Nitrogenase regulation in Vibrio natriegens differs from other γ-proteobacteria

Abstract

Regulation of nitrogenase, which converts nitrogen gas (N₂) into ammonium (NH₄⁺), typically involves a conserved set of regulatory proteins across diverse N₂-fixing (diazotrophic) bacteria. However, the interactions and relative influence of these regulators can vary between species. Thus, one cannot make assumptions about nitrogenase regulation when working with uncharacterized diazotrophs like Vibrio natriegens, a γ-proteobacterium of growing interest for synthetic biology. Little is known about V. natriegens nitrogenase regulation, which could be used to exploit inexpensive N₂ for various applications, including NH₄⁺ production. Here, we characterized the roles of several annotated V. natriegens nitrogenase regulatory proteins in response to NH₄⁺ versus N₂. Using functional genomics, targeted mutations, and reporter assays, we identified a typical regulatory hierarchy where the two-component system NtrBC governs a nitrogen-scavenging regulon that includes NifA, the transcriptional activator of nitrogenase genes. Unlike other diazotrophic γ-proteobacteria, NifA was sufficient to activate nitrogenase gene expression, as a mutant lacking NtrBC grew normally with N₂ after a lag phase. Thus, NtrBC was dispensable, but still important for timely nitrogenase expression. Furthermore, NtrBC was negatively regulated by the nitrogen-responsive P_II proteins GlnB and GlnK; disruption of both P_II proteins led to NtrBC-dependent nitrogenase overactivity, marked by NH₄⁺ excretion. The redundant repression of NtrBC by GlnB and GlnK more closely resembles that of non-diazotrophic E. coli than other diazotrophic γ-proteobacteria. Together, our findings provide a framework for V. natriegens nitrogenase regulation that can be leveraged for applications like NH₄⁺ production.

Graphical Abstract

INTRODUCTION

Nitrogenase is a critical enzyme that sustains life by converting abundant nitrogen gas (N₂) into ammonium (NH₄⁺) in a process called N₂-fixation or diazotrophy. Only certain bacteria and archaea are diazotrophs, and their combined activity represents a major input of nitrogen into ecosystems [1]. Currently, the largest terrestrial nitrogen input comes from synthetic agricultural fertilizers produced by the industrial Haber-Bosch process, which is energy-intensive, polluting, and inaccessibly expensive for some developing regions [1–3]. Despite these undesirable aspects, Haber-Bosch fertilizers are important to sustain half the world’s population [3]. Thus, there are incentives to produce clean and affordable alternative fertilizers by harnessing diazotrophs [4, 5]. To do so requires an understanding of how diazotrophs regulate nitrogenase in response to environmental signals like NH₄⁺. Nitrogenase regulation is well-defined for several diazotrophs, but other diazotrophs of emerging biotechnological interest, like the marine γ-proteobacterium Vibrio natriegens ATCC 14048, have not been characterized.

V. natriegens has become popular in the last decade as a chassis for synthetic biology due to its rapid growth and metabolism [6–8] and its genetic tractability [9–12]. As a diazotroph, V. natriegens can also use N₂ as the sole nitrogen source under anoxic conditions [13]. Accessing inexpensive N₂ could provide cost-savings for industrial fermentations [14], including for NH₄⁺ production. However, since first being described in 1988, V. natriegens N₂ fixation has received little attention beyond measuring responses in nitrogenase structural nifHDK gene expression to fixed nitrogen (e.g., NH₄⁺, peptone, etc) and oxygen (O₂) [15] and verification of functional activity from a predicted nitrogenase gene cluster [16]. Thus, it is largely unknown how V. natriegens nitrogenase is regulated. Moreover, little is known about nitrogenase regulation for any Vibrio species beyond a recent investigation with V. diazotrophicus [17], and for marine heterotrophic diazotrophs in general, despite their abundance and importance in marine ecosystems [18, 19].

Regulation of nitrogenase typically involves multiple interacting regulatory modules, ensuring tight control over its energetically expensive production and operation; at least 20 genes are required to assemble nitrogenase, which then requires 16 ATP and 8 electrons per 2 NH₄⁺ produced (Eq 1) [20]. In one of the best characterized γ-proteobacterial diazotrophs, Klebsiella pneumoniae, the two-component system NtrBC forms the first module, governing a diverse regulon in response to nitrogen starvation that includes nitrogenase and other N₂-fixation genes (nif genes). When fixed nitrogen is available, the sensor kinase NtrB dephosphorylates the NtrC response regulator, preventing expression of nitrogen scavenging genes (Fig 1A, left) [20]. As fixed nitrogen becomes limiting, NtrB phosphorylates NtrC, which then serves as a transcriptional activator of nitrogen scavenging genes [20] (Fig 1A, right). Among these genes is nifA, which encodes the master transcriptional activator of nif genes. In K. pneumoniae, nifA is co-transcribed with nifL, encoding an anti-activator that represses NifA in a redox-responsive manner [20–23]. NifLA thus forms the 2nd regulatory module (Fig 1A). Finally, a third module of P_II proteins GlnB and GlnK modulates the activity for the other tiers in response to fixed nitrogen availability (Fig 1A). In K. pneumoniae, GlnB is constitutively expressed and modulates NtrB activity [24]. GlnK, expressed under the control of NtrBC, relieves NifL inhibition of NifA when fixed nitrogen is scarce [25–27]. P_II protein activity can also be modulated by the addition or removal of UMP groups by GlnD [20]. However, in K. pneumoniae, GlnK modification with UMP does not affect its activity on NifL [25, 28, 29] (Fig 1A). Like K. pneumoniae, V. natriegens ATCC 14048 does not appear to encode a fourth module of post-translational regulation involving DraTG nor does it encode either alternative nitrogenase [20] (Fig 1B, Table S1).

Fig 1. Nitrogenase regulation in K. pneumoniae and the V. natriegens nitrogenase gene clusters.

(A) Nitrogenase regulation in the γ-proteobacterium, K. pneumoniae, as a comparative reference for characterizing V. natriegens nitrogenase regulation, which has a similar inventory. There are three regulatory modules: (i) the two-component system NtrBC, which activates a regulon of nitrogen scavenging genes in response to low fixed nitrogen availability, (ii) NifLA, where the anti-activator NifL determines the availability of NifA to directly activate transcription of nitrogenase genes (nif genes), and (iii) P_II-proteins GlnB and GlnK, which sense fixed nitrogen availability and respond by influencing the NtrBC activity (via GlnB) or NifL-NifA interactions (via GlnK). P_II-activity is mediated by post-translation modification with uridine monophosphate (UMP) via GlnD (Fig 1A) [20]. NifL-NifA interactions are also influenced by O₂ sensing via NifL. (B) Organization of nif genes on V. natriegens ATCC 14048 chromosome 2.

$${\text{N}}_2+16\text{ATP}+8{\text{e}}^{-}+10{\text{H}}^{+}\to 2{\text{NH}}_4^{+}+16\text{ADP}+16\text{Pi}+{\text{H}}_2$$ Eq 1, Nitrogenase:

Whereas the above regulatory proteins are commonly found across diverse diazotrophs, they can interact in different ways and exert different levels of control. For example, three different purple phototrophic bacteria with similar inventories of nitrogenase regulatory proteins required mutations in different regulatory proteins to achieve constitutive nitrogenase activity [30–33]. Even for V. diazotrophicus, where nif expression was shown to be NtrC-dependent [34], there is little shared synteny in the nif gene clusters with V. natriegens, suggesting different evolutionary histories and possibly distinct mechanisms of nif regulation [16, 35]. Thus, the influence that each regulatory protein has over nitrogenase cannot be reliably inferred from a genome sequence; it must be interrogated directly.

Here, we provide the first genetic characterization of the V. natriegens nitrogenase regulatory hierarchy. We find that unlike other γ-proteobacterial diazotrophs, NtrBC is not required for N₂ fixation. NtrBC facilitates early nif gene expression in response to low nitrogen, but wild-type diazotrophic growth rates can still be achieved without NtrBC. Instead, NifA autoactivation is sufficient for a high diazotrophic growth rate. Also differing from other diazotrophic γ-proteobacteria, V. natriegens P_II proteins play redundant roles in preventing unchecked nitrogenase activity via NtrBC. Mutants lacking both P_II proteins exhibited high nitrogenase activity, manifesting in NH₄⁺ excretion. Our work provides a foundational framework of V. natriegens nitrogenase regulation and exposes engineering targets for NH₄⁺ production.

RESULTS

Activation of NifA is required for V. natriegens nitrogenase activity.

To characterize V. natriegens nitrogenase regulation, we first verified that V. natriegens fixes N₂ when grown fermentatively in an anoxic minimal medium with glucose. Indeed, the diazotrophic growth rate (nitrogen source: N₂) was 0.33 ± 0.05 h⁻¹, 42% of the growth rate with NH₄⁺ (0.79 ± 0.03 h⁻¹; mean ± SD; Fig 2A, B; note different x-axis scales). We then addressed the essentiality of NifA, the canonical master transcriptional regulator of nif genes, for diazotrophic growth [20]. When we deleted nifA, the resulting ΔnifA mutant grew with NH₄⁺ but not with N₂ (Fig 2A, B). Thus, NifA is required for V. natriegens diazotrophic growth.

Fig 2. NifA is essential for growth on N₂.

(A, B) Growth curves for V. natriegens strains grown in anoxic minimal media with glucose and either NH₄⁺ (A) or N₂ (B) as the sole nitrogen source. Parent, NWH003 (Δdns::Km^R); ΔnifA, NWH036 (Δdns::Km^R). Points, mean; shading, SD; n=3. (C) A P_nifLA-lacZ transcriptional reporter was used to compare nifLA expression in the parent (NWH010; Δdns::Sp^R-P_nifLA-lacZ) when grown with NH₄⁺ or N₂. Points, biological replicates; bars, mean; error bars, SD; n=3.

nifA is often co-transcribed with nifL. NifL conventionally binds NifA and represses transcription in a redox-dependent manner [23]. In some γ-proteobacteria like Azotobacter vinelandii, nifLA is constitutively expressed [36, 37] whereas in K. pneumoniae and Pseudomonas stutzeri, nifLA is transcribed only when fixed nitrogen is scarce [38, 39]. To determine if V. natriegens nifLA expression was constitutive or responsive to NH₄⁺ availability, we measured nifLA promoter (P_nifLA) activity using a LacZ transcriptional reporter (P_nifLA-lacZ) in WT V. natriegens grown with NH₄⁺ versus N₂. We only observed LacZ activity under N₂-fixing conditions (Fig 2C). Thus, V. natriegens nifLA expression is responsive to NH₄⁺ availability.

NifL typically modulates nitrogenase activity by inhibiting NifA through protein-protein interactions in response to fixed nitrogen or O₂ [40]. To determine the extent to which NifL controls nitrogenase activity in V. natriegens, we first compared growth between the parent and a ΔnifL mutant during anaerobic growth with NH₄⁺ versus N₂. Growth trends were superimposable with NH₄⁺, whereas with N₂ the ΔnifL mutant exhibited a similar growth rate but a lower final cell density compared to the parent (Fig 3A). Based on these growth trends, we predicted that (i) nif genes would not be expressed with NH₄⁺ in the ΔnifL mutant and (ii) that there might be excessive nitrogenase expression with N₂, creating an energetic burden. Indeed, P_nifLA-lacZ reporter activity was not observed in either the parent or the ΔnifL mutant when grown with NH₄⁺. With N₂, P_nifLA-lacZ activity was 2-fold higher in the ΔnifL mutant compared to the parent (Fig 3B). This increase could be due to NifA, no longer inhibited by NifL, increasing nifA expression (autoactivation), which can occur in other diazotrophs like K. pneumoniae, according to reporter and biochemical assays [41–43]; we directly test this possibility later in the study. Similarly, when we measured expression of the nitrogenase structural genes using a P_nifHDK-lacZ reporter under diazotrophic conditions, LacZ activity was 1.7-fold higher in the ΔnifL mutant compared to the parent (Fig 3C). This increased nif gene expression translated to increased nitrogenase activity, measured as H₂ production, an obligate coproduct of nitrogenase activity (Eq 1); we previously verified that V. natriegens does not have hydrogenases to make H₂ and it does not produce H₂ when supplied with NH₄⁺, conditions where nitrogenase is repressed [44]. H₂ production from the ΔnifL mutant was 1.4-fold higher than the parent (Fig 3D). These data indicate that NifL alone does not control NifA expression or activity in response to NH₄⁺, but it might have a dampening effect on nif transcription under N₂-fixing conditions, where we infer that some NifA is always bound and inhibited by NifL.

Fig 3. NifL is a negative regulator of nitrogenase expression and activity.

(A) Growth curves for V. natriegens strains grown in anoxic minimal media with glucose and either NH₄⁺ or N₂ as the sole nitrogen source. Parent, NWH003 (Δdns::Km^R); ΔnifL, NWH037 (Δdns::Km^R). Points, mean; shading, SD; n=3. (B, C) A lacZ transcriptional reporter was used to compare expression from P_nifLA (B) or P_nifHDK (C) under N₂-fixing conditions. (D) Nitrogenase activity was measured in stationary phase (13 h) as accumulated H₂ (Eq 1). (B-D) Points, biological replicates; bars, mean; error bars, SD; n=3. Statistical differences were determined using unpaired, two-tailed t tests; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Although herein we primarily focus on nitrogenase regulation in response to NH₄⁺ versus N₂, we also verified whether NifL prevents nif gene expression in response to O₂. To induce nitrogenase expression during aerobic growth, we used glutamate as the sole nitrogen source. Glutamate can ‘trick’ some diazotrophs into expressing nitrogenase even though glutamate is available as a nitrogen source [45]. V. natriegens grew aerobically with glutamate, albeit with a growth rate that was 15% of that with NH₄⁺ (Fig S1A). To test whether glutamate induced nif gene expression, we first measured P_nifLA-lacZ reporter activity. Indeed, glutamate led to nifLA expression in the parent (Fig S1B). The ΔnifL mutant had P_nifLA-lacZ activity that was 2.7-fold higher than the parent (Fig S1B), again suggesting possible NifA autoactivation. We then determined if NifL ultimately prevents nitrogenase expression during aerobic growth by measuring P_nifHDK-lacZ reporter activity. Indeed, despite the expression from P_nifLA, no P_nifHDK-lacZ activity was observed in the parent. The ΔnifL mutant P_nifHDK-lacZ activity was about as high as ΔnifL mutant levels under N₂-fixing conditions (Fig. S1C vs Fig 3C). Thus, NifL prevents nitrogenase expression in response to O₂.

NtrBC activates nifLA expression but is dispensable for N₂ fixation.

The lack of P_nifLA expression with NH₄⁺, regardless of the presence of NifL, indicated that there is another regulatory tier governing P_nifLA. In various diazotrophic bacteria, nifLA expression is induced by a two-component system such as FixLJK, RegSR (RegBA/PrrAB), or NtrBC, with the latter expected for γ-proteobacteria [20]. Indeed, using BLASTp [46] to search for FixLJK (Accession: P23222–1, P29286) and RegSR (Accession: BAC46170–69) from Bradyrhizobium diazoefficiens USDA 110, and NtrBC from K. pneumoniae (Accession: CDO16416–7), we only identified homologs to NtrBC in V. natriegens (Table S1).

Across diverse bacteria, NtrBC governs a large regulon of nitrogen scavenging genes. In diazotrophs like K. pneumoniae, P. stutzeri, and V. diazotrophicus, this regulon includes nif genes (Fig 1A) [17, 20, 41, 47, 48], and deletion of ntrBC can dampen nif gene expression by an order of magnitude [17, 47]. We hypothesized that NtrBC controls nifLA in V. natriegens in response to NH₄⁺ availability. To test this hypothesis, we deleted ntrBC and examined the growth of the resulting ΔntrBC mutant with NH₄⁺ versus N₂. The ΔntrBC mutant grew like the parent when provided NH₄⁺ (Fig 4A), but a ~16 h lag phase occurred under N₂-fixing conditions (Fig 4B), after which the ΔntrBC mutant exhibited similar growth trends as the parent (Fig 4B, C). We verified that the ΔntrBC mutant growth was still dependent on NifA, as a ΔntrBCΔnifA strain was incapable of growth with N₂ (Fig 4B).

Fig 4. NtrBC is not required for diazotrophic growth.

Growth curves for V. natriegens strains grown in anoxic minimal media with glucose and either NH₄⁺ (A) or N₂ (B). Parent, NWH003 (Δdns::Km^R); ΔntrBC, NWH014 (Δdns::Km^R); ΔntrBCΔnifA, NWH043 (Δdns::Km^R). (A, B) The same parent growth curves were used as in Fig 2. (C) Growth rates of the parent and ΔntrBC mutant from panel B. NS, non-significant difference as determined using an unpaired, two-tailed t test. (D) Once N₂-fixing cultures reached stationary phase, cells were subcultured into oxic media with NH₄⁺, grown to stationary phase, then subcultured once more into anoxic minimal media with N₂, where growth was then monitored as shown in panel E. (A, B, E) Points, mean; shading, SD; n=3.

To explain the lag phase and high diazotrophic growth rate of the ΔntrBC mutant, we considered two possibilities: (i) suppressor mutations or (ii) nif gene expression is eventually activated by another protein(s) without mutation. We first tested the prediction that suppressor mutants, if present, would be enriched during growth with N₂ and thus eliminate the ΔntrBC lag phase upon subculturing. We thus grew ΔntrBC cultures with N₂ and then subcultured into aerobic conditions with NH₄⁺ to produce progeny that were free of nitrogenase (Fig 4D); O₂ irreversibly damages nitrogenase and NH₄⁺ represses nif transcription [15, 20]. We then subcultured the aerobically grown cultures back to N₂-fixing conditions (Fig 4D), where a similar lag phase was observed (Fig 4E), suggesting that the eventual ΔntrBC mutant growth was not due to suppressor mutations. We thus suspected that another protein(s) compensated for the loss of ntrBC without mutation.

The NifA regulon is a subset of the NtrC regulon.

A possible candidate that could compensate for the absence of NtrBC is NifA, which can positively regulate its own expression in some diazotrophs, including K. pneumoniae [41–43, 49–51]. We thus sought to define the NtrBC and NifA regulons by using RNA-seq. To circumvent complications with the ΔntrBC mutant lag phase and the inability of a ΔnifA mutant to grow with N₂, we performed the analysis on non-growing cell suspensions, by transferring NH₄⁺-grown cultures to nitrogen-free media under argon (Ar) to induce nitrogenase expression (Fig 5A). We verified that nitrogenase was expressed within 2 h by measuring H₂ (Eq 1).

Fig 5. The NifA regulon is within the NtrBC regulon.

(A) Cultures were grown anaerobically with glucose and NH₄⁺, then were washed and resuspended in nitrogen-free media under argon to induce nitrogen starvation. (B, C) Differential gene expression analysis was performed on nitrogen-starved cell suspensions, comparing the ΔntrBC mutant (NWH014, Δdns::Km^R) (B) or the ΔnifA mutant (NWH036, Δdns::Km^R) (C) to the parent (NWH003, Δdns::Km^R). Blue and red dots respectively indicate lower (log₂ fold change ≤ −2) or higher (log₂ fold change ≥ 2) mutant transcript levels (P ≤ 0.05). (D) Growth curves testing the essentiality of nitrogenase electron-transfer components encoded by the rnf operon and nifF in an anoxic minimal medium with N₂ as the sole nitrogen source. Δrnf, NWH038 (Δdns::Km^R); ΔnifF, NWH039 (Δdns::Km^R). Points, mean; shading, SD; n=3. (E) Comparison of the number of genes regulated by NtrBC and NifA, with some examples listed. The inner circle shows the number of genes that showd upstream binding by His-tagged NtrC as determined by ChIP-seq. (F) Putative NtrC DNA binding sequence based on binding sites from ChIP-seq. (G) A P_nifLA-lacZ transcriptional reporter was used to measure nifLA expression in strains grown in a minimal anoxic medium with NH₄⁺ or as non-growing cell suspensions (-N) after 2 h without nitrogen. Points, biological replicates; bars, mean; error bars, SD; n=3. The statistical difference was determined using an unpaired, two-tailed t test; **, P < 0.01.

Comparing the ΔntrBC mutant and parent transcriptomes revealed 303 differentially expressed genes (DEGs) (Fig 5B, Table S2, File S1), 291 of which were downregulated in the ΔntrBC mutant. These DEGs encompassed genes required for nitrogenase activity, including nifA, and for accessing other nitrogen sources. When comparing the ΔnifA mutant to the parent, a smaller regulon of 45 DEGs was observed, all of which were downregulated in the mutant (Fig 5C, Table S3, File S2). The NifA regulon consisted of all nif genes, and other genes needed for nitrogenase activity, such as those for an ABC molybdate transporter, an rnf-like electron transport complex, and the NifF flavodoxin; we verified that rnf and nifF are essential for N₂ fixation using deletion mutants (Fig 5D). All but two of the NifA-dependent DEGs were within the NtrBC regulon (Fig 5E), indicating that NtrBC regulates N₂ fixation via NifA.

The NtrBC regulon defined by RNA-seq contained other regulatory proteins. Thus, we sought to understand which genes are regulated through direct DNA-binding by NtrC. To identify NtrC binding sites, we ectopically expressed His-tagged NtrC in a ΔntrBC mutant and performed chromatin immunoprecipitation combined with massively parallel sequencing (ChIP-seq) for cultures grown with N₂ versus NH₄⁺ (Fig S2). NtrC bound sites upstream of 69 genes /operons. All these genes were part of the NtrBC regulon defined by RNA-seq (Fig 5E, Table S2), and included glnA, encoding glutamine synthase; genes for the uptake and catabolism of alternative nitrogen sources like amino acids and purines; glnK, encoding a P_II protein; and amtB1/amtB2, encoding two NH₄⁺ uptake transporters. The binding sites were common between N₂-fixing conditions, where NtrC is presumably phosphorylated (Fig S2A, B), and conditions with NH₄⁺, where NtrC is presumably unphosphorylated (Fig S2C, D). This observation is consistent with that from Salmonella typhimurium where DNA-binding by NtrC is independent of its phosphorylation state [52]. A putative V. natriegens NtrC DNA-binding motif was also identified (Fig 5F) that is present at least once in all NtrC-enriched regions and is similar to that in other bacteria [53, 54].

Despite NifA being within the NtrBC regulon according to RNA-seq, no NtrC binding was detected upstream of nifLA by ChIP-seq, nor was a binding motif identified upstream of nifLA or any of the other genes in the NifA regulon. To help resolve these conflicting results, we verified that nifLA expression was NtrC-dependent by measuring P_nifLA-lacZ reporter activity. No P_nifLA-lacZ activity was observed from the parent or the ΔntrBC mutant during anoxic growth with NH₄⁺, as expected. However, in non-growing cell suspensions that had been incubated without any nitrogen for 2 h, reporter activity was observed in the parent strain, but not in the ΔntrBC mutant (Fig 5G). Thus, although ChIP-enrichment of P_nifLA was not observed, our LacZ reporter data indicates that NtrBC is an activator of nifLA expression, at least within 2 h after transitioning to nitrogen-free conditions. Nonetheless, the dispensability of NtrBC for diazotrophic growth and the lack of binding sites suggests that it is not the only P_nifLA activator.

NifA can activate the expression of its own gene.

Evidence from diazotrophs like K. pneumoniae [41–43] and higher P_nifLA-lacZ activity in ΔnifL mutants (Fig 3) led us to hypothesize that NifA can activate nifLA expression (autoactivation). To test this hypothesis, we first examined the impact of NtrBC and NifA on P_nifLA-lacZ reporter activity in non-growing cell suspensions after 2 h without nitrogen. As before, NtrBC was required for P_nifLA expression within this time frame (Fig 5G and 6A). For the ΔnifA mutant, LacZ activity was still observed, but at 29% of parent levels. No reporter activity was detected from the ΔntrBCΔnifA mutant (Fig 6A). These results indicate that both NtrBC and NifA contribute to nifLA expression, but that NtrBC is required for expression within the first 2 h of nitrogen deprivation.

Fig 6. NifA activates its own expression.

(A) A P_nifLA-lacZ transcriptional reporter was used to measure nifLA expression in strains as nitrogen-starved cell suspensions with or without NifA expression from inducible P_tet-nifA. Parent, NWH010 (Δdns::Sp^R-P_nifLA-lacZ); ΔntrBC, NWH015 (Δdns::Sp^R-P_nifLA-lacZ); ΔnifA, NWH090 (Δdns::Sp^R-P_nifLA-lacZ); ΔntrBCΔnifA, NWH126 (Δdns::Sp^R-P_nifLA-lacZ); ΔntrBCΔnifA, P_tet-nifA, NWH127 (Δdns::Sp^R-P_nifLA-lacZ-P_tet-nifA-P_lacI-tetR) (B) Growth curves for strains grown in anoxic minimal media with glucose and or N₂. Points, mean; shading, SD; n=3. (C, D) Accumulation of H₂ in the headspace (limit of detection ~10 nmol) (C) and of NH₄⁺ in the supernatant (limit of detection ~6 μM) (D) in stationary-phase (27 h) using separate cultures from those in panel B. (A-D) NifA expression was induced with anhydrotetracycline (aTc, +) or was not induced by adding DMSO (−). (B-D) Parent, NWH004 (Δdns::Sp^R); ΔntrBC, NWH014 (Δdns::Km^R); ΔnifA, NWH036 (Δdns::Km^R); ΔntrBCΔnifA strains: NWH043 (Δdns::Km^R) and NWH108 (Δdns::Sp^R-P_tet-nifA-P_lacI-tetR). (A, C, D) Points, biological replicates; bars, mean; error bars, SD; n=3. Different letters indicate statistical differences as determined using a one-way ANOVA, P < 0.05. ND, not determined due to no growth.

Based on the above results, we surmised that initial activation by NtrC at P_nifLA would increase NifA levels for subsequent autoactivation. If so, then expressing NifA from an inducible promoter would bypass the early need for NtrBC and activate P_nifLA-lacZ expression. We tested this prediction by ectopically expressing nifA from an inducible promoter (P_tet) in ΔntrBCΔnifA mutant cell suspensions that had been incubated without nitrogen for 2 h. Reporter activity was 1.5-times that of parent levels (Fig 6A). A small amount of P_nifLA-lacZ activity was also observed in uninduced cultures (Fig 6A), likely due to leaky P_tet-nifA expression. We conclude that NifA is sufficient to activate expression at P_nifLA, thus serving as an autoactivator.

We hypothesized that NifA autoactivation could explain the ΔntrBC mutant’s eventual diazotrophic growth (Fig 4B). For example, the ΔntrBC mutant would be deficient for nifLA expression, but leaky or stochastic nifLA expression over the lag phase could eventually result in enough NifA in some cells to create a positive feedback loop of autoactivation. The time needed for this subpopulation to reach detectable levels would also contribute to the lag phase. Thus, we predicted that ectopic NifA expression should lessen the ΔntrBC mutant lag phase under N₂-fixing conditions. Even without induction, the ΔntrBC mutant lag phase was reduced to < 10 h from the presence of P_tet::nifA in the chromosome (Fig 6B), likely due to leaky P_tet-nifA expression triggering further expression at the native nifLA operon (Fig 6A). In agreement with this notion, the removal of chromosomal nifA (ΔntrBCΔnifA mutant) required P_tet-nifA induction for growth with N₂ (Fig 6B). Induction of P_tet-nifA in the ΔntrBC mutant further eroded the lag period but also resulted in a lower growth rate and final cell density relative to the parent (Fig 6B). We speculated that the ectopic nifA expression led to abnormal nitrogenase activity. Indeed, when examining nitrogenase products (Eq 1), H₂ levels were 2-fold higher in strains where P_tet-nifA was induced compared to the parent (Fig 6C), and extracellular NH₄⁺ was also observed (Fig 6D). Together, our results support a model where NtrC acts as an initial activator of nifLA expression in low nitrogen conditions, and then once expressed, NifA can further activate nifLA expression.

PII proteins redundantly and negatively regulate nitrogenase.

The above observations indicate that NtrBC makes a non-essential, but important, contribution to V. natriegens diazotrophic growth by initiating or amplifying early nifLA expression. However, it remained unclear what factors repress nitrogenase in response to NH₄⁺. P_II proteins are often involved in sensing and transmitting fixed nitrogen availability. In K. pneumoniae, GlnB regulates nitrogenase expression by mediating the phosphorylation activity of NtrB, and GlnK alleviates NifL repression of NifA [20] (Fig 1A). We therefore investigated the contribution of V. natriegens P_II proteins to nitrogenase regulation.

We first tested whether V. natriegens P_II proteins affect growth with N₂ versus NH₄⁺ using single glnB or glnK deletion mutants. Both ΔglnB and ΔglnK mutants, grew like the parent, regardless of the nitrogen source (Fig 7A, B). Thus, single deletions did not cause a regulatory disruption, at least that was evident from growth trends. In contrast, a ΔPII mutant lacking both P_II proteins (ΔglnBΔglnK), grew poorly with each nitrogen source (Fig 7C, D). Growth trends varied between biological replicates, suggesting that there was selective pressure for suppressor mutations.

Fig 7. P_II proteins are redundant negative regulators of nitrogenase.

(A-D) Growth curves for strains grown in anoxic minimal media with glucose and either NH₄⁺ or N₂. (A, B) Parent, NWH003 (Δdns::Km^R); ΔglnB, NWH042 (Δdns::Km^R), ΔglnK, NWH047 (Δdns::Km^R). Points, mean; shading, SD; n=3. (C, D) Parent, NWH004 (Δdns::Sp^R); ΔPII mutant, NWH050 (ΔglnBΔglnK, Δdns::Sp^R). All replicate cultures are shown to reveal the diverse growth trends exhibited by the ΔPII mutant. (E, F) H₂ was measured as an indicator of nitrogenase activity at different times for those cultures in panel C grown with NH₄⁺ (E) or those in panel D grown with N₂ (F). (G) NH₄⁺ in supernatant samples taken from N₂-fixing cultures at different times. (E-G) Points, biological replicates; bars, mean; error bars, SD; n=3. Statistical differences were determined between the parent and the ΔPII mutant at each time point using an unpaired, two-tailed t test; *, P < 0.05; **, P < 0.01.

We hypothesized that the poor ΔPII mutant growth was due to overactive nitrogenase activity. Indeed, the ΔPII mutant produced H₂ during growth with NH₄⁺ (Fig 7E). H₂ per cell was highest early in the growth curve, again supporting the notion that suppressor mutants with less nitrogenase activity were enriched during the experiment. Under N₂-fixing conditions, ΔPII mutant H₂ production was significantly higher than that of the parent (Fig 7F) and NH₄⁺ was detected in the supernatant, accumulating up to 3.6 mM/OD₆₆₀ (Fig 7G). These results indicate that V. natriegens P_II proteins have functional redundancy as negative regulators of nitrogenase.

P_II proteins can potentially affect broader aspects of nitrogen metabolism, for example by influencing the NtrBC-regulon. Thus, it was unclear if the ΔPII mutant growth defect was solely due to excessive nitrogenase activity or if other aspects of nitrogen metabolism were involved. To distinguish between these possibilities, we deleted either nifA or ntrBC in the ΔPII mutant and examined growth and nitrogenase activity via H₂ production. When grown with NH₄⁺, the ΔPIIΔnifA mutant displayed a similar growth defect as the ΔPII mutant (Fig 8A) but it did not produce H₂ (Fig 8B). This growth defect with NH₄⁺ was eliminated in the ΔPIIΔntrBC mutant; growth and H₂ production resembled the parent (Fig 8A, B). Similarly, in N₂-fixing conditions, ΔPIIΔntrBC mutant growth and H₂ production resembled that of the ΔntrBC mutant rather than the ΔPII mutant, including a ~16 h lag phase (Fig 8C, D). Thus, the ΔPII mutant growth defect is likely due to NtrBC contributing both to excessive nitrogenase and other aspects of nitrogen metabolism, which could include burdensome overexpression of nitrogen scavenging genes or dysregulation of assimilatory pathways.

Fig 8. ΔPII growth defects are NtrBC-dependent.

Growth curves (A, C) and final time-point H₂ measurements (B, D) for strains grown in anoxic minimal media with glucose and either NH₄⁺ (A, B) or N₂ (C, D). In N₂ conditions, strains lacking nifA were omitted due to an inability to grow. Parent, NWH004 (Δdns::Sp^R); ΔPII, NWH050 (ΔglnBΔglnK, Δdns::Sp^R); ΔntrBC, NWH014 (Δdns::Km^R); ΔPIIΔntrBC, NWH111 (Δdns::Sp^R); ΔPIIΔnifA, NWH112 (Δdns::Sp^R). (A, C) Points, mean; shading, SD; n=3. (B, D) Points, biological replicates; bars, mean; error bars, SD; n=3. (D) Different letters indicate statistical differences as determined using a one-way ANOVA, P < 0.05.

DISCUSSION

V. natriegens NifA autoactivation in comparison to other diazotrophs.

We defined the basic architecture of nitrogenase regulation in V. natriegens. In some ways, the regulatory network resembles that of other diazotrophic γ-proteobacteria (Fig 1A and Fig 9). NtrBC facilitates the transition to N₂-fixing conditions by responding to low nitrogen by activating early nifLA expression (Fig 6A and Fig 9). Without NtrBC, nifA can still be expressed by NifA autoactivation, but growth is delayed by ~16 h (Fig 4B). Although we have not ruled out the possibility of an intermediate regulator participating between NtrBC and NifA, this would represent a highly unusual deviation from every nitrogenase regulatory scheme involving NtrBC. We thus find NifA autoactivation alone to be more likely. Additionally, P_II proteins were important negative regulators of nitrogenase expression that ensured that nitrogenase activity was responsive to NH₄⁺ availability (Fig 8, 9).

Fig 9. Emerging picture of nitrogenase regulation in V. natriegens.

The NtrBC, NifLA, and P_II regulatory modules connect in ways that differ from other γ-proteobacteria. NtrBC is important for early nif gene expression but otherwise NifA autoactivation is sufficient for normal diazotrophic growth. P_II proteins play redundant negative roles, preventing NtrBC from contributing to excessive nitrogenase activity and broader growth defects. Several questions remain regarding how P_II proteins are regulated, including the role of post-translational modification by GlnD, and how else they might interact with the NtrBC and NifLA modules.

Although our work supports several findings from other diazotrophic γ-proteobacteria, it also highlights the importance of testing growth phenotypes. Early work on nitrogenase regulation, on which the field still heavily relies, often did not involve growth assays. For example, in K. pneumoniae, researchers described NifA-autoactivation from LacZ-reporter assays [42, 43] and from weak in-vitro NtrC binding to P_nifLA relative to another NtrC-activated promoter, P_glnA [41]. However, without growth phenotypes, one cannot truly know the relative contributions of regulators to nitrogenase activity. For example, in an α-proteobacterium where NifA autoactivation was suspected [30], an ΔntrBC mutant had an order-of-magnitude lower diazotrophic growth rate than the parent [55]. We recommend that mutants for other model diazotrophs be revisited where growth phenotypes are lacking. For example, non-growing cell suspensions of a P. stutzeri ntrB mutant showed nitrogenase activity only after 20 h, similar to the lag time we observed in the V. natriegens ΔntrBC mutant, and perhaps indicative of NifA autoactivation [47]. However, it is unknown if this trend translated to growth or if full nitrogenase activity was eventually achieved [47].

V. natriegens nitrogenase regulation differs from other γ-proteobacteria.

Although the V. natriegens nitrogenase regulatory network resembles that in other γ-proteobacteria, the contributions of the regulatory proteins differed. One example is the relatively minor impact of V. natriegens ΔntrBC mutant aside from a diazotrophic lag phase (Fig 4B). Although diazotrophy is not yet well-characterized in Vibrios, a V. diazotrophicus ΔntrC mutant exhibited 10% of the wild-type nitrogenase activity [17], a similar decrease as the α-proteobacterial mutant above [55]. This low activity still supported growth, but unfortunately no quantitative growth metrics are available to make comparisons [17]. Regardless, the defects of these mutants contrast our results where a V. natriegens ΔntrBC mutant had a wild-type diaztotrophic growth rate (Fig 4C) and H₂ production (Fig 6C), which should likely require wild-type nitrogenase activity.

Another way that V. natriegens nitrogenase regulation differs from other diazotrophic γ-proteobacteria is through the roles of P_II proteins. The V. natriegens P_II protein inventory of GlnB and GlnK resembles that of K. pneumoniae; Pseudomonas and Azotobacter, other well-characterized diazotrophic γ-proteobacteria, only have GlnK [56]. In K. pneumoniae, GlnB and GlnK play functionally opposing roles in nitrogenase regulation, while acting at different levels. GlnB negatively regulates nitrogenase by inhibiting NtrBC phosphorylation in response to fixed nitrogen (Fig 1); with fixed-nitrogen, a K. pneumoniae glnB mutant showed high P_glnK-lacZ reporter activity, which is controlled by NtrC [28]. On the other hand, K. pneumoniae GlnK positively regulates nitrogenase by facilitating the dissociation of NifL from NifA (Fig 1) [26, 28]; for a K. pneumoniae glnK mutant, NifH-LacZ reporter activity was ~25% of wild-type [28], even lower for a NifK-LacZ reporter [27], and diazotrophic growth was severely impaired [27]. K. pneumoniae GlnB also cannot substitute for GlnK, at least at physiological expression levels [25].

The above K. pneumoniae observations contrast those observed herein with V. natriegens where either P_II protein could be deleted with little consequence for diazotrophic growth, but deletion of both P_II proteins resulted in overactive nitrogenase activity that was dependent on NtrBC (Fig 7, 8). Thus, differing from the K. pneumoniae model, our results suggest that (i) each V. natriegens P_II protein plays a redundant negative role in regulating NtrBC and (ii) neither P_II protein is needed for dissociation of NifL from NifA. In these ways, V. natriegens P_II functions resemble those in the non-diazotrophic γ-proteobacterium E. coli, where GlnB and GlnK can play redundant regulatory roles and only disruption of both P_II proteins led to a NtrBC-dependent growth defect [57]. We also note that our ΔPII mutant resembles the α-proteobacterium Rhodobacter capsulatus, where deletion of both P_II proteins led to constitutive nitrogenase activity [33]. However, in R. capsulatus P_II protein regulation intersects with post-translational regulation of nitrogenase via DraTG, which V. natriegens ATCC 14048 does not possess. These differences from the expected K. pneumoniae model for nitrogenase regulation (Fig 1A) highlight the importance of characterizing nitrogenase regulation in diazotrophs of emerging interest. For V. natriegens, there is still much to understand about how the P_II proteins intersect with the other regulatory modules, including the effects of P_II modification with UMP via GlnD (Fig 9).

Applied avenues for NH₄⁺ production.

Our work also identified P_II proteins as targets for engineering NH₄⁺-excreting strains of V. natriegens. The highest level of NH₄⁺ excretion was observed for the ΔP_II mutant, where extracellular NH₄⁺ was ~75-times higher than an ΔntrBC mutant overexpressing NifA (Fig 7G vs Fig 6D). However, we suspect that the high nitrogenase activity, combined with dysregulation of broader nitrogen metabolism (Fig 8), made the desirable phenotype unstable; the poor growth created selective pressure for less NH₄⁺ excretion. High NH₄⁺ excretion is known to be unstable in other engineered NH₄⁺-excreting bacteria [58–60]. Thus, further optimization of the trade-off between V. natriegens growth rate and NH₄⁺ excretion is required. We anticipate that this pursuit will benefit from an improved understanding of interactions between P_II proteins and NtrBC and NifLA, along with a wholistic consideration of how NtrBC affects broader aspects of V. natriegens nitrogen metabolism.

MATERIALS and METHODS

Growth and cell suspension conditions.

V. natriegens strains stored as 25% glycerol frozen stocks were struck for isolation on 1.5% agar plates of lysogeny broth (Miller) supplemented with 20 g NaCl L⁻¹ (LB3), and antibiotics when appropriate (μg mL⁻¹): 100 carbenicillin, 100 kanamycin (Km), 250 spectinomycin (Sp). Antibiotics and 100 μM isopropyl β-D-1-thiogalactopuranoide (IPTG) were also added to LB3 agar or broth as appropriate during strain construction. All experiments used minimal VMDC175 medium, which is M9-derived coculture medium [61] plus 100 mM MOPS (pH 7) and an additional 175 mM NaCl. Anoxic conditions were established by aliquoting 10 mL of media into 27 mL anaerobic tubes and bubbling with N₂ or Ar. Tubes were sealed with rubber stoppers and aluminum crimps, then autoclaved. The following components were added after autoclaving via syringe (final concentrations): 25 mM glucose, 10 mM NH₄Cl, 1 mM MgSO₄, and 0.1 mM CaCl₂. NH₄Cl was omitted for N₂-fixing conditions. When appropriate, 200 nM anhydrotetracycline hydrochloride (aTc) was added from a 20 μM stock solution in 100% dimethyl sulfoxide (DMSO) to induce expression from P_tet; control cultures received the same volume of DMSO only. Single colonies were used to inoculate starter cultures in VMDC175 with either 1 or 10 mM NH₄Cl, depending on whether the starter culture would be used to inoculate test conditions with N₂ or NH₄⁺, respectively. After starter cultures reached late-exponential phase (0.65–0.85 OD₆₆₀), they were used to inoculate test conditions at an initial cell density of ~0.01 OD₆₆₀. For nitrogen-starved cell suspensions, cultures were first grown to exponential-phase (0.3–0.5 OD₆₆₀) in VMDC175 with 1 mM NH₄Cl, then the entire culture was pelleted by centrifugation and washed once in an equal volume of VMDC175. Cell pellets were then resuspended in 0.3 mL of VMDC175 and the entire volume was transferred to 10 mL of anoxic VMDC175 under Ar with all media components except NH₄Cl. All cultures and cell suspensions were incubated at 30°C, laid horizontally with shaking at 150 rpm for anoxic conditions or upright at a 45° slant with shaking at 250 rpm for oxic conditions. Assays involving cell suspensions were performed after 2 h, which we verified is adequate to consistently detect nitrogenase activity as H₂. All shaking used a ¾” stroke length.

Strain construction.

Strains and primers are in Table S4 and Table S5, respectively. V. natriegens parent strains NWH003 or NWH004 were made by replacing dns (PN96_RS00885) in the type strain TND1964 with a kanamycin (Δdns::Km^R) or spectinomycin resistance cassette (Δdns::Sp^R), respectively. TND1964 is the “Dalia, SAD1302, 2016” variant wild-type V. natriegens ATCC 14048 with pMMB-tfoX, allowing for IPTG-inducible competency (https://portal.cultivarium.org/communities/vnat-sequencing?tab=data&dataset=Pilot%20study). Mutants were made as described [9] by natural transformation with both a deletion construct plus a selectable marker to replace the parent dns antibiotic cassette with the alternative cassette. Selectable markers were amplified from V. natriegens SAD1306 and were flanked by ~1 kb upstream and downstream of dns. Deletion constructs were made by PCR amplifying ~3 kb upstream and downstream of the region to be deleted using NEB Q5 DNA polymerase. The two regions were then connected to a 48 bp universal linker (MiniFRT; generated by combining oligos of each strand) by splicing-by-overlap extension (SOE) PCR using the outermost primers. The resulting SOE PCR product (1 μg) was co-transformed into the recipient strain along with 50 ng of either Δdns::Sp^R or Δdns::Km^R. In all other cases, 200 ng of PCR-amplified DNA was transformed. To construct transcriptional reporters, lacZ plus 12 bp upstream of the start codon to capture the ribosomal binding site (RBS) was first amplified from pHRP309 [62]. Using SOE PCR, lacZ was then combined with an upstream Km^R cassette and flanked by ~1 kb of homology surrounding dns. The product was then transformed into TND1964 generating NWH006 with a promoterless lacZ (Δdns::Km^R-lacZ). SOE PCR was then used to combine the Sp^R cassette upstream of a desired promoter, minus the native RBS, with 1 kb flanking DNA to match the region upstream of dns (upstream) and the lacZ insertion in NWH006 (downstream). The product was then transformed into NWH006 to generate a functional transcriptional reporter. To generate C-terminal tagged NtrC, ntrBC plus the native promoter and RBS (P_ntrBC) was first amplified, fused with an upstream Km^R cassette and 1 kb flanking regions by SOE PCR as above to replace dns after transformation into TND1964, creating NWH060. The construct was then amplified and transformed into a ΔntrBC mutant (NWH064). A PCR SOE product was also generated with a Sp^R cassette fused upstream of constitutive glnA promoter (P_glnA) to replace the Km^R-P_ntrBC in NWH064, but preserving the native ntrBC RBS, creating NWH061. After verifying that the construct complemented diazotrophic lag phase, 1 kb upstream and downstream of the ntrC 3’ end were amplified and connected by a flexible Gly-Ser-Gly-Ser linker to a Hisx6 tag (generated by combining 30 bp oligos of each strand) directly upstream of the NtrC stop codon by SOE PCR. The final construct was then introduced into the ΔntrBC mutant to create NWH071. To create an inducible nifA construct, a Sp^R cassette was combined upstream of an anhydrotetracycline-inducible tet promoter (P_tet) and RBS amplified from pAJM.011 [63] and fused to nifA and a downstream constitutive P_lacI-tetR repressor cassette (also from pAJM.011) by SOE PCR, creating a product with flanking regions to replace dns; pAJM.011 was a gift from C. Voigt (Addgene plasmid # 108529; http://n2t.net/addgene:108529 ; RRID:Addgene_108529). In all cases, transformants were isolated on LB3 agar with the appropriate antibiotic and verified by colony PCR and Sanger or Nanopore sequencing.

β-galactosidase (LacZ) assays.

Transcriptional LacZ reporter assays were performed in Z-buffer, consisting of 60 mM Na₂HPO₄, 40 mM NaH₂PO₄, 10 mM KCl, 1 mM MgSO₄ pH adjusted to 7 with 1 M HCl before filter sterilization. Immediately prior to use, 2-mercaptoethanol was added to a final concentration of 50 mM. Lysis solution was prepared by diluting 10x FastBreak^™ Cell Lysis Reagent (Promega) in Z-Buffer with 10 mg mL⁻¹ lysozyme. Samples from exponential-phase cultures or cell suspensions (both 0.3–0.5 OD₆₆₀) were lysed by combining 20 μL of culture with 180 μL lysis solution in a 96-well plate. Lysis occurred over a 30 min incubation in a H1 Synergy microplate reader (BioTek) at 37°C with double-orbital shaking. A stock solution of o-nitrophenyl-β-D-galactopyranoside (ONPG) was prepared by dissolving ONPG in 100% DMSO, then diluting in Z-buffer to 4 mg mL⁻¹. Following lysis, 20 μL ONPG solution was added to each well. LacZ activity was then tracked by absorbance (A₄₂₀) for 1 h at 37°C in the plate reader. LacZ activity (V₄₂₀ = A₄₂₀ ÷ time) was quantified using the linear regression function in GraphPad Prism v.6 and expressed as Modified Miller units = ( V₄₂₀ • 1000 • 1.56 ) / (OD₆₆₀ • 0.02 mL lysate volume) [64]. Values from appropriate control strains with a promoterless LacZ were subtracted from those from transcriptional reporters to control for ONPG instability.

RNA extraction and sequencing.

Cell suspensions were incubated under nitrogen-free conditions for 2 h at 30°C, then were chilled on ice and pelleted by centrifugation at 4°C. Supernatants were discarded and pellets frozen using dry ice before storing at −80°C. RNA was extracted with TRIzol^™ (Thermo Fisher) /chloroform as described [65]. Samples were treated with 4 U Turbo DNase (Invitrogen) in 100 μL at 37°C for 1 h, then purified using a QIAGEN RNeasy MinElute Cleanup Kit. RNA was quantified using an Agilent 2200 TapeStation at the Indiana University Center for Genomics and Bioinformatics (IU CGB), then submitted to SeqCoast Genomics where rRNA was depleted and libraries prepared using the Illumina Stranded Total RNA Prep Ligation Kit with Ribo-Zero Plus Microbiome and the RNA sequenced on an Illumina NextSeq2000 platform to produce 2×150 bp paired reads. Differential gene expression analysis was performed as described [66] using the V. natriegens ATCC 14048 genome (NCBI RefSeq: GCF_001456255.1). Differentially expressed genes with an adjusted Pvalue < 0.05 and a |log₂(fold-change)| > 2.0 were considered significant.

Chromatin immunoprecipitation-sequencing (ChIP-seq).

ChIP-seq was performed as described [67, 68] with some modifications. Five 10-mL anaerobic cultures were grown to exponential phase (0.3–0.5 OD₆₆₀). Samples (2 ml) were then extracted and pooled for whole-genome sequencing to be used as the ‘input’ for ChIP-seq normalization. These pooled samples were pelleted by centrifugation at 4°C, supernatants discarded, pellets frozen in dry ice, and then stored at −80°C. gDNA was extracted using the QIAGEN DNeasy Blood and Tissue kit. The remaining anaerobic cultures received an injection of 37% formaldehyde (1% final) and were crosslinked for 30 min at room temperature. Tubes were then unsealed, cultures pooled, and crosslinking quenched with 125 mM glycine. Cells were then pelleted by as above, resuspended in 1 mL ice-cold TBS buffer (20 mM TRIS, 0.9% NaCl, pH 7.4), transferred to a 1.5 ml tube and pelleted again by centrifugation at 4 °C. This wash step was repeated, then cell pellets were frozen using dry ice, and stored at −80°C. Chromatin was immunoprecipitated from cell pellets as described [67], except Protein G Mag Sepharose^™ Xtra magnetic beads (Cytiva) and anti-His antibodies (Genscript) were used to pull down His-tagged NtrC. The strain NHW061, with untagged NtrC was used as a control for nonspecific enrichment. Library preparation for genomic and ChIP DNA was performed using the NEBNext Ultrall kit (NEB) and sequenced at the IU CGB using an Illumina NextSeq2000, generating paired-end reads. Reads were mapped to the V. natriegens ATCC 14048 genome (GCA_001456255.1) using CLC Genomics Workbench (CLC Bio, QIAGEN). ChIP and input sequences were normalized based on the total number of reads, and ChIP enrichment (ChIP/input) was plotted using customized R scripts (https://github.com/xindanwanglab/Haas-2026a).

Analytical procedures.

Cell density was measured as turbidity at 660 nm (OD₆₆₀) using a Genesys 20 spectrophotometer (Thermo-Fisher) directly in culture tubes. Growth rates were calculated using GraphPad Prism v.6 by fitting an exponential trendline to turbidity measurements between 0.1–0.8 OD₆₆₀. H₂ was quantified by sampling 0.1 mL of headspace using a gas-tight syringe and injecting into a Shimadzu GC-2014 gas chromatograph with a thermal conductivity detector as described [69]. The detection limit was 10 nmol H₂. NH₄⁺ was quantified with an indophenol colorimetric assay [61] using late-stationary phase cultures (0.65–0.85 OD₆₆₀) grown in VMDC175 but omitting MOPS and using 15 mM glucose to avoid acidic conditions that interfere with the assay. Culture samples (1 ml) were pelleted, and 550 μL of supernatant was combined with 50 μL of 1 M NaOH, 100 μL phenol nitroprusside (Sigma-Aldrich) and 100 μL of alkaline hypochlorite (Sigma-Aldrich). For ΔPII mutant samples only, 50 μL of supernatant plus 500 μL of VMDC175 (no MOPS) was used. Samples were vortexed and incubated at room temperature for 15 minutes, then absorbance measured (A₆₃₀) in 1 mL cuvettes. Standard curves used NH₄Cl in VMDC175 (no MOPS). The limit of detection was ~6 μM NH₄⁺.

Statistics.

GraphPad Prism v.6 was used for all statistical analyses for growth, LacZ-reporter, H₂, and NH₄⁺ assays.

HIGHLIGHTS.

• A genetic examination of Vibrio natriegens nitrogenase regulation is performed

• NtrBC is important for early nitrogenase gene expression but is not essential

• NifA autoactivation is sufficient for nitrogenase expression

• P_II proteins GlnB and GlnK are redundant negative regulators of nitrogenase

• Genetic targets are identified that result in excretion of NH₄⁺

Data availability

Processed files and raw reads for RNA-seq and ChIP-seq have been deposited in NCBI’s Genome Expression Omnibus [70] with GEO series accession numbers GSE333724 and GSE333729, respectively. NCBI Sequence Read Archive and under BioProject accession number PRJNA1472683 (https://www.ncbi.nlm.nih.gov/bioproject).