Functional analysis of the copy 1 of the fixNOQP operon of Ensifer meliloti under free-living micro-oxic and symbiotic conditions



María J. Delgado, Department of Soil Microbiology and Symbiotic Systems, Estación Experimental del Zaidin, CSIC, PO Box 419, 18080-Granada, Spain. E-mail:



In this work, phenotypic analyses of a Ensifer meliloti fixN1 mutant under free-living and symbiotic conditions have been carried out.

Methods and Results

Ensifer meliloti fixN1 mutant showed a defect in growth as well as in TMPD-dependent oxidase activity when cells were incubated under micro-oxic conditions. Furthermore, haem c staining analyses of a fixN1 and a fixP1 mutant identified two membrane-bound c-type cytochromes of 27 and 32 kDa, present in microaerobically grown cells and in bacteroids, as the FixO and FixP components of the E. meliloti cbb3 oxidase. Under symbiotic conditions, fixN1 mutant showed a clear nitrogen fixation defect in alfalfa plants that were grown in an N-free nutrient solution during 3 weeks. However, in plants grown for a longer period, fixNOQP1 copy was not indispensable for symbiotic nitrogen fixation.


The copy 1 of the fixNOQP operon is involved in E. meliloti respiration and growth under micro-oxic conditions as well as in the expression of the FixO and FixP components of the cbb3 oxidase present in free-living microaerobic cultures and in bacteroids. This copy is important for nitrogen fixation during the early steps of the symbiosis.

Significance and Impact of the Study

It is the first time that a functional analysis of the E. meliloti copy 1 of the fixNOQP operon is performed. In this work, the cytochromes c that constitute the cbb3 oxidase operating in free-living micro-oxic cultures and in bacteroids of E. meliloti have been identified.


Soil bacteria, collectively known as rhizobia, form nitrogen-fixing nodules on the roots of leguminous plants. They have been intensively studied because of their agronomic importance and the inherent biological interest of their complex interactions with their host plants. The establishment of an effective symbiotic association between rhizobia and legumes is a highly specific and complex developmental process, in which both partners undergo differentiation in a concerted way (reviewed by Jones et al. 2007; Oldroyd and Downie 2008; Oldroyd et al. 2011). Following invasion of the plant cells via a complex signalling pathway between bacteria and plant, rhizobia stop dividing and undergo differentiation into nitrogen-fixing bacteroids. The activity of the nitrogen-reducing enzyme nitrogenase requires a high rate of oxygen respiration to supply the energy demands of the nitrogen reduction process. However, oxygen irreversibly inactivates the nitrogenase complex. These conflicting demands are met by controlling oxygen flux to the infected plant cells through an oxygen diffusion barrier, which greatly limits permeability to oxygen (Minchin et al. 2008). Oxygen is then delivered to the bacteroids by the plant oxygen carrier, leghaemoglobin, present exclusively in the nodule (Downie 2005). To cope with the low ambient oxygen concentration in the nodule (10–50 nmol l−1 O2), nitrogen-fixing bacteroids induce a high-affinity cytochrome cbb3-type oxidase (Delgado et al. 1998). Genes encoding the cbb3 complex were initially isolated from rhizobia and named fixNOQP due to its requirement for symbiotic nitrogen fixation (Preisig et al. 1996). Since then, orthologous genes called ccoNOQP were identified in other Proteobacteria including photosynthetic and pathogenic bacteria (reviewed in Cosseau and Batut 2004; Bueno et al. 2012; Ekici et al. 2012).

Cytochrome cbb3 oxidases have been purified from several organisms, including Paracoccus denitrificans, Rhodobacter sphaeroides, Rhodobacter capsulatus and Bradyrhizobium japonicum (reviewed in Pitcher and Watmough 2004). Subunit I (FixN or CcoN) is a membrane-integral b-type and copper-containing cytochrome. Electrons are delivered to the haem/CuB site on subunit I via the membrane-anchored monohaem c- and dihaem c-type cytochromes FixO/CcoO and FixP/CcoP, which constitute subunits II and III, respectively (Buschmann et al. 2010). FixQ or CcoQ is required for optimal oxidase activity, because it stabilizes the interaction of CcoP with the CcoNO core complex, leading subsequently to the formation of the active 230-kDa complex (Peters et al. 2008). The biogenesis of this oxidase depends on the ccoGHIS gene products, which are proposed to be specifically required for cofactor insertion and maturation of cbb3-type cytochrome c oxidases (Kulajta et al. 2006; Pawlik et al. 2010). Several additional proteins including SenC (Swem et al. 2005), PCuAC (Banci et al. 2005; Abriata et al. 2008; Serventi et al. 2012), DsbA (Deshmukh et al. 2003) and CcoA (Ekici et al. 2012) might be also involved in cbb3 biogenesis.

Ensifer meliloti is an aerobic soil bacterium that establishes symbiotic N2-fixing associations with plants of the genera Medicago, Melilotus and Trigonella. The expression of E. meliloti genes required for nitrogen fixation and for microaerobic respiration is coordinated by fixLJ and fixK genes, which are conserved among rhizobia (Fischer 1994; Dixon and Kahn 2004). Under oxygen-limiting conditions, FixL autophosphorylates and transmits phosphate to the FixJ response regulator. Once phosphorylated, FixJ activates transcription of the nifA and fixK genes, which induce expression of nif and fix genes, respectively (Reyrat et al. 1993). A set of publications has demonstrated that fixT and fixM are also targets of FixJ (Ampe et al. 2003; Barnett et al. 2004; Becker et al. 2004; Bobik et al. 2006; Meilhoc et al. 2010). While fixT negatively affects expression of FixLJ-dependent genes by inhibiting FixL autophosphorylation (Garnerone et al. 1999), fixM encodes a flavoprotein that modulates inhibition by 5-aminoimidazole-4-carboxamide nucleotide (AICAR) or 5′adenosine monophosphate (5′AMP) of respiratory and nitrogen fixation genes expression in E. meliloti (Cosseau et al. 2002).

Inspection of the E. meliloti 1021 genome sequence shows a composite architecture, consisting of three replicons with distinctive structure and function: a 3·65-Mb chromosome and two megaplasmids, pSymA (1·35 Mb) and pSymB (1·68 Mb) (Galibert et al. 2001). pSymA contains a large fraction of genes known to be specifically involved in symbiosis such as genes involved in nodulation or in nitrogen fixation process, as well as genes involved in microaerobic metabolism or in denitrification (Barnett et al. 2001). A 290-kilobase (kb) region of pSymA contains nodulation genes as well as genes involved in nitrogen fixation (nif and fix) and it carries repeated sequences (Renalier et al. 1987). One of these reiterated sequences had been identified as part of a cluster of fix genes located 220 kb downstream of nifHDK genes and contains regulatory genes (fixLJ, fixT1, fixK1, fixM), fixGHIS genes, as well as a copy of the fixNOQP operon (fixNOQP1) (Fig. 1). The second fix cluster maps 40 kb upstream of the nifHDK genes and carries another copy of the fixNOQP operon (fixNOQP2), regulatory genes (fixT2, fixK2) as well as a nod locus (Renalier et al. 1987). Both fixNOQP copies are closely related because they encode for proteins having a 95% homology in their respective sequences. In the pSymA genome, there is a third copy of the fixNOQP operon (fixNOQP3), which only presents a 61% homology with the other two copies. Neither genes related with nodulation nor nitrogen fixation is located in the fixNOQP3 genomic context (Fig. 1,

Figure 1.

Genomic context of the three copies of the fixNOQP operon in the symbiotic plasmid pSymA of Ensifer meliloti

Recent transcriptomic analyses have shown that fixNOQP1 genes of E. meliloti are induced under microaerobic free-living and symbiotic conditions (Becker et al. 2004). Furthermore, fixNOQP1 genes have been identified as targets of FixK and FixJ in response to low-oxygen conditions (Bobik et al. 2006; Meilhoc et al. 2010). However, up-to-date functional analyses of these genes are missing. The involvement of fixNOQP1 genes in free-living respiration and symbiotic nitrogen fixation has been investigated in this work.

Materials and methods

Bacterial strains, growth conditions and recombinant DNA methods

Bacterial strains used in this work are E. meliloti wild-type strain 1021 (Meade et al. 1982), fixP1 mutant strain G1PELR32C12 RhizoGATE (Becker et al. 2009) and fixN1 mutant strain 2104 (this work). The fixN1 gene was mutated by performing gene-directed mutagenesis by marker exchange. A PCR fragment of 1·5 kb containing the fixN1 coding region of Sm1021 was subcloned into pK18 mobsac (Schäfer et al. 1994) to obtain plasmid pBG2104. Finally, the 2-kb fragment (Ω Spc/Sm interposon) of pHP45Ω (Prentki and Krisch 1984) was inserted at the unique NruI site located 683 bp downstream of the FixN1 start codon in the 1·5-kb PCR fragment. The resulting plasmid pBG2104Ω was transferred via conjugation into E. meliloti 1021 using Escherichia coli S17-1 carrying pBG2104Ω as donor. Double recombination events were favoured by growth on agar plates containing sucrose. Mutant strains resistant to spectinomycin/streptomycin but sensitive to kanamycin were checked by Southern hybridization experiments (data not shown) for correct replacement of the wild-type fragment by the Ω interposon. The mutant derivative 2104, used in this study, was obtained. Total and plasmid DNA isolation, digestion with restriction enzymes, cloning, agarose gel electrophoresis and Escherichia coli transformation were performed using standard protocols (Sambrook and Russell 2004). Enzymes used for DNA restriction and modification were purchased from Fermentas (Vilnius, Lithuania) and were used according to the instructions of the manufacturer. For Southern hybridizations, DNA was digested with appropriate restriction enzymes, electrophoresed in 1% (wt/vol) agarose gels and blotted onto nylon (Hybond N+). Hybridization was carried out under high stringency conditions using RapidHyb buffer (Amersham, Bucks, UK). Specific probes were normally obtained by PCR and were labelled with α32P-CTP by random priming, using Amersham's Rediprime system.

Ensifer meliloti strains were routinely grown in medium TY (Tryptone Yeast, Beringer 1974) at 30°C. For determinations of growth rates, respiratory activity and haem c staining in free-living conditions, cells were grown in minimal medium (Robertsen et al. 1981). Cell growth under micro-oxic conditions was performed by fluxing a gas mixture of 2% O2 and 98% Ar into the cultures. Initial optical density at 600 nm of the cultures was about 0·1. Antibiotics were added to E. meliloti cultures at the following concentrations (μg ml−1): spectinomycin, 200; streptomycin, 200, kanamycin, 100. Escherichia coli strains were cultured in Luria–Bertani medium (Miller 1972) at 37°C. Escherichia coli DH5α (Stratagene, Heidelberg, Germany) was used as host in standard cloning procedures and Escherichia coli S17-1 (Simon et al. 1983) served as the donor in conjugative plasmid transfer. The antibiotics used were (μg ml−1) ampicillin, 200; streptomycin, 20; spectinomycin, 20; and kanamycin, 25.

Plant growth conditions

Alfalfa (Medicago sativa, var. Aragón) seeds were surface-sterilized by immersing in 2·5% HgCl2 for 9 min. Then, seeds were washed with sterile water and germinated on wet filter paper in Petri dishes in darkness at 28°C for 36 h. Selected seedlings were planted in 1-l autoclaved Leonard jars (Leonard 1943) filled with vermiculite and containing nitrogen-free mineral solution (Rigaud and Puppo 1975). Seeds (eight per jar) were inoculated at sowing with 1 ml of a single bacterial strain (108 cells per ml). Plants were grown in controlled environmental chambers (night/day temperature, 19/25°C; photoperiod, 16/8 h; PPF, 400 μmol m−2 s−1; and relative humidity, 60–70%). For nodulation kinetics assays, germinated seeds were transferred into autoclaved glass tubes containing 5 ml of the N-free nutrient solution and inoculated with approximately 1 × 108 of a single bacterial strain. Each tube, covered with a cotton stopper, was incubated in a growth chamber. After inoculation, the number of nodulated plants and the number of nodules per plant were recorded daily.

Plant assays

Shoots (separated from roots at the cotyledonary node) were dried to a constant weight at 60°C. Dry weight on shoots (SDW), height on shoots (SH) and roots (RH) and nodule fresh weight (NFW) were determined per plant. Nodules were harvested from 7-week-old plants and were frozen into liquid nitrogen and stored at −80°C. Total nitrogen was measured in oven-dried shoots weighed and grounded in an IKA A 11 basic analytical mill (Rose Scientific Ltd., Alberta, Canada). Total nitrogen was determined using a LECO TruSpec CN Elemental Analyzer (LECO Corp., St Joseph, MI, USA).

Bacterial respiratory capacity

Oxygen uptake was determined as described by Marroqui et al. (2001). Cells were harvested after 48 h of growth at 30°C in minimum medium, washed and resuspended in 1 ml 25 mmol l−1 potassium phosphate buffer (pH 7·0). The oxygen uptake at 21°C was measured using a Clark-type oxygen electrode (Hansatech, Norkfolk, UK) after addition of 2 mmol l−1 N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) and 4 mmol l−1 sodium ascorbate to the cellular suspension (0·15–0·25 mg protein). The time taken to consume the oxygen present in the system was used to calculate the rate of TMPD-dependent oxygen consumption.

Membrane extraction, cells fractionation and haem c staining

Cells of E. meliloti grown aerobically in 150 ml TY medium were harvested by centrifugation at 12 000 g for 5 min, washed twice with minimal medium, resuspended in 500 ml of the same medium and finally incubated under low-oxygen conditions for 2 days. Bacteroids from nodules were prepared as previously described by Mesa et al. (2004). Briefly, 1 g of fresh nodules was ground in 7·5 ml Tris/HCl (pH 7·5) supplemented with 250 mmol l−1 mannitol. The homogenate was filtered through four layers of cheesecloth and was centrifuged at 250 g at 4°C for 5 min to remove nodule debris. The resulting supernatant was centrifuged twice at 12 000 g at 4°C for 10 min and was washed twice in 50 mmol l−1 potassium phosphate buffer (pH 7). Free-living cells and bacteroids were resuspended in 3 ml of 50 mmol l−1 potassium phosphate buffer (pH 7) containing 100 μmol l−1 4-(2-aminoethyl) benzene-sulfonyl fluoride hydrochloride (ABSF), RNase (20 μg ml−1), and DNase I (20 μg ml−1). Cells were disrupted using a French pressure cell (SLM Aminco, Jessup, MD, USA). The cell extract was centrifuged at 20 000 g for 20 min to remove unbroken cells, and the supernatant was then centrifuged at 140 000 g for 1 h. The membrane pellet was resuspended in 100 μl of the same buffer. Membrane protein aliquots (from free-living cells or bacteroids) were diluted in sample buffer [124 mmol l−1 Tris/HCl, pH 7·0, 20% glycerol, 4·6% sodium dodecyl sulfate (SDS) and 50 mmol l−1 2-mercaptoethanol] and incubated at room temperature for 10 min. Membrane proteins were separated at 4°C in SDS-12% polyacrylamide gel electrophoresis, transferred to a nitrocellulose membrane and stained for haem-dependent peroxidase activity as described previously (Vargas et al. 1993) using the chemiluminescence detection kit ‘SuperSignal’ (Thermo Fisher Scientific, Pierce, IL, USA).

Analytical methods

The protein concentration was estimated using the Bio-Rad assay (Bio-Rad Laboratories, Richmond, CA, USA) with a standard curve of varying bovine serum albumin concentrations.


Free-living growth rates and respiratory capacity

To investigate the involvement of the cbb3 oxidase encoded by the fixNOQP1 operon in free-living growth, a fixN1 mutant was incubated aerobically and microaerobically (2% O2) in minimum medium. Growth was determined by monitoring the optical density at 600 nm (OD600). After incubation under aerobic conditions, no significant differences were observed in growth rates of the wild-type and the fixN1 mutant (Fig. 2). However, the fixN1 mutant showed a defect in growth, reaching an OD600 of only 0·5 compared to that of 0·9 determined in wild-type (WT) cells after 72-h incubation under micro-oxic conditions (Fig. 2). Cytochrome c-dependent oxygen consumption was measured in the WT strain and the fixN1 deficient mutant using ascorbate-reduced TMPD as a nonphysiological electron donor (Fig. 3). Independently of the strain, TMPD oxidase activity observed in cells grown aerobically was lower than that observed in cells grown under oxygen-limiting conditions. This difference could be due to the induction under oxygen-limiting conditions of high-affinity cytochrome c oxidases, which have greater activity than those induced in aerobic cultures that have low affinity for oxygen. Alternatively, it might be possible that the affinity for TMPD/ascorbate is higher in micro-oxic cultures than in oxically grown cells. Under aerobic conditions, the fixN1 mutant showed levels of TMPD oxidase activity similar to those of the WT strain (Fig. 3). However, when cells were incubated under micro-oxic conditions, oxygen consumption rates of fixN1 cells were approximately about 37% lower than those of WT cells after 48 h growth (Fig. 3). The decrease in TMPD-dependent oxidase activity observed in the fixN1 mutant under oxygen-limiting conditions compared to that observed in the wild-type strain could explain the defect of fixN1 growth under these conditions (Figs 2 and 3).

Figure 2.

Growth of wild-type Ensifer meliloti 1021 (●, ○) and fixN1 mutant (■, □) strains in minimal medium under aerobic conditions (closed symbols) or micro-oxic conditions (open symbols). Error bars represent standard deviation of data from at least two different cultures assayed in triplicate.

Figure 3.

TMPD-dependent oxygen consumption capacity by whole cells of wild-type Ensifer meliloti 1021 (grey bars) and fixN1 mutant (white bars) after 48 h growth in minimal medium under aerobic or micro-oxic conditions. Error bars represent standard deviation of data from at least two different cultures assayed in triplicate.

Haem c staining analyses

Iron present in haem groups that are covalently bound to proteins, such as c-type cytochromes, can be visualized by using a sensitive chemiluminescence assay (Vargas et al. 1993). Haem c staining of electrophoretically fractionated membrane preparations of E. meliloti 1021 grown under aerobic conditions revealed two bands of about 40 and 33 kDa (Fig. 4a, lane 1). In the membrane fractions of WT cells grown under micro-oxic conditions, two additional stained bands at kDa values of 32 and 27 could be detected (Fig. 4a, lane 2). Profiles from the membrane fraction of microaerobically grown cells of the fixN1 mutant showed that it lacked the c-type cytochromes of about 32 and 27 kDa (Fig. 4a, lane 3). Similarly, we could not detect the 32-kDa band in membrane fractions of a fixP1 mutant (Fig. 4a, lane 4). However, the c-type cytochrome of about 27 kDa was present in membranes of the fixP1 mutant (Fig. 4a, lane 4). These results indicate that the 27-kDa and the 32-kDa c-type cytochromes only appear in E. meliloti cells grown under low-oxygen conditions and they correspond to the E. meliloti FixP1 and FixO1 components, respectively, of the cbb3-type cytochrome oxidase encoded by the fixNOQP1 operon. The identification of the haem-stainable bands of approximately 40 and 33 kDa present in membranes of wild-type cells grown under both aerobic and microaerobic conditions as well as in those of fixN1 and fixP1 cells grown under microaerobic conditions is at the moment unknown.

Figure 4.

(a) Haem c-stained proteins in membranes prepared from wild-type Ensifer meliloti 1021 (lanes 1 and 2), fixN1 mutant (lane 3) and fixP1 mutant (lane 4). Cells were incubated aerobically (lane 1) or microaerobically (lanes 2, 3 and 4) in minimal medium. (b) Haem c-stained proteins in membranes of bacteroids of wild-type E. meliloti 1021 (lane 1) and fixN1 mutant (lane 2). Nodules were collected from 7-week-grown alfalfa plants. In (a and b), each lane contains about 20 mg membrane proteins. Apparent masses of the proteins (kDa) are shown at the left margin.

In the membrane fraction of bacteroids, the total amount of haem-stained proteins was considerably higher than that seen with membranes from free-living cells (Fig. 4a,b). Apparent masses of these haem-stained proteins suggest that they correspond to the 40, 33, 32 and 27 c-type cytochromes detected in membranes of free-living microaerobically grown cells of the WT strain (Fig. 4a lane 2 and b, lane 1). These results indicate that, as observed in microaerobically grown cells, bacteroids from E. meliloti 1021 produce the 32 and 27-kDa c-type cytochromes, corresponding to FixP and FixO components of the cbb3 cytochrome oxidase. By contrast to free-living conditions, the band of about 33 kDa migrated together with FixP in membranes of bacteroids. Consequently, we could not demonstrate the absence of the 32-kDa FixP c-type cytochrome in membranes of fixN1 bacteroids (Fig. 4a,b). However, the c-type cytochrome of 27 kDa corresponding to FixO was not detected in membranes from bacteroids of the fixN1 mutant (Fig. 4b, lane 2).

Symbiotic phenotype of the fixN1 mutant

The fixN1 mutant was used to inoculate alfalfa plants that were grown in N-free nutrient solution. After 3 weeks, the alfalfa plants that were not inoculated with any E. meliloti strains were short and turning yellow. The alfalfa plants that were inoculated with E. meliloti 1021 strain were tall and green and therefore had established an efficient symbiosis. However, most of the alfalfa plants inoculated with the fixN1 mutant were shorter and lighter green than those inoculated with the wild-type strain showing signs of nitrogen deficiency (Fig. 5). To further confirm the symbiotic deficiency of the fixN1 mutant, physiological parameters, including nodulation capacity, shoot dry weight (SDW), total nitrogen content (N), and shoot and roots height (SH and RH), were measured in plants inoculated with the wild-type or the fixN1 mutant (Table 1). Noninoculated alfalfa plants had no nodules (data not shown). Plants inoculated with either the wild-type or the fixN1 mutant had an average of four nodules per plant after 3 weeks (Table 1). Similarly, 100% of the plants were nodulated by strains 1021 or fixN1 (Table 1). However, plants inoculated with the fixN1 mutant displayed significant decreases in SDW and [N] compared to those inoculated with the wild-type strain (38 and 30%, respectively) (Table 1). Similarly, SH and RH of plants inoculated with the fixN1 mutant were significantly lower than those of plants inoculated with E. meliloti 1021 (46 and 52%, respectively) (Table 1). Taken together, these results demonstrate that E. meliloti fixN1 mutant has not any defect in nodulation efficiency. However, this mutant shows a clear defect in symbiotic nitrogen fixation.

Table 1. Shoot dry weight (SDW), nitrogen content [N], shoot height (SH), root height (RH), nodules number per plant (NNP) and percentage of nodulated plants (NP) inoculated with the wild-type Ensifer meliloti 1021 strain and the fixN1 mutant derivative. Plants were grown for 3 weeks after inoculation. Data are means with the standard error in parentheses from at least 100 different plants assayed in at least three independent experiments. In the columns, values followed by the different lower-case letter are significantly different as determined by the Tukey HSD test at ≤ 0·05
 SDW (mg plant−1)[N] (mg N plant−1)SH (cm plant−1)RH (cm plant−1)NNPNP
WT13·9 (1·6)a0·570 (0·040)a9·31 (1·24)a32·95 (5·74)a4·36 (0·42)a100%a
fixN 8·7 (0·7)b0·399 (0·057)b5·07 (0·96)b15·99 (2·34)b3·82 (0·52)a100%a
Figure 5.

Nitrogen fixation–dependent growth of alfalfa plants noninoculated (a) or inoculated with Ensifer meliloti wild-type (c) or fixN1 mutant (b) 21 days after inoculation.

To confirm further the symbiotic nitrogen fixation deficiency of the fixN1 mutant, alfalfa plants inoculated with either the wild-type or the fixN1 mutant were grown over a longer period. Surprisingly, after 7 weeks, alfalfa plants inoculated with fixN1 were tall and green showing a similar aspect as those inoculated with the wild-type strain (Fig. 6). To confirm these observations, SDW, [N] and nodulation capacity measured as nodule fresh weight (NFW) were determined in alfalfa plants after 7 weeks growth. As shown in Table 2, plants inoculated with the fixN1 mutant had similar SDW, [N] and NFW than plants inoculated with the WT strain.

Table 2. Shoot dry weight (SDW), nitrogen content [N] and nodule fresh weight (NFW) of plants inoculated with the wild-type Ensifer meliloti 1021 and fixN1 mutant derivative. Plants were grown for 7 weeks after inoculation. Data are means with the standard error in parentheses from at least 70 different plants, assayed in at least three independent experiments. In the columns, values followed by the same lower-case letter are not significantly different as determined by the Tukey HSD test at ≤ 0·05
 SDW (mg plant−1)[N] (mg N plant−1)NFW (mg plant−1)
WT281·1 (47·9) a11·585 (0·564) a35·5 (3·5) a
fixN 264·6 (52·0) a11·616 (1·186) a42·0 (4·0) a
Figure 6.

Nitrogen fixation–dependent growth of alfalfa plants inoculated with Ensifer meliloti wild-type (a) or fixN1 mutant (b) 49 days after inoculation.


The actual physiological role of the high-affinity cbb3 oxidase encoded by the fixNOQP operon in symbiotic nitrogen fixation has been investigated in many rhizobial species such as B. japonicum (Preisig et al. 1993, 1996), Azorhizobium caulinodans (Mandon et al. 1993, 1994), Rhizobium leguminosarum (Schlüter et al. 1997) and Rhizobium etli (Girard et al. 2000; Granados-Baeza et al. 2007). While B. japonicum and A. caulinodans have only one copy of the fixNOQP operon, reiteration of these genes has been reported in Rh. leguminosarum bv. viciae (Schlüter et al. 1997), Rh. etli (Girard et al. 2000) and in Mesorhizobium loti (Uchiumi et al. 2004). In E. meliloti, three copies of the fixNOQP operon have been identified (Fig. 1, The first copy is located in a DNA region containing also the whole set of regulatory genes (fixLJ, fixK, fixT and fixM) required for microaerobic respiration and nitrogen fixation. These observations suggest that copy 1 of E. meliloti fixNOQP genes is the potential candidate to support respiration under free-living and symbiotic conditions.

Free-living experiments have demonstrated the involvement of fixNOQP1 genes in respiration and growth of E. meliloti cells under low-oxygen conditions. Furthermore, chemiluminescent staining analyses used to visualize proteins that contain c-type cytochromes have demonstrated by the first time that the two membrane-bound c-type cytochromes, with molecular masses of 27 and 32 kDa, detected in microaerobically grown cells, correspond to the FixO and FixP components of the E. meliloti cbb3 oxidase. The absence of these cytochromes in the fixN1 mutant suggests that copy 1 of fixNOQP operon is the sole functional copy required to express the cytochrome cbb3 terminal oxidase under free-living micro-oxic conditions. Supporting our findings, it has been recently proposed that fixNOQP1 genes are regulated by FixJ under both micro-oxic free-living and symbiotic conditions, whereas copy 2, which is located next to the NifA regulon, was only detectable in bacteroids (Bobik et al. 2006). With respect to fixNOQP3 operon, it was not induced under either microaerobic or symbiotic conditions (Bobik et al. 2006). Two genes from the fixNOQP3 operon have been showed partially phoB regulated (Krol and Becker 2004). It has been proposed that the three copies of fixNOQP operon undergo differential regulation in E. meliloti (Bobik et al. 2006) suggesting a different physiological role for them. It has been shown in M. loti that the fixNOQP copy located out of the symbiotic island functions preferentially in microaerobic environments, whereas bacteroids likely use two copies for symbiotic respiration (Uchiumi et al. 2004).

The haem-stained band of about 33 kDa also present in membranes of E. meliloti microaerobically grown cells is the predicted size for cytochrome c1 that is a component of bc1 complex. In B. japonicum (Thony-Meyer et al. 1989) and in Rh. leguminosarum (Wu et al. 1996), it has been demonstrated that bc1 complex transfers electrons to the cbb3 oxidase and is essential for symbiotic nitrogen fixation. The 40-kDa protein band had been previously detected in E. meliloti membranes by Yurgel et al. (2007). A search for E. meliloti genes predicted to produce proteins that contain the CXXCH haem-binding motif and are in this molecular mass range allowed these authors to propose SMb21367 (cycA) or SMc02858 (a 41-kDa DnaJ-type protein) as potential candidates.

In this work, we have also investigated the function of the fixNOQP1 operon under symbiotic conditions. Nodulation kinetics, plant dry weigh and total nitrogen results in plants inoculated with the fixN1 mutant and grown for 21 days in N-free nutrient solution clearly suggest that this copy is required for optimal fixation of nitrogen. However, symbiotic performance of alfalfa plants inoculated with the fixN1 mutant and grown for 49 days was very similar to the plants inoculated with the wild-type strain. It might be possible that for longer growth periods, the other copies of fixNOQP are functional in symbiotic conditions. These results agree with those published previously by Trzebiatowski et al. (2001) where a E. meliloti strain carrying a Tn5-1063 insertion within fixN was symbiotically proficient, suggesting that a second functional copy of fixN could be involved in symbiotic nitrogen fixation. In Rh. leguminosarum bv. viciae, both copies of fixNOQP genes are required for optimal nitrogen fixation (Schlüter et al. 1997). However, in Rh. etli, a mutation in the fixN of plasmid d (but not in that of plasmid f) was severely affected, indicating a differential role for these reiterations in nitrogen fixation (Granados-Baeza et al. 2007).

The absence of the 27-kDa c-type cytochrome corresponding to FixO in membranes of bacteroids from nodules of 7-week-old plants inoculated with the fixN1 mutant suggests that the copy one of fixNOQP genes is the sole functional copy responsible for expression of FixP and FixO proteins in bacteroids. By contrary to our observations, transcriptomic analyses have demonstrated expression of the copy 2 of E. meliloti fixNOQP operon in bacteroids (Bobik et al. 2006). However, these authors found that levels of induction of fixNOQP2 genes by low-oxygen conditions under free-living conditions is only twofold compared to fivefold induction of fixNOQP1 genes relative to expression levels under oxic conditions (Bobik et al. 2006). Hence, we do not exclude that the lack of detection of FixO in fixN1 bacteroids where fixNOQP2 genes might be expressed could actually be due to a technical limitation, given that overall sensitivity of the arrays is better than the haem-staining protein detection. Therefore, on the long term, other copies could replace copy one through lower expression rates. Alternatively, after 7 weeks, some recombination could take place with the remaining copies that complement a native-like fixNOQP1 operon. It might be also possible that other terminal oxidases such as the high-affinity bd-type oxidase or the cyo quinol oxidase are also involved in supporting nitrogenase activity in 7-week-old plants. In this context, the E. meliloti chromosome contains smc02254 and smc02255 genes (, which encode a high-affinity quinol oxidase that is likely to contribute to micro-oxic respiration in addition to or in the absence of the cbb3 oxidase. Furthermore, recent transcriptional studies have reported that cyo genes encoding a cytochrome o ubiquinol oxidase were induced under microaerobic conditions (Bobik et al. 2006).


This work was supported by Fondo Europeo de Desarrollo Regional-cofinanced grant AGL2010-18607 from Ministerio de Economía y Competitividad (Spain) and by Ciencia y Tecnologıa para el Desarrollo [grant number 107PICO312]. Support from the Junta de Andalucía to Group BIO-275 is also acknowledged. M.J.T. was supported by a fellowship from the Consejo Superior de Investigaciones Cientificas I3P Programme.