Fine-tuned regulation by oxygen and nitric oxide of the activity of a semi-synthetic FNR-dependent promoter and expression of denitrification enzymes in Paracoccus denitrificans

Jirí Mazoch1, Michal Kunák1, Igor Kucera1 and Rob J. M. van Spanning2

1 Department of Biochemistry, Faculty of Science, Masaryk University, Kotlárská 2, CZ-61137 Brno, Czech Republic
2 Department of Molecular Cell Physiology, Faculty of Biology, BioCentrum Amsterdam, Vrije Universiteit, NL-1081 HV Amsterdam, The Netherlands

Correspondence
Igor Kucera
ikucera{at}chemi.muni.cz


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In Paracoccus denitrificans at least three fumarate and nitrate reductase regulator (FNR)-like proteins [FnrP, nitrite and nitric oxide reductases regulator (NNR) and NarR] control the expression of several genes necessary for denitrifying growth. To gain more insight into this regulation, {beta}-galactosidase activity from a plasmid carrying the lacZ gene fused to the Escherichia coli melR promoter with the consensus FNR-binding (FF) site was examined. Strains defective in the fnrP gene produced only very low levels of {beta}-galactosidase, indicating that FnrP is the principal activator of the FF promoter. Anoxic {beta}-galactosidase levels were much higher relative to those under oxic growth and were strongly dependent on the nitrogen electron acceptor used, maximal activity being promoted by N2O. Additions of nitrate or nitroprusside lowered {beta}-galactosidase expression resulting from an oxic to micro-oxic switch. These results suggest that the activity of FnrP is influenced not only by oxygen, but also by other factors, most notably by NO concentration. Observations of nitric oxide reductase (NOR) activity in a nitrite-reductase-deficient strain and in cells treated with haemoglobin provided evidence for dual regulation of the synthesis of this enzyme, partly independent of NO. Both regulatory modes were operative in the FnrP-deficient strain, but not in the NNR-deficient strain, suggesting involvement of the NNR protein. This conclusion was further substantiated by comparing the respective NOR promoter activities.


Abbreviations: FNR, fumarate and nitrate reductase regulator (E. coli); FnrP, NarR, nitrate reductase regulators (P. denitrificans); FF site, FNR-binding site; NAR, nitrate reductase; NIR, nitrite reductase (cytochrome cd1); NNR, nitrite and nitric oxide reductases regulator (P. denitrificans); NOR, membrane-bound nitric oxide reductase (cytochrome bc); SNP, sodium nitroprusside


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Paracoccus denitrificans is a facultative anaerobe capable of reducing nitrate to dinitrogen gas via nitrite, NO and N2O in order to generate metabolic energy in the absence of molecular oxygen. The synthesis of four enzymes involved in denitrification is controlled at the transcriptional level and depends upon the availability of oxygen and nitrogenous electron acceptors (Baker et al., 1998). At least three proteins belonging to the superfamily of fumarate and nitrate reductase (NAR) regulator (FNR)-type transcription factors seem to play an especially important role in this environmental adaptation process. The FnrP protein activates anaerobic expression of the membrane-bound NAR (van Spanning et al., 1997). As is the case for the Escherichia coli FNR protein, the N-terminal domain of FnrP harbours three cysteine residues which, together with a cysteine residue located in the central domain, can carry a labile [4Fe–4S] cluster (Hutchings et al., 2002). The remaining proteins, nitrite and nitric oxide reductases regulator (NNR) and NarR, differ markedly from FnrP in that they lack the N-terminal cysteine ligands. NNR is involved in induction of genes in the nir and nor operons encoding nitrite reductase (NIR) and nitric oxide reductase (NOR), respectively (van Spanning et al., 1995), while NarR elevates the expression of NAR in response to nitrate and/or nitrite (Wood et al., 2001).

As yet we do not know exactly how the biological activity of the FNR-like proteins of P. denitrificans is modulated. The [4Fe–4S] cluster of the E. coli FNR protein exhibits great sensitivity to molecular oxygen, which rapidly converts it to the [2Fe–2S] form in a direct reaction; this is accompanied by dissociation of the FNR dimer into monomers with a loss of the ability to bind DNA (Khoroshilova et al. 1995, 1997; Green et al., 1996; Jordan et al., 1997). An analogous Fe–S cluster could be reconstituted into the overexpressed FnrP in vitro, confirming that this protein is a true orthologue of FNR from E. coli with a similar mechanism of oxygen sensing (Hutchings et al., 2002). By studying the {beta}-galactosidase activity in P. denitrificans strains with the nirS and norC promoters fused to the 'lacZ reporter gene, NNR-dependent transcription was shown to be controlled by NO or a related species (van Spanning et al., 1999), analogous to earlier findings in Rhodobacter sphaeroides (Tosques et al., 1996; Kwiatkowski et al., 1997). Primer extension analysis of the nor promoter region revealed the presence of two transcription start sites. A transcript that initiated downstream of the putative NNR-binding site was identified as the major contributor to NNR-dependent anaerobic nor expression. Besides this, a weak transcription from an upstream start site still occurred both aerobically and anaerobically (Hutchings & Spiro, 2000).

An understanding of the regulatory network in P. denitrificans was initially developed, based upon the use of an artificial FF-melR class II promoter, in which the consensus FNR-binding (FF) site is centred at -41·5 with respect to the transcription start site (Lodge et al., 1990). This promoter could be activated in P. denitrificans during the switch from oxic to anoxic growth conditions (Spiro, 1992), provided that the culture contained a sufficient amount of iron (Kucera & Mat'chová, 1997). An alternative mode of FF-melR activation upon the addition of respiratory inhibitors to highly aerated cultures forced us to consider the role of changes in the intracellular redox state in transcriptional regulation besides the direct interaction of oxygen with the [4Fe–4S] cluster(s) (Kucera et al., 1994). Here we decided to broaden the studies using FF-melR to include fnrP and nnr mutant strains and a range of culture conditions, i.e. using various terminal electron acceptors. In this way, we attempted to answer two questions. (a) Which of the FNR-like proteins present in P. denitrificans cells activates FF-melR? (b) Are there factors other than oxygen concentration that modulate the extent of this activation?

The second objective of the present study concerns the true role of nitrogenous electron acceptors in the expression of denitrification enzymes. The newly proposed essential activation of NNR by NO has to be reconciled with previous studies showing that oxygen limitation alone was the dominant regulatory factor (Kucera et al., 1984; Boublíková et al., 1985). A possible trivial explanation for the previous results could be the presence of traces of nitrate in the culture media used. To check this possibility, we followed the formation of NOR in cells under carefully controlled growth conditions, minimizing the formation of endogenous NO.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Strains and growth conditions.
The pertinent bacterial stains used in this study are listed in Table 1. P. denitrificans strains were grown at 30 °C on a basal salts medium with 50 mM succinate as carbon source (Kucera et al., 1990). When required, rifampicin, tetracycline and kanamycin were added to concentrations of 20, 1 and 30 µg ml-1, respectively. Growth under oxic conditions was carried out with shaking at 200 r.p.m. in 250 ml or 1 litre open conical flasks filled to one-tenth of their volume. Anoxic growth proceeded statically in sealed flasks previously sparged with oxygen-free argon for 5 min; the terminal electron acceptors used were nitrate (10 mM), nitrite (5 mM) or N2O (sparging with N2O gas for 5 min). The growth of cells was followed by measuring OD600 (1 cm light path). An OD600 of 1·0 corresponds to 0·31 mg dry wt (l culture)-1. Micro-oxic incubation was performed by suspending the oxically grown cells in 50 ml growth medium supplemented with 0·6 mM sodium molybdate at 1·15–1·25 mg dry wt ml-1 and agitating the suspension in 250 ml open conical flasks at 120 r.p.m. Under these conditions the growth of cultures is severely limited, but the de novo synthesis of denitrification enzymes proceeds optimally (Boublíková et al., 1985). The concentration of nitrate present as a contaminant in growth media was determined by nitration of salicylic acid (Nedoma, 1983). Ten millilitres of medium were mixed with 1 ml 0·5 % sodium salicylate and evaporated to dryness at 110 °C. The cold residue was left to stand with 1 ml concentrated sulphuric acid for 10 min and then NaOH was added to raise the pH above 12 before measuring A410 against the reagent blank. A detection limit of about 1 µM could be decreased further by taking a larger aliquot of the medium. When appropriate, most of the nitrate contamination was removed from the medium biologically, exploiting the high affinity of the anoxically grown P. denitrificans cells for nitrate (Parsonage et al., 1985). Freshly harvested cells were washed twice in 0·1 M sodium phosphate buffer, pH 7·3, and introduced at a final concentration of ~1·1 mg dry wt ml-1. Following flushing with argon, the suspension was incubated for 22 h at 30 °C and then centrifuged.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this work

 
Plasmid isolation and transfer.
The bacterial plasmid used was pRW2A/FF, a derivative of the broad-host-range plasmid pRK250 carrying a fusion of the melR promoter to lacZ downstream of an FNR-binding sequence (Table 1). It was purified on a silica column (Qiagen) and then transformed into E. coli SM10 (Table 1) by electroporation as described by Ausubel et al. (1995) by using a Gene Pulser II (Bio-Rad) at a field strength of 2·5 kV cm-1. The plasmid-containing strain E. coli SM10 served as a donor in agar-plate matings with P. denitrificans strains. The donor culture was grown in Luria–Bertani (LB) medium with tetracycline (15 µg ml-1) to stationary phase; the recipient culture was grown in LB with rifampicin (20 µg ml-1) to late exponential phase. Cells were harvested by centrifugation, resuspended in LB, mixed on LB plates at a donor/recipient ratio of 1 : 10 and incubated at 30 °C for 16 h. After mating, the cells were washed from the agar surface in sterile 0·9 % NaCl, diluted appropriately and plated on LB agar supplemented with rifampicin (40 µg ml-1), neomycin (30 µg ml-1; not added to the wild-type culture) and tetracycline (1 µg ml-1) to obtain the exconjugants.

Consideration of the amount of plasmid DNA in cells.
Because a plasmid system was used to study FnrP-dependent transcription, we considered the possibility that the various levels of {beta}-galactosidase would emerge as a function of the varying plasmid copy number. Therefore, the amount of plasmid DNA per unit of total cellular protein was estimated for several culture samples and correlated with the corresponding specific activity of {beta}-galactosidase. Cells in 25–50 ml cultures were pelleted and subjected to the boiling miniprep procedure (Ausubel et al., 1995). Electrophoresis of the thus obtained plasmid DNA samples was performed in a horizontal 0·5 % agarose gel (10x8x0·5 cm) at 70 V for 2 h. The photographs of the ethidium-bromide-stained gels were scanned using a Personal Densitometer SI and evaluated by the ImageQuant Version 1.2 program (Molecular Dynamics). The absence of positive correlation indicated that the extent of {beta}-galactosidase overproduction was related to an increased transcription efficiency rather than uncontrolled plasmid copy number variations.

Enzyme assays.
NAR activity was measured spectrophotometrically by following the oxidation of reduced methyl viologen coupled to the reduction of 10 mM nitrate to nitrite in cells made permeable by including 0·1 % Triton X-100 in the reaction mixture (Kucera & Kaplan, 1996). NIR activity was measured by determining the amount of nitrite consumed by a suspension of intact cells in the presence of 20 mM succinate as electron donor (Kucera et al., 1990). NOR activity was measured amperometrically using a Clark-type electrode sensitive to NO. The reaction mixture contained 5 mM ascorbate in conjunction with 0·2 mM N,N,N',N'-tetramethylphenylene-1,4-diamine as electron donor and redox mediator, respectively; the initial concentration of NO being 50 µM (Kucera, 1992). When the sensitivity to antimycin was examined, the artificial electron donor was replaced by 7 mM succinate. {beta}-Galactosidase activity was measured at 30 °C in toluene-permeabilized cells by monitoring the hydrolysis of ONPG (Miller, 1972). Activities were expressed in terms of cell density (OD600) by using a modified formula of Miller (1972): Miller units=1000x(A420-1·57xA550)/(A600xVxt), in which A{lambda} denotes an absorbance at a wavelength {lambda} (nm), V is the volume of the added cell suspension (ml) and t is the reaction time (min). The activity data presented are means±SEM (n>=3).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
FF-lacZ expression in the fnrP and nnr mutant strains
The primary aim of this work was to determine which of the FNR-type transcription factors mediates the anoxic activation of the FF site in P. denitrificans. For this analysis, plasmid pRW2A/FF was conjugated into fnrP, nnr and fnrP/nnr mutant strains and the cells were harvested from oxically grown cultures shifted to incubation under micro-oxic conditions in the presence of nitrate. As shown in Table 2, an increase in {beta}-galactosidase reporter activity was apparent in the nnr mutant strain, which contains an intact fnrP gene. On the other hand, much lower transcriptional activation of the FF site took place in the fnrP mutant, in spite of the fact that NNR was probably present and active as shown by the concomitant synthesis of NIR. The fnrP/nnr double mutant strain lacked both {beta}-galactosidase and the denitrification enzymes, NAR and NIR.


View this table:
[in this window]
[in a new window]
 
Table 2. Formation of {beta}-galactosidase and denitrification enzymes in the P. denitrificans wild-type and fnrP/nnr mutant strains

Measurements were made on strains harbouring plasmid pRW2A/FF. Cells were grown oxically and then subjected to micro-oxic conditions for 3 h in succinate medium containing 10 mM nitrate.

 
Effect of terminal electron acceptors on FF-lacZ expression
Having established the main responsibility of FnrP for FF site activation, we started to investigate the effects of various terminal acceptors. Strain Pd1222(pRW2A/FF) was therefore grown in succinate medium either oxically or anoxically with a nitrogenous terminal acceptor (nitrate, nitrite or N2O) until the OD600 rose to the required value and then assayed for {beta}-galactosidase activity. As Table 3 indicates, the activity of the melR promoter was strongly influenced not only by oxygen, but, under anoxic conditions, also by the type of nitrogenous electron acceptor present. An exceptionally high level of {beta}-galactosidase was found in the cells grown with N2O. Interestingly, when N2O was supplied to the culture growing on nitrate, the resulting {beta}-galactosidase activities were always much lower than for nitrate or N2O alone.


View this table:
[in this window]
[in a new window]
 
Table 3. Effect of terminal electron acceptors on FF-lacZ expression

P. denitrificans PD1222(pRW2A/FF) cells were cultivated in succinate minimal medium either oxically (see Methods) or anoxically on the specified nitrogenous terminal acceptor. The initial OD600 was 0·05 in all cases. Inocula were grown to the late exponential phase under the same conditions as the main cultivation. The cells were allowed to grow until the suspension reached an OD600 of 0·4 (except growth on N2O, where the final OD600 was 0·25) and the activity of {beta}-galactosidase was assayed. Experiments were performed three times and results are from one typical experiment. The means±SEM are shown; each activity was measured in three to four replicates.

 
Parallel activation of the FF-melR promoter and appearance of the membrane-bound NAR
Earlier work by Boublíková et al. (1985) established that some NAR activity is present during micro-oxic incubation of a wild-type strain of P. denitrificans even in the absence of added nitrate. Formation of NAR by the Pd1222 (pRW2A/FF) ex-conjugant under such conditions was accompanied by substantial increases in {beta}-galactosidase (Fig. 1). The addition of nitrate had two effects. First, the synthesis of NAR was significantly promoted. Second, the final activity of {beta}-galactosidase was reduced. A plausible explanation for these observations would be that activation by microaerobiosis of a transcriptional factor measured with the FF-lacZ fusion (probably FnrP) alone suffices to trigger increased synthesis of NAR. Nitrate seems to be sensed independently, but it or its metabolite may interfere with the oxygen-sensing system to some extent.



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 1. Expression of the FF-lacZ fusion and total NAR upon shift from oxic to micro-oxic conditions. Washed oxically grown cells of P. denitrificans PD1222(pRW2A/FF) were resuspended in succinate medium without (triangles) or with (circles) 10 mM nitrate at 1·2 mg dry wt ml-1 and incubated under limited aeration for up to 3 h. Samples taken at various times were assayed for {beta}-galactosidase (a) and NAR (b). Results of a typical experiment are shown as the means±SEM; each activity was measured in three to four replicates.

 
The data in Fig. 1 are derived from whole cells and consequently do not discriminate between the membrane-bound and periplasmic NARs. Since these enzymes are known to differ in their sensitivities toward azide (Sears et al., 1993), it was felt worthwhile to examine the dependence of the enzyme activity upon azide concentration. For the cells adapted to micro-oxic conditions without nitrate, a monophasic inhibition curve was obtained corresponding to an I50() value (the concentration of a given inhibitor which allows an enzyme-catalysed reaction to proceed at 50 % of the uninhibited rate) of 18 µM at 10 mM nitrate (results not shown). This figure agrees well with the kinetic constants Km (nitrate) and Ki () of 280 µM and 0·55 µM, respectively, reported for the membrane-bound NAR of P. denitrificans (Craske & Ferguson, 1986).

Inhibition of FF-lacZ expression by nitroprusside
In heterologous E. coli systems, the NNR- and FnrP-dependent anoxic transcriptions from the FF-melR promoter were reported to be activated and inhibited, respectively, by sodium nitroprusside (SNP) as a source of NO+ (Hutchings et al., 2000, 2002). These findings prompted us to investigate which type of response would occur in P. denitrificans. Washed oxically grown cells of P. denitrificans PD1222(pRW2A/FF) were resuspended in succinate growth medium without nitrate at 1·2 mg dry wt ml-1 and incubated with various concentrations of SNP under limited aeration for 1 h. If we regarded the {beta}-galactosidase activity increments in control samples without SNP (153 Miller units) as 100 %, for 0·001, 0·01, 0·1 and 1 mM SNP the respective relative increments (means±SEM, n=3) were 99±4, 83±3, 40±3 and 13±1 %. These results show that SNP significantly inhibits the increase of {beta}-galactosidase production under oxygen-limiting conditions, indicating a ‘FnrP-type’ response.

NOR synthesis in response to oxygen withdrawal and to the availability of nitrogenous compounds
As mentioned in the Introduction, the synthesis of denitrification enzymes observed under micro-oxic conditions could be due to the reduction of trace amounts of nitrate by bacteria, thereby producing the regulatory NO signal. This possibility was minimized in three ways. First, the growth media were pre-incubated with the washed anoxically grown cells to ensure that nitrate concentration was decreased to submicromolar levels where it could not be metabolized easily. Second, we employed a mutant strain devoid of the NO-forming NIR. Third, haemoglobin was included in some experiments as an NO trap. The results obtained using this combined approach are exemplified by Figs 2 and 3. Comparison of panels (a) and (b) in Fig. 2 indicates that a strain containing an NIR mutation behaved like the wild-type strain in its ability to express NOR in response to oxygen limitation without nitrogenous terminal acceptors. The presence of SNP, an NO-generating compound, stimulated a further NOR increase in both strains, at least after 1·5 h treatment, while nitrite was only effective in the wild-type strain. The absence of FnrP influenced neither low-oxygen nor NO response significantly (Fig. 2c). On the contrary, we found no NOR activity in the NNR-deficient strain (results not shown). Haemoglobin (10 µM) reversed the effects of exogenous NO, but not the effect of hypoxia (Fig. 3). We also asked whether the observed NOR activity does indeed reflect the presence of the respiratory-chain-linked NOR. This could be unequivocally proven by the finding that with succinate as electron donor, the respiratory inhibitor antimycin at a concentration of 3 µg (mg dry wt)-1 reduced the rate of NO consumption to a value below 5 %. Taken together, these results support the idea that both NO-independent and NO-dependent signals are required for the full expression of NOR.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 2. Effect of nitrite and SNP on NOR expression under micro-oxic conditions. Oxically grown cells of wild-type strain Pd1222 (a), nirS- (b) or fnrP- (c) were harvested and exposed to a limited aeration in the presence of 1 mM nitrite (hatched bars) or 10 µM SNP (light grey bars) for up to 3 h. White bars, control. The activity of NOR formed was measured in washed cells amperometrically using ascorbate as electron donor and N,N,N',N'-tetramethylphenylene-1,4-diamine as mediator. The results are expressed as the means±SEM of three individual experiments.

 


View larger version (23K):
[in this window]
[in a new window]
 
Fig. 3. Haemoglobin abolishes the enhancing effect of SNP on NOR expression under micro-oxic conditions. Oxically grown cells of the nirS- strain were harvested and exposed to a limited aeration in the presence of 10 µM SNP (dark grey bars), 10 µM haemoglobin (hatched bars) or a combination of both (light grey bars) for up to 3 h. White bars, control. Results of a typical experiment are shown as the means±SEM; each activity was measured in three replicates.

 
Activity of the norC promoter
In an earlier study (van Spanning et al., 1999), expression of NOR was evaluated from the activity of a norC-lacZ promoter fusion in wild-type and the NNR-deficient strains. {beta}-Galactosidase levels resulting from cultures growing under vigorous shaking were the same in both strains (96 Miller units). A lower frequency of shaking and a greater volume of culture in the cultivation flasks caused an increase in the wild-type (142 Miller units), while there was a decrease in the nnr mutant (38 Miller units). These experiments were repeated in exactly the same manner with the fnrP mutant. The induction patterns of the norC promoter in this strain (70 and 140 Miller units for the aerated and oxygen-deficient culture, respectively, and a further increase up to 3100 Miller units after adding 100 mM nitrate to the latter) turned out to be similar to the values found in the wild-type cells, adding further evidence for the dispensability of FnrP in NOR expression. It is appropriate to note here that the oxygen limitation conditions originally used in the publication of van Spanning et al. (1999) were not optimized for the production of denitrification enzymes and differ from the micro-oxic incubation protocol applied here. For this reason, we did not attempt to correlate the above-reported norC promoter activities with the enzyme activities of NOR found in other experiments.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
We conclude that the transcription from the FF-melR promoter in P. denitrificans is mediated chiefly by the FnrP protein based on the following observations. (i) A decrease in promoter activity caused by SNP (this work) resembles a similar result obtained with FnrP in the heterologous E. coli expression system (Hutchings et al., 2002). The reduced sensitivity to SNP inactivation observed by us possibly bears on the fact that the experiments with E. coli were performed under anoxic conditions while the work with P. denitrificans required limited aeration and consequently some of the NO generated from SNP could be trapped by oxygen. (ii) An fnrP knock-out leads to a clear-cut loss of {beta}-galactosidase production (Table 2), showing that the main regulator is encoded by the fnrP gene. However, since the double fnrP/nnr mutant strain manifested somewhat lower {beta}-galactosidase activity than the single fnrP mutant strain, NNR may also have some activation effect, but this is very small in comparison with FnrP. The activity of the FF-melR promoter close to the background value in the double mutant strain indicates that NarR does not activate transcription from this promoter. The possibility that fnrP mutant strains lack active NarR because of regulation of its expression by FnrP is rendered improbable by the results of recent direct measurements of narR promoter activity on the fnrP- background (R. J. M. van Spanning, unpublished data). In Pseudomonas aeruginosa, which possesses proteins ANR and DNR in place of FnrP and NNR, respectively, the FF promoter also became silent as a result of an anr gene mutation (Galimand et al., 1991), whereas a mutation of the dnr gene had no effect (Hasegawa et al., 1998). However, a clear difference with respect to the P. denitrificans system is that the anr mutant strain is devoid not only of ANR, but also of DNR due to the control exerted by ANR over dnr gene transcription (Arai et al., 1997). As a result, its phenotype bears a notable resemblance to that of the fnrP/nnr double mutant of P. denitrificans, i.e. none of the denitrification enzymes is formed. High transcriptional activity of the FF site after transformation with a plasmid expressing the dnr gene indicates that DNR can activate the consensus FNR promoter equally as well as ANR (Hasegawa et al., 1998). Such an overlapping recognition specificity may be less pronounced in P. denitrificans. The low capability of NNR to activate transcription from the FF promoter in P. denitrificans cannot be due to a failure to bind to this region of DNA since the activation takes place in E. coli with the expressed P. denitrificans NNR (Hutchings et al., 2000). A physiological explanation for the different behaviour in both systems is not readily apparent. A challenging, but speculative hypothesis may be that NNR works with an alternative sigma factor in P. denitrificans (cf. the discussion in van Spanning et al., 1997) and that the RNA polymerase containing that sigma factor binds only weakly to the FF promoter.

Although subject to the possible limitations of our plasmid system, the current data argue for the existence of additional signals besides oxygen concentration, capable of modulating FnrP function. One of the factors affecting the activity of FnrP under anoxic conditions may be NO, one of the free intermediates in denitrification. This role of NO is supported by the recently demonstrated sensitivity of the [4Fe–4S] cluster to low concentrations of NO (Wu et al., 2000; Cruz-Ramos et al., 2002), by the observed inactivation of FnrP in vivo by artificial NO donors (Hutchings et al., 2002; this work), by the retarding effect of nitrate on the transcriptional activation of the FF site under microaerobiosis (Fig. 1) and especially by the notable difference between reporter enzyme activities in cells grown with nitrate or nitrite and in cells grown with N2O (Table 3), which possibly arises from the ability of the former electron acceptors to produce NO. Since the nnr mutation causes nitrite accumulation via reduced expression of NIR (van Spanning et al., 1995) and nitrite in turn inhibits the activity of NOR (Kucera, 1992), a negative modulation by NO may also underlie the lower activity of the FF-melR promoter in the nnr mutant compared to the wild-type strain (Table 1). The mechanism by which nitrogenous terminal acceptors forming NO (nitrate and nitrite) exert their effect can be based on (i) a direct degradation of the [4Fe–4S] cluster by NO (Cruz-Ramos et al., 2002; Hutchings et al., 2002; Wu et al., 2000) and/or (ii) activation of other transcriptional regulators, which may compete with FnrP for promoter binding, yet only weakly activate expression. Whatever direct or indirect inactivation of FnrP by NO would adversely affect the first reaction of the denitrification pathway, thus providing a negative feedback mechanism that counters the destabilizing positive feedback effect of NO to enhance the expression of NIR and its own production.

Previous measurements of the appearance of enzyme activities, following adaptation of oxically grown batch cultures to the conditions of restricted aeration, have indicated that oxygen limitation alone is the dominant regulatory factor even if nitrate also has some inducing effect on NAR (Kucera et al., 1984; Boublíková et al., 1985). Our new data lend further support to this original view by demonstrating a partial induction of NAR (Fig. 1) and NOR (Fig. 2) under micro-oxic conditions. Further induction seen upon the addition of nitrogenous compounds suggests the operation of dual control mechanisms involving separate oxygen and nitrogen oxides sensors. For NAR this is in accord with FnrP and NarR acting in concert to regulate the NAR operon (Wood et al., 2001), although some uncertainty persists as to the relative roles of both transcriptional factors. For NOR, the involvement of the NNR-mediated NO signal is well documented, while the molecular basis for the NO-independent part of NOR expression remains unclear at present. One possibility is that a metabolic signal other than NO can also activate NNR-dependent transcription. Physiologically, the dual control of denitrification enzymes possibly reflects a strategy of the cell to get ready for denitrification when oxygen tension becomes limiting.

An alternative approach to study the expression of denitrification enzymes is to measure individual mRNA levels. In this way, results at variance with the dual control model were obtained by Baumann et al. (1996), who failed to detect any induction during the oxic-to-anoxic switch when working with a nitrate-free medium. We think that the discrepancy between our results and those of Baumann et al. (1996) arises essentially from the different concentrations of oxygen in the culture medium used. A number of authors (e.g. Payne et al., 1971; Kucera et al., 1984; Boublíková et al., 1985; Aida et al., 1986) have already demonstrated that the anoxic adaptation of strictly respiring (non-fermenting) bacteria requires a residual respiration of O2 to produce the metabolic energy inevitable for biosynthesis of the denitrification pathway. This condition was possibly met for our cultures agitated in open flasks but not for Baumann's chemostat culture sparged with helium. Using an experimental set-up for oxygen limitation similar to ours, Härtig & Zumft (1999) observed a transient accumulation of denitrification gene transcripts in cultures of Pseudomonas stutzeri subjected to a low oxygen tension in the absence of N oxides. Further work is required to characterize such mRNA changes elicited by an appropriately reduced oxygen supply in P. denitrificans.


   ACKNOWLEDGEMENTS
 
We are grateful to Dr Stephen Spiro for kindly supplying the plasmid pRW2FF and critical reading of the first version of the manuscript. The work was supported by the Grant Agency of the Czech Republic (203/01/1589) and partly by the Ministry of Education, Youth and Sport (MSM14310 0005).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Aida, T., Hata, S. & Kusunoki, H. (1986). Temporary low oxygen conditions for the formation of nitrate reductase and nitrous oxide reductase by denitrifying Pseudomonas sp. G59. Can J Microbiol 32, 543–547.[Medline]

Arai, H., Kodama, T. & Igarashi, Y. (1997). Cascade regulation of the two CRP/FNR-related transcriptional regulators (ANR and DNR) and the denitrification enzymes in Pseudomonas aeruginosa. Mol Microbiol 25, 1141–1148.[Medline]

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (1995). Short Protocols in Molecular Biology. New York: Wiley.

Baker, S. C., Ferguson, S. J., Ludwig, B., Page, M. D., Richter, O.-M. H. & van Spanning, R. J. M. (1998). Molecular genetics of the genus Paracoccus: metabolically versatile bacteria with bioenergetic flexibility. Microbiol Mol Biol Rev 62, 1046–1078.[Abstract/Free Full Text]

Baumann, B., Snozzi, M., Zehnder, A. J. B. & van der Meer, J. R. (1996). Dynamics of denitrification activity of Paracoccus denitrificans in continuous culture during aerobic-anaerobic changes. J Bacteriol 178, 4367–4374.[Abstract]

Boublíková, P., Kucera, I. & Dadák, V. (1985). The effect of oxygen and nitrate on the biosynthesis of denitrification enzymes in Paracoccus denitrificans. Biologia (Bratislava) 40, 357–363.

Craske, A. & Ferguson, S. J. (1986). The respiratory nitrate reductase from Paracoccus denitrificans. Molecular characterization and kinetic properties. Eur J Biochem 158, 429–436.[Abstract]

Cruz-Ramos, H., Crack, J., Wu, G., Hughes, M. N., Scott, C., Thomson, A. J., Green, J. & Poole, R. K. (2002). NO sensing by FNR: regulation of the Escherichia coli NO-detoxifying flavohaemoglobin, Hmp. EMBO J 21, 3235–3244.[Abstract/Free Full Text]

de Boer, A. P., Reijnders, W. N., Kuenen, J. G., Stouthamer, A. H. & van Spanning, R. J. (1994). Isolation, sequencing and mutational analysis of a gene cluster involved in nitrite reduction in Paracoccus denitrificans. Antonie Van Leeuwenhoek 66, 111–127.[Medline]

de Vries, G. E., Harms, N., Hoogendijk, J. & Stouthamer, A. H. (1989). Isolation and characterization of Paracoccus denitrificans mutants with increased conjugation frequencies and pleiotropic loss of a (nGATCn)-DNA-modifying property. Arch Microbiol 152, 52–57.

Galimand, M., Gamper, M., Zimmermann, A. & Haas, D. (1991). Positive FNR-like control of anaerobic arginine degradation and nitrate respiration in Pseudomonas aeruginosa. J Bacteriol 173, 1598–1606.[Medline]

Green, J., Bennett, B., Jordan, P., Ralph, E. T., Thomson, A. J. & Guest, J. R. (1996). Reconstitution of the [4Fe–4S] cluster in FNR and demonstration of the aerobic–anaerobic transcription switch in vitro. Biochem J 316, 887–892.[Medline]

Härtig, E. & Zumft, W. G. (1999). Kinetics of nirS expression (cytochrome cd1 nitrite reductase) in Pseudomonas stutzeri during the transition from aerobic respiration to denitrification: evidence for a denitrification-specific nitrate- and nitrite-responsive regulatory system. J Bacteriol 181, 161–166.[Abstract/Free Full Text]

Hasegawa, N., Arai, H. & Igarashi, Y. (1998). Activation of a consensus FNR-dependent promoter by DNR of Pseudomonas aeruginosa in response to nitrite. FEMS Microbiol Lett 166, 213–217.[CrossRef][Medline]

Hutchings, M. I. & Spiro, S. (2000). The nitric oxide regulated nor promoter of Paracoccus denitrificans. Microbiology 146, 2635–2641.[Abstract/Free Full Text]

Hutchings, M. I., Shearer, N., Wastell, S., van Spanning, R. J. M. & Spiro, S. (2000). Heterologous NNR-mediated nitric oxide signalling in Escherichia coli. J Bacteriol 182, 6434–6439.[Abstract/Free Full Text]

Hutchings, M. I., Crack, J. C., Shearer, N., Thompson, B. J., Thompson, A. J. & Spiro, S. (2002). Transcription factor FnrP from Paracoccus denitrificans contains an iron–sulfur cluster and is activated by anoxia: identification of essential cysteine residues. J Bacteriol 184, 503–508.[Abstract/Free Full Text]

Jordan, P. A., Thomson, A. J., Ralph, E. T., Guest, J. R. & Green, J. (1997). FNR is a direct oxygen sensor having a biphasic response curve. FEBS Lett 416, 349–352.[CrossRef][Medline]

Khoroshilova, N., Beinert, H. & Kiley, P. (1995). Association of a polynuclear iron–sulfur center with a mutant FNR protein enhances DNA binding. Proc Natl Acad Sci U S A 92, 2499–2503.[Abstract]

Khoroshilova, N., Popescu, C., Münck, E., Beinert, H. & Kiley, P. (1997). Iron–sulfur cluster disassembly in the FNR protein of Escherichia coli by O2: [4Fe–4S] to [2Fe–2S] conversion with loss of biological activity. Proc Natl Acad Sci U S A 94, 6087–6092.[Abstract/Free Full Text]

Kucera, I. (1992). Oscillations of nitric oxide concentration in the perturbed denitrification pathway of Paracoccus denitrificans. Biochem J 286, 111–116.[Medline]

Kucera, I. & Kaplan, P. (1996). A study on the transport and dissimilatory reduction of nitrate in Paracoccus denitrificans using viologen dyes as electron donors. Biochim Biophys Acta 1276, 203–209.

Kucera, I. & Mat'chová, I. (1997). Iron as a possible mediator of the oxic-to-anoxic transition in Paracoccus denitrificans. Biochem Mol Biol Int 43, 305–310.[Medline]

Kucera, I., Boublíková, P. & Dadák, V. (1984). Function of terminal acceptors in the biosynthesis of denitrification pathway components in Paracoccus denitrificans. Folia Microbiol 29, 108–114.

Kucera, I., Mat'chová, I. & Dadák, V. (1990). Respiratory rate as a regulatory factor in the biosynthesis of the denitrification pathway of the bacterium Paracoccus denitrificans. Biocatalysis 4, 29–37.

Kucera, I., Mat'chová, I. & Spiro, S. (1994). Respiratory inhibitors activate an FNR-like regulatory protein in Paracoccus denitrificans: implications for the regulation of the denitrification pathway. Biochem Mol Biol Int 32, 245–250.[Medline]

Kwiatkowski, A. V., Laratta, W. P., Toffanin, A. & Shapleigh, J. P. (1997). Analysis of the role of the nnR gene product in the response of Rhodobacter sphaeroides 2.4.1 to exogenous nitric oxide. J Bacteriol 179, 5618–5620.[Abstract]

Lodge, J., Williams, R., Bell, A., Chan, B. & Busby, S. (1990). Comparison of promoter activities in Escherichia coli and Pseudomonas aeruginosa: use of a new broad-host-range promoter-probe plasmid. FEMS Microbiol Lett 67, 221–226.[CrossRef]

Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Nedoma, J. (1983). Determination of nitrates in waters using the salicylate method and chromotropic acid. Chem Listy 77, 638–641 (in Czech).

Parsonage, D., Greenfield, A. J. & Ferguson, S. J. (1985). The high affinity of Paracoccus denitrificans cells for nitrate as an electron acceptor. Analysis of possible mechanisms of nitrate and nitrite movement across the plasma membrane and the basis for inhibition by added nitrite of oxidase activity in permeabilised cells. Biochim Biophys Acta 807, 81–95.

Payne, W. J., Riley, P. S. & Cox, C. D., Jr (1971). Separate nitrite, nitric oxide, and nitrous oxide reducing fractions from Pseudomonas perfectomarinus. J Bacteriol 106, 356–361.[Medline]

Saunders, N. F. W., Houben, E. N. G., Koefoed, S., de Weert, S., Reijnders, W. N. M., Westerhoff, H. V., de Boer, A. P. N. & van Spanning, R. J. M. (1999). Transcription regulation of the nir gene cluster encoding nitrite reductase of Paracoccus denitrificans involves NNR and NirI, a novel type of membrane protein. Mol Microbiol 34, 24–36.[CrossRef][Medline]

Sears, H. J., Ferguson, S. J., Richardson, S. J. & Spiro, S. (1993). The identification of a periplasmic nitrate reductase in Paracoccus denitrificans. FEMS Microbiol Lett 113, 107–112.[CrossRef]

Simon, R., Priefer, U. & Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram negative bacteria. Bio/Technology 1, 1–6.

Spiro, S. (1992). An FNR-dependent promoter from Escherichia coli is active and anaerobically inducible in Paracoccus denitrificans. FEMS Microbiol Lett 98, 145–148.[CrossRef]

Tosques, I. E., Shi, J. & Shapleigh, J. P. (1996). Cloning and characterization of nnr, whose product is required for the expression of proteins involved in nitric oxide metabolism in Rhodobacter sphaeroides 2.4.3. J Bacteriol 178, 4958–4964.[Abstract]

van Spanning, R. J. M., de Boer, A. P. N., Reijnders, W. N. M., Spiro, S., Westerhoff, H. V., Stouthamer, A. H. & van der Oost, J. (1995). Nitrite and nitric oxide reduction in Paracoccus denitrificans is under the control of NNR, a regulatory protein that belongs to the FNR family of transcriptional activators. FEBS Lett 360, 151–154.[CrossRef][Medline]

van Spanning, R. J. M., de Boer, A. P. N., Reijnders, W. N. M., Westerhoff, H. V., Stouthamer, A. H. & van der Oost, J. (1997). FnrP and NNR of Paracoccus denitrificans are both members of the FNR family of transcriptional activators but have distinct roles in respiratory adaptation in response to oxygen limitation. Mol Microbiol 23, 893–907.[CrossRef][Medline]

van Spanning, R. J. M., Houben, E., Reijnders, W. N. M., Spiro, S., Westerhoff, H. V. & Saunders, N. (1999). Nitric oxide is a signal for NNR-mediated transcription activation in Paracoccus denitrificans. J Bacteriol 181, 4129–4132.[Abstract/Free Full Text]

Wood, N. J., Alizadeh, T., Bennett, S., Pearce, J., Ferguson, S. J., Richardson, D. J. & Moir, J. W. B. (2001). Maximal expression of membrane-bound nitrate reductase in Paracoccus is induced by nitrate via a third FNR-like regulator named NarR. J Bacteriol 183, 3606–3613.[Abstract/Free Full Text]

Wu, G., Cruz-Ramos, H., Hill, S., Green, J., Sawers, G. & Poole, R. K. (2000). Regulation of cytochrome bd expression in the obligate aerobe Azotobacter vinelandii by CydR (Fnr). Sensitivity to oxygen, reactive oxygen species, and nitric oxide. J Biol Chem 275, 4679–4686.[Abstract/Free Full Text]

Received 6 June 2003; revised 20 August 2003; accepted 15 September 2003.