Azotobacter vinelandii AmrZ is a global regulator linking alginate production and c-di-GMP homeostasis

TL;DR

The AmrZ protein in Azotobacter vinelandii controls alginate production and cell movement by regulating c-di-GMP levels and gene expression.

Contribution

This study identifies AmrZ as a global regulator linking alginate biosynthesis and c-di-GMP homeostasis in Azotobacter vinelandii.

Findings

AmrZ activates algD transcription and increases c-di-GMP, promoting alginate synthesis.

AmrZ positively autoregulates its expression via an AlgU-dependent feedback loop.

AmrZ influences motility and broader cellular processes like metabolism and iron homeostasis.

Abstract

Azotobacter vinelandii AmrZ positively regulates alginate production by directly activating algD transcription and enhancing intracellular c-di-GMP accumulation, which, in turn, activates the alginate polymerization complex. Two diguanylate cyclases (DGCs) were identified as AmrZ targets, likely accounting for the reduced c-di-GMP levels observed in the absence of AmrZ. Transcription of avGReg, which encodes the main DGC during vegetative growth, was not under AmrZ control; however, post-translational regulation of AvGReg activity by AmrZ cannot be ruled out. AmrZ also positively autoregulates its own expression through an AlgU-dependent feedback loop. Finally, AmrZ positively modulates swimming motility, and our transcriptomic data suggest that this effect is indirect, occurring via its positive influence on c-di-GMP levels.

Azotobacter vinelandii, a member of the Pseudomonadaceae, produces the exopolysaccharide alginate during vegetative growth; however, the circuitry linking alginate biosynthesis to lifestyle transitions remains poorly defined. Here, we show that the Ribbon-Helix-Helix transcription factor AmrZ coordinates alginate production, intracellular c-di-GMP levels and motility. Deletion of amrZ abolished alginate synthesis, whereas chromosomal complementation restored it. A PalgD-gusA fusion and RT-qPCR demonstrated that algD, the first gene in the alginate biosynthetic cluster, depends on AmrZ for expression. Motif analysis identified multiple AmrZ sites upstream of algD, and electrophoretic mobility-shift assays (EMSAs) confirmed specific binding to these regions. AmrZ also positively autoregulates: PamrZ-gusA activity decreased in ΔamrZ, and purified AmrZ bound the amrZ promoter in EMSA. Moreover, PamrZ activity required the sigma factor AlgU, consistent with the presence of an AlgU promoter; this positive, AlgU-dependent feedback may stabilize AmrZ under alginate-inducing conditions. To probe AmrZ control of c-di-GMP, we implemented a riboswitch-based biosensor in A. vinelandii. The ΔamrZ strain showed a markedly reduced signal, similar to a diguanylate cyclase (DGC) mutant, whereas a phosphodiesterase mutant displayed elevated output, validating the assay. RNA-seq and RT-qPCR identified two DGC genes, AVAEIV_RS11610 and AVAEIV_RS18795, as AmrZ-activated targets; EMSA verified direct binding at the RS11610 regulatory region. By contrast, transcription of the principal vegetative DGC AvGReg was not AmrZ-regulated. Lower c-di-GMP in ΔamrZ correlated with larger swimming halos. Collectively, these genetic, biochemical and transcriptomic data support a model in which AmrZ directly activates algD and elevates c-di-GMP via selected DGCs, thereby promoting alginate synthesis while reducing motility. RNA-seq data also indicate that AmrZ influences broader cellular programmes, including metabolism and iron homeostasis, positioning AmrZ as a central regulator that links c-di-GMP homeostasis to coordinated exopolysaccharide production in A. vinelandii. This work contributes to our understanding of the regulatory networks controlled by AmrZ outside the Pseudomonas genus and reveals important differences in its targets and regulatory mechanisms.

Figures

AmrZ is necessary for alginate production and algD expression. (a) Alginate quantification (mg alginate mg⁻¹ protein) for cultures grown for 48 h in BS medium. (b) β-Glucuronidase activity from an algD-gusA transcriptional fusion in the WT (black) and ΔamrZ (grey). Specific activity is shown in U mg⁻¹ protein. (c) Relative algD mRNA levels in cultures grown for 24 h in BS medium (RT-qPCR). Bars denote mean±sd from three independent experiments. Statistical significance (two-tailed Student’s t-test) is indicated: P<0.01; *P<0.001; P<0.0001.

AmrZ directly binds the algD regulatory region. (a) Schematic of putative AmrZ-binding sites (BS1–BS5) within the algD regulatory region. The promoter fragments tested in panels (b–e) are indicated: 679 bp (BS1–BS5), 208 bp (BS1–BS2), 194 bp (BS3–BS4) and 206 bp (BS5). (b–e) EMSAs with increasing concentrations of AmrZ and the indicated algD promoter fragments. Shifted bands indicate AmrZ–DNA complex formation. Negative controls (C−) were a 251 bp fragment from ORF AVAEIV_RS16850 (panel b) and a 110 bp fragment from gyrA (AVAEIV_RS08140) (panels c–e). DNA was visualized by ethidium bromide staining.

amrZ transcription is controlled by AlgU and AmrZ. (a) Activity of a PamrZ-gusA transcriptional fusion in the WT and the ΔamrZ and ΔalgU backgrounds during growth in liquid BS medium. (b) EMSA showing AmrZ binding to the amrZ regulatory region. A promoter fragment was incubated with increasing concentrations of purified His-AmrZ. A 251 bp fragment from ORF AVAEIV_RS16855 served as a negative control. DNA was visualized by ethidium bromide staining. Bars denote mean±sd from three independent experiments. Statistical significance (two-tailed Student’s t-test) is indicated: *P<0.05; P<0.01; *P<0.001; P<0.0001.

AmrZ regulates intracellular c-di-GMP levels. (a) c-di-GMP biosensor readout in WT, ΔamrZ, ΔavGReg and ΔmucG strains. Accumulation was monitored with the TurboRFP riboswitch biosensor (Bc3–Bc5). RFI=TurboRFP fluorescence/total protein. Bars show mean±sd from three independent experiments. Statistical significance (two-tailed Student’s t-test): P<0.01; *P<0.001; P<0.0001. (b) RT-qPCR of AVAEIV_RS11610, AVAEIV_RS18795 and AVAEIV_RS25725 mRNA in the indicated strains grown for 24 h in BS medium. Domain organizations of the encoded proteins are shown. (c) EMSA demonstrating AmrZ binding to the AVAEIV_RS11610 promoter region. DNA was incubated with increasing concentrations of His-AmrZ. A 251 bp fragment from AVAEIV_RS16855 served as a negative control. DNA was visualized by ethidium bromide staining. RFI, relative fluorescence intensity.

Analysis of the transcriptomic data from the ΔamrZ mutant versus WT. (a) Volcano plot of differential expression. Vertical lines mark log₂-fold change ≥2, and the horizontal line marks Padj<0.01; red points denote genes meeting both criteria. (b) Global functional profile of the DEG in the △amrZ mutant. The per cent of each GO term is indicated. MFS, Major Faclitator Superfamily.

AmrZ negatively regulates swimming motility. Representative swimming motility of the indicated strains on soft agar after 24 h of incubation. The graph shows the quantification of swimming halo diameters. Statistical significance (two-tailed Student’s t-test) is indicated: *P<0.05.

Full text

Data Availability

The Azotobacter vinelandii WT AEIV genome sequence is available at GenBank (GCA_030506185.2). All data generated or analysed during this study are included in this article and its supplementary information files.

Introduction

The free-living bacterium Azotobacter vinelandii, a member of the Pseudomonadaceae family, is strictly aerobic and motile by peritrichous flagella [1]. Vegetative cells undergo morphological and physiological differentiation to form cysts that are resistant to desiccation [2]. Alginate, a linear exopolysaccharide composed of β-d-mannuronic and α-l-guluronic acid residues, is the major structural component of the cyst layers [23]. This polymer is also synthesized in vegetative, non-encysting cells of A. vinelandii [45]. Because alginate has distinctive physicochemical properties that can be exploited as a bio-based material [67], alginate production in A. vinelandii has been studied to better understand its genetic regulation, with the ultimate goal of optimizing microbial production [458].

Except for algC, the genes required for alginate biosynthesis in A. vinelandii are organized in a 12-gene chromosomal cluster (algD-alg8-alg44-K-J-G-X-l-I-V-F-A) that encodes the proteins forming the periplasm-spanning biosynthetic machinery [589]. This cluster is headed by algD, whose transcription is finely tuned by several regulators, including the sigma factors AlgU and RpoS [1013], and by the bis (3',5')-cyclic dimeric guanosine monophosphate (c-di-GMP)-responsive regulator FleQ, which binds the algD regulatory region to inhibit transcription [14]. The GacS/GacA-Rsm system controls algD expression post-transcriptionally [1516]. As in Pseudomonas aeruginosa, alginate production in A. vinelandii is also post-translationally regulated: c-di-GMP binds the PilZ domain of the co-polymerase Alg44, thereby activating the alginate polymerase Alg8 [1720]. Moreover, in A. vinelandii, c-di-GMP strongly influences alginate’s physicochemical properties by modulating both molecular mass and monomer composition [2022].

Although alginate biosynthesis is largely conserved between A. vinelandii and P. aeruginosa, there are notable regulatory differences that likely reflect the polymer’s distinct functions in these organisms. In P. aeruginosa, AlgR, AlgB and AmrZ are required for transcriptional activation of the algD promoter [2326]. In A. vinelandii, however, AlgR and AlgB are not required for algD expression or alginate production [2728]. In a previous study, a screen of a transposon insertion library in A. vinelandii revealed that AmrZ is required for alginate production [28].

AmrZ is a 108-amino-acid protein with a flexible N-terminus, a central Ribbon-Helix-Helix (RHH) DNA-binding domain and a C-terminal domain required for oligomerization [2931]. Initially described in P. aeruginosa as a regulator of alginate and twitching motility [3234], AmrZ is now recognized as a global regulator for environmental sensing and adaptation in several Pseudomonas species. AmrZ is implicated in iron acquisition, flagellar structure, type III and type VI secretion systems, rhizosphere colonization, virulence and biofilm formation, among other processes [25,3541]. AmrZ is a key regulator of exopolysaccharide production. In P. aeruginosa and Pseudomonas syringae pv. tomato, AmrZ promotes alginate production by activating algD [4042]. Four AmrZ-binding sites have been identified in the P. aeruginosa algD promoter, and their occupancy promotes formation of a higher order DNA–AmrZ complex activating transcription [313243]. Conversely, AmrZ represses genes for the synthesis of Pea, Psl, Pel and cellulose in different Pseudomonas [254044]. Many of these pathways are also regulated by the second messenger c-di-GMP [2541]. Indeed, in different Pseudomonas, AmrZ modulates c-di-GMP levels by regulating genes encoding diguanylate cyclases (DGCs) and phosphodiesterases (PDEs), which synthesize and degrade c-di-GMP, respectively [2545].

The aim of this study is to define the role of AmrZ in A. vinelandii physiology, representing the first characterization of this regulator outside the Pseudomonas genus. This was achieved using genetic, biochemical and transcriptomic analyses. We show that AmrZ is essential for alginate production because it is required for transcriptional activation of algD. Our results further indicate that AmrZ positively influences c-di-GMP levels by regulating specific DGCs and that this control prevents swimming motility.

Methods

Bacterial strains and growth conditions

Bacterial strains and plasmids used in this study are listed in Table S1, available in the online Supplementary Material. The A. vinelandii AEIV WT strain (also named E strain) [46] and derivatives were grown diazotrophically in minimal Burk’s medium supplemented with sucrose (20 g l⁻¹). The composition of Burk’s medium has been described previously [21]. Cultures were grown at 30 °C with shaking at 200 r.p.m. in 250 ml Erlenmeyer flasks containing 50 ml of medium, for the times indicated. Cultures were inoculated with 400 µg of cells from 20 to 24 h liquid Burk’s-sucrose (BS) pre-cultures. The final antibiotic concentrations (µg ml⁻¹) used for A. vinelandii and Escherichia coli, respectively, were as follows: tetracycline (Tc), 30 and 15; gentamicin (Gm), 1 and 10; spectinomycin (Sp), 100 and 100; and ampicillin (Ap), not used and 200.

AEIV genome sequencing

Three library types were prepared. Short-read paired-end libraries (2×75 bp) were sequenced on an Illumina NextSeq 500, while long-read libraries were sequenced using a PacBio RS II and an Oxford Nanopore MinION. Raw Illumina reads were adapter- and quality-trimmed with Atropos v1.1.32 [47]. PacBio reads, together with the processed Illumina reads, were used for de novo hybrid assembly with Unicycler v0.5.1 [48] under default settings.

Following the initial assembly, loci showing ambiguities or breaks were manually curated. Nanopore reads were mapped to the draft assembly and examined using the visualization tool Tablet v1.21.02.08; problematic regions were then edited to restore contiguity and ensure accurate sequence representation. Assembly integrity was subsequently evaluated by remapping all three datasets (Illumina, PacBio and Nanopore). The curated assembly was submitted to the National Center for Biotechnology Information (NCBI), where gene prediction and feature annotation were carried out using the Prokaryotic Genome Annotation Pipeline (PGAP v6.9) [49]. Project metadata are available under BioProject accession no. PRJNA809843 and BioSample accession no. SAMN26203234. The complete assembled and annotated genome of A. vinelandii strain AEIV is available in GenBank under assembly accession no. GCA_030506185.2.

Standard techniques

Oligonucleotides used for PCR (Table S2) were designed from the A. vinelandii AEIV genome sequence (GenBank: CP092752.2). PCRs were performed with Phusion High-Fidelity DNA Polymerase (Thermo Fisher Scientific). Amplicons were verified by DNA (Sanger) sequencing.

Analytical methods

β-Glucuronidase activity in A. vinelandii was measured as described [50]. Protein concentration was determined by the Lowry assay [51]. Alginate was quantified using the carbazole-based spectrophotometric assay for uronic acids [52]. All experiments were performed in biological triplicate (n=3), each with three technical replicates; results are reported as the mean of the biological replicates. Statistical significance was assessed with Student’s t-test (two-tailed; P=0.05). Swimming motility was assayed as described previously [20].

Construction of A. vinelandii mutants

Details of mutant construction and plasmids are listed in Table S1. Competent A. vinelandii cells were prepared as described, exploiting the naturally competent state under iron-limited conditions [2153]. Competent cells were transformed with 5 µg of linearized plasmid DNA carrying the desired mutation to promote double-crossover allelic exchange, and transformants were selected on the appropriate antibiotic.

Identification of potential AmrZ-binding sites

AmrZ recognition sequences from P. aeruginosa [25] were used to build a position-specific scoring matrix and motif logo with Multiple EM for Motif Elicitation (MEME) in ‘one occurrence per sequence’ mode [54]. The matrix was queried against A. vinelandii AEIV 5′-UTR sequences using Find Individual Motif Occurrences (FIMO) [55]. The 5′-UTR sequences were extracted with bedtools getfasta from the BEDTools toolkit [56], comprising 400 nt upstream and 50 nt downstream of the annotated start coordinate of each gene heading predicted transcriptional units in the AEIV genome. Operon prediction was conducted as described [57].

RNA isolation and quantitative reverse transcription PCR (RT-qPCR)

Total RNA from A. vinelandii was extracted with TRIzol reagent (Invitrogen) following the manufacturer’s instructions. RNA integrity was verified by agarose gel electrophoresis, and concentrations were measured with a NanoDrop spectrophotometer. Residual genomic DNA was removed by treatment with RNase-free DNase I (Invitrogen) at 37 °C for 30 min, followed by heat inactivation at 65 °C for 10 min. First-strand cDNA was synthesized with the Maxima Reverse Transcriptase kit (Thermo Fisher Scientific) from 2 µg of total RNA using gene-specific reverse primers (Table S2). cDNA was quantified and diluted to a working concentration of 50 ng µl⁻¹ with nuclease-free water. Quantitative real-time PCR was performed with Maxima SYBR Green/ROX qPCR Master Mix (Thermo Fisher Scientific) on a QuantStudio^TM^ 5 Real-Time PCR System (Applied Biosystems) in 96-well plates. Each 10 µl reaction contained 50 ng cDNA template (5 ng µl⁻¹), 1× Master Mix and gene-specific primers (Table S2). Cycling conditions were an initial denaturation, followed by 35 cycles of 95 °C for 15 s, 58 °C for 30 s and 72 °C for 30 s. Primer specificity was assessed by agarose-gel analysis of conventional PCR products and by melt-curve analysis. The recA (AVAEIV_RS18860) or the gyrA (AVAEIV_RS08140) gene served as the internal reference for normalization. Relative expression was calculated using the 2^−ΔΔCt^ method with the WT strain as the calibrator (expression level=1) [58]. Experiments were performed with three biological replicates, each measured in triplicate.

Expression and purification of His-AmrZ

AmrZ was produced as an N-terminal His₆-tagged protein (His-AmrZ). A single colony of E. coli BL21(DE3) carrying pET-AmrZ was inoculated into 5 ml LB with kanamycin (Km) and grown overnight at 37 °C, 200 r.p.m. One millilitre of this culture was used to inoculate 250 ml LB (Km); cells were grown ~3 h to OD₆₀₀≈0.6, induced with 1 mM IPTG, and incubated for an additional 3 h. Cells were harvested (4,000 g, 10 min, 4 °C), resuspended in lysis buffer (50 mM NaH_₂_PO_₄_, 300 mM NaCl, 10 mM imidazole, pH 7.0) and sonicated on ice (3×15 s). The lysate was clarified (4,000 g, 20 min, 4 °C) and applied to Ni-NTA resin (Thermo Fisher Scientific) pre-equilibrated in lysis buffer; binding was carried out for 30 min at 4 °C with gentle mixing. The column was washed with wash I (50 mM NaH_₂_PO_₄_, 300 mM NaCl, 50 mM imidazole, pH 7.0) and wash II (same salts, 250 mM imidazole, pH 7.0), and protein was eluted with elution buffer (same salts, 500 mM imidazole, pH 7.0). The eluate was concentrated using Amicon Ultra-0.5, 10 kDa MWCO devices. Protein concentration was determined by the Lowry assay [51]. Purity and apparent molecular mass (~15 kDa) were assessed by 10% SDS–PAGE.

EMSA

Electrophoretic mobility-shift assays (EMSAs) were performed using a non-radioactive protocol. Four PCR fragments spanning the algD regulatory region were used: (a) 679 bp (AmrZ sites S1–S5), primers palgD-EMSA-F1.21/palgD R; (b) 204 bp (S1, S2), palgD-EMSA-F1.21/palgD-EMSA-R1; (c) 194 bp (S3, S4), palgD-EMSA-F2/palgD-EMSA-R2 and (d) 206 bp (S5), palgD-EMSA-F3/palgD R. For testing binding to the amrZ regulatory region, pAmrZ-XbaF/pAmrZ-EcoR-Rv amplified PamrZ.

Binding reactions (20 µl) contained 100 ng DNA, increasing concentrations of His-AmrZ and binding buffer (10 mM Tris-HCl, pH 8.0; 50 mM KCl; 1 mM DTT; 0.5 mM EDTA; 5% glycerol; 10 µg ml⁻¹ BSA). After 20 min at room temperature, samples were resolved on 6% nondenaturing polyacrylamide gels in Tris/borate/EDTA (TBE; 45.5 mM Tris base, 45.5 mM boric acid, 1 mM EDTA, pH 8.3) and visualized by ethidium bromide staining under UV light.

Construction of a c-di-GMP biosensor suitable for A. vinelandii

A 2,228 bp fragment was PCR-amplified from plasmid pFY4357 [59] using primers pMMB-Rv and pMMBdual-secF. The amplicon carries a c-di-GMP biosensor consisting of a tandem dual riboswitch (Bc4–5) from Bacillus thuringiensis and the TurboRFP reporter gene, whose expression is repressed in the absence of c-di-GMP [60]. The PCR product was cloned into pJET1.2/Blunt, yielding pJB2. A BamHI fragment (1,103 bp) was then excised and ligated into pLA65 (BamHI-digested), downstream of a constitutive σ^70^ promoter, to generate pLA66. The entire cassette, including the promoter, was subsequently excised with BglII and cloned into pUMA-Km(5′–3′), producing pLA68. This vector, pre-linearized with ScaI, enables chromosomal integration of the biosensor into a neutral locus (melA) of A. vinelandii via double-homologous recombination.

c-di-GMP quantification

Biosensor strains

The AEIV WT, ΔmucG, ΔavGReg and ΔamrZ strains were transformed with plasmid pLA68, yielding strains CLAM36, CLAM37, MG07 and MG08, respectively. These strains carry a chromosomally integrated biosensor for intracellular c-di-GMP. CLAM37 (ΔmucG) and MG07 (ΔavGReg) served as controls because they exhibit increased and decreased c-di-GMP levels, respectively, when compared with the WT background (CLAM36).

Fluorescence readout

For c-di-GMP quantification, cultures were sampled over a time course (12, 18, 24 and 48 h). At each time point, 1 ml of culture was centrifuged, the pellet was resuspended in 1 ml of water, and cells were washed twice with 10 mM MgSO_₄_ to remove alginate. Aliquots (200 µl) were transferred to black 96-well plates, and fluorescence was recorded (Ex 550 nm/Em 580 nm) on an Agilent BioTek Synergy H1 multimode microplate reader. Fluorescence values were normalized to total protein content.

RNA-seq

The A. vinelandii AEIV WT strain and its ΔamrZ derivative were grown diazotrophically in BS medium for 18 h. Aliquots of these cultures (corresponding to 400 µg of protein) were used to inoculate 50 ml of fresh BS medium and incubated for 24 h at 30 °C. Cells were harvested by centrifugation (4,000 g, 10 min, 4 °C). Total RNA was extracted with the RiboPure RNA Purification Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. For each strain (WT and ΔamrZ), RNA was isolated from three independent cultures (biological replicates). RNA quality was assessed by capillary electrophoresis on a 2100 Bioanalyzer (Agilent Technologies). Ribosomal RNA was depleted using the RiboMinus Bacteria 2.0 Transcriptome Isolation Kit (Thermo Fisher Scientific).

Library preparation and sequencing

Libraries were prepared following the TruSeq Stranded mRNA Sample Preparation Guide (Illumina). cDNA libraries were sequenced as 2×75 bp paired-end reads on an Illumina NextSeq 500 using the NextSeq 500/550 High Output Kit v2.5 (150 cycles). A total yield of 43,253,228 paired-end reads was obtained for the AEIV WT strain and 30,955,718 paired-end reads for the ΔamrZ derivative.

Processing of RNA-seq data

Quality control and read mapping: Read quality was assessed with FastQC v0.10.0 (Babraham Bioinformatics; https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). All libraries showed high per-base Phred scores (>Q30), with no evidence of adapter contamination. Reads were aligned to the A. vinelandii AEIV reference genome (GenBank assembly GCA_030506185.2) using SMALT v0.7.6 (https://www.sanger.ac.uk/science/tools/smalt-0). Feature-level coverages/counts were generated with BEDTools v2.4.0 (coverageBed), yielding a count table for downstream analysis [61].

Differential gene expression analysis

Counts per million (CPM) were calculated using the cpm function in edgeR v3.12.1 [62]. Genes were retained if they had ≥1 count in at least three samples. Differential expression was assessed with DESeq2 v1.10.1 using the filtered count matrix and default parameters. P-values were adjusted by the Benjamini–Hochberg procedure. Genes with log_₂_-fold change >2 or <−2 and adjusted P (Padj) <0.01 were considered differentially expressed and classified as AmrZ-responsive.

Results

The transcriptional regulator AmrZ is required for alginate production

To assess the role of AmrZ (AVAEIV_RS17290) in A. vinelandii alginate production, we quantified polysaccharide levels in the ΔamrZ mutant and compared them with the WT strain. After 48 h in BS medium, alginate production was abolished in ΔamrZ: values were at or below the limit of detection and comparable to those of the alginate-negative ΔalgU mutant [63] (Fig. 1a). Genetic complementation with a chromosomally integrated WT amrZ allele restored alginate synthesis, indicating that the AmrZ DNA-binding protein is required for alginate production.

*AmrZ is necessary for alginate production and algD expression. (a) Alginate quantification (mg alginate mg⁻¹ protein) for cultures grown for 48 h in BS medium. (b) β-Glucuronidase activity from an algD-gusA transcriptional fusion in the WT (black) and ΔamrZ (grey). Specific activity is shown in U mg⁻¹ protein. (c) Relative algD mRNA levels in cultures grown for 24 h in BS medium (RT-qPCR). Bars denote mean±sd from three independent experiments. Statistical significance (two-tailed Student’s t-test) is indicated: **P<0.01; ***P<0.001; ***P<0.0001.

Effect of AmrZ on algD expression

To assess the role of AmrZ in algD transcription, we introduced the ΔamrZ mutation into strain AED-gusA, which carries an algD-gusA transcriptional fusion, generating strain JG522. In the absence of AmrZ, algD transcription was strongly reduced relative to the WT at early time points (12 and 24 h). By 48 h, reporter activity in ΔamrZ increased slightly, reaching ~30% of the WT level (Fig. 1b). Consistent with these data, algD mRNA measured by RT-qPCR was nearly undetectable in ΔamrZ at 24 h and comparable to that of the ΔalgU mutant. As expected, algD mRNA levels were restored in the complemented strain (ΔamrZ/amrZ^+^), confirming that AmrZ is essential for algD transcriptional activation (Fig. 1c).

AmrZ regulates algD by directly binding its regulatory region

A MEME/FIMO search (Methods) was performed to identify potential AmrZ targets in the A. vinelandii AEIV genome (Table S3). In P. aeruginosa, AmrZ directly binds the algD regulatory region to activate transcription [42]. Consistent with a conserved mechanism in A. vinelandii, five candidate AmrZ-binding sites (BS1–BS5) were identified at −565, –548, −426, –388 and −62, respectively, relative to the algD start codon (Figs 2a and Supplementary Material 1).

AmrZ directly binds the algD regulatory region. (a) Schematic of putative AmrZ-binding sites (BS1–BS5) within the algD regulatory region. The promoter fragments tested in panels (b–e) are indicated: 679 bp (BS1–BS5), 208 bp (BS1–BS2), 194 bp (BS3–BS4) and 206 bp (BS5). (b–e) EMSAs with increasing concentrations of AmrZ and the indicated algD promoter fragments. Shifted bands indicate AmrZ–DNA complex formation. Negative controls (C−) were a 251 bp fragment from ORF AVAEIV_RS16850 (panel b) and a 110 bp fragment from gyrA (AVAEIV_RS08140) (panels c–e). DNA was visualized by ethidium bromide staining.

EMSAs were used to test these predictions. AmrZ caused a mobility shift of a 679 bp PalgD fragment, indicating DNA–protein complex formation (Fig. 2b). With increasing AmrZ concentration, at least three discrete retarded bands were observed, consistent with occupancy of multiple sites and formation of higher order complexes. Binding was specific, as no shift was detected with a negative-control DNA fragment.

To delimit the sites required for binding, three shorter PCR fragments were tested: 208 bp (BS1–BS2), 194 bp (BS3–BS4) and 206 bp (BS5). AmrZ bound the BS1–BS2 and BS3–BS4 fragments (Fig. 2c, d) but not the BS5 fragment (Fig. 2e). Together, these data indicate that AmrZ directly recognizes the algD regulatory region and acts as a transcriptional activator in A. vinelandii.

Transcriptional regulation of amrZ

AmrZ is an RHH transcription factor reported to function as both an activator and a repressor; in P. aeruginosa, it negatively autoregulates its own transcription [64]. We therefore examined whether this occurs in A. vinelandii, noting a putative AmrZ-binding motif in the amrZ regulatory region (PamrZ) (Table S3 and Fig. S2). amrZ transcription was monitored using a PamrZ-gusA fusion. In the WT, PamrZ activity increased gradually over the growth curve, reaching a maximum at ~48 h. In the ΔamrZ mutant, PamrZ activity was clearly reduced from 12 h onward (Fig. 3a), indicating that, contrary to what has been described in some Pseudomonas species, AmrZ positively regulates its own transcription in A. vinelandii. This observation is consistent with the location of the predicted AmrZ-binding site, which lies ∼150 nt upstream of the AlgU-dependent promoter (Fig. S2). In P. aeruginosa, by contrast, AmrZ binds near the transcription start site of repressed genes, including its own amrZ promoter [25].

*amrZ transcription is controlled by AlgU and AmrZ. (a) Activity of a PamrZ-gusA transcriptional fusion in the WT and the ΔamrZ and ΔalgU backgrounds during growth in liquid BS medium. (b) EMSA showing AmrZ binding to the amrZ regulatory region. A promoter fragment was incubated with increasing concentrations of purified His-AmrZ. A 251 bp fragment from ORF AVAEIV_RS16855 served as a negative control. DNA was visualized by ethidium bromide staining. Bars denote mean±sd from three independent experiments. Statistical significance (two-tailed Student’s t-test) is indicated: *P<0.05; **P<0.01; ***P<0.001; ***P<0.0001.

To test whether this regulation is direct, EMSAs were performed with a DNA fragment encompassing the amrZ promoter region and purified A. vinelandii His-AmrZ. A concentration-dependent mobility shift was observed; one retarded band increased with added protein, indicating formation of a DNA–protein complex (Fig. 3b). Together with the reporter data, these results support a model in which AmrZ positively autoregulates transcription by directly binding its own regulatory region.

AlgU is required for amrZ transcription

Previous work showed that AmrZ is under positive control of the sigma factor AlgU in A. vinelandii: the AmrZ protein was undetectable in the proteome of an AlgU-deficient strain, and an AlgU-dependent promoter motif was identified upstream of amrZ [50]. To test this directly, as also reported for P. aeruginosa [65], we introduced a PamrZ-gusA transcriptional fusion into a ΔalgU mutant. PamrZ activity was markedly reduced in ΔalgU relative to WT, confirming positive regulation of amrZ by AlgU (Fig. 3a). Notably, loss of either AlgU or AmrZ produced a similar temporal profile, near-complete loss of promoter activity at early time points, followed by a gradual increase, suggesting that AmrZ activates transcription from the AlgU-dependent promoter and that an additional promoter contributes later in growth. Consistent with this, a σ⁷⁰-type promoter is predicted within PamrZ, ~205 bp upstream of the AlgU-dependent promoter (Fig. S2).

AmrZ affects the pool of c-di-GMP in A. vinelandii

In Pseudomonas spp., AmrZ modulates intracellular c-di-GMP levels through the regulation of enzymes involved in the synthesis and degradation of this second messenger. In P. aeruginosa, AmrZ represses the DGC AdcA, thereby limiting the cellular c-di-GMP pool, whereas in Pseudomonas ogarae (formerly known as Pseudomonas fluorescens), AmrZ controls the expression of several genes involved in c-di-GMP metabolism, and its absence results in a pronounced reduction of intracellular c-di-GMP levels [2545]. Consistent with these observations, we examined whether AmrZ also influences the c-di-GMP pool in A. vinelandii. We used a previously characterized biosensor in which TurboRFP expression is controlled by three tandem c-di-GMP riboswitches (Bc3–Bc5) [60].

To validate the sensor in A. vinelandii, we integrated it into ΔmucG and ΔavGReg mutants, which exhibit ~fourfold higher and ~threefold lower c-di-GMP levels, respectively, relative to WT AEIV [20]. TurboRFP fluorescence was measured and normalized to total protein content. As shown in Fig. 4, the fluorescence of the TurboRFP protein increased, suggesting that c-di-GMP concentration in AEIV increased over the growth curve, peaking at 48 h. In the ΔmucG background (PDE-deficient), a similar trajectory was observed but with consistently higher values than WT, whereas ΔavGReg (DGC-deficient) showed a sustained reduction throughout growth. These results validate the biosensor in A. vinelandii, enabling detection of both increased and decreased c-di-GMP relative to WT (Fig. 4a).

*AmrZ regulates intracellular c-di-GMP levels. (a) c-di-GMP biosensor readout in WT, ΔamrZ, ΔavGReg and ΔmucG strains. Accumulation was monitored with the TurboRFP riboswitch biosensor (Bc3–Bc5). RFI=TurboRFP fluorescence/total protein. Bars show mean±sd from three independent experiments. Statistical significance (two-tailed Student’s t-test): **P<0.01; ***P<0.001; ***P<0.0001. (b) RT-qPCR of AVAEIV_RS11610, AVAEIV_RS18795 and AVAEIV_RS25725 mRNA in the indicated strains grown for 24 h in BS medium. Domain organizations of the encoded proteins are shown. (c) EMSA demonstrating AmrZ binding to the AVAEIV_RS11610 promoter region. DNA was incubated with increasing concentrations of His-AmrZ. A 251 bp fragment from AVAEIV_RS16855 served as a negative control. DNA was visualized by ethidium bromide staining. RFI, relative fluorescence intensity.

We then introduced the biosensor into the ΔamrZ mutant. TurboRFP fluorescence was markedly reduced across the growth curve, approaching the ΔavGReg levels, indicating that AmrZ is a key positive determinant of the intracellular c-di-GMP pool in A. vinelandii.

AmrZ modulates c-di-GMP via targets other than avGReg

Because ΔavGReg and ΔamrZ displayed comparably low c-di-GMP levels across the growth curve, we first tested whether AmrZ regulates avGReg. This was plausible given that AvGReg is the principal DGC during vegetative growth in A. vinelandii [20]. Expression of avGReg was monitored using a PavGReg-gusA transcriptional fusion. At 24 h, no significant difference was observed between the WT and the ΔamrZ mutant (Fig. S3), and a slight increase in expression was detected in ΔamrZ after 24 h. These results indicate that the reduced basal c-di-GMP in the absence of AmrZ is unlikely to be explained by a transcriptional control of avGReg and instead suggest that AmrZ modulates c-di-GMP homeostasis through additional DGCs and/or PDEs.

The AmrZ regulon in A. vinelandii

To identify genes directly or indirectly controlled by this transcriptional regulator, we performed RNA-seq-based transcriptional profiling of the WT strain AEIV and its ΔamrZ derivative (JG521) grown for 24 h in BS medium. To accurately define the AmrZ regulon, the AEIV genome was first sequenced (Methods); reads were then mapped to this reference (GenBank GCA_030506185.2).

Principal component analysis revealed a clear separation between WT and ΔamrZ transcriptomes (Fig. S4). Differential expression analysis with DESeq2 (threshold log_₂_-fold change ≥2 and Padj <0.01) identified 835 differentially expressed genes (DEGs): 498 upregulated and 337 downregulated in ΔamrZ relative to WT (Table S4 Supplementary Material 2). The predominance of genes showing increased mRNA accumulation in the mutant suggests a prominent role for AmrZ as a transcriptional repressor, a trend clearly reflected in the volcano plot (Fig. 5a). A search for AmrZ-binding sites in the regulatory region of the 835 DEGs was conducted, identifying 166 genes that are potentially directly regulated by AmrZ (Table S5).

Analysis of the transcriptomic data from the ΔamrZ mutant versus WT. (a) Volcano plot of differential expression. Vertical lines mark log₂-fold change ≥2, and the horizontal line marks Padj<0.01; red points denote genes meeting both criteria. (b) Global functional profile of the DEG in the △amrZ mutant. The per cent of each GO term is indicated. MFS, Major Faclitator Superfamily.

Gene Ontology (GO) term analysis was then performed on the proteins encoded by the DEGs (Fig. 5b). This analysis revealed enrichment of proteins associated with several functional categories, primarily electron transfer/redox processes, iron and siderophore acquisition, ABC transporters and transcription. Notably, a substantial fraction of the DEGs (17.4%) could not be assigned to a known function, implying that a significant portion of the AmrZ regulon comprises proteins of unknown or uncharacterized function.

Closer inspection of the affected genes highlighted specific biological processes, including ribosome biogenesis and modification, iron homeostasis, aliphatic sulphonate metabolism, nitrogen fixation and aromatic compound degradation. In particular, several genes involved in iron and siderophore acquisition were upregulated in the absence of AmrZ, including genes encoding FecR domain-containing proteins (AVAEIV_RS11105, AVAEIV_RS11520, AVAEIV_RS20470), FecI-related sigma factors (AVAEIV_RS09105, AVAEIV_RS20475, AVAEIV_RS23745) and the PvdS sigma factor (AVAEIV_RS12905). These data suggest that, as reported for other Pseudomonas species [2539], AmrZ plays a role in the regulation of iron homeostasis in A. vinelandii.

Transcriptomic analysis confirmed that AmrZ positively regulates algD (log_₂_-fold change, ΔamrZ versus WT=−3.79) and other alginate-biosynthetic genes, including algL (log_₂_-fold change, ΔamrZ versus WT=−2.33; Table S4). The dataset also revealed DEGs involved in c-di-GMP metabolism, consistent with the altered levels of this second messenger in the ΔamrZ mutant.

AmrZ controls genes for the metabolism of c-di-GMP

AVAEIV_RS11610 and AVAEIV_RS18795, both encoding proteins with conserved c-di-GMP synthesis (GGDEF) domains, showed reduced mRNA accumulation (log_₂_-fold change=−2.04 and −2.87, respectively) in the ΔamrZ mutant, whereas AVAEIV_RS25725, predicted to encode a hybrid DGC/PDE protein, was upregulated (log_₂_-fold change=2.38) according to the transcriptome analysis.

RT-qPCR using RNA from the WT, ΔamrZ and the complemented strain (ΔamrZ/amrZ^+^) confirmed downregulation of the DGC-encoding genes in the mutant and restoration upon complementation, indicating positive regulation by AmrZ. In contrast, the RNA-seq upregulation of AVAEIV_RS25725 was not validated by RT-qPCR: its transcript levels were slightly diminished in ΔamrZ (Fig. 4b).

MEME/FIMO analysis identified a putative AmrZ-binding site upstream of AVAEIV_RS11610 (Table S3 and Fig. S5). Consistent with direct regulation, EMSA showed specific binding of AmrZ to this promoter region (Fig. 4c), suggesting that AVAEIV_RS11610 is directly activated by AmrZ, whereas regulation of AVAEIV_RS18795 is likely indirect. These regulatory data provide a plausible explanation for the reduced c-di-GMP pool observed in the ΔamrZ mutant but do not rule out an AmrZ post-translational control of AvGReg activity, the main DGC under vegetative growth conditions.

AmrZ impacts swimming motility via c-di-GMP

In Pseudomonas species, AmrZ regulates swimming motility, largely through c-di-GMP-mediated pathways [384566]. Notably, the regulatory outcome of this phenotype is species dependent. Whereas AmrZ exerts a negative effect on swimming motility in P. aeruginosa and P. ogarae, it acts as a positive regulator in P. syringae and Pseudomonas stutzeri [67]. We therefore investigated whether AmrZ affects swimming motility in A. vinelandii.

Transcriptomic analysis revealed downregulation of several flagellar genes in the ΔamrZ mutant, suggesting a positive transcriptional role for AmrZ (Table S4). However, transmission electron microscopy showed that ΔamrZ cells remained flagellated (data not shown). Consistent with this observation, deletion of amrZ resulted in a significantly larger swimming halo compared with the parental strain, and complementation (ΔamrZ/amrZ^+^) restored the WT phenotype (Fig. 6). This enhanced motility correlates with the reduced intracellular c-di-GMP levels observed in the ΔamrZ mutant, indicating an overall negative influence of AmrZ on swimming motility.

*AmrZ negatively regulates swimming motility. Representative swimming motility of the indicated strains on soft agar after 24 h of incubation. The graph shows the quantification of swimming halo diameters. Statistical significance (two-tailed Student’s t-test) is indicated: P<0.05.

The increased swimming ability of the ΔamrZ strain was not attributable to alginate deficiency, as deletion of algA did not suppress the hypermotile phenotype (data not shown).

Discussion

Our findings demonstrate that the transcriptional regulator AmrZ is essential for alginate biosynthesis in A. vinelandii, acting mainly through direct activation of algD, encoding the key enzyme in this pathway. Loss of AmrZ completely abolished alginate production, a phenotype restored by genetic complementation, underscoring its central role. This function parallels that of AmrZ in P. aeruginosa [42]. The positive effect of AmrZ on algD transcriptional activation has also been demonstrated in P. syringae pv. tomato [40], suggesting a conserved regulatory mechanism among different genera of the Pseudomonadaceae.

EMSA results confirmed that AmrZ binds multiple sites within the algD promoter region. The four predicted AmrZ-binding sites are organized into two clusters separated by 110 bp and located upstream of the AlgU-dependent promoter. This arrangement suggests that, as in P. aeruginosa, AmrZ bridges multiple binding sites in PalgD, promoting the formation of higher order AmrZ–DNA complexes that stimulate transcription. We recently demonstrated that FleQ represses algD via two sites, but near the RpoS-dependent promoter [14], indicating that algD expression integrates positive regulation by AmrZ and negative regulation by FleQ. Given that algD transcription ultimately requires c-di-GMP, these opposing activities are likely coordinated by intracellular levels of this second messenger. Regulation of the fleQ gene by AmrZ was not detected under the conditions tested, based on the transcriptomic analysis of the △amrZ mutant. This observation was further supported by analysis of a PfleQ-gusA transcriptional fusion, which exhibited similar activity in both the WT and △amrZ backgrounds (Fig. S6).

Interestingly, AmrZ positively regulates its own expression in A. vinelandii, in contrast to the autorepression described in P. aeruginosa [64]. This positive feedback, dependent on AlgU, may stabilize AmrZ expression under conditions that favour alginate production.

Beyond alginate regulation, our transcriptomic and biochemical data demonstrate that AmrZ exerts broad control over c-di-GMP metabolism. Notably, the principal vegetative DGC, AvGReg, was not transcriptionally affected by AmrZ: PavGReg activity was not diminished, yet cellular c-di-GMP was markedly reduced in ΔamrZ. This suggests that, in the ΔamrZ background, AvGReg activity may be attenuated post-transcriptionally, potentially via signals sensed by its globin domain [21], although it remains to be tested. We identified two additional DGC genes that are positively regulated by AmrZ; their decreased expression in ΔamrZ likely contributes to the diminished c-di-GMP pool. This phenotype resembles that of P. ogarae, where AmrZ promotes c-di-GMP accumulation by regulating the expression of multiple genes involved in c-di-GMP metabolism [2545]. In A. vinelandii, low c-di-GMP in the ΔamrZ strain also helps explain the complete absence of alginate production, given the well-established requirement of this messenger for polysaccharide synthesis [20].

Phenotypically, loss of AmrZ in A. vinelandii resulted in increased swimming motility, a feature commonly associated with reduced c-di-GMP signalling and reminiscent of phenotypes described in P. ogarae and P. aeruginosa [3445]. In A. vinelandii, however, RNA-seq data suggest a positive effect of AmrZ on the transcription of several flagellar genes, some of which harbour putative AmrZ-binding sites (Table S3). Despite this transcriptional trend, the △amrZ mutant exhibited enhanced swimming motility. The basis for these apparently contrasting effects remains unclear and may reflect post-transcriptional regulation linked to reduced c-di-GMP levels.

The consequences of AmrZ loss extended beyond polysaccharide, motility or c-di-GMP regulation. RNA-seq analysis revealed over 800 DEGs, underscoring the pleiotropic role of this regulator. Affected functional categories included iron and siderophore acquisition and ABC transporters, consistent with observations reported for other Pseudomonas species [2539]. Although we validated several direct AmrZ targets in A. vinelandii using binding motifs defined in P. aeruginosa, this approach has inherent limitations. The large number of predicted AmrZ targets in A. vinelandii (Tables S3 and S5) indicates species-specific AmrZ recognition sequences, highlighting the need for dedicated binding-site analyses in future studies.

Collectively, these findings reveal that AmrZ integrates multiple regulatory layers to coordinate polysaccharide synthesis, motility and global cellular functions through modulation of c-di-GMP homeostasis.

Supplementary material

10.1099/mic.0.001686Supplementary Material 1.

10.1099/mic.0.001686Supplementary Material 2.