The Caulobacter NtrB-NtrC two-component system bridges nitrogen assimilation and cell development

Abstract

A suite of molecular sensory systems enables Caulobacter to control growth, development, and reproduction in response to levels of essential elements. The bacterial enhancer binding protein (bEBP) NtrC, and its cognate sensor histidine kinase NtrB, are key regulators of nitrogen assimilation in many bacteria, but their roles in Caulobacter metabolism and development are not well defined. Notably, Caulobacter NtrC is an unconventional bEBP that lacks the σ⁵⁴-interacting loop commonly known as the GAFTGA motif. Here we show that deletion of C. crescentus ntrC slows cell growth in complex medium, and that ntrB and ntrC are essential when ammonium is the sole nitrogen source due to their requirement for glutamine synthetase (glnA) expression. Random transposition of a conserved IS3-family mobile genetic element frequently rescued the growth defect of ntrC mutant strains by restoring transcription of the glnBA operon, revealing a possible role for IS3 transposition in shaping the evolution of Caulobacter populations during nutrient limitation. We further identified dozens of direct NtrC binding sites on the C. crescentus chromosome, with a large fraction located near genes involved in polysaccharide biosynthesis. The majority of binding sites align with those of the essential nucleoid associated protein, GapR, or the cell cycle regulator, MucR1. NtrC is therefore predicted to directly impact the regulation of cell cycle and cell development. Indeed, loss of NtrC function led to elongated polar stalks and elevated synthesis of cell envelope polysaccharides. This study establishes regulatory connections between NtrC, nitrogen metabolism, polar morphogenesis, and envelope polysaccharide synthesis in Caulobacter.

Introduction

Nitrogen exists in a multitude of forms in the environment, and bacteria have a variety of molecular mechanisms to assimilate this essential element. Accordingly, bacterial cells commonly express sensory transduction proteins that detect environmental nitrogen and regulate the transcription of genes that function in nitrogen assimilation. The conserved NtrB-NtrC two-component system (TCS) is among the most highly studied of these regulatory systems. The NtrB-NtrC TCS has been broadly investigated, particularly in Enterobacteriaceae where it is well established that the NtrB sensor histidine kinase controls phosphorylation state of the DNA-binding response regulator, NtrC, in response to intracellular nitrogen and carbon status (1–4). Phospho-NtrC (NtrC~P) activates transcription of multiple genes involved in inorganic nitrogen assimilation and adjacent physiologic processes.

The preferred inorganic nitrogen source for many bacteria is ammonium (NH₄⁺) (5), and NtrC~P commonly activates transcription of glutamine synthetase (glnA) (6), which functions to directly assimilate NH₄⁺ in the process of glutamine synthesis. In the freshwater- and soil-dwelling bacterium, Caulobacter crescentus (hereafter Caulobacter) (7), glutamine levels per se are a key indicator of intracellular nitrogen status and impact cell differentiation and cell cycle progression via the nitrogen-related phosphotransferase (PTS^Ntr) system (8). The deletion of ntrC results in a nitrogen deprivation response in Caulobacter (8) and it is expected that this is due, at least in part, to reduced glnA transcription. However, NtrC belongs to a broadly conserved class of transcriptional regulators known as bacterial enhancer binding proteins (bEBPs) that can function as global regulators of gene expression (9), so NtrC is predicted to regulate expression of more than just glnA in Caulobacter. Indeed, ChIP-seq and transcriptomic studies in Escherichia coli demonstrated that NtrC binds dozens of sites on the chromosome (10, 11) and affects transcription of »2% of the genome (12). Given the importance of cellular nitrogen status as a cell cycle and developmental regulatory cue in Caulobacter, we sought to define the NtrC regulon and to assess the role of the NtrBNtrC TCS in the regulation of cell development and physiology.

Caulobacter NtrC is a standard (Group 1) bEBP and thus possesses, a) a receiver (REC) domain, b) an ATPase associated with cellular activity (AAA+) domain, and c) a DNA-binding/helix-turn-helix (HTH) domain (9) (Fig S1). Phosphorylation of a conserved aspartate residue in the REC domain regulates oligomerization of Group 1 bEBPs as transcription factors, which activate transcription from s⁵⁴-dependent promoters through a mechanism that requires ATP hydrolysis by the AAA+ domain (9). However, there are limited examples of NtrC proteins that lack a critical series of amino acids in the AAA+ loop 1 (L1) domain known as the GAFTGA motif, which is necessary for interaction with the s⁵⁴ N-terminal regulatory domain (13). Some bEBP proteins lacking the L1 GAFTGA motif are reported to regulate transcription from s⁷⁰-dependent promoters through a mechanism that does not require ATP hydrolysis (14–16). Species in the genus Caulobacter harbor a distinct eight residue deletion in AAA+ L1 of NtrC that encompasses the GAFTGA motif (Fig S1). This observation raised the question of whether Caulobacter NtrC has functional/regulatory properties that are distinct from orthologs of other genera.

We conducted a molecular genetic analysis of the Caulobacter NtrB-NtrC TCS. A main objective of this study was to determine the functional roles of ntrB and ntrC during growth in media containing inorganic and organic nitrogen sources. Using transcriptomic and ChIP-seq approaches, we defined the NtrC regulon, revealing its dual function as both an activator and a repressor. Our ChIP-seq analysis identified dozens of NtrC binding sites across the Caulobacter chromosome, many of which directly overlap with binding sites for the essential nucleoid-associated protein, GapR (17, 18), and the cell cycle regulator, MucR1 (19). Deletion of ntrC led to slow growth in complex medium and an inability to grow when NH₄⁺ was the sole nitrogen source, due to a lack of glnBA transcription. Random transposition of a conserved Caulobacter IS3-family mobile genetic element into the promoter of the glnBA operon was a frequent and facile route to rescue the growth defect of ntrC mutants; IS3 transposition effectively rescued glnBA transcription, enabling growth of the ΔntrC strain. Caulobacter is a prosthecate bacterium that elaborates a thin stalk structure from its envelope at one cell pole, and we further discovered that loss of ntrC resulted in hyper-elongated stalks and a hyper-mucoid phenotype. These phenotypes were complemented by either glutamine supplementation to the medium or by ectopic glnBA expression. Our study provides a genome-scale view of transcriptional regulation by a NtrC protein with distinct structural features and defines a regulatory link between NtrC and nitrogen assimilation, polar morphogenesis, and envelope polysaccharide synthesis in Caulobacter.

Results

The nitrogen assimilation defect of Caulobacter DntrC is not genetically complemented by Escherichia coli or Rhodobacter capsulatus ntrC.

Given the well-established role for the NtrB-NtrC TCS in inorganic nitrogen assimilation (20), we predicted that a Caulobacter mutant harboring an in-frame deletion of ntrC (DntrC) would exhibit growth defects in a defined medium with NH₄⁺ as the sole nitrogen source (M2 minimal salts with glucose; M2G). As expected, the DntrC mutant failed to grow in M2G and this growth defect was genetically complemented by restoring ntrC at an ectopic locus (Fig 1A). The sole predicted route of NH₄⁺ assimilation in Caulobacter is via glutamine synthetase (8), and we, therefore, predicted that replacement of NH₄⁺ with glutamine as the nitrogen source would restore growth of DntrC in M2G. As expected, replacement of NH₄⁺ with molar-equivalent levels of glutamine (9.3 mM final concentration) restored ΔntrC growth in M2G (Fig 1A). We conclude that ntrC is required for NH₄⁺ assimilation in a defined medium.

Figure 1. ntrC is required for growth in defined medium in which NH4+ is the sole nitrogen source.

(A) Terminal culture densities of WT, ΔntrC, and ΔntrC carrying a complementing copy (ntrC⁺) or empty vector control (EV). Culture growth was measured spectrophotometrically at 660 nm (OD₆₆₀) after 24 hours (h) of growth in M2G or M2G in which NH₄⁺ was replaced with molar-equivalent (9.3 mM final concentration) glutamine (gln). Data represent mean ± standard deviation of three replicates. (B) Terminal densities of WT and ΔntrC containing empty vector (EV) or expressing Caulobacter ntrC from its native promoter (ntrC_Cc⁺) or E. coli ntrC or R. capsulatus ntrC expressed from P_xyl (ntrC_Ec⁺⁺ or ntrC_Rc⁺⁺). Culture growth was measured spectrophotometrically at OD₆₆₀ after 24 h of growth in M2G supplemented with 0.15% xylose. Data represent mean ± standard deviation of three independent replicates.

The functional conservation of ntrC between phylogenetically proximal (21–23) and distal (24) species has been demonstrated by heterologous genetic complementation. Caulobacter NtrC shares 40% sequence identity with the highly studied Escherichia coli NtrC (Fig S1), but expression of E. coli ntrC from a xylose-inducible promoter did not restore growth of Caulobacter ΔntrC in M2G (Fig 1B) even though E. coli NtrC was stably produced in Caulobacter (Fig S2A). Inspection of NtrC primary sequences revealed that the AAA+ domain from Caulobacter species lacks the conserved GAFTGA motif (Fig S1), which is important for the promoter remodeling activity of the AAA+ domain and for coupling promoter conformation information to s⁵⁴-RNAP (13). Rhodobacter capsulatus, like Caulobacter, is in the class Alphaproteobacteria. NtrC from this species and others in the order Rhodobacterales also harbor a deletion of the L1 loop containing the GAFTGA motif (Fig S1); R. capsulatus NtrC is reported to activate gene expression through s⁷⁰ rather than s⁵⁴ (15). Expression of R. capsulatus ntrC from a xylose-inducible promoter also failed to restore growth of Caulobacter ΔntrC in M2G (Fig 1B), though the protein was stably produced (Fig S2A). The L1 deletion surrounding the GAFTGA motif in R. capsulatus NtrC differs – and is larger than – the deletion in Caulobacter NtrC (Fig S1). These results provide evidence that Caulobacter NtrC has distinct structural and functional features, which merit further investigation.

Mutation of ntrB and ntrC has disparate effects on growth in defined versus complex medium.

We demonstrated that ntrC is essential in M2G defined medium. Glutamine levels in peptone yeast extract (PYE) – a complex medium – are reported to be low (8), and we have confirmed a previous report by Ronneau and colleagues (8) that ΔntrC has a growth defect in PYE that is complemented by expression of ntrC from an ectopic locus (Fig 2A) or by addition of glutamine to the medium (Fig 2B). We predicted that deletion of the gene encoding NtrB, the sensor kinase that phosphorylates NtrC in vitro (25), would result in similar defects as deletion of ntrC. We created an in-frame deletion of ntrB (ΔntrB) and observed no effect on growth rate in complex medium relative to wild type (WT) (Fig 2A).

Figure 2. Mutants of the ntrB-ntrC system have disparate effects on growth in defined versus complex medium.

(A) Growth of WT, ΔntrC, and ΔntrB possessing empty vector (EV) or a genetic complementation vector (⁺) in which indicated genes were expressed from their native promoters (ectopically integrated at the xylX locus); growth was measured spectrophotometrically at 660 nm (OD₆₆₀) in PYE complex medium without and (B) with supplemented 9.3 mM glutamine (gln). (C) Growth curves of WT, ΔntrC, ΔntrB, ntrC^D56A, and ntrC^D56E in PYE and (D) PYE supplemented with 9.3 mM gln. (E) Growth curves of WT, ΔntrC, ntrC^ΔREC (residues deleted: 17–125), ntrC^ΔAAA (residues deleted: 159–363), ntrC^ΔHTH (residues deleted: 423–462) in PYE and (F) PYE supplemented with 9.3 mM gln. Plotted points for A-F represent average OD₆₆₀ ± standard deviation of three independent replicates. (G) Terminal OD₆₆₀ of WT, ΔntrC, ΔntrB, ntrC^D56A, ntrC^D56E, ntrC^ΔREC, ntrC^ΔAAA, and ntrC^ΔHTH after 24 hours (h) of growth in M2G defined medium and (H) M2G in which NH₄⁺ was replaced with molar-equivalent (9.3 mM) gln. Data represent mean ± standard deviation of three independent replicates.

Given this result, we explored the possibility that phosphorylation is not required for NtrC-dependent growth regulation in complex medium. To assess the functional role of NtrC phosphorylation, we mutated the conserved aspartyl phosphorylation site in the receiver domain of NtrC to either alanine (ntrC^D56A), which cannot be phosphorylated, or glutamic acid (ntrC^D56E), which functions as a “phosphomimetic” mutation in some cases (26). Like ΔntrB, the growth rates of ntrC^D56A and ntrC^D56E strains were indistinguishable from WT in complex medium, though ntrC^D56A cultures had reduced terminal density (Fig 2C) that was complemented by glutamine supplementation to the medium (Fig 2D). Both NtrC point mutants were stably produced in Caulobacter as determined by western blot (Fig S2B). In fact, steady-state levels of NtrC were elevated in ΔntrB and ntrC^D56A compared to WT and ntrC^D56E (Fig S2B), indicating that these proteins are either more stable, more highly expressed, or both. We further investigated growth of these mutants in M2G defined medium. The ΔntrB and ntrC^D56A strains failed to grow in M2G, while ntrC^D56E grew like WT (Fig 2G). Like ΔntrC, the growth defect of ΔntrB and ntrC^D56A in M2G was rescued by replacing NH₄⁺ with molar-equivalent glutamine (Fig 2H). We conclude that while NtrC phosphorylation does not greatly impact growth in complex medium, it is essential for growth when NH₄⁺ is the sole nitrogen source.

To extend our structure-function analysis of Caulobacter ntrC, we engineered mutant strains harboring ntrC alleles in which either the receiver domain (ntrC^DREC; residues 17–125), the s⁵⁴-activating/AAA ATPase domain (ntrC^DAAA; residues 159–363), or the helix-turn-helix DNA-binding domain (ntrC^DHTH; residues 423–462) were removed. Growth of all three mutants (ntrC^DREC, ntrC^DAAA and ntrC^DHTH) was slower than WT in PYE complex medium, though the growth defects of ntrC^DREC and ntrC^DAAA were more extreme than ntrC^DHTH and DntrC (Fig 2E). The growth defects of all domain mutants in PYE were complemented by glutamine supplementation to the medium (Fig 2F). Each of these domain truncation alleles was stably expressed in Caulobacter (Fig S2C). Again, steady-state levels of NtrC^DHTH, NtrC^DREC, and NtrC^DAAA were elevated, indicating that these mutant proteins are either more stable, more highly expressed, or both. All ntrC domain mutants failed to grow in M2G defined medium (Fig 2G). Replacement of NH₄⁺ with molar-equivalent glutamine in M2G fully rescued the culture yield (i.e. terminal density) of ntrC^ΔHTH, although, yields of ntrC^ΔREC and ntrC^ΔAAA were only partially rescued (Fig 2H). Altogether, these results provide evidence that each of the NtrC domains is required for proper NH₄⁺ assimilation, though the culture yield defects of NtrC^ΔREC and NtrC^ΔAAA are not solely linked to nitrogen availability.

Having shown that the AAA+ domain of NtrC is required for growth in defined medium, we next investigated the role of ATP binding and ATP hydrolysis by this domain in NH₄⁺ assimilation. To probe the impact of ATP binding on NtrC function, we mutated the conserved lysine (K178) in the Walker A motif of AAA+, which is necessary for ATP binding in bEBPs (27) (Fig S1). To evaluate ATP hydrolysis, we mutated the conserved aspartate residue (D235) within the Walker B motif of AAA+ (Fig S1). This residue is vital for ATP hydrolysis but not for ATP binding (27, 28). Strains solely expressing either the ntrC^K178A or ntrC^D235A alleles did not grow in M2G defined medium (Fig S3), though these mutant proteins were stably expressed in Caulobacter (Fig S2D). As observed in other null NtrC mutants, steady state levels of NtrC^K178A and NtrC^D235A were increased, suggesting that these proteins are either more stable, more highly expressed, or both. These results provide evidence that conserved residues in the NtrC AAA+ domain known to impact ATP binding and ATP hydrolysis are required for NH₄⁺ assimilation in defined medium.

IS3 rescue of glnBA transcription restores growth of ΔntrC.

During our investigation of ΔntrC, we noticed occasional instances of robust bacterial growth in M2G defined medium, indicating the possibility that spontaneous mutation(s) could bypass the growth defect of ΔntrC. Indeed, in four independent cases in different ntrC mutant backgrounds, we isolated suppressor mutants that exhibited growth in M2G (Table S1 and Fig 3A). Whole genome sequencing revealed that in three of these strains, an IS3-family (IS511/ISCc3) insertion element had integrated into the promoter of the glnBA operon. In the ΔntrC parent strain, an IS3 family insertion element inserted 8 bp upstream of glnB (ΔntrC P_glnBA::IS3) (Fig 3B); this insertion was accompanied by a large deletion of sequence in the adjacent operon (CCNA_02043-02045). We also identified two independent IS3-family (IS511/ISCc3) insertions upstream of glnBA (16 bp and 51 bp upstream of glnB) that rescued the growth defect of ntrC^ΔHTH mutants. In diverse bacteria, NtrC~P is known to activate transcription of glnA (6), which encodes glutamine synthetase. This enzyme directly assimilates NH₄⁺ by synthesizing glutamine from NH₄⁺ and glutamate. glnB encodes a conserved PII protein that regulates GlnA (6). We observed a fourth growth rescue mutation in ntrC^D56A, where a non-synonymous intragenic transversion resulting in a N94Y mutation rescued growth of the non-phosphorylatable NtrC^D56A mutant (Table S1).

Figure 3. Spontaneous transposition of an IS3-family insertion element restores glnBA expression in ΔntrC and rescues the ΔntrC growth defect.

(A) Terminal optical density (OD₆₆₀) of WT, ΔntrC, a spontaneous suppressor of ΔntrC (ΔntrC P_glnBA::IS3), and ΔntrC expressing glnBA from an inducible promoter (ΔntrC glnBA⁺⁺) grown for 24 hours (h) in M2G defined medium. (B) Site of the spontaneous lesion upstream of glnBA in the ΔntrC suppressor strain as determined by whole-genome sequencing. The insertion sequence (IS) element in inserted such that the 3’ end of the transposase (matching the 3’ end of the transposases CCNA_00660 and CCNA_02814) is positioned at nucleotide 2192508, which is 8 nucleotides upstream of the glnB start codon. In addition, a 3285 bp deletion eliminated most of CCNA_02043–45 operon. (C) RNAseq counts per million (CPM) of glnB (left) and glnA (right) transcripts in WT, ΔntrC, and ΔntrC P_glnBA::IS3 exponential phase cells grown in PYE complex medium. (D) Aligned RNAseq read counts (blue) corresponding to the 5’ end of the glnBA operon from WT, ΔntrC, and ΔntrC P_glnBA::IS3 cells. Annotated regions are diagramed below the x-axis for each strain.

To determine the transcriptional consequences of IS3 insertion at P_glnBA, we assessed global transcript levels in WT, ΔntrC, and the ΔntrC P_glnBA::IS3 suppressor strain (sup 1; Table S1). As expected, the ΔntrC strain had negligible glnBA transcripts compared to WT (Fig 3C and Table S2). However, glnB and glnA transcription was restored in ΔntrC P_glnBA::IS3 (Fig 3C and Table S2). Mapped reads demonstrated transcription originating from the IS3 element that extended into glnBA (Fig 3D). This provides evidence that sequences within the IS511-ISCc3 mobile element promote transcription of glnBA independent of NtrC, thereby enabling growth of the Caulobacter ΔntrC mutant in M2G. To test if glnBA transcription alone is sufficient to restore ΔntrC growth, we expressed glnBA from a xylose-inducible promoter (ΔntrC glnBA⁺⁺). We observed similar growth restoration in M2G in this strain (Fig 3A). These findings demonstrate that the inability of ΔntrC to grow when NH₄⁺ is the sole nitrogen source is from the lack of glnBA transcription, and that this transcriptional (and growth) defect can be rescued by insertion of mobile DNA elements into the glnBA promoter.

Considering that strains with mutations affecting NtrC phosphorylation (e.g., ΔntrB, ntrC^D56A) do not grow in M2G (Fig 2G), we examined the effect of ntrB and ntrC mutations on glnBA expression using a fluorescent P_glnBA transcriptional reporter (P_glnBA-mNeonGreen). Reporter activity was significantly reduced in ΔntrB, ntrC^D56A, and ntrC^D56E when cultivated in complex medium, although ntrC^D56E had higher P_glnBA-mNeonGreen transcription than ntrC^D56A (Fig S4). These results provide evidence that an intact phosphorylation site in the NtrC receiver domain (D56) is important for the activation of glnBA transcription by NtrC. The lack of P_glnBA activity in ΔntrB supports a model in which NtrB functions as the NtrC kinase in vivo.

Defining the Caulobacter NtrC regulon.

NtrC belongs to a class of proteins known as bEBPs, which often function as global regulators of transcription in bacteria. We sought to comprehensively define the NtrC regulon in Caulobacter. To this end, we used RNA sequencing (RNA-seq) and chromatin immunoprecipitation sequencing (ChIP-seq) approaches. Deletion of ntrC significantly changed transcript levels for nearly one-quarter of genes in the Caulobacter genome relative to WT (RNA-seq; P < 10⁻⁴) when strains were cultivated in PYE complex medium (Fig 4A and Table S2). To distinguish genes directly regulated by NtrC from indirectly regulated genes, we performed ChIP-seq using a 3xFLAG-tagged NtrC fusion. This experiment identified 51 significantly enriched peaks (Fig 4D and Table S3), which represent direct NtrC binding sites. From these peaks, we identified a common DNA sequence motif (Fig 5A) that is significantly related to the multifunctional DNA-binding protein Fis of E. coli, and with the NtrC motif of E. coli, though there are features that clearly distinguish the Caulobacter NtrC motif from E. coli NtrC (Fig S5).

Figure 4. NtrC globally regulates gene expression in Caulobacter.

In PYE complex medium, 473 genes exhibit differential transcript abundance in the ΔntrC mutant compared to WT based on the following criteria: fold change > 1.5, FDR P < 0.000001 and maximum group mean RPKM > 10. (A-C) Log₂ fold change in abundance for these 473 transcripts for the following comparisons: (A) ΔntrC vs. WT cultures grown in PYE, (B) ΔntrC vs. WT cultures grown in PYE supplemented with 9.3 mM glutamine (gln). Differentially regulated genes in the ΔntrC mutant are largely restored to WT-like levels upon supplementation with glutamine; exceptions are highlighted in colored diamonds. (C) ΔntrC P_glnBA::IS3 vs. WT cultures grown in PYE, where each symbol represents a gene. The x-axis represents WT transcript abundance in PYE (counts per million; CPM) for each gene. (D) NtrC ChIP-seq peaks (q-value < 0.05, area under the curve (AUC) > 20) across the Caulobacter genome plotted as log₂ fold enrichment in read counts compared to the input control. Peaks highlighted in color are in the promoter of genes highlighted in (A-C), or, in the case of cdzCDI, overlapping the coding region. Colors correspond to the following genes: pink, cdzCDI; green, glnK-CCNA_01399; blue, CCNA_01813-ntrB; orange, CCNA_02044-45; red, glnBA; yellow, CCNA_02727.

Figure 5. NtrC binding sites are often co-located with the binding sites of select chromosome structuring proteins and cell cycle regulators.

(A) DNA motif enriched in NtrC ChIP-seq peaks, as identified by MEME (79). (B) Distribution of the relative position of NtrC ChIP-seq summits to the nearest GapR, MucR1, CtrA, or SciP summit as calculated by ChIPpeakAnno (78). NtrC summits >1,000 bp away from the nearest cell cycle regulator summits were excluded from the plots. Frequency distributions were plotted as histograms with 50 bp bins.

As expected, the data indicate that NtrC directly activates glnBA: a major NtrC peak was identified in the glnBA promoter region (Fig 4D and Table S3). NtrC also directly binds the promoter region of the glnK-CCNA_01399 operon (Fig 4A and Table S3). glnK encodes a PII protein homologous to GlnB, which has been shown to similarly regulate GlnA in bacteria (6, 29), while CCNA_01399 is an annotated as an AmtB-family NH₄⁺ transporter. Transcript levels of glnK and CCNA_01399 are decreased 8–15 fold (P < 10⁻⁶¹), respectively, in ΔntrC relative to WT (Fig 4A and Table S2); we conclude that NtrC directly activates transcription of these genes. We further observed an NtrC peak in the promoter regions of two genes in the nitrate assimilation locus, which is transcriptionally activated by nitrate (30) and functions to reduce nitrate to ammonium. Specifically, NtrC peaks are present at the 5’ end of the nitrate response regulator NasT and in the promoter region of the MFS superfamily nitrate/nitrite transporter NarK (Table S3). RNA-seq measurements were conducted in the absence of nitrate so, as expected, we did not observe differential transcription of this locus. Transcription of genes residing in the same operon as ntrC, including ntrB and a predicted tRNA-dihydrouridine synthase (CCNA_01813), are increased in ΔntrC by approximately 20-fold (Fig 4A and Table S2). NtrC directly binds the promoter of its operon (Fig 4D and Table S3), providing evidence that it functions as an autorepressor. This is consistent with our western blot data showing that ntrB-ntrC loss-of-function mutants (e.g., ΔntrB, ntrC^D56A, ntrC^ΔREC, ntrC^ΔAAA, ntrC^ΔHTH, ntrC^K178A, ntrC^D235A) have increased levels of NtrC protein (Fig S2B–D), indicating loss of autorepression at this genetic locus.

We have further identified genes in our datasets that are not known to be directly involved in nitrogen assimilation. In fact, 9 of the 51 NtrC binding sites are located within a mobile genetic element (MGE) (CCNA_00460–00482) that is known to spontaneously excise from the Caulobacter genome at low frequency (31). This MGE is responsible for biosynthesis of a capsular polysaccharide (31) that is differentially regulated across the cell cycle and confers resistance to the caulophage fCr30 (32). Select genes within this locus have enhanced transcription in ΔntrC (P < 10⁻⁵), including those encoding GDP-L-fucose synthase (CCNA_00471), GDP-mannose 4,6 dehydratase (CCNA_00472), and a P4 family DNA integrase (CCNA_00480) (Table S2). Two NtrC-binding sites also flank a second capsule biosynthesis and regulatory locus (CCNA_00161-CCNA_00167) outside of the MGE (Table S3), and deletion of ntrC results in significantly enhanced expression of several genes within this locus, including the capsule restriction factor, hvyA (32) (3-fold; Table S2). In all cases, NtrC binding sites within the MGE directly overlap reported binding sites of the nucleoid associated protein, GapR (17, 18) and either overlap or are adjacent (within 200 bp) with binding sites for the cell cycle regulators MucR1/2 (19) (Fig 5B and Table S3). Thirty-seven of the 51 total NtrC binding sites that we have identified directly overlap with one of the 599 reported GapR binding sites across the Caulobacter genome (18) (Fig 5B and Table S3). gapR itself is significantly downregulated by 2-fold in the ΔntrC mutant (Table S2). These results suggest that NtrC has a chromosome structuring role in addition to its direct role in transcriptional regulation of nitrogen assimilation genes.

The promoter region of the cell cycle regulator, sciP (CCNA_00948) (33, 34), contains an NtrC binding site (Table S3), and the transcription of sciP and adjacent flagellar genes, flgE and flgD, is significantly increased in ΔntrC (Table S2), indicating that NtrC represses transcription from this site. NtrC also directly binds the promoter of mucR1 (CCNA_00982) (Table S3); this regulator, along with SciP, has been implicated in controlling the S^®G1 cell cycle transition (19). Like sciP, deletion of ntrC results in enhanced expression of mucR1 (Table S2), and we conclude that NtrC also represses transcription from this site. We assessed overlap of NtrC binding sites with SciP binding sites across the genome, but observed no significant overlap (Fig 5B). We note occasional overlap between NtrC and binding sites for the essential cell cycle regulator, CtrA (Fig 5B), including sites within the promoters of sciP and hvyA (Table S3). An additional cell cycle gene that is regulated by NtrC is hdaA, which is reported to inactivate DnaA after replication initiation (35). NtrC binds the chromosome upstream of hdaA (Table S3), and deletion of ntrC results in significantly diminished transcription of hdaA (Table S2). Conversely, the region upstream of the DNA replication inhibitor toxin socB (36) (within the socA gene) is bound by NtrC (Table S3), and deletion of ntrC results in significantly enhanced transcription of socB (2-fold) without corresponding induction of the socA antitoxin (Table S2). Together, these results provide support for a model in which NtrC can function to modulate expression of key cell cycle/replication regulators in Caulobacter.

Transcripts corresponding to the contact-dependent inhibition by glycine zipper proteins (cdzCDI) system (37) are highly elevated in ΔntrC relative to WT (15-22-fold) (Fig 4A and Table S2), although the nearest NtrC ChIP-seq peak resides downstream of the promoter of this operon, within cdzI, itself (Fig 4D and Table S3). It is unclear whether expression of these genes is directly impacted by NtrC, but this NtrC binding site overlaps with a reported GapR binding site (18). CCNA_02727, encoding an uncharacterized PhoH family protein (38, 39), provides yet another example of gene with overlapping NtrC and GapR binding sites (18) in its promoter that exhibits strongly increased transcription in ΔntrC relative to WT (10-fold) (Fig 4A&D and Fig 7A–B).

Figure 7. Transcriptional regulation and functional impact of the phoH-family gene, CCNA_02727.

(A) Transcript levels of CCNA_02727 measured by RNAseq in different genetic backgrounds and conditions: WT and ΔntrC strains grown in PYE complex medium or PYE supplemented with 9.3 mM glutamine (gln), and the ΔntrC P_glnBA::IS3 strain grown in PYE. Data represent mean ± standard deviation of three replicate samples. (B) NtrC chromatin immunoprecipitation sequencing (ChIP-seq) revealed a binding peak upstream of an operon containing the small hypothetical gene, CCNA_03973, and CCNA_02727. Data represent log₂ fold enrichment sequence reads in the NtrC immunoprecipitation samples compared to total input sample. Positions of annotated genes are represented by gray bars above the plot. The genomic positions in the reference genome (Genbank accession CP001340) are indicated. (C) Summary of stalk length data, comparing different strains: WT and ΔntrC strains containing an empty vector (EV), a genetic complementation vector (ΔntrC ntrC⁺), ΔntrC ΔCCNA_02727, and CCNA_02727 overexpressed in WT from a xylose-inducible promoter (CCNA_02727⁺⁺). The data represent the median and interquartile range; a minimum length of 0.6 μM was used for stalk segmentation. Statistical significance was assessed using the Kruskal-Wallis test followed by Dunn’s test, comparing each condition to WT EV (*** P < 0.0001). WT EV: n=330; ΔntrC EV: n=1,481; ΔntrC ntrC⁺: n=366; CCNA_02727⁺⁺ n=238; ΔntrC ΔCCNA_02727: n=1,261.

Glutamine and glnBA activation rescue the ΔntrC transcriptional defect.

Glutamine supplementation rescued the growth defect of ΔntrC in PYE complex medium (Fig 2B), which raised the question of whether glutamine supplementation would also restore the global transcriptional defect of ΔntrC in PYE. Indeed, glutamine supplementation broadly restored transcription of genes dysregulated in the ΔntrC mutant to WT levels (Fig 4B, Table S2, and Fig S6). However, genes directly regulated by NtrC that are involved in nitrogen assimilation remained significantly dysregulated when glutamine was added to the medium (Fig 4B, Table S2, and Fig S6). For example, glnB and glnA transcript levels remained 15- and 8-fold lower in ΔntrC than in WT in the presence of 9.3 mM glutamine, while glnK and CCNA_01399 remained 6- and 4-fold lower, respectively (Fig 4B and Table S2). Transcripts from the ntrC locus, which is autorepressed, also remained significantly elevated in ΔntrC, as did genes of the cdz locus (Fig 4B and Table S2).

We further analyzed transcription in the suppressor mutant, ΔntrC P_glnBA::IS3, which permitted us to assess the transcriptome in a strain that lacks ntrC but that expresses glnBA (Fig 3C). Restoration of glnBA expression in this background restored transcription to WT levels for a subset of the loci that were dysregulated in ΔntrC, though transcription of many dysregulated genes was only partially rescued or remained unchanged (Fig 4C, Table S2, and Fig S6). Again, NtrC-regulated genes directly involved in nitrogen assimilation (e.g., glnK-CCNA_01399) remained significantly dysregulated in this strain (Fig 4C and Table S2). Furthermore, while gapR transcription is significantly reduced in ΔntrC, its transcription is significantly increased (3-fold) above WT in ΔntrC P_glnBA::IS3 to a level that is congruent to WT cultivated in the presence of 9.3 mM glutamine (Table S2). This same effect is observed for the iron-dependent Fur regulon (40) (e.g., CCNA_00027, CCNA_00028) (Table S2). Thus, for a subset of genes, IS3 insertion at P_glnBA results in a transcriptional effect that mimics media supplementation with 9.3 mM glutamine.

Loss of the ntrB-ntrC system results in stalk elongation.

Caulobacter has a dimorphic life cycle wherein each cell division produces two morphologically and developmentally distinct cells including 1) a flagellated, motile swarmer cell and 2) a sessile stalked cell. The Caulobacter stalk is a thin extension of the cell envelope and its length is known to be impacted by phosphate limitation (41) and sugar-phosphate metabolism imbalances (42). We observed that ΔntrC mutant cells develop elongated stalks when cultivated in PYE complex medium (Fig 6A–B). ΔntrB and ntrC^D56A mutants displayed an intermediate stalk elongation phenotype, while stalks of ntrC^D56E mutants did not differ from WT (Fig 6B). We conclude that loss of ntrC function results in development of elongated stalks in complex medium.

Figure 6. Deletion of the ntrB-ntrC two-component system results in development of hyper-elongated stalks.

(A) Representative phase-contrast images showing the stalk elongation phenotype of a ΔntrC strain compared to WT; strains were cultivated in PYE complex medium (top). The elongated stalk phenotype is chemically complemented by the addition of 9.3 mM glutamine (gln) to the medium (bottom). Scale bar (white; top left) equals 5 μm. Example stalks in the ΔntrC panel are marked with black arrows. (B) Summary of stalk length measurements for WT, ΔntrC, ntrC^D56A, ΔntrB, ntrC^D56A, ΔntrC P_glnBA::IS3, and ΔntrC glnBA⁺⁺ cultivated without (−/black) and with (+/gray) gln. Data represent median ± interquartile range. Minimum length for stalk segmentation was 0.6 μm. Statistical significance assessed by Kruskal-Wallis test followed by Dunn’s post-test comparison to WT (*** P < 0.0001). WT: n=314(− gln) n=207(+ gln); ΔntrC: n=1020(−) n=338(+); ΔntrB: n=440(−) n=75(+); ntrC^D56A: n=849(−) n=204(+); ntrC^D56E: n=339(−) n=177(+); ΔntrC P_glnBA::IS3: n=218(−); ΔntrC glnBA⁺⁺: n=503(−).

Our transcriptomic data showed no evidence of a phosphate limitation response upon ntrC deletion, nor did we observe changes in manA or spoT/rsh expression (Table S2), which have been implicated in stalk elongation (42). We did observe that expression of the phoH-like gene, CCNA_02727, was elevated 10-fold in ΔntrC compared to WT (Fig 7A). This gene has an NtrC peak in its promoter (Fig 7B), suggesting it is directly repressed by NtrC. PhoH proteins have been implicated in phosphate starvation responses in other bacteria (43, 44), so we tested whether de-repression of this gene in ΔntrC impacted stalk development, which is known to be stimulated by phosphate starvation in Caulobacter (41). Overexpression of CCNA_02727 from a xylose-inducible promoter in WT resulted in significantly longer stalks compared to WT (Fig 7C). However, deletion of CCNA_02727 in the ΔntrC strain did not ablate stalk elongation (Fig 7C). We conclude that elevated expression of CCNA_02727 is sufficient to promote stalk elongation, but that enhanced expression of CCNA_02727 in ΔntrC does not solely explain the long stalk phenotype.

Supplementation of PYE with 9.3 mM glutamine fully complemented the stalk length phenotype of ΔntrC (Fig 6A–B) and restoration of glnBA expression, either in the suppressor (ΔntrC P_glnBA::IS3) or in the glnBA overexpression strain (ΔntrC glnBA⁺⁺), restored ΔntrC stalk length to WT levels (Fig 6B). Similarly, glutamine supplementation complemented stalk length defects of ΔntrB and ntrC^D56A (Fig 6B). Altogether, these results indicate that stalk elongation phenotype of ntrB-ntrC mutants results from an explicit lack of usable nitrogen or a nutrient imbalance due to the reduced availability of usable nitrogen. We note that expression of CCNA_02727 in ΔntrC is restored to WT levels when PYE is supplemented with glutamine (Fig 7A). This result indicates that regulation of CCNA_02727 by NtrC is not via a simple, direct repressive mechanism.

ΔntrC forms mucoid colonies.

The Caulobacter swarmer and stalked cell types differ not only in cellular morphology, but also in their capsulation state (32). The swarmer cell is non-capsulated, while the stalked cell elaborates an exopolysaccharide (EPS) capsule composed of a repeating tetrasaccharide (45). Capsulation results in enhanced buoyancy which is apparent during centrifugation (32). When centrifuged, ΔntrC cells cultivated in PYE displayed a “soft” or “fluffy” pellet compared to WT (Fig 8A), which suggested that ΔntrC had altered EPS. Over-production of EPS results in colonies that appear mucoid (i.e., glossy) on solid medium containing abundant sugar (31, 32), and ΔntrC displayed a mucoid phenotype on PYE supplemented with 3% sucrose, a condition that has been shown to enhance Caulobacter mucoidy (31) (Fig 8B). We conclude that loss of ntrC impacts the production or composition of envelope polysaccharides.

Figure 8. The hyper-mucoid phenotype of ΔntrC in PYE complex medium is suppressed by either glutamine supplementation or glnBA expression.

(A) Cell pellets of WT and ΔntrC carrying an empty vector (EV) or vectors expressing ntrC (ntrC⁺) or glnBA (glnBA⁺⁺). Strains were grown overnight in PYE complex medium or PYE supplemented with 9.3 mM glutamine (gln). Overnight cultures were normalized to OD₆₆₀ = 0.5 and cells from 10 ml were centrifuged at 7,197 × g for 3 min at 4°C, and pellets were photographed. (B) Growth of WT EV, ΔntrC EV, ΔntrC ntrC⁺, and ΔntrC glnBA⁺⁺ on PYE agar supplemented with 3% sucrose (PYE + sucrose) or PYE agar supplemented with 3% sucrose and 9.3 mM glutamine (PYE + sucrose + gln). Plates were incubated for 4 days at 30°C and photographed.

We again tested whether glutamine supplementation could restore a phenotype of ΔntrC to that of WT. Centrifugation of ΔntrC cultures grown in PYE supplemented with 9.3 mM glutamine resulted in a compact pellet like WT (Fig 8A). Furthermore, ΔntrC cultivated on PYE supplemented with sucrose and 9.3 mM glutamine had a WT appearance (Fig 8B). Ectopic expression of glnBA in ΔntrC grown in PYE similarly complemented ΔntrC mucoidy (Fig 8A–B). These results support a model in which genetic or chemical restoration of intracellular glutamine levels complements the mucoid phenotype of ΔntrC.

The ΔntrC mucoidy phenotype requires the MGE.

The mucoid appearance of ΔntrC aligns with transcriptomic and ChIP-seq data that show ntrC-dependent repression of EPS synthesis genes, including those located within the Caulobacter mobile genetic element (MGE) (e.g., CCNA_00471, CCNA_00472) (Table S2). Considering the numerous NtrC binding sites within the MGE and its role in capsular polysaccharide biosynthesis (31), we tested whether the mucoid phenotype of ΔntrC required the MGE. Specifically, we deleted ntrC from a Caulobacter crescentus NA1000 strain that had spontaneously lost the MGE (31), resulting in a Caulobacter ΔMGE ΔntrC mutant. When cultivated in PYE, the NA1000 ΔMGE ΔntrC strain did not display a “fluffy” pellet or exhibit a mucoid phenotype on solid medium (Fig S7A). We further deleted ntrC in C. crescentus CB15 strain (CB15 ΔntrC), which similarly lacks the MGE (31). CB15 ΔntrC had WT phenotype in pellet and plate growth assays (Fig S7B). These results provide evidence that the mucoid phenotype of ΔntrC is dependent on the EPS synthesis genes of the MGE. We conclude that transcriptional dysregulation due to loss of NtrC impacts cell envelope polysaccharide production via the MGE and, perhaps, through other genes involved in EPS biosynthesis (Table S3).

Discussion

ntrB-ntrC differentially impacts growth in defined and complex medium

Environmental nitrogen is an important cell cycle and developmental regulatory cue in Caulobacter (8), which motivated us to explore the function of the NtrB-NtrC TCS, a broadly conserved regulator of nitrogen metabolism (6). We characterized the population-level growth phenotypes of ntrB and ntrC mutants under media conditions containing distinct nitrogen sources and demonstrated that the sensor kinase gene, ntrB, and the AAA+-type response regulator gene, ntrC, are essential for growth in a defined medium in which NH₄⁺ is the sole nitrogen source (Fig 1 and Fig 2). Strains expressing a ntrC allele harboring a mutated aspartyl phosphorylation site in its receiver domain (ntrC^D56A) also failed to grow in this defined medium (Fig 2G). These data support an expected model in which phosphorylation of NtrC by NtrB is necessary for NH₄⁺ assimilation. An additional histidine kinase/response regulator pair, ntrY-ntrX, is part of the ntrBC genetic locus in Caulobacter and is postulated to have arisen from gene duplication (25). The inability of ΔntrB to grow in M2G indicates that NtrB (and not NtrY) is the major histidine kinase for NtrC in vivo. Each of the three NtrC domains − 1) Receiver, 2) AAA+ ATPase, and 3) HTH DNA binding domain – are required for growth in ammonium-defined medium (M2G) (Fig 2G). Replacement of NH₄⁺ with glutamine in M2G fully rescued growth of ntrC^ΔHTH but not ntrC^ΔREC and ntrC^ΔAAA mutants (Fig 2H). This suggests that some component of the growth defect of ntrC^ΔREC and ntrC^ΔAAA in defined medium is independent of cellular nitrogen status. Considering the overlap of NtrC binding sites with GapR and MucR1, variants of NtrC without the AAA+ or REC domains might exhibit dominant-negative effects. These truncated alleles could interfere with interactions involving GapR and MucR1, thereby disrupting gene expression at multiple chromosomal loci.

A strain lacking ntrC is viable in PYE complex medium but has a reduced growth rate, a phenotype that is complemented by addition of glutamine (Fig 2A–B) (8). Surprisingly, ΔntrB and ntrC^D56A had no growth rate defect, but did exhibit a growth yield defect in PYE (i.e., final culture density) (Fig 2C) that was rescued by the addition of glutamine to the medium (Fig 2D). From these results, we conclude that NtrC~P is less important in complex medium during log phase growth and becomes more important at higher cell density when organic nitrogen becomes more limited and waste products accumulate. NtrC domain truncation mutants, ntrC^DREC, ntrC^DAAA, and ntrC^DHTH, grew slower in PYE (Fig 2E), though the ntrC^DREC and ntrC^DAAA strains had more severe defects than ntrC^DHTH, which phenocopied ΔntrC. As discussed above, the ntrC^DREC and ntrC^DAAA yield phenotypes in PYE may be due to dominant-negative effects of expressing these truncated NtrC polypeptides in vivo, though glutamine supplementation to PYE did complement the defects of all NtrC domain truncation mutants in PYE (Fig 2F).

IS3 transposition repeatedly rescued the growth defect of ntrC mutants

The role of NtrC in activating glutamine synthase expression and facilitating ammonium assimilation is well-established in various species (6). We demonstrated that Caulobacter ntrC is essential in ammonium-defined medium (M2G) and made the surprising observation that cultures of Caulobacter ΔntrC occasionally showed robust growth in M2G; this suggested there was a route for spontaneous genetic rescue of the ΔntrC growth defect. We discovered that these “jackpot”-like cultures (46) were a consequence of random insertion of an IS3-family mobile genetic element at the glnBA promoter (P_glnBA) of ΔntrC that restored glnBA transcription (Fig 3). IS3 elements are present in multiple copies in the Caulobacter NA1000 genome (31), and the IS3-dependent transcriptional rescue phenotype we observe is consistent with a report that IS3 insertion elements can function as mobile promoters (47). We also identified two independent IS3-family (IS511/ISCc3) insertions upstream of glnBA at nucleotide 2192500 (16 bp upstream of the glnB start codon) and nucleotide 2192465 (51 bp upstream of the glnB start codon) that rescued growth of ntrC^ΔHTH mutants in M2G defined medium (Table S1), indicating that this is a facile evolutionary route to rescue loss of ntrC function under particular conditions.

Caulobacter insertion elements were previously shown to be transcriptionally activated in mutants that accumulate the alarmone (p)ppGpp (48), and Ronneau et al (8) have reported that glutamine limitation results in (p)ppGpp accumulation via activation of the PTS^Ntr system in Caulobacter. Furthermore, (p)ppGpp accumulates in Caulobacter starved for NH₄⁺ in defined medium (48). We postulate that in the absence of ntrC, decreased levels of intracellular glutamine result in (p)ppGpp accumulation and IS3 activation; an NtrC binding peak within an IS3-family element (adjacent to CCNA_02830) could contribute to IS3 regulation (Table S3).

The NtrC regulon in Caulobacter: more than just nitrogen metabolism

NtrC binds to multiple sites on the Caulobacter chromosome, playing a role in both activating and repressing gene expression. As expected, NtrC directly activates transcription of nitrogen assimilation genes such as glnBA, glnK, and the putative NH₄⁺ transporter CCNA_01399. Conversely, NtrC represses its own operon demonstrating autoregulation, which is well-established for this class of regulators (49). Our study also identified genes not directly involved in nitrogen assimilation in the NtrC regulon. Nine of the 51 NtrC binding sites are located within a mobile genetic element responsible for biosynthesis of a capsular polysaccharide that is differentially regulated across the cell cycle and confers resistance to a caulophage (32). The impact of ntrC on envelope polysaccharide is discussed below.

Thirty-seven of 51 NtrC binding sites (>70%) directly overlap with one of the 599 reported GapR binding sites (18) across the Caulobacter genome (Fig 5B and Table S3). GapR is a nucleoid-associated protein that binds positively supercoiled DNA and supports DNA replication (17), suggesting a possible connection between NtrC and chromosome organization/maintenance in Caulobacter. In addition, we observed significant overlap in binding sites of NtrC and the cell cycle regulator, MucR1 (19). Beyond mucR1, NtrC directly bound upstream and modulated transcription of other genes that impact cell cycle processes, including sciP, hdaA, and socB (33–36). NtrC appears to repress transcription of sciP and mucR1, which have been implicated in controlling the cell cycle transition from S to G1 upon compartmentalization of the nascent swarmer cell and also represses transcription of socB, a DNA replication inhibitor toxin. The exact mechanism of repression at these promoters remains undefined. These findings suggest that NtrC directly impacts regulation of the cell cycle in Caulobacter.

NtrC also regulates the cdzCDI operon that encodes a bacteriocin cell killing system activated in stationary phase (37). Loss of ntrC results in increased expression of the Cdz system; this transcriptional phenotype is not fully complemented by glutamine supplementation to the medium (Fig 4B and Table S2). Thus, repression of this locus by NtrC is not solely determined by nitrogen availability.

ntrC is a stalk elongation factor

Caulobacter cell division results in the production of a swarmer cell and a stalked cell. The swarmer cell differentiates into a reproductive stalked cell by shedding its polar flagellum, producing an adhesive holdfast at the same cell pole, and forming and a stalk that extends from that pole. Stalk length is regulated, and phosphate limitation was previously believed to be the only factor that determined Caulobacter stalk length (41). However, recent studies have demonstrated that metabolic imbalances in sugar-phosphate metabolism influence stalk elongation (42).

We have demonstrated that stalk elongation is genetically linked to the ntrB-ntrC TCS. The deletion of ntrC, ntrB, or replacement of ntrC with a non-phosphorylatable allele (D56A) resulted in hyper-elongated stalks in PYE (Fig 6B). Supplementation of PYE with glutamine or ectopic glnBA expression restored stalk lengths of ntrB and ntrC mutants to WT. We conclude that the stalk lengthening phenotype of ntrB and ntrC mutants is a consequence of decreased intracellular glutamine and that stalk elongation is linked to loss of ntrB-ntrC and possibly nitrogen limitation. Notably, limiting NH₄⁺ in defined growth medium does not result in increased stalk length in Caulobacter (50, 51). Furthermore, excess NH₄⁺ in combination with high pH restricts stalk elongation even when phosphorus is limited (52). These findings indicate that, while a connection between nitrogen availability and stalk length exists, not all nitrogen limitation conditions impact stalk development. Links between nitrogen availability, phosphorus availability, starvation signals such as (p)ppGpp (42), and stalk length are clearly complex and require further research.

Stalk elongation was previously postulated to enhance diffusive surface area, allowing for increased uptake of nutrients (53, 54), but subsequent work indicated that this is unlikely due to diffusion barriers within the stalk (51, 55). A recent model is that stalk lengthening allows Caulobacter in surface-attached communities to reach beyond its neighbors to better access available nutrients, thereby outcompeting other attached microbes and assisting in releasing progeny into the environment (52, 55). We predict that when nitrogen becomes limiting in surface-attached communities, the NtrB-NtrC system can cue the cell, perhaps through intracellular glutamine, to lengthen its stalk to better access nitrogen or other nutrients.

NtrC strongly represses transcription of CCNA_02727, a gene encoding a PhoH-family protein, and overexpression of CCNA_02727 in WT cells results in increased stalk length (Fig 7). However, deletion of CCNA_02727 in a ΔntrC background did not affect the stalk length of ΔntrC. PhoH family proteins typically possess ATPase and ribonuclease activity (38, 39, 44) and are often activated by the Pho regulon under phosphate starvation conditions in bacteria (43, 44). CCNA_02727 is not regulated by the Pho regulon in Caulobacter (56) but is strongly upregulated under other environmental conditions, such as carbon limitation (57) and heavy metal stress (57) in addition to glutamine deprivation via loss of ntrC as described here (Fig 7A). Crosstalk between different sensing systems to balance nutrient levels is well described in bacteria (58) and, therefore, it is possible that regulation of CCNA_02727 has a general role in controlling nutrient balance or stress response in Caulobacter.

ntrC regulates envelope polysaccharide production

Caulobacter ΔntrC displays a hyper-capsulation phenotype (Fig 8). NtrC orthologs are reported to regulate biofilm formation and EPS production in other bacteria, including P. aeruginosa, V. vulnificus, and B. cenocepacia, where loss of the ntrB-ntrC TCS decreases biofilm and EPS production (59–61). In V. cholerae, loss of ntrC increases biofilm formation and increases expression of EPS gene regulators (21).

Transcriptomic and ChIP-seq data presented in this study identified an NtrC peak in the promoter of hvyA, a gene encoding a transglutaminase homolog that prevents capsulation of swarmer cells (32). Although deletion of hvyA increases Caulobacter capsulation, its transcription is increased in ΔntrC by 3-fold. The link between hvyA expression and the ΔntrC capsule/mucoid phenotype, if any, remains undefined. We further observed an NtrC peak in the promoter region of the operon containing the CCNA_00471 (fcI; GDP-L-fucose synthase) and CCNA_00472 (GDP-mannose 4,6 dehydratase) genes (Table S3), which reside in the MGE of the Caulobacter NA1000 genome. The transcription of these two genes increased 2-fold and 3-fold in ΔntrC relative to WT, respectively. These enzymes function in the two-step synthesis of fucose, which is one of the sugars comprising the tetrasaccharide capsule of Caulobacter. It is reported that loss of these genes leads to a significant reduction in EPS production (62). The upregulation of CCNA_00471–00472 in ΔntrC may contribute to an increase in EPS production and, consequently, the hyper-mucoid and buoyancy phenotypes of ΔntrC. This is supported by the observation that Caulobacter ΔntrC strains lacking the MGE (i.e., NA1000 ΔMGE ΔntrC; CB15 ΔntrC) are not mucoid (Fig S7). However, EPS production is a complex process that involves multiple pathways and other genetic and physiological factors could also contribute to the envelope polysaccharide phenotype of ΔntrC. Indeed, EPS production is apparently linked to changes in intracellular glutamine levels independent of NtrC, given that adding glutamine to the medium represses EPS gene expression in ΔntrC (Table S2). The effect of glutamine on EPS gene transcription is congruent with our observation that either addition of glutamine to PYE or the ectopic expression of glnBA complements the mucoid phenotype of ΔntrC (Fig 8).

An unconventional NtrC

Caulobacter NtrC is lacks a GAFTGA motif within its primary structure (Fig S1), which is necessary for interaction with s⁵⁴ (13). Consistent with previous reports of NtrC orthologs lacking a GAFTGA motif (14, 15), our data indicate that NtrC regulates s⁷⁰-dependent promoters. For example, NtrC-repressed genes such as hvyA and sciP are activated by CtrA, a s⁷⁰-dependent transcriptional regulator (32–34). The NtrC binding peak summits within P_hvyA and P_sciP reside 4 bp and 55 bp from the CtrA peak summit at these promoters, respectively (Table S3), indicating that NtrC may directly compete with CtrA at these sites to repress transcription. We also identified NtrC-activated genes that possess s⁷⁰ promoters such as hdaA, which is also activated by DnaA (35), a s⁷⁰-dependent regulator (63, 64). The mechanism by which Caulobacter NtrC functions at s⁷⁰ promoters remains unclear.

Mutation of the conserved NtrC aspartyl phosphorylation site (D56) results in reduced transcriptional activation of the glnBA locus (Fig S4), highlighting the important role of this residue in NtrC-mediated transcriptional activation (at glnBA). Similarly, in R. capsulatus NtrC, which also lacks GAFTGA, aspartyl phosphorylation is required for transcriptional activation (14, 15). V. cholerae VspR, a bEBP that lacks GAFTGA and regulates s⁷⁰ promoters, does not require phosphorylation but utilizes the conserved aspartyl phosphorylation site for phosphate sensing (16). Whether D56 phosphorylation differentially affects NtrC function at binding sites across the Caulobacter chromosome is not known. In R. capsulatus NtrC, ATP binding rather than hydrolysis by the AAA+ domain is essential for transcriptional activity (15), while VspR does not require ATP to function (65). We have shown that conserved residues of the Walker A and Walker B motifs in the Caulobacter NtrC AAA+ domain are required for NH₄⁺ utilization in defined medium (Fig S3), providing evidence that ATP binding and ATP hydrolysis by NtrC are necessary for controlling the gene expression program that underlies NH₄⁺ assimilation. More generally, we predict that ATP binding and hydrolysis by Caulobacter NtrC contribute to the regulation of σ⁷⁰-dependent promoters, distinguishing it from other unconventional σ⁷⁰-regulating bEBPs. The Caulobacter genome (31) encodes four bEBPs: NtrC, NtrX, FlbD, and TacA. Unlike NtrC and NtrX, FlbD and TacA possess the GAFTGA motif. Notably, TacA regulates stalk biogenesis by controlling expression of s⁵⁴-dependent genes, including staR (66). Our study establishes a genetic link between the ntrB-ntrC TCS and the Caulobacter stalk. Thus, development of the polar stalk structure is controlled by at least two distinct bEBPs, NtrC and TacA, which are regulated by different environmental stimuli and have distinct primary structural and regulatory properties.

Materials and Methods

Growth conditions

E. coli strains were cultivated in Lysogeny Broth (LB) [10 g tryptone, 5 g yeast extract, 10 g NaCl per L] or LB solidified with 1.5% (w/v) agar at 37°C. LB was supplemented with appropriate antibiotics when necessary. Antibiotic concentrations for selection of E. coli were as follows: kanamycin 50 μg/ml, chloramphenicol 20 μg/ml, carbenicillin 100 μg/ml. Caulobacter strains were cultivated in peptone yeast extract (PYE) [2 g/L peptone, 1 g/L yeast extract, 1 mM MgSO₄, 0.5 mM CaCl₂] complex medium or PYE solidified with 1.5% (w/v) agar at 30°C or 37°C. Antibiotic concentrations for selection of Caulobacter were as follows: kanamycin 25 μg/ml (in solid medium), 5 μg/ml (in liquid medium), chloramphenicol 1.5 μg/ml. Nalidixic acid (20 μg/ml) was added to counterselect E. coli after conjugations. For glutamine supplementation experiments in PYE, 9.3 mM (final concentration) glutamine was added. For experiments in defined medium, Caulobacter strains were grown in M2 minimal salts medium with glucose (M2G) [6.1 mM Na₂HPO₄, 3.9 mM KH₂PO₄, 9.3 mM NH₄Cl, 0.25 mM CaCl₂, 0.5 mM MgSO₄, 10 uM ferrous sulfate chelated with EDTA (Sigma), and 0.15% glucose]. For glutamine supplementation experiments in M2G, NH₄⁺ was replaced with molar-equivalent (9.3 mM final concentration) glutamine.

Strains and plasmids

Strains, plasmids, and primers used in this study are presented in Table S4. To generate plasmid constructs for in-frame deletions and other allele replacements, homologous upstream and downstream fragments (~500 bp/each) were PCR-amplified and joined via overlap extension PCR (67). PCR products were cloned into plasmid pNPTS138 by restriction enzyme digestion and ligation. Similarly, to create genetic complementation constructs, target genes were amplified and fused to their upstream promoters (~500 bp fragment immediately upstream of the start of the annotated operon) via overlap extension PCR and these fused PCR products were purified and cloned into pXGFPC-2 (pMT585) (68), a plasmid that integrates into the xylX locus in Caulobacter. For complementation, the genes with their native promoters were cloned in the opposite orientation of the P_xylX promoter in this plasmid. For xylose-inducible expression, target genes were PCR-amplified and ligated into pMT585 in the same orientation as (i.e., downstream of) the P_xylX promoter. To create the glnBA transcriptional reporter construct, the target promoter (~500 bp fragment upstream of the start of the glnBA operon) was PCR-amplified and cloned into pPTM056 (69), which resulted in the fusion of P_glnBA to mNeonGreen. All ligations were transformed into E. coli TOP10. All plasmids were sequence confirmed.

Plasmids were transformed into Caulobacter via electro-poration or triparental mating from TOP10 using FC3 as a helper strain (70). In-frame deletion and allele replacement strains were generated via two-step recombination using sacB counterselection using an approach similar to that described by Hmelo et al (71). Briefly, primary recombinants bearing pNPTS138-derived allele-replacement plasmids were selected on solidified PYE containing kanamycin. Single colonies were then grown in PYE broth without selection for 6–18 hours (h) before secondary recombinants were selected on PYE containing 3% sucrose. The resulting clones were screened to confirm kanamycin sensitivity. Then allele replacement was confirmed by PCR for in-frame deletion alleles or PCR amplification and Sanger sequencing for point mutation alleles.

Measurement of growth in PYE complex medium

Starter cultures were grown overnight in PYE or PYE plus 9.3 mM glutamine shaken at 30°C. Overnight cultures were diluted to OD₆₆₀ 0.1 in the same media and incubated shaking for 2 h at 30°C to bring cultures to a similar (logarithmic) phase of growth. Cultures were then diluted to OD₆₆₀ 0.025 in the same media and shaken at 30°C. Optical density at 660 nm was measured at the timepoints indicated.

Measurement of growth in M2G defined medium

Starter cultures were shaken overnight in PYE at 30°C. Starter cultures were pelleted and washed three times with M2G or M2G in which NH₄⁺ was replaced with molar equivalent glutamine (9.3 mM final concentration) before dilution to OD₆₆₀ 0.025 in the respective medium. These cultures were incubated at 30 °C with shaking for 24 h and culture density was measured optically (OD₆₆₀).

Selection of ΔntrC and ntrC^ΔHTH suppressors

When ΔntrC, ntrC^ΔHTH, or ntrC^D56A cultures incubated in M2G overnight at 30°C exhibited visible turbidity, cultures were spread on PYE to isolate individual colonies bearing suppressing mutations. These putative suppressor strains were re-inoculated into M2G to confirm growth in the absence of a functional ntrC allele. Strains that grew rapidly - similar to WT - were saved and genomic DNA was purified and sequenced. Briefly, genomic DNA was extracted from 1 ml of saturated PYE culture using guanidinium thiocyanate (72). Genomic DNA was sequenced (150 bp paired-end reads) at SeqCenter (Pittsburgh, PA) using an Illumina NextSeq 2000. DNA sequencing reads were mapped to the Caulobacter NA1000 genome (Genbank accession CP001340) (31) and polymorphisms were identified using breseq (73).

RNA extraction, sequencing, and analysis

Starter cultures were grown for 18 h at 30°C in PYE or PYE plus 9.3 mM (final concentration) glutamine. Cultures were then diluted to OD₆₆₀ 0.1 in their respective medium and grown for 2 h to get the cultures in similar (logarithmic) phase of growth. Once again, cultures were diluted to OD₆₆₀ 0.1 in their respective medium and grown another 3.25 h (OD₆₆₀ < 0.4) to capture mRNA in a similar log phase of the growth curve. 6 ml of each culture were pelleted via centrifugation (1 min at 17,000 × g). Pellets were immediately resuspended in 1ml TRIzol and stored at −80°C until RNA extraction. To extract RNA, thawed samples were incubated at 65°C for 10 min. After addition of 200 μl of chloroform, samples were vortexed for 20 s and incubated at room temperature (RT) for 5 min. Phases were separated by centrifugation (10 min at 17,000 × g). The aqueous phase was transferred to a fresh tube and an equal volume of isopropanol was added to precipitate the nucleic acid. Samples were stored at 80°C (1 h to overnight), then thawed and centrifuged at 17,000 × g for 30 min at 4°C to pellet the nucleic acid. Pellets were washed with ice-cold 70% ethanol then centrifuged for at 17,000 × g for 5 min at 4°C. After discarding the supernatant, pellets were air-dried at RT, resuspended in 100 μl RNAse-free water, and incubated at 60°C for 10 min. Samples were treated with TURBO DNAse (Invitrogen) following manufactures protocol for 30 min at RT and then column purified using RNeasy Mini Kit (Qiagen). RNA samples were sequenced at SeqCenter (Pittsburgh, PA). Briefly, sequencing libraries were prepared using Illumina’s Stranded Total RNA Prep Ligation with Ribo-Zero Plus kit and custom rRNA depletion probes. 50 bp paired end reads were generated using the Illumina NextSeq 2000 platform (Illumina). RNA sequencing reads are available in the NCBI GEO database under series accession GSE234097. RNA sequencing reads were mapped to the Caulobacter NA1000 genome (31) using default mapping parameters in CLC Genomics Workbench 20 (Qiagen). To identify genes regulated by NtrC, the following criteria were use: fold change > 1.5, FDR P < 0.000001 and maximum group mean RPKM > 10. Gene expression data were hierarchically clustered in Cluster 3.0 (74) using an uncentered correlation metric with average linkage. The gene expression heatmap was generated using Java TreeView (75).

Chromatin immunoprecipitation with sequencing (ChIP-seq)

Caulobacter ntrC was PCR-amplified and inserted into pPTM057–3xFLAG expression vector via restriction digestion and ligation to generate a 3xFLAG-NtrC fusion expressed from a cumate-inducible promoter. This suicide plasmid was propagated in E. coli TOP10 and conjugated into Caulobacter ΔntrC to integrate at the xylose locus. For ChIP-seq experiments, the ΔntrC xylX::pPTM057–3xFLAG-ntrC strain was grown overnight in PYE at 30°C. The overnight culture was diluted to OD₆₆₀ 0.1 in PYE and outgrown for 2 h at 30°C. This culture was back-diluted to OD₆₆₀ 0.1 in PYE supplemented with 50 μM cumate and grown for 3.25 h at 37°C to induce 3xFLAG-ntrC during log phase growth. To crosslink 3xFLAG-NtrC to DNA, formaldehyde was added to 125 ml of culture to a final concentration of 1% (w/v) and shaken at 37°C for 10 min. The crosslinking was quenched using a final concentration of 125 mM glycine and shaken at 37°C for 5 min. Cells were pelleted by centrifugation at 7,196 × g for 5 min at 4°C. Supernatant was removed and the pellet was washed 4 times with ice-cold PBS pH 7.5. To lyse the cells, the washed pellet was resuspended in 1 ml lysis buffer [10 mM Tris pH 8, 1 mM EDTA, protease inhibitor tablet (Roche), 1 mg/ml lysozyme]. After a 30 min incubation at 37°C for 0.1% (w/v) sodium dodecyl sulfate (SDS) was added. To shear the genomic DNA to 300–500 bp fragments, the lysate was sonicated on ice for 10 cycles (20% magnitude for 20 sec on/off pulses using a Branson Sonicator). Cell debris was cleared by centrifugation (15,000 × g for 10 min at 4°C). Supernatant was transferred to a clean tube and Triton X-100 was added to a final concentration of 1% (v/v). The sample was pre-cleared via incubation with 30 μl of SureBeads Protein A magnetic agarose beads (BioRad) for 30 min at RT. The supernatant was transferred to a clean tube and 5% of the total lysate was saved as the input DNA reference sample. Pulldown was performed as previously described (69). Briefly, 100 ul magnetic agarose anti-FLAG beads (Pierce / Thermo) were pre-equilibrated in binding buffer [10 mM Tris pH 8 at 4°C, 1 mM EDTA, 0.1% (w/v) SDS, 1% (v/v) Triton X-100] supplemented with 1% (w/v) bovine serum albumin (BSA) overnight at 4°C, washed with binding buffer and incubated in the lysate for 3 h at RT. Beads were cleared from the lysate with a magnet, and washed with a low-salt buffer [50 mM HEPES pH 7.5, 1% (v/v) Triton X-100, 150 mM NaCl], followed by a high-salt buffer [50 mM HEPES pH 7.5, 1% (v/v) Triton X-100, 500 mM NaCl], and then LiCl buffer [10 mM Tris pH 8 at 4°C, 1 mM EDTA, 1% (w/v) Triton X-100, 0.5% (v/v) IGEPAL CA-630, 150 mM LiCl]. Finally, beads were incubated with 100 ul elution buffer [10 mM Tris pH 8 at 4°C, 1 mM EDTA, 1% (w/v) SDS, 100 ng/μl 3xFLAG peptide] for 30 min at RT. After pulldown, the input sample was brought to equal volume as the output/pulldown sample using elution buffer [10 mM Tris pH 8, 1 mM EDTA pH 8, 1% SDS, 100 ng/μl 3xFLAG peptide]. Input and output samples were supplemented with 300 mM NaCl and 100 μg/ml RNAse A and incubated at 37°C for 30 min. Proteinase K was added to samples at a final concentration of 200 μg/ml and samples were incubated overnight at 65°C to reverse crosslinks. Samples were purified using the Zymo ChIP DNA Clean & Concentrator kit. ChIP DNA was sequenced at SeqCenter (Pittsburgh, PA). Briefly, sequencing libraries were prepared using the Illumina DNA prep kit and sequenced (150 bp paired end reads) on an Illumina Nextseq 2000. ChIP-seq sequence data have been deposited in the NCBI GEO database under series accession GSE234097.

ChIP-seq analysis

Paired-end reads were mapped to the C. crescentus NA1000 reference genome (GenBank accession number CP001340) with CLC Genomics Workbench 20 (Qiagen). Peak calling was performed with the Genrich tool (https://github.com/jsh58/Genrich) on Galaxy; peaks are presented in Table S3. Briefly, PCR duplicates were removed from mapped reads, replicates were pooled, input reads were used as the control dataset, and peak were called using the default peak calling option [Maximum q-value: 0.05, Minimum area under the curve (AUC): 20, Minimum peak length: 0, Maximum distance between significant sites: 100].

To identify promoters that contained NtrC peaks, promoters were designated as 300 bp upstream and 100 bp downstream of the transcription start sites (TSS) annotated for each operon (76, 77). For genes/operons that did not have an annotated TSS, the +1 nucleotide of the first gene in the operon was designated as the TSS. Promoters were defined as containing an NtrC peak if there was any overlap between the NtrC ChIP-seq peak and the indicated promoter. To compare the relative location of NtrC binding sites with various cell cycle regulators, ChIPpeakAnno (78) was used to determine distance from the summit of the NtrC peaks to the nearest CtrA, SciP, MucR1, and GapR peak summit. To compare the relative location of NtrC binding sites with various cell cycle regulators, ChIPpeakAnno (78) was used to determine distance from the summit of the NtrC peaks to the nearest CtrA, SciP, MucR1, and GapR peak summit. ChIP-seq peaks (50 bp windows) for CtrA, SciP, and MucR1 were derived from (19) and the summits were considered the center of the 50 bp window. ChIP-seq summits for GapR were derived from (18). For motif discovery, sequences of the ChIP-seq peaks were submitted to MEME suite (79). Sequences were scanned for enriched motifs between 6 and 30 bp in length that had any number of occurrences per sequence.

NtrC protein purification

Caulobacter ntrC was PCR-amplified and inserted into a pET23b-His6-SUMO expression vector using classical restriction digestion and ligation, such that ntrC was inserted 3’ of the T7 promoter and the His6-SUMO coding sequence. After sequence confirmation, pET23b-His6-SUMO-ntrC was transformed into chemically competent E. coli BL21 Rosetta (DE3) / pLysS. This strain was grown in 1 L of LB at 37°C. When the culture density reached approximately OD₆₀₀ ≈ 0.4, expression was induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) overnight at 16°C. Cells were harvested by centrifugation (10,000 × g for 10 min) and resuspended in 20 ml lysis buffer [20 mM Tris pH 8, 125 mM NaCl, 10 mM imidazole] and stored at −80°C until purification.

For protein purification, resuspended cell pellets were thawed at RT. 1 mM phenylmethylsulfonyl fluoride (PMSF) was added to inhibit protease activity and DNase I (5 μg/ml) was added to degrade DNA after cell lysis. Cells incubated on ice were lysed by sonication (Branson Instruments) at 20% magnitude for 20 sec on/off pulses until the suspension was clear. The lysate was cleared of cell debris by centrifugation (30,000 × g for 20 min) at 4°C. The cleared lysate was applied to an affinity chromatography column containing Ni-nitrilotriacetic acid (NTA) superflow resin (Qiagen) pre-equilibrated in lysis buffer. Beads were washed with wash buffer [20 mM Tris pH 8, 125 mM NaCl, 30 mM imidazole]. Protein was eluted with elution buffer [20 mM Tris pH 8, 125 mM NaCl, 300 mM imidazole]. The elution fractions containing His6-SUMO-NtrC (~52 kDa) were pooled and dialyzed in 2 L dialysis buffer [20 mM Tris pH8, 150 mM NaCl] for 4 h at 4°C to dilute the imidazole. Purified ubiquitin-like-specific protease 1 (Ulp1) was added to the eluted His6-SUMO-NtrC containing solution which was then dialyzed overnight at 4°C in 2 L fresh dialysis buffer to cleave the His6-SUMO tag. Digested protein was mixed with 3 ml of NTA superflow resin (Qiagen) that had been pre-equilibrated in wash buffer. After incubation for 30 min at 4°C, the solution was placed onto a gravity drip column at 4°C. Flowthrough containing cleaved NtrC was collected and used to generate α-NtrC polyclonal antibodies (Pacific Immunology).

Western blotting

To prepare cells for analysis, overnight PYE cultures of Caulobacter strains in Fig S2B and S2C were diluted in fresh PYE to OD₆₆₀ 0.1 and grown 2 h at 30°C. These outgrown cultures were then re-diluted in fresh PYE to OD₆₆₀ 0.1 and grown for 3.25 h at 30°C to capture exponential growth phase. Cells from 1 ml of each culture were collected by centrifugation (12,000 × g for 1 min). After discarding the supernatant, cell pellets were stored at −20°C until western blot analysis. Strains in Fig S2A were grown as above except that the outgrowth medium was supplemented with 0.15% xylose and upon re-dilution in xylose supplemented medium, cultures were grown for 24 h at 30°C to capture stationary growth phase (OD₆₆₀ > 0.6). Cells from 1 ml of each stationary phase culture were harvested as above and stored at −20°C until western blot analysis. Strains for Fig 2SD were grown in PYE overnight. Cells from 1 ml of each overnight culture were collected by centrifugation as described above and the pellets were placed at −20°C until western blot analysis.

For western blot analysis, cell pellets were thawed and resuspended in 2X SDS loading buffer [100 mM Tris-Cl (pH 6.8), 200 mM dithiothreitol, 4% (w/v) SDS, 0.2% bromophenol blue, 20% (v/v) glycerol] to a concentration of 0.0072 OD₆₆₀ • ml culture / μl loading buffer. After resuspension, genomic DNA is digested by incubation with 1 μl Benzonase per 50 μl sample volume for 20 min at RT. Samples then were denatured at 95°C for 5 min. 10 μl of each sample was loaded onto a 4–20% mini-PROTEAN precast gel (Bio-Rad) (Fig S2C) or a 7.5% mini-PROTEAN precast gel (BioRad) (Fig S2A, S2B, and S2D) and resolved at 180 V at RT. Separated proteins were transferred from the acrylamide gel to a PVDF membrane (Millipore) using a semi-dry transfer apparatus (BioRad) at 10 V for 30 min at RT [1X Tris-Glycine, 20% methanol]. The membrane was blocked in 10 ml Blotto [1X Tris-Glycine with 0.1% Tween 20 (TBST) + 5% (w/v) powdered milk] for 1 h to overnight at 4°C. The membrane was then incubated in 10 ml Blotto + polyclonal rabbit α-NtrC antiserum (1:1,000 dilution) 1 h to overnight at 4°C. The membrane was washed in TBST three times. The membrane was then incubated in 10 ml Blotto + goat α-rabbit poly-horseradish peroxidase secondary antibody (Invitrogen; 1:10,000 dilution) for 1–2 h at RT. The membrane was then washed three times with TBST and developed with ProSignal Pico ECL Spray (Prometheus Protein Biology Products). Immediately upon spraying, the membrane was imaged using BioRad ChemiDoc Imaging System (BioRad).

Caulobacter stalk length measurement and analysis

To prepare stationary phase cells, starter cultures were grown in PYE overnight at 30°C and diluted to OD₆₆₀ 0.1 in fresh PYE or PYE plus 9.3 mM glutamine. After a 2 h outgrowth at 30°C cultures were re-diluted to OD₆₆₀ 0.1 in fresh medium and grown for 24 h at 30°C to capture stalk lengths in stationary phase (> OD₆₆₀ 0.6). 2 μl of each stationary phase culture were spotted on an agarose pad [1% agarose dissolved in water] on a glass slide and covered with a glass cover slip. Cells were imaged using a Leica DMI 6000 microscope using phase contrast with an HC PL APO 63x/1.4 numeric aperture oil Ph3 CS2 objective. Images were captured with an Orca-ER digital camera (Hamamatsu) controlled by Leica Application Suite X (Leica). Stalk length was measured using BacStalk (80) with a minimum stalk length threshold of 0.6 microns.

Transcriptional reporter assay

Overnight starter cultures grown in PYE supplemented with chloramphenicol (1.5 μg/ml) to maintain the replicating plasmid were diluted to OD₆₆₀ 0.1 in the same medium and outgrown for 2 h at 30°C. Outgrown cultures were rediluted to OD₆₆₀ 0.1 in the same medium and grown at 30°C for 24 h to capture expression in stationary phase. 200 μl of each culture was transferred to a Costar flat bottom, black, clear bottom 96-well plate (Corning). Cell density assessed by absorbance (660 nm) and mNeonGreen fluorescence (excitation = 497 ± 10 nm; emission = 523 ± 10 nm) were measured in a Tecan Spark 20M plate reader.

Importance.

Bacteria balance cellular processes with the availability of nutrients in their environment. The NtrB-NtrC two-component signaling system is responsible for controlling nitrogen assimilation in many bacteria. We have characterized the effect of ntrB and ntrC deletion on Caulobacter growth and development and uncovered a role for spontaneous IS element transposition in the rescue of transcriptional and nutritional deficiencies caused by ntrC mutation. We further defined the regulon of Caulobacter NtrC, a bacterial enhancer binding protein, and demonstrate that it shares specific binding sites with essential proteins involved in cell cycle regulation and chromosome organization. Our work provides a comprehensive view of transcriptional regulation mediated by a distinctive NtrC protein, establishing its connection to nitrogen assimilation and developmental processes in Caulobacter.