Impact of Na⁺-Translocating NADH:Quinone Oxidoreductase on Iron Uptake and nqrM Expression in Vibrio cholerae

Investigating strategies of iron acquisition, storage, and delivery in Vibrio cholerae is a prerequisite to understand how this pathogen thrives in hostile, iron-limited environments such as the human host. In addition to highlighting the maturation of the respiratory complex NQR, this study points out the influence of NQR on iron metabolism, thereby making it a potential drug target for antibiotics.

ABSTRACT

The Na⁺ ion-translocating NADH:quinone oxidoreductase (NQR) from Vibrio cholerae is a membrane-bound respiratory enzyme which harbors flavins and Fe-S clusters as redox centers. The NQR is the main producer of the sodium motive force (SMF) and drives energy-dissipating processes such as flagellar rotation, substrate uptake, ATP synthesis, and cation-proton antiport. The NQR requires for its maturation, in addition to the six structural genes nqrABCDEF, a flavin attachment gene, apbE, and the nqrM gene, presumably encoding a Fe delivery protein. We here describe growth studies and quantitative real-time PCR for the V. cholerae O395N1 wild-type (wt) strain and its mutant Δnqr and ΔubiC strains, impaired in respiration. In a comparative proteome analysis, FeoB, the membrane subunit of the uptake system for Fe²⁺ (Feo), was increased in V. cholerae Δnqr. In this study, the upregulation was confirmed on the mRNA level and resulted in improved growth rates of V. cholerae Δnqr with Fe²⁺ as an iron source. We studied the expression of feoB on other respiratory enzyme deletion mutants such as the ΔubiC mutant to determine whether iron transport is specific to the absence of NQR resulting from impaired respiration. We show that the nqr operon comprises, in addition to the structural nqrABCDEF genes, the downstream apbE and nqrM genes on the same operon and demonstrate induction of the nqr operon by iron in V. cholerae wt. In contrast, expression of the nqrM gene in V. cholerae Δnqr is repressed by iron. The lack of functional NQR has a strong impact on iron homeostasis in V. cholerae and demonstrates that central respiratory metabolism is interwoven with iron uptake and regulation.

IMPORTANCE Investigating strategies of iron acquisition, storage, and delivery in Vibrio cholerae is a prerequisite to understand how this pathogen thrives in hostile, iron-limited environments such as the human host. In addition to highlighting the maturation of the respiratory complex NQR, this study points out the influence of NQR on iron metabolism, thereby making it a potential drug target for antibiotics.

INTRODUCTION

Iron is an essential nutrient that pathogenic bacteria acquire to support the function of a number of enzymes crucial for survival, such as ribonucleotide reductase for the synthesis of DNA precursors or cofactors for respiratory and tricarboxylic acid (TCA) cycle enzymes for energy conservation (1). Iron is the fourth-most abundant element in the earth’s crust, but it is poorly soluble at physiological pH in the presence of oxygen and therefore not readily bioavailable. Since the concentration of ferric iron (Fe³⁺) in tissues is extremely low (10⁻¹⁸ M) (2), bacteria use various strategies to acquire Fe³⁺ to compete with the host’s iron-scavenging systems. The main strategies for iron acquisition include direct extraction of Fe³⁺ from specific iron-containing complexes synthesized by their hosts, such as lactoferrin, transferrin, hemoglobin, or heme (3, 4), and/or production of siderophores, small Fe³⁺-chelating molecules with high affinity and selectivity for Fe³⁺ ions (5, 6). Iron can also be detrimental to cells due to the reactive oxygen species produced in the presence of this element (7). Thus, the uptake and intracellular processing of iron are tightly regulated. Iron can combine with elemental sulfur to form iron-sulfur (Fe-S) centers bound to proteins (8). In particular, Fe-S clusters are among the earliest iron cofactors on Earth (9). These clusters confer a number of functions to proteins, ranging from electron transfer to catalysis and regulatory processes. However, both iron and sulfide are highly reactive and toxic in vivo, which points out the essentiality of the Fe-S cluster biosynthetic pathways in cells to be well coordinated (8).

Genomic analyses revealed that the number and types of operons coding for Fe-S assembly systems vary from one microorganism to another (10). However, all of these systems share the same basic principles and similar key molecular actors, which work together to produce Fe-S clusters. Vibrio cholerae, the causative agent of the severe diarrheal disease cholera, has multiple iron transport systems that may allow optimal acquisition of this essential element in each of its diverse environments (11, 12). Once inside the cytoplasm, iron is incorporated into the various Fe-S center-containing proteins of V. cholerae. However, the pathways for Fe-S cluster assembly in V. cholerae have not been studied yet.

The Na⁺-translocating NADH:quinone oxidoreductase (NQR) is a redox-driven sodium pump operating in the respiratory chain of V. cholerae. NQR catalyzes electron transfer from NADH to ubiquinone, coupled with Na⁺ translocation across the membrane (13). NQR is a six-subunit membrane protein complex encoded by the consecutive structural nqrABCDEF genes. The peripheral subunit NqrF catalyzes NADH oxidation and hosts one flavin adenine dinucleotide (FAD) and a 2Fe-2S cluster as cofactors (14). The membrane-bound subunits NqrD and NqrE ligate an Fe center within the membrane part of the NQR complex (13). Subunits NqrB and NqrC each contain one covalently attached flavin mononucleotide (FMN), which requires ApbE, a flavin insertase encoded by apbE, the gene immediately downstream of nqrF (15). In 2015, Kostyrko et al., identified another open reading frame downstream of apbE which, like apbE, was essential for production of functional NQR in a heterologous host (16). It was therefore designated nqrM (maturation of NQR) and was proposed to code for a protein participating in Fe insertion into NQR (16).

A V. cholerae mutant strain lacking the structural nqrABCDEF genes is impaired in the generation of a transmembrane potential (17), but it is able to grow by shifting metabolism from respiratory toward fermentative ATP formation (18). A comparative, quantitative proteome study of the V. cholerae reference (wild type [wt]) and the nqr deletion strains revealed a 2.7-fold increase in abundance of the predicted Fe²⁺ transporter (FeoB) in the nqr mutant strain (18). Here, we study the impact of iron on growth and expression of feoB in the V. cholerae reference strain and nqr deletion strain and on a mutant defective in ubiquinone biosynthesis, the ΔubiC strain. We show that the predicted NQR assembly genes are part of the nqr operon and that their expression is differentially regulated by iron in the V. cholerae wt and nqr deletion strains. Our findings indicate the presence of an operator region between the promoter and the structural nqr genes, which is regulated in an iron-dependent manner.

RESULTS AND DISCUSSION

Proteins participating in iron metabolism in V. cholerae wt and Δnqr strains.

The V. cholerae genome has been completely sequenced (19), and iron transport systems are encoded on both of its chromosomes. In our recent proteome study (18), we compared the abundance of proteins in V. cholerae O395N1 (reference strain) with that in its mutant Δnqr derivative strain, where the six structural genes nqrABCDEF of the operon encoding the membrane-bound NQR respiratory complex are deleted (20). About 2,000 proteins were identified in each strain, and most of the iron transporters were present in both the wt and Δnqr strain in similar abundances. Consistent with the fundamental role that iron plays in microbial growth and its relative scarcity, microbes have developed an arsenal of techniques to procure it. The different energy-dependent iron transporters in the inner and outer membranes of V. cholerae facilitate the uptake of this essential nutrient in different forms: in a free form as Fe³⁺/Fe²⁺ (FbpABCD/FeoABC), as bound by siderophores (FhuABCD, ViuAPCDG, and IrgA/VctAPCDG), or in complex as heme (HasR, HutR, and HutABCD) (11). Once inside the cytoplasm, the iron is liberated from the complex, and ferric iron is reduced to the ferrous form by the ferric reductases and then either stored in the iron storage protein bacterioferritin or transported by different iron carriers (such as IscX and YggX) (21) toward the Fe-S biogenesis system to be incorporated as the Fe-S cofactors in different proteins of the respiratory and metabolic pathways, including NQR (Fig. 1). A proteomic study revealed that most of the iron homeostasis proteins were present in similar abundances in both strains. These included proteins predicted to participate in the assembly of Fe-S clusters, in the transport and storage of iron, and in the binding and transport of heme (Table 1). However, interestingly, some of these proteins were differentially regulated. While the ferric siderophore transporter (Fhu) and the ferrous iron transporter (Feo) involved in iron uptake were increased in abundance, the ferric iron transporter (Fbp) was decreased in abundance in the nqr deletion mutant (Fig. 1). Figure 1 shows all the iron transport proteins that were identified by the proteomic analysis as well as those that are present in the genome of V. cholerae but could not be identified by the proteomic study. It is important to note that peptides from hydrophobic, membrane-bound proteins were only rarely detected despite the use of SDS as the extraction detergent. A large number of identified proteins could not be assigned to a specific function by KEGG. In these cases, we searched for homologous sequences in genomes from Gram-negative bacteria with known functions using the BLAST tool (22).

FIG 1.

Schematic illustration of proteins participating in iron uptake and metabolism in Vibrio cholerae. The cellular compartments (C, cytoplasm; iM, inner membrane; P, periplasm; oM, outer membrane) are indicated. The iron and heme transporters identified in the proteome study are indicated in yellow, the putative Fe-S assembly proteins are in green, and subunits of NQR are in peach. Proteins which were not identified by proteome analyses but are present in the genome are shown in black and white stripes. The boxes represent the transporter systems. Bold lines, upregulated systems; dashed lines, downregulated systems; thin lines, transporter systems in similar abundance in both wt and Δnqr strains. The box with dotted lines represents the proteins encoded by the NQR operon.

TABLE 1.

Proteins predicted to participate in iron metabolism identified in the proteome analysis of Vibrio cholerae O395N1

Protein category and IDa	Locusb	Genec	Functiond
Fe-S cluster assembly
A0A0H3AH94	VC0395_A0276	iscR	Transcriptional regulator
A0A0H3AHI2	VC0395_A0279	iscA	A-type Fe-S cluster carrier
A5F3G4	VC0395_A0277	iscS	Cysteine desulfurase
A0A0H3AJF7	VC0395_A0278	iscU	Fe-S assembly scaffold protein
A5F3G8	VC0395_A0281	hscA	Chaperone protein
A5F3G7	VC0395_A0280	hscB	Chaperone protein
A0A0H3AI85	VC0395_A0282	fdx	Ferredoxin
A0A0H3AL23	VC0395_A0283	iscX	Fe-S assembly protein
A5F4R9	VC0395_A2293	nfuA	Fe-S biogenesis protein
A0A0H3AJA8	VC0395_A0555	mrp (apbC)	Fe-S cluster carrier protein
A0A0H3AFC3	VC0395_A1631	grxD	Glutaredoxin
A5F947	VC395_0644	erpA	Fe-S cluster insertion protein
A5F9G8	VC0395_A0002	yggX	Putative iron trafficking protein
A0A0H3AG44	VC0395_A1896	csdA	Cysteine desulfurase
A0A0H3AMG0	VC0395_A1898	csdL	Sulfur acceptor protein
A0A0H3AJK4	VC0395_A1897	csdE	Sulfur acceptor protein
A5F6G4	VC0395_A1690/VC395_2220	fur	Iron response regulator
A0A0H3ALU6	VC0395_A2411	ryhB	Fur regulated, small mRNA
Iron binding, storage, and transport
A0A0H3AIK0	VC0395_A2580	fhuA	Outer membrane transporter of Fe3+-ferrichrome
A0A0H3AJ03	VC0395_A2582	fhuD	Periplasmic Fe3+-ferrichrome/vibriobactin binding protein
A0A0H3AIS3	VC0395_A2583	fhuB	Inner membrane Fe3+-ferrichrome ABC transporter
A0A0H3AKG5	VC0395_A2581	fhuC (+4.1)	Nucleotide binding domain of Fe3+-ferrichrome ABC transporter
A5F661	VC0395_A1803	viuA	Outer membrane transporter of Fe3+-vibriobactin
A0A0H3AHM4	VC0395_A0305	viuP	Periplasmic Fe3+-vibriobactin/enterobactin binding protein
A0A0H3AII9	VC0395_A0306	viuD	Inner membrane Fe3+-vibriobactin ABC transporter
A0A0H3AGN1	VC0395_A0307	viuG	Inner membrane Fe3+-vibriobactin ABC transporter
A0A0H3ANY0	VC0395_A0308	viuC	Nucleotide binding domain of Fe3+-vibriobactin ABC transporter
A5F660	VC0395_A1802	viuB	Fe3+-ferrichrome interacting protein (as a ferric reductase)
A0A0H3ADQ9	VC0395_0996	vctA	Outer membrane transporter of Fe3+-enterobactin
A0A0H3ACL3	VC0395_1001	vctP	Periplasmic Fe3+-vibriobactin/enterobactin binding protein
A0A0H3AFD8	VC0395_1000	vctD	Inner membrane Fe3+-vibriobactin/enterobactin ABC transporter
A0A0H3AFY8	VC0395_0999	vctG	Inner membrane Fe3+-vibriobactin/enterobactin ABC transporter
A0A0H3AEA2	VC0395_0998	vctC	Nucleotide binding domain of Fe3+-enterobactin ABC transporter
A5F9G0	VC0395_A0028	irgA	Outer membrane transporter of Fe3+-enterobactin
Fe3+ transporter (Fbp system)
A0A0H3AFK0	VC395_0625	fbpA (−5.6)	Periplasmic Fe3+-binding protein
ACP08644	VC395_0626	fbpB	Inner membrane Fe3+-ABC transporter
A0A0H3ADT9	VC0395_0627	fbpC	Nucleotide binding domain of Fe3+-ABC transporter
Fe2+ transporter (Feo system)
A0A0H3AM36	VC0395_A1665	feoA	Nucleotide binding domain of Fe2+ transporter
A0A0H3AFT4	VC0395_A1664	feoB (+2.7)	Inner membrane Fe2+-transporter
A0A0H3AJN6	VC0395_A1663	feoC	Nucleotide binding domain of Fe2+ transporter
A0A0H3AID2	VC0395_A2661	vciB	Mediates iron reduction, Fe3+ to Fe2+
A0A0H3AID4	VC0395_A2776	bfr	Bacterioferritin, iron storage protein
Heme storage and transport (Hut system)
A0A0H3AGI5	VC0395_0519	hutA	Outer membrane heme transporter
A0A0H3AEH2	VC0395_0074	hutR	Outer membrane heme transporter
A0A0H3ADZ8	VC0395_0566	hasR	Outer membrane heme transporter
A0A0H3AE70	VC0395_0324	hutB	Periplasmic heme binding protein
A0A0H3AFH2	VC0395_0323	hutC	Inner membrane heme ABC transporter
A0A0H3ADP8	VC0395_0322	hutD	Nucleotide binding domain of heme ABC transporter
A0A0H3AGE3	VC0395_0330	hutZ	Heme degradation protein
A5F4I9	VC0395_A2396	cyaY	Heme binding protein for iron transport
A0A0H3AGB6	VC0395_0328	hutW	Heme utilization protein

^aV. cholerae O395N1 protein identifiers (IDs) refer to the UniProt or NCBI database accession number.

^bFrom the V. cholerae O395N1 classical biotype strain.

^cValues in parentheses represent corresponding fold changes. A positive value indicates an increase, and a negative value indicates a decrease in relative protein abundance.

^dFunctional assignment is based on UniProt and KEGG database annotation and is refined by NCBI-based BLAST results.

Deletion of nqr does not affect the regulation of proteins involved in Fe-S biogenesis pathways.

There are distinct Fe-S biosynthesis pathways in bacteria which are utilized under different environmental conditions via differential gene regulation (23). Three different systems involved in the biogenesis of bacterial Fe-S proteins have been identified: the nitrogen fixation (Nif) system, which is responsible for the specific maturation of nitrogenase in azototrophic bacteria (24, 25), and the iron-sulfur cluster (Isc) and sulfur mobilization (Suf) systems, which generate housekeeping Fe-S proteins under normal and stress conditions, respectively (10, 26). The Isc system appears to be the housekeeping Fe-S cluster assembly pathway, while the Suf system appears to be important for the adaptation of Fe-S cluster formation under iron starvation and oxidative-stress conditions (23, 27, 28). E. coli harbors the Isc, Suf, and Csd-Suf hybrid systems for Fe-S assembly biogenesis (21). Homologous proteins predicted to participate in the biosynthesis of Fe-S clusters were present in similar abundances in the V. cholerae wt and Δnqr strains, indicating constitutive expression of these essential proteins in the two strains (Table 1).

The genome predicts that V. cholerae possesses a single Isc system for assembly of Fe-S clusters. This situation is reminiscent of that in Pseudomonas aeruginosa and Acinetobacter baumannii, which also possess only the Isc system (21, 29, 30). This absence of an additional system for Fe-S assembly could lead to increased susceptibility of V. cholerae to oxidative stress or iron-limiting conditions. Even though V. cholerae does not encode an additional Fe-S cluster assembly system, it possesses iron-responsive regulators, such as Fur, RyhB, and OxyR, that are responsible for tight Fe-S homeostasis regulation. The Fur-regulated, small noncoding RNA, RyhB, which was identified to maintain iron homeostasis in a V. cholerae El Tor biotype strain by microarray analysis (31), was also identified in genomic analysis of the V. cholerae O395N1 classical biotype strain studied here. Also, other putative Fe-S assembly proteins such as CsdA, CsdL, NfuA, Mrp, and GrxD and one homolog of SufE identified in the proteome of V. cholerae point toward Fe-S assembly via different routes which may replace part of the Isc machinery under stress conditions. In Escherichia coli, a potential CsdAE-SufBCD hybrid system has been reported; however, this hybrid system is not found in V. cholerae, which lacks the suf operon. Interestingly, a sufE homolog (VC0395_A1897) in V. cholerae is located in the csd operon, between the csdA (VC0395_A1896) and csdL (VC0395_A1898) genes, rather than on an independent sufABCDSE operon as observed in E. coli. V. cholerae SufE exhibits only 35% sequence identity with SufE of E. coli but 49% identity with the CsdE sulfur acceptor protein, indicating that it is wrongly annotated and that it should rather be annotated as CsdE, the sulfur acceptor protein.

The high level of sequence identity of the V. cholerae ISC components with the well-characterized E. coli proteins allowed the assignment of putative functions to them (Table 1) (21). IscR which is encoded by the first gene of the iscRSUA-hscBA-fdx operon, is a 2Fe-2S cluster-containing transcriptional regulator that has been reported to regulate both Isc and Suf systems, the cysteine desulfurase (IscS), a scaffold protein for Fe-S cluster assembly (IscU), an A-type Fe-S cluster carrier (IscA) which is proposed to act as a conduit between IscU and empty apoenzymes in the cell, the HscAB ATPase that aids transfer of completed clusters, and a ferredoxin (fdx) with a putative role in electron transfer during cluster biogenesis (32, 33).

The absence of a second assembly pathway such as the Suf system indicates that in V. cholerae, the assembly of the Fe-S centers of the proteins including the membrane-bound NQR respiratory complex ultimately depends on Fe-S delivery via the Isc system (Fig. 1). While the cytosolic Nqr subunits and ApbE were detected in the proteome of wt V. cholerae (18, 19), the putative Fe-S maturation factor NqrM, encoded by the gene located downstream of the nqr operon, was not detected. NqrM is a rather small protein, with only 77 amino acids, which does not have many tryptic cleavage sites. Therefore, only a small number of peptides are produced upon a tryptic digest, which likely makes identification of NqrM by mass spectrometry difficult. NqrM expression was therefore confirmed on the mRNA level, as described below.

Effect of iron limitation on growth of V. cholerae wild type, the nqr deletion strain, and a mutant defective in ubiquinone biosynthesis.

The V. cholerae cells used for the proteome analysis were grown in a glucose minimal medium without added transition metals (18). We first asked if these cells had suffered from iron limitation, which resembles the situation in the human host when iron is restricted. Analysis by inductively coupled plasma mass spectrometry (ICP-MS) showed that the minimal medium contained 11.8 μg/liter of iron (0.2 μM), 0.9 μg/liter of copper (0.014 μM), and less than 0.5 μg/liter manganese (0.009 μM).

We next examined if this residual iron content of the glucose minimal medium (0.2 μM) was limiting for growth of V. cholerae wt and the Δnqr and ΔubiC mutants. In the ΔubiC mutant, the ubiC gene (VC0395_A2420), which encodes the chorismate pyruvate-lyase, a ubiquinone precursor synthesis enzyme (34), was deleted. Ubiquinone is predominantly used in aerobic respiration in bacteria (34, 35). The small intestine might provide both microaerobic and -anaerobic environments (36), so the ubiquinone-dependent respiratory pathway might be relevant during infections. Iron in the M9 minimal medium was limiting for the wild type and the two mutants, while supplementation with ferric chloride (FeCl₃) (260 μM) stimulated growth of all of the strains (Fig. 2). Without iron addition, the strains reached only about 60% of the final cell density compared to that with iron supplementation. This effect can be attributed to an increased energetic burden of cells in the medium with low iron content due to synthesis, secretion, and (re)uptake of siderophores. With iron-supplemented medium, we observed a steep fall in the cell density after maximum optical density (OD) values were reached, indicating cell lysis. Analysis of glucose concentration after 10 h and 35 h under all growth conditions and with every strain revealed that around 64% of the glucose added at the beginning (10 mM) was consumed after 10 h whereas 87% of the glucose was consumed after 10 h in medium supplemented with iron. Under both iron-limited and iron-supplemented conditions, glucose was completely consumed after 35 h. This indicates that V. cholerae undergoes an adaptive process for complete glucose consumption which is preceded by a growth arrest and partial cell lysis. In addition to NQR, V. cholerae possesses a so-called alternative, noncoupled NADH:quinone oxidoreductase (NDH-2) but does not have NDH-I, the homolog of the proton-translocating complex I (37). As V. cholerae does not encode a complex I homolog, NQR functions as the main respiratory enzyme (17). Therefore, loss of NQR will impair ATP production, as demonstrated by the decreased growth yield of the Δnqr strain compared to that of V. cholerae wt. A similar decrease in growth yield was observed with a ΔubiC deletion strain, in accord with a central role of ubiquinone for energy conservation in V. cholerae (Fig. 2).

FIG 2.

Growth of V. cholerae wt and mutant strains impaired in respiration under iron-limited and iron-replete conditions. The V. cholerae wild-type, Δnqr, and ΔubiC strains were grown without iron addition in M9 medium or supplemented with 260 μM FeCl₃. Growth was monitored at 595 nm. Averages and standard deviations from three biological replicates are shown.

We then asked if the oxidation state of iron (Fe²⁺ and Fe³⁺) also affected the growth pattern of the V. cholerae wt and Δnqr strains, using ferrous sulfate (FeSO₄) or ferric chloride (FeCl₃) for supplementation, with an emphasis on the first 20 h of growth. Cells grown in M9 minimal medium showed a significant increase in the growth yield upon addition of Fe^2+/3+ (Fig. 3). While the final growth yield was not affected by the oxidation state of iron added to the medium, there were marked differences in growth rates observed with the two strains depending on the iron cation. It was observed that addition of ferrous sulfate led to a 1.5-fold increase in growth rate in the Δnqr strain compared to growth of the wt during 1 to 5 h of growth. This was followed by partial cell lysis in the Δnqr strain or growth arrest in the wt, which clearly was not caused by glucose limitation and requires further investigation.

FIG 3.

Effect of the iron oxidation state on growth of V. cholerae wt and the Δnqr deletion strain. The V. cholerae wild-type strain and the Δnqr strain were grown without iron addition in M9 medium (squares) or supplemented with either 260 μM FeCl₃ (triangles) or FeSO₄ (circles). Growth was monitored at 595 nm. Averages and standard deviations from two biological replicates are shown.

The increased growth in the Δnqr mutant in the presence of Fe²⁺ is in line with the observations from the proteome study and from quantitative PCR (qPCR) studies (see below), which show that in V. cholerae Δnqr, the Feo system for Fe²⁺ uptake is upregulated compared to that in the wt (18). In E. coli, the Feo system is the preferred iron uptake system during anaerobic growth (38). Deletion of NQR in V. cholerae leads to a metabolic shift toward fermentative reactions under oxic conditions which is accompanied by upregulation of Feo in the Δnqr strain (18). Our finding of an increased growth rate of V. cholerae Δnqr in the presence of ferrous sulfate compared to that of the wt further substantiates the proteome study and is in full accordance with increased FeoB abundance in the Δnqr strain. Note that final growth yields observed with Fe²⁺ and Fe³⁺ were comparable. We expect continuous conversion of Fe²⁺ to Fe³⁺ (39) by chemical oxidation during prolonged incubation, suggesting that toward the end of growth, Fe³⁺ was the dominant species under all conditions.

Increased expression of feoB in V. cholerae strains impaired in respiration under iron-depleted conditions.

In the proteome study, we observed a 2.7-fold increase in abundance of FeoB in V. cholerae Δnqr compared to that in the wt (18) in an iron-limited growth medium. To confirm this upregulation on the level of mRNA and to study the influence of iron on feoB expression, we performed qPCR on feoB with cDNA synthesized from mRNA isolated from the wt and Δnqr strains grown in glucose minimal medium with and without iron (260 μM FeCl₃) addition. In addition to the influence of iron on feoB expression in the Δnqr strain, we also analyzed feoB expression in the ΔubiC strain (Fig. 4). The ubiC gene is responsible for the synthesis of ubiquinone, which is the electron acceptor of the major respiratory complex NQR and shuttles the electrons to the downstream respiratory enzymes during aerobic respiration (35). Expression levels of feoB were compared to the expression level of recA, the control gene whose expression was not influenced by iron, as indicated by threshold cycle (C_T) values observed with recA in the absence and presence of iron (data not shown). Comparing the relative expression levels in a given strain, we observed that the expression of feoB was increased by approximately 2-fold in the Δnqr strain and around 3-fold in ΔubiC deletion mutants under iron limitation compared to expression in the wild type (Fig. 4). The increase in feoB mRNA under iron-limiting, aerobic conditions in the Δnqr and ΔubiC strains studied here might reflect an adaptation to some pseudoanaerobic metabolism associated with a metabolic shift toward fermentative respiration even under oxic conditions (18). Deletion of ubiC also led to increased expression of feoB, similar to the situation observed in the nqr deletion strain. This is in accord with impaired NQR activity also in the V. cholerae ΔubiC strain since NQR requires ubiquinone-8 as a bound cofactor (40). The confidence intervals (indicated by the error bars in Fig. 4) for both of these deletion mutants confirm that feoB expression was consistently higher than that in the wild type (Fig. 4). We conclude that increased feoB expression in V. cholerae Δnqr is associated with impaired overall respiratory activity.

FIG 4.

Expression of feoB in V. cholerae wt and deletion strains impaired in respiration under iron-limiting conditions. A comparison of V. cholerae strains and the relative fold change of feoB expression under conditions of iron limitation in the Δnqr and ΔubiC strains compared to that in the wt are shown. The expression of the housekeeping gene recA was the reference. Representative results (mean values) from three biological replicates each measured in three technical replicates are presented. The mean values are reported here with error bars representing the confidence intervals for the fold changes. The numbers over the columns are P values obtained from a comparison of the respective ΔC_T values using Student's t test.

As a further explanation, we should consider a direct link of increased feoB expression in the nqr deletion strain and the absence of functional NQR in this strain. In the presence of NQR, VciB was proposed by Peng and Payne in 2017 (41) to stimulate the reduction of Fe³⁺ to Fe²⁺, which subsequently is taken up by the Feo system. It was suggested that Fe³⁺ acts as an acceptor for electrons deriving from the respiratory chain under participation of the NQR. The ferrous iron thus produced is a substrate for the Feo transport system. An impaired respiratory activity in the nqr deletion strain may thus lead to reduced Fe²⁺ levels in the periplasmic space, which could be counteracted in V. cholerae Δnqr by upregulation of the Feo transport system. Upregulation of feoB expression in the ΔubiC strain suggests that ubiquinone shuttles redox equivalents between NQR and VciB to reduce the Fe³⁺ to Fe²⁺. Hence, the absence of ubiquinone would lead to inactive NQR and consequently reduced Fe²⁺ levels in the periplasmic space. As observed in the Δnqr strain, the upregulation of the Feo transport system would counteract these low Fe²⁺ levels in the periplasm of V. cholerae ΔubiC.

nqrM is coexpressed with the nqr genes.

Kostyrko et al. showed in 2015 (16) that a functional sodium NQR can be expressed in a heterologous host only in the presence of the nqrM gene. The NqrM protein was identified in the V. cholerae O1 biovar El Tor N16961 strain in a proteome study that investigated the blue light-induced photooxidative stress response (42). However, NqrM was not identified in our proteome study of V. cholerae O395N1 (18), and as a consequence, its expression required independent confirmation. In bacterial genomes, genes of functionally related polypeptides are often clustered together (43). Thus, the six genes encoding the NQR subunits are commonly organized into a single nqrABCDEF gene cluster (44, 45) followed by apbE, which codes for an enzyme that inserts the flavin cofactor into the NqrB and NqrC subunits, and by nqrM (15). Using the operon prediction database OperonDB (46, 47), we found that nqrM is located in the same transcriptional orientation as apbE in 23 bacterial genomes. The probability that apbE and nqrM are in the same operon amounts to a confidence value of 63. Organisms carrying apbE and nqrM in the same direction in their genome are, among others, Vibrio parahaemolyticus, Neisseria gonorrhoeae, and the gastroenteritis pathogen Aeromonas hydrophila. Most species of Vibrionaceae possess two slightly different putative nqrM genes, one (nqrM₁) adjacent to the nqr operon and the other (nqrM₂) usually adjacent to a gene whose product is annotated as a dinitrogenase iron-molybdenum cofactor domain protein. Yet V. cholerae possesses only one copy of nqrM adjacent to the structural nqr genes and apbE. The intergenic regions between nqrF and apbE and between apbE and nqrM have lengths of 47 and 17 bp, respectively (Fig. 5A).

FIG 5.

Organization of the NQR gene cluster. (A) Schematic illustration and arrangement of the nqrF, apbE, and nqrM genes (true to scale) and the position of primer binding sites. (B) Agarose gel (2%) showing the PCR products of cDNA with different primer pairs in the wt and Δnqr strains. The three primer pairs gave the PCR products nqrF-apbE (1,090 bp; lane 1), apbE-nqrM (241 bp; lane 2), and nqrF-nqrM (2,200 bp; lane 3).

To examine coexpression of nqrM with the structural nqr genes, the genes were amplified from cDNA with different primers. cDNA from the wt and Δnqr mutant strains was used. Cells were grown in minimal medium without or with added FeCl₃ (260 μM) and harvested at an OD at 600 nm (OD₆₀₀) of 0.4, corresponding to early exponential phase. Different sets of primer pairs for amplifications were designed: nqrF (for the nqrF gene, VC0395_A1879) forward and apbE (for the apbE gene, VC0395_A1878) reverse primers, apbE forward and nqrM (for the nqrM gene, VC0395_A1877) reverse primers, and nqrF forward and nqrM reverse primers (Fig. 5A). In the case of coexpression of the structural nqr genes with apbE and nqrM, the mRNA of V. cholerae would contain all the three genes, nqrF, apbE and nqrM, on the same transcript, which upon amplification with the above primers would result in three expected PCR products. With V. cholerae wt cDNA, we obtained three PCR products of sizes corresponding to the pairs nqrF-apbE, apbE-nqrM, and nqrF-nqrM (Fig. 5B). As expected, only one PCR product (apbE-nqrM) was obtained with cDNA from V. cholerae Δnqr which lacks the structural nqrABCDEF genes. We conclude that nqrM in V. cholerae O395N1 colocalizes and coexpresses with the apbE and the structural nqr genes.

Increased expression of the nqr operon under iron-replete conditions in V. cholerae wild type.

We studied the influence of iron on nqrM expression in the V. cholerae wt and Δnqr strains grown in glucose minimal medium without or with iron (260 μM FeCl₃) addition. Again, expression levels of nqrM were compared to the expression level of recA. For recA and nqrM, threshold cycles were reached between cycle 20 and 26 of the qPCR (see Fig. S1 in the supplemental material). Interestingly, there was no significant difference between the housekeeping gene (recA) and nqrM in the amplification chart, indicating that nqrM is approximately expressed in amounts similar to those of the housekeeping gene. V. cholerae wt showed a 1.9-fold increase in the expression of nqrM in the presence of 260 μM Fe³⁺ compared to the level under the iron-limited condition. This response enables V. cholerae to make optimal use of the iron source to increase overall respiratory activity. In marked contrast, V. cholerae Δnqr showed a 0.71-fold decrease in the nqrM expression upon iron addition (Fig. 6). It should be noted that V. cholerae Δnqr contains the promoter binding region (−35 and −10 region) and the first 399 bp of nqrA (20), yet expression of the remaining apbE and nqrM genes in V. cholerae Δnqr is repressed by iron. This suggests that a functional NQR is required to promote the iron-stimulated expression of the nqr operon in V. cholerae wt.

FIG 6.

Expression of nqrM in V. cholerae wt and the Δnqr strain under iron-limiting and iron-replete conditions. Data represent the change in relative fold expression of nqrM in a given V. cholerae strain (wt or Δnqr strain) under iron-replete conditions (260 μM Fe³⁺) compared to that under iron limitation (0.2 μM iron). The expression of the housekeeping gene recA was used as a reference. Representative results (mean values) from three biological replicates each measured in three technical replicates are presented. The mean values are reported here with error bars representing the confidence intervals for the fold changes. The numbers over the columns are P values obtained from a comparison of the respective ΔC_T values using Student's t test.

Conclusion.

The V. cholerae strains impaired in respiration show upregulation of the Fe²⁺ transport system, Feo, under iron limitation. V. cholerae strains impaired in respiration show abundances of proteins participating in Fe-S cluster biogenesis similar to those in the wild-type strain. The putative Fe-S biogenesis protein NqrM coded by nqrM is cotranscribed with the structural nqr genes, forming a single operon comprising also the flavin insertion gene, apbE, required for NqrB and NqrC maturation. The nqr operon, which includes apbE and nqrM, is upregulated by iron. This suggests the presence of an operator region between the promoter and the structural nqr genes, where a thus far unknown regulator might bind in an iron-responsive manner.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Vibrio cholerae O395N1 (48) was used as the reference strain (designated the wild type [wt]) and was compared to its derivative Δnqr strain (20) lacking the six nqrABCDEF genes and the ΔubiC strain lacking the gene for chorismate-pyruvate lyase. The ΔubiC mutant strains were constructed by homologous recombination (20). The ΔubiC mutant (with a deletion from bp 138 to 1416) was confirmed by sequencing using the primers 1,000 bp up- and downstream of the gene VC0395_A2420 (forward and reverse primers for ubiC) (Table 2). Strains were grown in the presence of 50 μg/liter streptomycin at 37°C overnight in a shaker incubator on mixed amino acids (LB medium, consisting of 10 g/liter tryptone, 5 g/liter yeast extract, 10 g/liter NaCl) or in phosphate-buffered M9 minimal medium containing 6.28 g/liter Na₂HPO₄, 3 g/liter KH₂PO₄, 0.5 g/liter NaCl, and 1 g/liter NH₄Cl. In addition, the medium contained 0.1 mM CaCl₂, 0.1 mM MgSO₄, 40 mg/liter each of methionine, threonine, histidine, and leucine, and 2 mg/liter thiamine with 0.2% (wt/vol) glucose as a carbon source (49). This medium had no iron addition to maintain an iron-limiting condition. Growth was monitored with a diode array spectrophotometer (HP 8452A) at 600 nm or with a Tecan Infinite F200 Pro plate reader at 595 nm using 24-well plates (Nunclon 24-well flat-bottom plates; Thermo Fisher Scientific). For the growth experiments monitored by the plate reader, cells grown overnight in LB medium at 37°C were washed and resuspended in M9 minimal medium to a normalized OD₆₀₀ of 1. The growth was monitored after setting an OD₆₀₀ of 0.005 as the starting point. The final volume was 1 ml per well. FeSO₄ or FeCl₃ at a final concentration of 260 μM was added to M9 minimal medium with glucose. Growth was followed in the plate reader at 37°C with linear shaking four times for 60 s each approximately every 10 min. Two wells per condition were analyzed with four reads per well. Mean values and standard deviations of these eight data points are presented.

TABLE 2.

Sequences of primer pairs designed for the different genes

Gene target (locus)	Primer direction	Sequence (5′–3′)
ubiC (VC0395_A2420)	Forward	AGCTCTATGACCATGTCGCTATGT3
	Reverse	ATTCAAGAAAGCTTGTGTCAGGTTG
nqrM (VC0395_A1877)	Forward	CCGTGATTGCCGCCATG
	Reverse	GCATCACAAGGCTCAGGG
recA (VC0395_A0070)	Forward	ACCATTTCGACCGGTTCTCT
	Reverse	GAGAAACCAGCAGCTCATCG
apbE (VC0395_A1878)	Forward	CCAGCCAGCAAGAGAAGAGA
	Reverse	CCAGCCAGCAAGAGAAGAGA
nqrF (VC0395_A1879)	Forward	GCGTGTCAGGTTGCTGTAAA
feoB (VC0395_A1664)	Forward	CGGTTACATGTCTCGTGCTG
	Reverse	CGTTCACGCTCTTGATCCAG

Glucose in supernatants from cell cultures was quantified with a d-glucose detection kit (Boehringer, Mannheim, Germany).

Proteome analysis.

Cell growth, protein extraction, and in-solution digestion of proteins for proteome analysis were performed as described by Toulouse et al. (18). Experiments were performed in triplicates. Methods for mass spectrometry, protein quantification, and data analysis are also reported in Toulouse et al. (18). The mass spectrometry proteomics data of V. cholerae grown in M9 minimal medium with glucose have been deposited in the ProteomeXchange Consortium via the PRIDE (50) partner repository under the data set identifier PXD009379.

Quantification of metal ions in growth medium by ICP-MS.

Samples from medium were freshly prepared for inductively coupled plasma mass spectrometry (ICP-MS) and filtered (0.2-μm pore size). Aliquots from complete medium and from individual medium components were analyzed so that any possible metal ion contamination source could be identified. The instrument, NexION 300 (PerkinElmer), was equipped with a Scott type Ryton spray chamber and a quadrupole mass filter. Rhodium (Rh; 10 μg/liter) served as an internal standard. The samples were diluted in 2% HNO₃ and measured against external calibration curves in duplicates.

Vibrio cholerae cell lysis for RNA isolation.

Cells for RNA isolation were retrieved in the exponential growth phase. The cells were grown in M9 medium supplemented with streptomycin (M9/Strep) without or with added iron (260 μM FeCl₃) at 37°C at 170 rpm for 9 h and harvested by centrifugation at 5,000 rpm for 5 min. Different methods of cell lysis for efficient RNA isolation were compared. Typically, 5 ml of cells normalized to an OD₆₀₀ of 0.5 was harvested by centrifugation. For mechanical cell lysis, cells were resuspended in 600 μl of buffer provided in an RNeasy Plus Mini kit (Qiagen) and vortexed for 12 min in a cell disrupter or vortexed with the addition of 0.3 g of glass beads (3-μm diameter [B. Braun Melsungen], twice washed in nuclease-free buffer). For enzymatic cell lysis, the cell pellet was resuspended in 0.5 ml of TE buffer (30 mM Tris-HCl, 1 mM EDTA, pH 8) with 5 mg/ml lysozyme, 1 mM MgCl₂, 1 mM CaCl₂, and one tip of a spatula of DNase I. The sample was incubated for 30 min at 30°C and then vortexed for 12 min in a cell disrupter (Disruptor Genie; Scientific Industries). Ultrasonic treatment was another method for cell lysis. The cell pellet was resuspended in 600 μl of buffer and incubated in an ultrasonic bath (Sonorex Super RK 102 H; Bandelin) for 2 min. In another approach, cells were lysed by heating samples to 70°C, followed by vortexing for 12 min with a cell disrupter. All cell disruption methods resulted in a higher total RNA yield than that with simple vortexing (see Fig. S2 in the supplemental material). The cell lysis method of heating gave a 5-fold-higher yield than vortexing alone. However, heating resulted in fragmentation of RNA (Fig. S3). The yield and quantity of RNA were evaluated by 0.75% Tris-acetate-EDTA (TAE)–formaldehyde gel electrophoresis (Fig. S3). The RNA gel showed smears of RNA in the preparations obtained by heating and enzymatic treatment whereas 16S and 23S rRNA bands were clearly visible after sonication or glass bead treatment. We chose sonication as the method for cell lysis because it is less prone to contamination than the glass bead method. Total RNA was isolated from cells lysed by sonication using an RNeasy Mini Plus kit (Qiagen). Notably, we could not extract RNA from cells grown in the presence of Fe²⁺, probably due to precipitates formed in the culture supernatants. Therefore, Fe³⁺ was used as an iron source to study iron-dependent gene expression by qPCR.

cDNA synthesis.

DNase I treatment of total RNA and complementary DNA (cDNA) synthesis were performed using a First Strand cDNA synthesis kit (Thermo Scientific), according to the manufacturer’s standard protocol, using random hexamer primers. The amount of total RNA used to synthesize cDNA was 1 μg. The thermal cycling parameters for the reverse transcriptase (RT) reactions were as follows: 25°C for 5 min, 37°C for 60 min, and finally 70°C for 5 min to inactivate the reverse transcriptase. To confirm that the samples were free of DNA contamination, the RT step was omitted, and only DNA polymerase was used in the reaction mixture; no PCR products were obtained in these reactions. PCR was done using cDNA and RT-negative controls as the template and gene-specific primers. Primers were designed using NCBI sequence sources and the Primer3 (51) online primer design tool. The corresponding PCR products were confirmed by sequencing (Eurofins). Primers were designed for nqrM (VC0395_A1877), recA (VC0395_A0070), apbE (VC0395_A1878), nqrF (VC0395_A1879), and feoB (VC0395_A1664) (Table 2).

In silico operon search.

The promoter region of nqrM was examined by BPROM promoter prediction software (52) considering a window of ±500 bp upstream and downstream of the nqrM sequence. The promoter region was manually checked for the +1 transcription start site, the −10 box, and the −35 box. The operon prediction database OperonDB (46, 47) was used to investigate the correlation of the positions of nqrM, apbE, and nqr genes in various microorganisms.

Quantitative PCR analysis of feoB and nqrM expression.

The expression levels of specific target genes were determined by reverse transcription-quantitative PCR (RT-qPCR) using a Platinum SYBR green qPCR SuperMix-UDG kit (Invitrogen) on a CFX 96 optical thermocycler with the CFX Manager software (Bio-Rad). Samples from three biological replicates were measured in three technical replicates each. Mean values and standard deviations of these data were used to compare expression by the 2^−(ΔΔCT) method (where C_T is the threshold cycle) (53). The three equations used for the calculation of expression differences are as follows: (i) ΔC_T = C_{T(gene of interest)} − C_{T(housekeeping gene)}; (ii) ΔΔC_T = ΔC_T(sample) − ΔC_T(control); (iii) fold change in expression = 2^−(ΔΔCT).

A two-step cycling program (95°C for 15 s and 60°C for 60 s) was used. The PCR mixture contained 0.2 μM each primer in a total volume of 10 μl. A cDNA dilution series was used for calibration of the assays. Ultimately, the expression of the gene of interest was normalized to the expression of recA.

Basu et al. (54) introduced recA as the control gene for qPCR studies in V. cholerae since it is a constitutively expressed DNA repair and recombination gene which is not regulated by iron. Often genes associated with ribosomal assembly are chosen as a reference, but the in silico analyses showed that the formerly used rssA gene (49) is less suited due to its multilocus appearance in the V. cholerae genome. In contrast, the gene recA exists in only one allele in many Vibrio species including V. cholerae (55).