild-type, △glnA1, and △glnA strains grew rapidly at 16 h after inoculation when they were cultivated in a broth medium containing 0 mM NH₄Cl or 100 NH₄Cl (Fig. S2B). When grown in a broth medium containing 0 mM NH₄Cl for 16 h, the wild-type strain accumulated the highest level of intracellular glutamate, but △glnA1 and △glnA mutants almost did not have an accumulation of intracellular glutamate. When grown in a broth medium containing 100 mM NH₄Cl for 16 h, wild-type, △glnA1, and △glnA strains have similar levels of intracellular glutamate (Fig. S3B). When grown in a broth medium containing 0 mM NH₄Cl for 16 h, the wild-type strain accumulated the highest level of the intracellular glutamine, but △glnA1 and △glnA mutants almost did not have an accumulation of intracellular glutamine (Fig. S3C), suggesting that deletion of either glnA1 or glnA led to reduced GS activity. However, △glnA mutant had the highest level of intracellular glutamine among wild-type, △glnA1, and △glnA strains when grown in a medium containing 100 mM NH₄Cl for both 0 h and 16 h (Fig. S3C), suggesting that the activity of GS1 encoded by glnA1 gene in △glnA mutant was not inhibited by NH₄⁺ and GS1 activity encoded by glnA1 was regulated by glnA within glnRA operon. By contrast, the intracellular glutamine of △glnA1 mutant did not show great change under both 0 mM and 100 mM NH₄⁺ conditions (Fig. S3C), indicating that GS encoded by glnA within glnRA operon was not affected by GS1 encoded by glnA1.

Both P. polymyxa GS and GS1 possess enzyme activity

It is known that the GS of B. subtilis is synthesized during nitrogen limitation and the enzymatic activity of the B. subtilis GS enzyme is subjected to feedback inhibition by the reaction product glutamine in vivo (31, 32, 38). Here, the GS activity in wild-type (P. polymyxa WLY78), ΔglnA, andΔglnA1 strains grown on glucose minimal medium containing low nitrogen (glutamate) or with excess nitrogen (glutamine or glutamate plus NH₄⁺ or NH₄⁺) was measured using the biosynthetic assay described in Materials and methods (Fig. 2). Wild-type cells grown with glutamate as the nitrogen source showed the highest level of GS activity, while the lowest level of GS activity was obtained in wild-type cells grown on glutamine. Wild-type cells contained intermediate levels of GS activity when the nitrogen source was glutamate plus NH₄⁺ or NH₄⁺. The data are consistent with the levels of GS activity of B. subtilis grown with glutamate, glutamate plus NH₄⁺ and glutamine as nitrogen sources (55). Both ΔglnA and ΔglnA1 mutants grown with glutamate (nitrogen limitation) had reduced GS activity compared to wild-type strain (Fig. 2), suggesting that both GS and GS1 enzymes are required for maintaining the high GS enzyme activity during nitrogen limitation. In contrast to the wild-type strain, ΔglnA1 mutant grown with all nitrogen sources produced the GS activities at similar and modest levels, suggesting that the enzyme activity of GS encoded by glnA within glnRA in ΔglnA1 mutant was inhibited by excess nitrogen (glutamine, glutamate plus NH₄⁺, and NH₄⁺). However, like the wild-type strain, ΔglnA mutant grown with high nitrogen (glutamine or glutamate plus NH₄⁺ or NH₄⁺) produced the GS activities at variable levels, indicating that the GS1 encoded by glnA1 in ΔglnA mutant was not regulated by glutamine. The data are in agreement with the results that P. polymyxa GS and GS1 belong to GSI-α and GSI-β, respectively, and that the GSI-α enzyme was inhibited by glutamine and NH₄⁺ (38).

Fig 2.

The GS enzyme activity of wild-type, ΔglnA, andΔglnA1 strains grown on glucose minimal medium containing low nitrogen (glutamate) or excess nitrogen (glutamine or glutamate plus NH₄⁺ or NH₄⁺). The results are the mean from at least three independent experiments. Error bars indicate standard deviation. Significance was determined by one-way analysis of variance. Not significant (ns P > 0.05), *P < 0.05, **P < 0.01.

GS encoded by glnA within glnRA operon regulates the transcription of glnA1, glnA, glnR (glnRA) and amtB

The transcript levels of glnA, glnR, and glnA1 in wild-type, ΔglnA, and ΔglnA1 strains grown in medium containing 2 mM glutamate without NH₄⁺ (nitrogen limitation) or with 100 mM NH₄⁺ (excess nitrogen) were detected by qRT-PCR (Fig. 3). The glnA1 in wild-type strain was highly transcribed under nitrogen limitation (0 mM NH₄⁺) and was expressed at basal level under nitrogen excess condition (100 mM NH₄⁺). By contrast, the glnA1 in ΔglnA and ΔglnRA mutants was highly transcribed under nitrogen excess condition and was expressed at basal level under nitrogen limitation (Fig. 3A), consisting with the above results that △glnA mutant had a much higher level of the intracellular glutamine in 100 mM NH₄⁺ than that in 0 mM NH₄⁺ after 16 h of growth (Fig. S3C) and suggesting that glnA1 expression is regulated by GS encoded by glnA within glnRA operon. As observed in the wild-type strain, the glnA within glnRA operon in ΔglnA1 was highly transcribed under nitrogen limitation and was transcribed at the basal level under nitrogen excess condition (Fig. 3B), indicating that the glnAR transcription is not regulated by GS1 encoded by glnA1.

Fig 3.

qRT-PCR analysis of transcript levels of glnA, glnA1, and glnR in wild-type, ΔglnA, and ΔglnA1 strains grown in a medium containing 2 mM glutamate without NH₄⁺ or with 100 mM NH₄⁺. (A) Relative glnA1 transcript levels. (B) Relative glnA2 transcript levels. (C) Relative glnR transcript levels. (D) Relative transcript levels of amtB. The results are the mean from at least three independent experiments. Error bars indicate standard deviation. Significance was determined by one-way analysis of variance. Not significant (ns P > 0.05), *P < 0.05, **P < 0.01.

As observed in wild-type strain, the glnR in ΔglnA1 mutant was highly transcribed under low nitrogen and it was transcribed at the basal level in ΔglnA1 mutant under excess nitrogen. The glnR in ΔglnA was constitutively expressed under both nitrogen-limited and -excess conditions (Fig. 3C), suggesting that glnR expression is regulated by GS but not by GS1. Our results are consistent with the report that GS represses the glnRA expression in B. subtilis (56, 57)

The transcript levels of amtB in △glnA1 mutant were similar to those in wild-type strain under both nitrogen conditions, suggesting that GS1 is not involved in regulation of amtB expression. In contrast to the wild-type strain, the transcript levels of amtB in △glnA were reduced under nitrogen limitation and increased under excess nitrogen, suggesting that GS is involved in the regulation of amtB expression.

Taken together, the results have revealed that P. polymyxa GS encoded by glnA within the glnRA operon regulates the transcription of glnAR , glnA1, and amtB, but GS1 encoded by glnA1 does not have a regulatory role. Our results support that P. polymyxa GS and GS1 belong to GSI-α and GSI-β, respectively.

GlnR represses the transcription of glnRA under excess nitrogen

To investigate whether the glnRA in P. polymyxa WLY78 is regulated by GlnR, the glnRA promoter is analyzed. The 5′-rapid amplification of cDNA end (5′-RACE) experiment showed that the transcriptional start site (TSS) of glnRA operon in P. polymyxa WLY78 is adenine (A) located 29 bp upstream of the translation initiation site of GlnR (Fig. 4A). The typical −35 (TTGACA) and −10 (TACAAT) regions showed the characteristics of a σ^A-dependent promoter. The glnRA promoter region contains a GlnR-binding site that is seated 42 bp upstream of the TSS of glnRA and lies upstream of −35 region (Fig. 4A). Then, the binding of GlnR to the GlnR-binding site was determined by SPR assay. There was no binding signal in the absence of GlnR and the binding signal was enhanced with the increase in GlnR protein concentration (Fig. 4B), indicating that GlnR directly binds to the DNA site. The association rate (K_A) and the dissociation rate (K_D) of GlnR to GlnR-binding site were 9.35 × 10⁷ and 1.07 × 10⁻⁸, respectively.

Fig 4.

GlnR-binding site in the glnRA promoter of P. polymyxa WLY78 and transcription analysis of glnR and glnA in wild-type, ΔglnR, and ΔglnA strains. (A) TSS and position of GlnR-binding site. (B) SPR analysis of GlnR binding to GlnR-binding site. (C) Transcript levels of glnA in WT (wild-type strain) and ΔglnR under 0 mM NH₄⁺ and 100 mM NH₄⁺ conditions determined by qRT-PCR. (D) Transcript levels of glnR in WT (wild-type strain) and ΔglnA under 0 mM NH₄⁺ and 100 mM NH₄⁺ conditions determined by qRT-PCR. The results are the mean from at least three independent experiments. Error bars indicate standard deviation. Significance was determined by student’s t-test. Not significant (ns P > 0.05), *P < 0.05, **P < 0.01.

The transcript levels of glnA and glnR in wild-type and ΔglnR strains under different nitrogen concentrations were determined by qRT-PCR. The glnA within glnRA operon in the wild-type strain was highly expressed under nitrogen limitation (0 mM NH₄⁺), but it was expressed at the basal level under excess nitrogen (100 mM NH₄⁺), whereas the glnA in ΔglnR strain was constitutively transcribed under both 0 mM and 100 mM NH₄⁺ conditions (Fig. 4C). Similarly, the glnR in wild-type strain was highly expressed under nitrogen limitation and was expressed at the basal level under excess nitrogen, whereas the glnR in ΔglnA strain was constitutively transcribed under both nitrogen-limited and -excess conditions (Fig. 4D). The results indicate that GlnR represses the transcription of glnRA operon under nitrogen-excess condition, consistent with the reports that GlnR represses the glnRA expression in B. subtilis under excess nitrogen (19, 33, 56).

Our results have shown that GlnR represses the glnRA expression in P. polymyxa WLY78 under excess nitrogen by binding to the single GlnR-binding site in the glnRA promoter. The current results are consistent with the reports that there is a single GlnR-binding site in the glnRA promoter region and FBI-GS stabilizes GlnR binding to the glnRA promoter in P. riograndensis SBR5T (47). By contrast, the expression of the glnRA operon in B. subtilis is repressed by both GlnR and TnrA (33, 37). The glnRA promoter of B. subtilis contains two GlnR/TnrA-binding sites (glnRAo1 and glnRAo2). The glnRAo1 site is located 50 bp upstream of the glnRA transcription start site and lies immediately upstream of the −35 promoter region and the glnRAo2 site is seated 26 bp upstream of the glnRA transcription site and overlaps the −35 promoter region (33, 58). The glnRAo2 site was required for regulation by both GlnR and TnrA, while the glnRAo1 site is only involved in GlnR-mediated regulation. TnrA binds to glnRAo2 site to repress expression under nitrogen limitation (33). GlnR binds simultaneously to both sites and blocks RNA polymerase binding to the promoter region and thus represses transcription of B. subtilis glnRA under excess nitrogen (33, 58). These results suggest that the regulatory mechanisms show a great difference between P. polymyxa glnRA and B. subtilis glnRA.

GlnR activates the transcription of glnA1 under nitrogen limitation

5′-RACE assay showed that the TSS of glnA1 is adenine (A) located 76 bp upstream of the translation initiation site (Fig. 5A). The glnA1 has a σ^A promoter with the common consensus sequences [−10 (TATAAT) and −35 (TTGACC) regions]. The glnA1 promoter has a GlnR-binding site that is seated 63 bp upstream of the glnA1 transcription start site and lies upstream of the −35 region. (Fig. 5A). SPR assay has shown that the affinity of GlnR to site was very strong and the binding of GlnR to site dissociated slowly (Fig. 5B). The corresponding values of K_A and K_D of GlnR to site were 2.57 × 10⁷ and 3.89 × 10⁻⁸, respectively.

Fig 5.

Position and binding affinity of GlnR-binding site in the glnA1 promoter and transcription analysis of glnA1 in P. polymyxa WLY78. (A) TSS and the GlnR-binding site. (B) SPR analysis of GlnR binding to the GlnR-binding site. (C) Transcript levels of glnA1 in WT (wild-type strain) and ΔglnR under 0 mM NH₄⁺ and 100 mM NH₄⁺ conditions determined by qRT-PCR. The results are the mean from at least three independent experiments. Error bars indicate standard deviation. Significance was determined by student’s t-test. Not significant (ns P > 0.05), *P < 0.05, **P < 0.01.

As shown in Fig. 5C, the transcript level of glnA1 in ΔglnR mutant under nitrogen limitation (0 mM NH₄⁺) was significantly reduced compared to that of wild-type strain, while the glnA1 in ΔglnR mutant was transcribed at low level as wild-type strain did under excess nitrogen (100 mM NH₄⁺). The results show that GlnR activates the expression of glnA1 gene under nitrogen limitation. This study has revealed that the regulation mechanisms of glnA1 and glnRA operons mediated by GlnR are different.

GlnR simultaneously activates and represses the transcription of amtBglnK (nrgAB) in response to nitrogen availability

5′-RACE assay showed that the TSS of amtBglnK is adenine (A) located 50 bp upstream of the translation initiation site (ATG) of AmtBGlnK in P. polymyxa WLY78 (Fig. 6A). The amtBglnK operon possesses a σ^A promoter with the common consensus sequences (−10 TATAAT and −35 TTGTCA regions). The amtBglnK promoter has two GlnR-binding sites and the two sites have a 93 bp interval. GlnR-binding site I is located 43 bp upstream of the −35 region, while GlnR-binding site II is seated downstream of the −10 region (Fig. 6A).

Fig 6.

Characteristics of the amtBglnK promoter and effects of mutation of GlnR-binding site(s) on the transcription of amtBglnK. (A) Two GlnR-binding sites in the amtBglnK promoter. (B) SPR analysis of GlnR binding to two GlnR-binding sites. (C) The binding affinity of GlnR to two sites. (D) A schematic site-specific mutation of the GlnR-binding sites. (E) EMSA verification of the binding of GlnR to the native GlnR-binding sites, the mutated GlnR-binding site I and GlnR-binding site II. (F) Transcription analysis of amtB and glnK in wild-type strain and mutants (PamtBT1, PamtBT2, and PamtBT12) under 0 mM and 100 mM NH₄⁺ by qRT-PCR. (G) Transcription analysis of amtB in wild-type strain under 0 mM and 100 mM NH₄⁺ by qRT-PCR. The results are the mean from at least three independent experiments. Error bars indicate standard deviation. Significance was determined by a one-way analysis of variance. Not significant (ns P > 0.05), *P < 0.05, **P < 0.01.

SPR assay revealed that GlnR binds to the two sites with different affinity abilities (Fig. 6B). The binding signal of GlnR to site I appeared when GlnR protein concentration was higher than 7.8 nM, while the binding signal of GlnR to site II appeared when GlnR protein concentration was higher than 3.9 nM. The affinity of GlnR for site II is stronger than that for site I (Fig. 6C). The data are consistent with the affinity pattern of the two GlnR-binding sites in the nif promoter of P. polymyxa WLY78 (40).

To investigate where the two GlnR-binding sites are the targets of GlnR, the site-specific mutagenesis on the two GlnR-binding sites was performed (Fig. 6D). The consensus sequences TGACAT in site I were replaced with GGTACC (the restriction site of KpnI), generating mutant PamtBT1. The consensus sequences TAACAT in site II were replaced with ATCGAT (the restriction site of ClaI), yielding mutant PamtBT2. Furthermore, a double mutant PamtBT12 with a mutation on both GlnR-binding sites was generated. EMSA confirmed that GlnR did not bind to the mutated sites (Fig. 6E).

qRT-PCR analysis showed that the transcript levels of amtB and glnK in PamtBT1 mutant (mutated site I) were slightly lower than those in wild-type strain under nitrogen-limited condition (0 mM NH₄⁺) (Fig. 6F). By contrast, the transcript levels of amtB and glnK in PamtBT2 mutant (mutated site II) were constitutively expressed under both nitrogen-limited and -excess conditions (100 mM NH₄⁺) (Fig. 6F). The transcript levels of amtB and glnK in double mutant PamtBT12 (mutation of both site I and site II) were nearly abolished under both nitrogen-limited and -excess conditions. The data suggest that site I is involved in activating the expression of amtBglnK under nitrogen limitation and site II is involved in repressing the transcription of amtBglnK under nitrogen-excess condition. Also, qRT-PCR analysis showed that amtB in wild-type strain under nitrogen limitation was highly transcribed, but amtB was nearly not transcribed under excess nitrogen (Fig. 6G).

Our results indicate that P. polymyxa GlnR activates the amtglnK expression under nitrogen limitation by binding to GlnR-binding site I, while GlnR represses the amtglnK expression under excess nitrogen by binding to GlnR-binding site II. The data are consistent with the findings that GlnR positively and negatively regulates transcription of the nif operon by binding to two GlnR-binding sites according to nitrogen availability in P. polymyxa WLY78 (40). However, amtglnK (nrgAB) of B. subtilis grown under nitrogen-limited condition is activated by TnrA by binding to the TnrA-biding site with a consensus sequence (TGTNAN7TNACA) that is centered 49–51 bp upstream of the transcriptional start sites in the amtBglnK promoter (28, 35). This study has revealed that the regulation mechanisms between P. polymyxa amtglnK and B. subtilis amtglnK show great variation.

GlnR activates and represses the transcription of gcvH (glycine cleavage system) in response to nitrogen availability

Glycine is an amino acid composed of two carbons and one amino group. The glycine cleavage system (also called glycine synthase) catabolizes glycine to CO₂, NH₄⁺, and a methylene group (–CH₂–) in response to glycine (59). The glycine cleavage system is composed of four proteins, namely P, H (GcvH), T, and L proteins. GcvH protein is encoded by gcvH and the other three proteins are encoded by the gcvT operon composed of gcvT, gcvPA, and gcvPB (Fig. 7A) (58, 59).

Fig 7.

The glycine cleavage system and transcription analysis of gcvH, gcvT, and gcvPB in WT (wild-type strain) and ΔglnR under both 0 mM NH₄⁺ and 100 mM NH₄⁺ conditions. (A) The glycine cleavage system is encoded by gcvH and gcvT operons (gcvT, gcvPA, and gcvPB). (B) The GlnR-binding sites in the gcvH promoter. (C) EMSA analysis of GlnR binding to DAN site in the gcvH promoter. DNA fragment carrying a GlnR-binding site was synthesized and biotin labeled. The biotin-labeled DNA fragments were incubated with His-GlnR supplemented without or with FBI-GS (5 mM glutamine and 500 nM His-GS). Each lane contained 0.15 nM biotin-labeled DNA. Lane 1 contained no GlnR. Lanes 2–10 contained increasing concentrations of His-tagged GlnR (7.8, 15.6, 31.2, 62.5, 125, 250, 500, 1,000, and 1,500 nM). (D) Transcript levels of gcvH, gcvT, and gcvPB in WT strain and ΔglnR mutant under 0 mM and 100 mM NH₄⁺ conditions determined by qRT-PCR. The results are the mean from at least three independent experiments. Error bars indicate standard deviation. Significance was determined by student’s t-test. Not significant (ns P > 0.05), *P < 0.05, **P < 0.01.

The gcvH promoter contains a GlnR-binding site located 84 bp upstream of the gcvH translation initiation codon (ATG) (Fig. 7B). Here, EMSA showed that GlnR protein bound to GlnR-binding site, and FBI-GS (Glutamine +GS) enhanced this binding (Fig. 7C). qRT-PCR revealed that the gcvH in the wild-type strain was only highly transcribed under nitrogen limitation. However, gcvH in ΔglnR mutant was constitutively expressed under both nitrogen-limited and -excess conditions (Fig. 7D). The results indicate that GlnR activates the transcription of gcvH under nitrogen-limited condition and GlnR together with FBI-GS inhibits the gcvH expression under nitrogen-excess condition by binding to the same GlnR-binding site in the gcvH promoter. The findings that a DNA site is involved in both activation and repression are also observed in the regulation of the tnrA expression in B. subtilis (33). The GlnR/TnrA site 1 lies immediately upstream of the −35 region in the tnrA promoter of B. subtilis and TnrA bound at this site activates transcription by directly interacting with RNA polymerase. Also, GlnR bound to the same site represses tnrA transcription (33).

In this study, we also show that the transcript levels of gcvT and gcvPB in ΔglnR mutant were significantly increased under excess nitrogen compared to those in the wild-type strain, suggesting that GlnR represses the expression of gcvT and gcvPB under excess nitrogen (Fig. 7D). It is known that there is no GlnR-binding site in the promoter of the gcvT operon composed of gcvT, gcvPA, and gcvPB. Thus, the results indicate that GlnR indirectly represses the expression of the gcvT operon under excess nitrogen.

GlnR regulates the expression of other genes involved in nitrogen metabolism

Our recent studies have revealed that GlnR is a global regulator that regulates nitrogen metabolism genes/operons in P. polymyxa WLY78 (48). In addition to glnA1, glnRA, amtBglnK, and gcvH, the nitrogen metabolism genes (nasA, nasD1, nasD2, nasT, glnQHMP, and glnS) have GlnR-binding site(s) in their promoter regions (Fig. S4A; Table S1).

The transcriptions of nasA (nitrate transporter), nasT (a regulator of nasAB), nasD1 (nitrite reductase), nasD2 (nitrite reductase), glnQHMP (glutamine uptake), and glnS (glutamine synthetase) in wild-type and △glnR strains under both 0 mM NH₄Cl and 100 mM NH₄Cl were determined by qRT-PCR (Fig. S4B). The transcript levels of nasA, nasD1, and nasD2 in △glnR mutant were much higher than those in wild-type strain under both nitrogen-limited and -excess conditions (Fig. S4B). The results are consistent with the above results that deletion of both glnA and glnA1 leads to a defect of nitrate assimilation. The transcript level of the glnS in △glnR mutant was much decreased compared to that in the wild-type strain under nitrogen limitation, while the transcript level of the glnS in △glnR mutant was similar to that in wild-type strain under excess nitrogen (Fig. S3B), suggesting that GlnR activates the glnS expression under nitrogen limitation. Notably, the transcript levels of glnQ (glnQMHP operon) and nasT in △glnR mutant were much higher than those in wild-type strain under excess nitrogen, suggesting that GlnR represses the expression of glnQMHP and nasT under excess nitrogen. It was reported that TnrA activates the expression of glnQHMP and nasDEF in B. subtilis under limited nitrogen conditions (5, 28, 36). As shown in Table S1, there are two GlnR-binding sites in the intervening region between GM005626 and glnQ or between nasA and nasD1 or between nasT and nasD2, we deduce that the GlnR-binding site that is the closest to ATG of glnQ or nasA or nasD1 or nasD2 or nasT is the target DNA of GlnR for regulating transcription of these genes. Comparison of the P. polymyxa GlnR-binding site sequences with B. subtilis GlnR motif and B. subtilis TnrA motif has shown that a common consensus sequence (5′-TGTNAN7TNACA-3′) in GlnR-binding sites is conserved (Fig. S5).

In summary, this study has revealed that GlnR, a global nitrogen regulator, uses multiple mechanisms to regulate nitrogen metabolism in P. polymyxa WLY78. GlnR represses the glnRA expression under excess nitrogen and activates the glnA1 expression under nitrogen limitation by binding to a single GlnR-binding site in the promoter regions of these genes. GlnR activates and represses the expression of amtBglnK and gcvH by binding to one or two GlnR-binding sites in response to nitrogen availability. Also, GlnR regulates the expression of other genes involved in nitrogen metabolism (e.g., nasA, nasD1, nasD2, nasT, glnQMHP, and glnS). GS encoded by glnA within glnRA and GS1 encoded by glnA1 belong to GSI-α and GSI-β, respectively. Both GS and GS1 play roles in the synthesis of glutamine from NH₄⁺ and glutamate, but only GS is involved in the repression mediated by GlnR. Taken together, our study has revealed that GlnR as an activator or repressor or simultaneously as an activator and repressor regulates nitrogen metabolism in P. polymyxa WLY78. Our results are distinct from B. subtilis GlnR that only represses the expression of glnRA, ureABC, and tnrA under excess nitrogen. The general regulatory model of nitrogen fixation and nitrogen metabolism mediated GlnR in P. polymyxa WLY78 is shown in Fig. 8.

Fig 8.

GlnR-mediated regulation of nitrogen metabolism in P. polymyxa WLY78.

MATERIALS AND METHODS

Strains, plasmids, and growth conditions

Bacterial strains used in this study are listed in Table 1 and plasmids are shown in Table S2. P. polymyxa strains were grown in a nitrogen-limited medium containing 2 mM glutamate or nitrogen-excess medium containing 2 mM glutamate plus 100 mM NH₄⁺ (45). E. coli strains JM109 and BL21 (DE3) were used as routine cloning and protein expression hosts, respectively. Thermo-sensitive vector pRN5101 was used for gene disruption in P. polymyxa (45). pET-28b(+) (Novagen) was used for expressing recombinant His6-tagged protein in E. coli. When appropriate, antibiotics were added in the following concentrations: 100 µg/mL ampicillin, 25 µg/mL chloramphenicol, 12.5 µg/mL tetracycline, 50 µg/mL kanamycin, and 5 µg/mL erythromycin for maintenance of plasmids.

TABLE 1.

Strains used in this study

Strains	Characteristics	Sources
Paenibacillus polymyxa
WLY78	Wild-type strain	(43)
ΔglnA	glnA in-frame deletion mutant	(45)
ΔglnR	glnR in-frame deletion mutant	(45)
ΔglnRA	Double mutant of glnR and glnA	(45)
PamtBT1	A P. polymyxa WLY78 derivative with a site-specific mutation on GlnR-binding site I in the amtB-glnK promoter	This study
PamtBT2	A P. polymyxa WLY78 derivative with a site-specific mutation on GlnR-binding site II in the amtB-glnK promoter	This study
PamtBT12	A P. polymyxa WLY78 derivative with site-specific mutations on both GlnR-binding sites in the amtB-glnK promoter	This study
Escherichia coli
JM109	General cloning host; recA1, endA1, gyrA96, thi-1, hsdR17, supE44, relA1, Δ(lac-proAB)/F′ (traD36, proAB+, lacIq, lacZΔM15)	Sangon Biotech Co.
BL21 (DE3)	Host for protein overexpression; F–, ompT, gal, dcm, lon, hsdSB (rB-mB--), λ (DE3 [lacI, lacUV5-T7 gene 1, ind1, sam7, nin5])	Sangon Biotech Co.

Construction of amtB mutants containing the specific mutagenesis on GlnR-binding site(s)

Three amtB mutants (PamtB T1, PamtB T2, and PamtB T12) containing the specific mutagenesis on GlnR-binding site(s) in the amtB promoter were constructed (Table S3 and S4). The last six base pairs in GlnR-binding site I and GlnR-binding site II were, respectively, replaced with KpnI restriction site (GGTACC) and ClaI restriction site (ATCGAT) that were designed in primers used for amplification of the amtB promoter region. Each of the PamtB T1, PamtB T2 and PamtB T12 mutants was selected from the initial Em^r transformants after several rounds of nonselective growth at 39°C and identified by PCR amplification and DNA sequencing analysis (40).

RNA preparation and qRT-PCR analysis

Transcript levels of genes were determined by quantitative real-time RT-PCR (qRT-PCR) analysis as described by Wang et al. (45). Total RNAs were extracted with RNAiso Plus (Takara, Japan) according to the manufacturer’s protocol. Removal of genome DNA and synthesis of cDNA were performed using the PrimeScript RT reagent Kit with gDNA Eraser (Takara, Japan). 16S rRNA was set as an internal control. Each experiment was performed in triplicate. Primers used for qRT-PCR are listed in Table S5.

Transcription start site identification

The 5′-RACE method was used to determine the TSS by 5′ RACE using the SMARTer RACE cDNA Amplification Kit (Clontech). Gene-specific primers are listed in Table S6.

Electrophoretic mobility shift assay

EMSA was performed as described by Wang et al. (45) using a DIG Gel Shift Kit (2nd Generation; Roche, USA). Two DNA fragments corresponding to the sequences of the first strand and the complementary DNA strand of the promoter fragments containing the putative GlnR-binding site were synthesized by Sangon Biotech Co., Ltd (Shanghai, China) (Table S7). The two strands were annealed and then labeled with digoxigenin (DIG) using terminal transferase, and used as probes in EMSA. Reaction mixtures contained poly [d(A-T)], labeled probe, and the purified His6-GlnR from E. coli and then were analyzed by electrophoresis using native 5% polyacrylamide gel after a certain time of incubation. Labeled DNAs were detected by chemiluminescence according to the manufacturer’s instructions, and recorded on X-ray film.

SPR detection

SPR experiments were carried out using Biacore 3,000 SPR sensor (Biacore AB, Uppsala, Sweden) in Beijing of China, according to the methods described by Fernandes et al. (47) and Wang et al. (45). DNA oligomer used for SPR assays (Table S8) were designed and synthesized based on the promoter region harboring the putative GlnR-binding sites and containing a single-stranded overhang complementary to the linker. The purified His6-GlnR from E. coli was used in SPR.

GS activity

GS biosynthetic activity was assayed by a modification of the radiochemical method to measure phosphate release (46, 60). The bacterial cell culture was collected and broken with the ultrasonic instrument. 100 µL cell-free extract was added to the assay solution that contains the following materials: 100 mM Imidazole-HCl buffer (pH 7.0), 50.0 mM NH₄Cl, 50 mM Na-glutamate, 10.0 mM ATP, and 20 mM MnCl₂. Biosynthetic GS activity was quantified in terms of nm PO₄^3-/assay.

The intracellular concentrations of NH₄⁺, glutamate, and glutamine

The intracellular concentrations of NH₄⁺, glutamate and glutamine were determined as described by Li et al. (46).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00139-23.

Fig. S1 to S5 and Table S1 to S8. aem.00139-23-s0001.docx.

Figures and legends.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.