Expression of the Escherichia coli yfiD gene responds to intracellular pH and reduces the accumulation of acidic metabolic end products

Neil R. Wyborn1, Sarah L. Messenger1, Robin A. Henderson1, Gary Sawers2, Ruth E. Roberts1, Margaret M. Attwood1 and Jeffrey Green1

The Krebs Institute for Biomolecular Research, Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield, S10 2TN, UK1
Department of Molecular Microbiology, John Innes Centre, Norwich NR4 7UH, UK2

Author for correspondence: Jeffrey Green. Tel: +44 114 222 4403. Fax: +44 114 272 8697. e-mail: jeff.green{at}sheffield.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The YfiD protein of Escherichia coli has been reported to be an acid-inducible protein. Here it is shown that expression of a yfiD::lac reporter fusion is enhanced up to 3·5-fold during acidic growth. The anaerobic transcription factor FNR was confirmed as the major regulator of yfiD expression, and ArcA was found to enhance anaerobic yfiD expression, probably by displacing a repressing FNR dimer in the -93·5 region of the promoter. Moreover, the pyruvate sensor PdhR was shown to act as a minor anaerobic repressor of yfiD expression. On the basis of its strong homology to the C-terminal region of pyruvate formate-lyase (PFL) it was predicted that YfiD would be a radical-containing enzyme. The YfiD radical was found to be introduced by the PFL-activase enzyme, but unlike PFL, AdhE did not deactivate radicalized YfiD. The extent of radical activation of YfiD was enhanced by low intracellular pH, and thus it was concluded that both yfiD expression and YfiD activity are affected by growth at low pH. The yfiD mutant strain JRG4033 excreted increased levels of organic acids compared to the parental strain when grown in chemostat culture under oxygen-starved conditions, consistent with the acid-inducibility of yfiD expression and the recently reported ability of YfiD to rescue the activity of oxygenolytically cleaved PFL.

Keywords: acid stress, radical protein, pH homeostasis, pyruvate formate-lyase, formate

Abbreviations: dO2, dissolved oxygen tension; FNR, fumarate and nitrate reduction regulator; GST, glutathione S-transferase; PFL, pyruvate formate-lyase


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Much of the success of bacteria depends on the ability to survive and thrive in adverse conditions. Escherichia coli is an adaptable bacterium that has evolved mechanisms to allow it to cope with conditions as diverse as the gut and the sewer. These different environments present particular challenges such as wide variations in oxygen availability, pH and osmolarity. The ability to adapt to these ever changing environmental factors is ultimately rooted in altered patterns of gene expression. E. coli is a facultative anaerobe that uses combinations of transcription factors to control gene expression in response to environmental oxygen. Predominant amongst these transcription factors is FNR (fumarate and nitrate reduction regulator), a direct oxygen sensor that functions mainly to activate the expression of genes required during anaerobic growth (Kiley & Beinert, 1999 ). FNR senses anaerobiosis through the acquisition of an oxygen-labile [4Fe–4S] cluster, which promotes dimerization of the protein and enhances site-specific DNA binding (Lazzazera et al., 1996 ; Jordan et al., 1997 ; Popescu et al., 1998 ). Analysis of the cytosolic polypeptides produced during anaerobic growth of E. coli revealed that the yfiD gene product (YfiD) was a member of the FNR regulon (Green & Baldwin, 1997 ). Searches for other E. coli proteins with amino acid sequences related to YfiD revealed a strong homology (77% identity over a 64 aa overlap) to the C-terminal region of pyruvate formate-lyase (PFL), the product of another FNR-regulated gene, pflB. As well as yfiD, E. coli possesses three more genes whose products are predicted to be similar to PFL: tdcE, pflD and f810 (Sawers & Watson, 1998 ). In E. coli the yfiD gene is located in the srmBung intergenic region and genes potentially encoding YfiD proteins have been identified in a wide range of Gram-negative and -positive bacteria including Serratia sp., Serratia liquefaciens and Haemophilus influenzae, as well as the bacteriophages T4 and T5. The H. influenzae yfiD gene is, like that of E. coli, close to the ung genes and for bacteriophage T4 and T5 it is close to the thymidine kinase gene and the tRNA gene cluster, respectively. In both Serratia species the yfiD gene is associated with the phlA gene, which encodes extracellular phospholipase A1. All the encoded proteins contain a highly conserved pentapeptide motif (RVXGY) around the C-terminal glycyl residue (G734 in PFL), which is the site of the glycyl radical produced during the activation of PFL (Sawers & Watson, 1998 ). The PFL-like proteins also contain the adjacent cysteinyl residues characteristic of the PFL active site. However, YfiD contains only one cysteine residue and an active site cannot be confidently predicted on the basis of its amino acid sequence. The yfiD promoter has an unusual architecture and has proved a useful tool yielding new insights into FNR-mediated repression (Green et al., 1998 ; Green & Marshall, 1999 ; Marshall et al., 2001 ). Expression of yfiD is activated by FNR acting at -40·5, a conventional activation position, but is downregulated by a second FNR dimer at -93·5, which leads to maximal yfiD expression during microaerobic growth (Green et al., 1998 ; Marshall et al., 2001 ).

The yfiD gene product has appeared in five proteomic screens. Firstly, as a polypeptide upregulated by expression of the heterologous oxygen-responsive FNR protein, HlyX, in anaerobically grown E. coli (Green & Baldwin, 1997 ). In the second analysis, YfiD was induced to high levels when the cytoplasm was acidified, i.e. during growth at pH 4·4, and upon the addition of propionate at pH 6 (Blankenhorn et al., 1999 ). Interestingly, the effects of pH on gene expression are known to overlap with many other environmental factors, but particularly with oxygen availability (Slonczewski & Foster, 1996 ). It has been noted that the acid-induced expression of amino acid decarboxylases is enhanced under anaerobic conditions (Meng & Bennett, 1992 ; Neely et al., 1994 ) and that expression of cytochrome o is repressed by acid stress (Cotter et al., 1990 ). It was therefore suggested that there is a complex relationship between YfiD, pH and oxygen (Blankenhorn et al., 1999 ). Thirdly, YfiD was decreased in gel-entrapped E. coli, and on the basis of its transient expression pattern it was suggested that YfiD is an early stationary phase protein (Perrot et al., 2000 ). In the fourth analysis, enhanced yfiD expression was noted in strains harbouring heterologous polyhydroxyalkanoate biosynthesis genes (Han et al., 2001 ). It was again suggested that YfiD was present under these particular conditions because the cytoplasm becomes more acidic. Finally, YfiD was detected as a phosphorylated protein in L-forms of E. coli (Freestone et al., 1998 ). Thus, although the exact physiological role of YfiD has not yet been assigned, this protein has emerged from a range of studies related to anaerobic and acidic growth. Therefore, because yfiD expression is regulated by FNR and the YfiD gene product is acid-inducible, we have investigated the effects of anaerobiosis, pH and growth phase on yfiD expression and activation of the YfiD protein. We find that the major regulator of yfiD expression is FNR, but that yfiD expression is induced at the level of transcription during growth under acidic conditions. Furthermore, the activation of YfiD by PFL activase is enhanced by growth at low pH in the presence of propionate or benzoate. In addition, accumulation of lactate, succinate, pyruvate and formate is greater for the yfiD mutant strain than for the parent, consistent with a role for YfiD in reducing exposure to acidic metabolic end products and in directing carbon flux.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids and growth conditions.
E. coli strains used were: the parental strain, MC1000 [{Delta}(lacIPZOYA)-X74 galU galK strr thi araD139 {Delta}(araABC–leu)]; an fnr strain, JRG1728 [MC1000 {Delta}(tyrR–fnr–rac–trg)17 zdd-230::Tn9]; and a pdhRc strain JRG2547 (Quail & Guest, 1995 ). The PFL activase (pflA) and deactivase (adhE) deficient strains were MC4100 [{Delta}ara {Delta}(argF–lac) rpsL relA1 flbB deoC1 ptsF25] derivatives, 234M11 (Heßlinger et al., 1998 ) and RM235, respectively. The arcA strain was BW3218 {Delta}lac-169 {Delta}(serB–trpR–arcA–thr), and this strain was complemented with pRB38, a multicopy plasmid encoding ArcA. E. coli strain W3110 was used as a prototroph to construct the yfiD null mutant (JRG4033) by homologous allelic exchange of chromosomal yfiD with yfiD::kanr. This was achieved using a pMAK705 (Hamilton et al., 1989 ) derivative, pGS1193, containing 1010 bp of 5' flanking DNA (beginning 44 bp downstream of the ATG start codon of yfiD and extending 966 bp upstream), a kanr cassette (isolated from pUC4-KIXX, Amersham-Pharmacia) and 1262 bp of 3' flanking DNA (beginning 193 bp upstream of the yfiD TAA stop codon and extending 1069 bp downstream of the stop codon). Thus pGS1193 contained an insert in which 147 bp of yfiD coding sequence was replaced by a kanr cassette. The direction of transcription of the kanr cassette was shown to be colinear with yfiD by diagnostic restriction digests. Once integrated it is unlikely that the disrupted yfiD gene could cause any polar effects because it is not part of an operon, its transcriptional terminators remain intact and the flanking genes (yfiK and ung) are transcribed in the opposite direction to yfiD. E. coli W3110 was transformed with pGS1193 and grown at non-permissive (42 °C) and permissive (30 °C) temperatures in the presence or absence of antibiotic(s) as appropriate to effect plasmid integration and subsequent resolution of the mutant (Hamilton et al., 1989 ). A representative strain, designated JRG4033, was identified after plasmid-curing as a yfiD null mutant on the basis of its stable kanr phenotype, the size of DNA fragments produced from PCR amplification of the disrupted yfiD gene from genomic DNA (not shown), Southern hybridization (not shown) and Western blotting (see Fig. 1). E. coli BL21 [F- ompT hsdS ( ) gal] transformed with pGS1011, constructed by ligating the yfiD-containing NcoI–SalI fragment from pGS1036, a pUC118 derivative containing a PCR-generated NcoI–SalI flanked yfiD coding region, into pGEX-KG (Guan & Dixon, 1991 ), was used to overproduce the GST–YfiD fusion protein. The same NcoI–SalI fragment was transferred into ptac85 to create the yfiD expression plasmid pGS1207, which was used to assess the effects of mutations in the PFL activase and deactivase genes on YfiD. For monitoring expression of yfiD under a variety of conditions, the low copy number (~5 per chromosome) pRW50-based yfiD::lac reporter plasmid pGS1000 was used (Green et al., 1998 ).



View larger version (7K):
[in this window]
[in a new window]
 
Fig. 1. Production of the YfiD protein is FNR-dependent. Western blot of stationary phase anaerobic cultures of the yfiD mutant, JRG4033, the parental strain, W3110 and an fnr mutant, JRG1728. YfiD was not detected in JRG4033 and was barely detectable in JRG1728. Two cross-reacting polypeptides were present in the extracts of the parental strain, equivalent to the full-length YfiD protein (YfiD) and a smaller product (YfiD’). The amount of YfiD’ detected in these stationary phase cultures was low because activation of YfiD was found to be optimal in the exponential phase of growth (see Fig. 4).

 
The media used to test the effects of pH on yfiD expression and YfiD activation was based on LBK broth (10 g tryptone l-1, 5 g yeast extract l-1, 6·4 g KCl l-1) containing an appropriate buffer (100 mM: LBK/MOPS, pH 7·2; LBK/MES, pH 6·3; and LBK/HOMOPIPES, pH 4·5) (Blankenhorn et al., 1999 ). These media were supplemented with propionic acid (50 mM) or benzoic acid (20 mM) as indicated to further depress the intracellular pH (Blankenhorn et al., 1999 ). Cultures were grown in the indicated buffered media supplemented with appropriate antibiotics (200 µg ampicillin ml-1, 35 µg tetracycline ml-1) to maintain plasmids. Anaerobic growth was in sealed bottles at 37 °C for 16 h for most expression studies. When cultures were analysed at different points in the growth cycle, samples were taken in a Mk3 Anaerobic Workstation (Don Whitley). Aerobic cultures were grown in 250 ml conical flasks containing 10 ml medium with vigorous shaking at 37 °C for 16 h. ß-Galactosidase activities were measured in duplicate for at least two independent cultures, as described by Miller (1972) .

Protein purification.
YfiD was initially purified as a glutathione S-transferase (GST)–YfiD fusion protein amplified in aerobically grown E. coli BL21(pGS1011). Cultures (500 ml) were grown at 37 °C with vigorous shaking in 2xYT medium (Sambrook et al., 1989 ) for 3 h, at which point the expression of the gst–yfiD fusion was initiated by addition of IPTG to a final concentration of 30 mg l-1. Incubation was continued for a further 3 h before the cultures were harvested by centrifugation. Clarified French pressure cell extracts (up to 150 mg total protein, in 10 mM Tris/HCl, pH 8·0 containing 10 mM benzamidine and 0·1 mM PMSF) were applied to a column (50x15 mm) of glutathione–agarose (Sigma) equilibrated with 50 mM Tris/HCl, pH 6·8 containing 0·5 mM DTT. The column was washed with 10 vols of the same buffer before eluting the GST–YfiD fusion with 50 mM Tris/HCl, pH 8·0 containing 10 mM reduced glutathione. Aliquots of the fusion protein were dialysed to remove glutathione and the pure YfiD protein was obtained by cleavage of the desalted GST–YfiD fusion with thrombin (14 U) for 2 h at 37 °C, followed by chromatography on glutathione–Sepharose to remove the released GST. The product contains an extra 15 aa, which are derived from the thrombin-sensitive linker, whose presence was confirmed by the total amino acid analysis and N-terminal sequencing. Protein concentrations were estimated with the Bio-Rad protein reagent.

Characterization of YfiD.
Total metal ion contents and electrospray mass spectrometry measurements were done with redissolved freeze-dried samples of isolated YfiD that had been dialysed against several changes of 1 mM ammonium acetate, pH 6·0. The native molecular mass of YfiD was also estimated by gel filtration on a Superdex 200 HR 10/30 column (Amersham-Pharmacia) equilibrated with 50 mM sodium phosphate buffer, pH 7·0 containing 0·15 M sodium chloride, and calibrated with a 6500–66000 molecular weight marker kit for gel filtration (Sigma). N-terminal amino acid sequencing was achieved by Edman degradation using an Applied Biosystems protein sequencer. Standard technologies were used for Western blotting. Polyclonal YfiD anti-sera were raised in rabbits (dwarf lop) and used at a final dilution of 1:20000. Detection of cross-reacting material was achieved using the Amersham-Pharmacia ECL detection system with horseradish peroxidase conjugated secondary antibodies, or by nitro blue tetrazolium/X-phosphate for alkaline phosphatase conjugated secondary antibodies, and was quantified using ImageMaster software (Amersham-Pharmacia).

Continuous culture.
Chemostat cultures were grown in LH Fermentation chemostat vessels (either 1 l nominal capacity, 750 ml working volume or 2 l nominal capacity, 1130 ml working volume) maintained at 37 °C. The minimal medium had the following composition (g l-1): (NH4)2SO4, 2·0; K2HPO4, 1·0; NaH2PO4, 1·0; MgSO4.7H2O, 0·2; supplemented with trace element solution (Vishniac & Santer, 1957 ), 2 ml; and glucose, 30 mM. The pH was adjusted to 7·0±0·2 by automatic titration with sterile 2 M KOH. Dissolved oxygen levels were monitored with galvanic oxygen electrodes. For aerobic cultures the culture agitation speed (400–500 r.p.m.) and air flow were adjusted to maintain a dissolved oxygen tension (dO2) >50% air saturation or <5% for microaerobic cultures. For oxygen-starved cultures, the air supply was replaced by oxygen-free nitrogen. In some experiments air and nitrogen mixtures were used to gas the cultures. Inocula were grown as batch cultures overnight in the continuous culture medium.

Glucose in feed medium and culture filtrates was estimated enzymically with glucose oxidase (Bergmeyer & Bernt, 1968 ). Low molecular mass overmetabolites in culture filtrates were monitored by HPLC using a Shodex Ion pak KC-811 column (8x300 mm) at 50 °C. Samples were eluted with 0·1% (v/v) phosphoric acid at a flow rate of 2·0 ml min-1 through detectors for UV (210 nm) and refractive index (Waters 490E and Waters 410, respectively) coupled to a data module (Millennium 2010). Peak areas were proportional to concentrations, which were determined using standard solutions of each organic acid. Acetate, formate and ethanol were also estimated using enzyme-based Boehringer Mannheim kits, which also served to confirm these peak assignments in the HPLC elution profiles; the estimates obtained by this method were similar to those obtained by calibrated HPLC.

To investigate whether the yfiD null mutant is physiologically disadvantaged compared to the parent, a chemostat containing carbon-limited medium was inoculated with approximately equal numbers of W3110 and JRG4033 bacteria by combining equivalent overnight cultures of these strains. The mixed culture was allowed to grow as a batch culture in the fermentation vessel for ~8 h before switching to continuous growth with a dilution rate of 0·2 h-1. Samples (~5 ml) were taken immediately after switching to continuous growth and subsequently every 5–10 generations. The proportion of each strain in the culture vessel was determined by plating serial dilutions (in triplicate) of the culture on carbon-limited agar plates with and without kanamycin (30 mg l-1). The plates were incubated for 40 h at 37 °C before counting the number of c.f.u. recovered. The numbers were averaged and the percentage of JRG4033 (kanr bacteria) was calculated. The competition experiments were performed over 50 generations under aerobic (50% dO2) and microaerobic (5% dO2) conditions.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Expression of ß-galactosidase from a yfiD::lac reporter is acid-inducible
The starting point for this work was the report that the yfiD gene product is induced by acid stress (Blankenhorn et al., 1999 ). To test if this was caused by enhanced expression of yfiD during acidic growth we used our existing low copy number yfiD::lac reporter, pGS1000 (Green et al., 1998 ), to monitor the expression of yfiD both aerobically and anaerobically at various pH values. The pH of batch cultures was set by buffering at the indicated pH, irrespective of the addition of 50 mM propionic or 20 mM benzoic acids, which act as membrane-permeant weak acids lowering the intracellular pH of E. coli (Blankenhorn et al., 1999 ). The data obtained revealed that, at pH 7·2, yfiD::lac expression was induced >10-fold under anaerobic (3910 Miller U) compared to aerobic conditions (310 Miller U) (Table 1). Lowering the pH of the culture medium, or the intracellular pH by adding the weak acids propionate or benzoate, enhanced yfiD expression up to a further 3·5-fold anaerobically (3910 Miller U at pH 7·2, compared to 13590 Miller U at pH 6·3 in the presence of propionate) and 2·5-fold aerobically (310 Miller U at pH 7·2, compared to 790 Miller U at pH 6·3 in the presence of propionate). The ß-galactosidase activities of cultures with the lowest intracellular pH values (i.e. those grown at pH 6·3 plus benzoate, or at pH 4·5) were unexpectedly low, especially when grown anaerobically, indicating that ß-galactosidase is unstable, yfiD expression is reduced, or that the reporter plasmid copy number has fallen to such an extent that ß-galactosidase activity no longer faithfully reflects transcription of yfiD at these low pH values. To test which of these possibilities might be correct, the YfiD protein was purified (see below) and antibodies were raised to determine the amount of YfiD protein in the cultures directly by Western blotting. These experiments revealed that for anaerobic cultures the cellular content of YfiD protein followed the expected trend (YfiD content increased as yfiD::lac expression increased) except at the lowest pH values (pH 4·5 and at pH 6·3 in the presence of benzoate) where, contrary to expectations based on the yfiD::lac reporter data, the intracellular levels of YfiD were the greatest (Table 1). Thus it was concluded that at all but the lowest pH values, the ß-galactosidase activities arising from the yfiD::lac reporter plasmid (pGS1000) reflected the production of YfiD protein. The possibility that the difference between the two measures of YfiD production at the lowest pH values may arise from a reduction in reporter plasmid copy number under these conditions cannot be excluded, but because it is known that ß-galactosidase is sensitive to pH values below 6·0 (Degraeve et al., 1996 ), the low reporter activities for cultures grown at the lowest pH values are most simply explained by reduced stability of ß-galactosidase. An alternative explanation for the observations would require a reduction in yfiD expression combined with enhanced stability of YfiD protein in acidic cultures.


View this table:
[in this window]
[in a new window]
 
Table 1. Expression of a yfiD::lac fusion and intracellular YfiD content during acid stress

 
That FNR is the major transcription regulator of yfiD expression was confirmed by Western blotting, which showed that extracts of an fnr mutant strain (JRG1728) grown anaerobically at pH 7·2 contained very little YfiD protein compared to the parent strain, indicating that anaerobic expression of yfiD under these conditions is FNR dependent (Fig. 1). Immunoblots also revealed that the cellular content of YfiD was enhanced by growth at low pH under aerobic conditions (Table 1). In aerobic cultures the activity of the anaerobic transcription factor FNR is minimal, and consequently yfiD::lac expression is low, but still pH responsive (Table 1). This suggests either that acidic conditions can enhance FNR-mediated expression of yfiD under aerobic conditions or that an alternative acid-responsive regulator may be involved. Nevertheless, it can be concluded that acid stress does increase expression from the yfiD promoter (Table 1), and that FNR is the major regulator of yfiD expression (Fig. 1).

It has long been recognized that there are overlaps in bacterial responses to anaerobiosis and acid stress at the level of gene regulation (Cotter et al., 1990 ; Meng & Bennett, 1992 ; Neely et al., 1994 ; Slonczewski & Foster, 1996 ; Blankenhorn et al., 1999 ). Therefore, we considered whether FNR and/or ArcAB, the major oxygen-sensing systems of E. coli contributed to the acid enhancement of yfiD expression. This could be achieved by enhancement of the activities of either, or both, regulatory systems at lower intracellular pH values. The data presented in Table 2 and Fig. 1 confirmed that FNR is responsible for the anaerobic induction of yfiD expression with dramatically reduced anaerobic yfiD::lac expression reported for the fnr mutant strain (200 Miller U at pH 7·2) compared to the parent (2990 Miller U at pH 7·2). As expected from the data in Table 1, lowering the pH of the cultures increased anaerobic yfiD::lac expression for the parent (from 2990 Miller U at pH 7·2 to 4210 Miller U at pH 6·3) but there was only a small increase for anaerobic cultures of the fnr mutant (from 200 Miller U at pH 7·2 to 230 Miller U at pH 6·3). To test the possibility that FNR may be directly involved in the observed acid response, expression from the simple FNR-regulated synthetic promoter FF-41.5 (Wing et al., 1995 ) was investigated at pH 7·2 and 6·3. The experiments revealed that although FNR-dependent expression from this promoter was subject to a small but reproducible enhancement after growth at low pH, it is unlikely that FNR activity is modulated by intracellular pH (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2. Effect of acidity on FNR, ArcA and PdhR activities

 
The suggestion that the acid-responsive yfiD expression is independent of FNR was supported by Western blot analyses of cell-free extracts of MC1000(pGS1000) (fnr+)and JRG1728(pGS1000) ({Delta}fnr) prepared after aerobic and anaerobic growth at pH 7·2 in the presence or absence of benzoate. The data obtained indicated that the intracellular content of YfiD was increased in response to the presence of benzoate. This enhancement was independent of FNR and dissolved oxygen tension, because even in the fnr strain, JRG1728, grown aerobically in the presence of benzoate, the intracellular content of YfiD increased ~fourfold compared to cultures grown in the absence of benzoate (not shown). Furthermore, the FNR-independent aerobic increase in YfiD protein content at low pH correlated with enhanced expression of yfiD::lac in these cultures (110 Miller U for cultures without added benzoate, compared to 460 Miller U for those with benzoate), suggesting that although FNR may contribute to increasing yfiD expression during growth under acidic conditions it is not essential for acid induction of yfiD expression and that another factor is mediating this effect.

Further analysis of the yfiD promoter revealed the absence of consensus sites for regulators such as PhoP and OmpR that are known to be involved in the response of E. coli to acid. However, two overlapping putative ArcA sites (7/10 bases matching the consensus centred at -97, AGTTgATgtA, and -94, TGTaAAacAA; non-consensus bases in lower case) were detected within the far upstream (-93·5) FNR site that is responsible for the anaerobic downregulation of yfiD expression (Green et al., 1998 ), suggesting that ArcA might prevent FNR binding at this location and thereby increase yfiD expression. Therefore, the effects of an arcA mutation on yfiD::lac expression were tested (Table 2). Anaerobic expression of yfiD was 2–2·7-fold lower in an arcA mutant strain (2540 Miller U compared to 6810 Miller U for the same strain expressing arcA from the plasmid pRB38 at pH 7·2, and 4470 Miller U compared to 9210 Miller U at pH 6·3), indicating that ArcA has a positive effect on yfiD expression. The simplest explanation for this data, based on our knowledge of FNR-mediated regulation of yfiD expression (Green et al., 1998 ; Green & Marshall, 1999 ; Marshall et al., 2001 ), is that ArcA relieves the repression mediated by the far upstream FNR dimer by competing for occupation of the -93 region of the yfiD promoter. Expression of the solely FNR-regulated model promoter FF-41.5 (Wing et al., 1995 ) was tested in the arcA background to determine whether there was any effect of removing ArcA on a simple FNR-driven promoter. Deletion of arcA had no significant effect on FNR-mediated expression from FF-41.5 at both pH 7·2 and pH 6·3, and thus the ArcA-mediated enhancement of yfiD expression (Table 2) is probably mediated by ArcA competing with FNR for occupancy of the -93 region of the yfiD promoter, where FNR can act negatively and ArcA positively, and that the positive effects of ArcA are magnified by the presence of arcA in multiple copies. However, this mechanism is not directly responsible for the anaerobic acid induction of yfiD expression because the yfiD promoter remains pH-responsive in the arcA background (2540 Miller U at pH 7·2, compared to 4470 Miller U at pH 6·3) (Table 2).

Analysis of the yfiD promoter region revealed the presence of a near consensus (cAaTGGTTTTACCAATT, 12/14, non-consensus bases in lower case) PdhR binding site (Quail & Guest, 1995 ) centred at +17 relative to the transcript start. The PdhR protein is responsible for repressing the expression of the pyruvate dehydrogenase complex, and thus plays a key role in the regulation of the central metabolic pathways in E. coli. PdhR senses the status of the intracellular pyruvate pool by binding the co-effector, pyruvate, and because the PdhR–pyruvate complex cannot bind target DNA, expression of the pdhR–aceEF–lpd operon, encoding pyruvate dehydrogenase, is derepressed when pyruvate is available (Quail & Guest, 1995 ). The expression of the yfiD::lac reporter was tested in JRG2547, a strain that carries the pdhRc mutation that prevents PdhR binding to target DNA (Quail & Guest, 1995 ). The data show that, as might be predicted from the position of the PdhR box, PdhR acts as a weak anaerobic repressor of yfiD expression (2990 Miller U at pH 7·2 for the parent compared to 3740 Miller U for the pdhRc mutant). Furthermore, because acidic growth conditions still enhance yfiD expression in a pdhRc background (3740 Miller U at pH 7·2 compared to 4950 Miller U at pH 6·3), it would appear that PdhR does not have a major role in acid-induced yfiD expression (Table 2).

A yfiD null mutant accumulates increased levels of acidic metabolic end products
To attempt to define a role for the YfiD protein, a yfiD null mutant strain (JRG4033) was created by internal deletion and insertion of a kanamycin resistance cassette into the yfiD gene of E. coli W3110. After Western blotting to confirm that JRG4033 was unable to produce YfiD protein (Fig. 1), a range of phenotypic tests were applied to the mutant, including: monitoring growth and viability of cultures grown aerobically and anaerobically on a variety of carbon sources; sensitivity to UV irradiation, ethanol and freeze–thaw cycling; and morphological differences under light and electron microscopy. However no obvious phenotype was indicated by these tests. To investigate whether YfiD plays a role in resisting acid stress, the responses of the parent (W3110) and the yfiD mutant (JRG4033) to acid shock were monitored. Under aerobic and anaerobic conditions the parent and mutant responded similarly in viability tests to acid shock at pH 2·5 (inorganic acid) or at pH 4·5 (weak organic acid), after growth at pH 7·2 or pH 5·7 (not shown), indicating no fundamental role for YfiD in resisting, or adapting to, acid stress. Therefore, it was concluded that the role of YfiD must be more subtle than could be revealed by the simple tests employed.

Because the tests described above failed to reveal a phenotype, and because yfiD expression is maximal under microaerobic conditions (Marshall et al., 2001 ), the properties of the yfiD mutant (JRG4033) and parent (W3110) strains were assessed in aerobic (50% dO2) and microaerobic (5% dO2) continuous culture competition experiments. When grown aerobically, the proportions of parent and mutant in the culture were essentially constant over 50 generations. However, when cultured microaerobically (<5% dO2) the proportion of the yfiD null mutant, as estimated by the relative number of kanamycin-resistant (yfiD null mutant) c.f.u., was reduced from 49% to 29%, and from 45% to 23%, in two independent experiments, within 35 generations. These data indicate that the yfiD mutant strain has a microaerobic growth defect. To investigate this further, the properties of the mutant (JRG4033) and parent (W3110) were compared in chemostat cultures under both aerobic and oxygen-starved conditions. The measured growth parameters (µmax and Ymax, max growth rate and max growth yield respectively) indicated that the disruption of the yfiD gene did not have a profound effect on the growth of E. coli under oxygen-starved conditions, confirming the observations made in batch culture. However, analysis of the culture medium of the yfiD mutant strain showed that it contained elevated levels of lactate, succinate, pyruvate and formate compared to the parental strain (Table 3). The concentrations of other metabolic end products (carbon dioxide, acetate and ethanol) were similar for both strains. During growth in the presence of oxygen there was no significant accumulation of any of the overmetabolites by either strain (not shown). These observations indicate that, directly or indirectly, YfiD acts to prevent the accumulation of lactate, succinate, pyruvate and formate during oxygen-starved growth, either by routing carbon flux away from these products or by removing them once they have formed. Such a role is in accord with the anaerobic- and acid-inducibility of yfiD expression, and also with the roles of other acid-inducible genes whose products appear to remove, or to prevent the accumulation of, acidic compounds.


View this table:
[in this window]
[in a new window]
 
Table 3. Physiological parameters of oxygen-starved chemostat cultures

 
Aerobic overproduction of the YfiD protein
The lac reporter work described above shows that yfiD expression is dramatically increased during anaerobiosis in an FNR-dependent manner and that the imposition of acidic growth conditions further enhances anaerobic yfiD expression. However, in a previous proteomic analysis of acid-inducible E. coli proteins it was reported that although YfiD was acid induced (in accord with the transcription data reported here) it was present at similar levels in both aerobically and anaerobically grown cultures (Blankenhorn et al., 1999 ). One possible explanation for this original observation is based on the prediction that under anaerobic conditions, YfiD is a radical-containing protein, and that exposure to oxygen would cause the activated protein to be cleaved in a manner analogous to that observed for PFL. Alternatively, intracellular pH may be influencing the extent of radical incorporation and hence activation of YfiD. Either of these possibilities could cause the YfiD spot on the 2D gels to appear similar under aerobic and anaerobic conditions. To test these possibilities it was necessary to isolate the YfiD protein to allow the predicted radical-activation mechanism to be investigated and to raise antibodies to determine the extent of YfiD activation under acidic growth conditions. Therefore, the YfiD protein was overproduced as a GST fusion protein from pGS1011. After induction with IPTG, the GST–YfiD fusion formed up to 16% of the soluble cell protein and the yield of pure YfiD was 19 mg (l culture)-1. Isolated overproduced YfiD had a subunit Mr of 15000 and was estimated to be >=95% pure after densitometric analysis of Coomassie brilliant blue stained SDS-polyacrylamide gels (Fig. 2). Because of the position of the thrombin cleavage site in the GST–YfiD fusion protein, and the presence of an engineered NcoI site overlapping the initiating ATG codon, the released YfiD protein has an additional 15 N-terminal amino acids (which derive from the vector-encoded linker) and an I2V substitution. The N-terminal amino acid sequence (GSPGISGGGGGILDSMVTGIQ) of the YfiD protein confirmed the presence of the additional amino acids and that the purified protein was YfiD.



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 2. SDS-PAGE analysis of overproduced YfiD protein. Left, molecular mass markers; right, purified YfiD protein (10 µg). Polypeptides were visualized by Coomassie blue staining.

 
Properties of aerobically isolated YfiD
The native molecular mass of YfiD estimated by gel filtration after purification from aerobic cultures indicated that it was a monomeric protein of Mr ~15000. Electrospray mass spectrometry of YfiD revealed a molecular mass of 15486 Da, which closely matches the 15482 Da predicted from the nucleotide sequence of yfiD plus the linker sequence, and indicates the absence of any post-translational modification. The UV-visible and EPR (electron paramagnetic resonance) spectra of these preparations were featureless, except for a maximum at 276 nm ({epsilon}276 15·7 mM-1 cm-1) in the absorption spectra. Furthermore, total metal ion analysis revealed that YfiD was metal-ion free except for a trace of Zn (0·1 mol Zn per mol of YfiD).

Expression of yfiD is enhanced in stationary phase
Previous studies indicated that yfiD is expressed throughout the anaerobic growth cycle but that expression increased a further 1·7-fold in stationary phase cultures compared to exponential phase (Green et al., 1998 ). A more thorough analysis of the growth phase-dependency of yfiD::lac expression was undertaken following the suggestion that YfiD is an early stationary phase protein (Perrot et al., 2000 ). The data revealed that in cultures buffered at pH 7·2, aerobic expression of yfiD::lac was low throughout the growth cycle, but increased as growth proceeded (Fig. 3). During the anaerobic growth cycle of similarly buffered cultures, yfiD::lac expression was much higher than observed aerobically, but followed the same pattern, with maximum ß-galactosidase activity being observed in stationary phase. Accordingly, under these conditions, YfiD protein was at the limits of detection by Western blotting of whole cell samples obtained at intervals during aerobic growth (<1000 molecules per cell), but accumulated in anaerobically grown cultures reaching a concentration of approximately 100000 molecules per cell (or 3·4% of total protein) after 7 h growth (not shown). This level of YfiD protein was maintained even after 48 h, indicating that YfiD is not an early stationary phase protein under these growth conditions, but that it accumulates throughout the growth cycle and is present for at least 48 h after growth has ceased.



View larger version (13K):
[in this window]
[in a new window]
 
Fig. 3. Temporal expression of yfiD::lac during aerobic and anaerobic growth of E. coli. Cultures of MC1000 ({Delta}lac) carrying the yfiD::lac reporter plasmid, pGS1000, were grown aerobically and anaerobically in medium buffered at pH 7·2 for 2 days at 37 °C. At intervals the cultures were sampled, the extent of growth estimated by measuring the OD600 and yfiD promoter activity was determined by measuring ß-galactosidase activity (Miller, 1972 ). Aerobic cultures (solid lines): growth ({blacktriangleup}) and ß-galactosidase activity ({bullet}). Anaerobic cultures (broken lines): growth ({blacksquare}) and ß-galactosidase activity ({diamondsuit}).

 
YfiD is a radical protein that shares the PFL activase
In initial experiments, using samples from anaerobic cultures grown at pH 7·2, two cross-reacting species were observed in Western blots; YfiD itself (Mr ~14000), and a second, higher mobility, YfiD' species (Mr~11000), (Figs 1 and 4). This species accounted for 27–34% of cross-reacting material in the 2–16 h growth period but then fell to 9–12% in the 24–31 h section of the anaerobic growth cycle. By analogy with PFL and the size of YfiD', it was reasoned that YfiD' is one of the products arising from oxidative cleavage of active, radical-containing YfiD protein. This interpretation was supported by the recent demonstration that YfiD is a glycyl radical protein (Wagner et al., 2001 ) and by the failure to detect YfiD' in a PFL activase deficient strain of E. coli (Fig. 5), indicating that insertion of the radical is required to generate YfiD'. Moreover, this latter observation indicates that YfiD does not have a unique activase but shares that of PFL. It would also appear that in anaerobic cultures at pH 7·2 during exponential growth and early stationary phase, only one third of the YfiD protein present in the cell is in the active state and that this falls to about 10% in late stationary phase. Thus, although the level of YfiD protein is maintained upon cessation of growth it is very likely that its activity changes. Furthermore, the extent of YfiD activation was affected by lowering the intracellular pH of the cultures by the addition of propionate (Fig. 4). After 7 h of growth in the presence of propionate, YfiD' now represented 57% of the total YfiD protein detected in Western blots compared to only 19% in the absence of propionate. Even in stationary phase when the proportion of YfiD' falls, it remained twice as high in the propionate-exposed cultures (Fig. 4). Thus it is concluded that lower intracellular pH not only enhances yfiD expression, but also enhances YfiD activity.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 4. The effect of intracellular pH and growth phase on cytoplasmic YfiD levels and radical activation. Cultures of MC1000 were grown anaerobically in LBK/MOPS, pH 7·2 in the absence or presence of propionate (50 mM). The cultures were sampled in exponential (7 h) and stationary phase (24, 28, 31 h) of the growth cycle, lysed in SDS-PAGE loading buffer and applied to SDS-polyacrylamide gels (20%) such that the same number of bacterial cells were present in each track. YfiD protein was detected by Western blotting with polyclonal anti-YfiD serum.

 


View larger version (14K):
[in this window]
[in a new window]
 
Fig. 5. The effect of the PFL activase (PflA) and deactivase (AdhE) on YfiD. Cultures of 234M11 (pflA, lacking PFL activase) and RM235 (adhE, lacking PFL deactivase) transformed with pGS1207 were grown anaerobically in LBK/MES, pH 6·3, containing 40 mM nitrate as a terminal electron acceptor, 50 mM propionate, 200 µg ampicillin ml-1 and 100 µg IPTG ml-1, to ensure maximal YfiD expression and activation. Samples were removed from the cultures in exponential (24 and 28 h for 234M11; 8 and 24 h for RM235) and stationary (31 and 48 h for 234M11; 28 and 31 h for RM235) phases of the growth cycle, lysed in SDS-PAGE loading buffer and applied to SDS-polyacrylamide gels (20%) such that the same numbers of bacterial cells were present in each track. A sample taken from a stationary phase culture of MC1000 (labelled WT; expressing both PFL activase and AdhE) is shown for comparison. YfiD protein was detected by Western blotting with polyclonal anti-YfiD serum.

 
Having established that YfiD is a radical-containing protein that shares the PFL activase it was of interest to investigate whether YfiD, like PFL, could be deactivated by the multifunctional alcohol and acetaldehyde dehydrogenase, AdhE (Kessler et al., 1992 ). Therefore, the extent of YfiD activation was estimated in an adhE mutant strain of E. coli after incubation in the presence of nitrate to allow anaerobic respiratory growth. The data obtained showed that the extent of oxygenolytic cleavage of YfiD to YfiD' was similar in the presence or absence of AhdE suggesting that, unlike PFL, YfiD is not deactivated by AdhE (Fig. 5).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The investigations reported here were prompted by a proteomic analysis of acid-inducible proteins in E. coli (Blankenhorn et al., 1999 ). This indicated that the yfiD gene product was upregulated both aerobically and anaerobically after exposure to acid stress (Blankenhorn et al., 1999 ). Here it has been shown that the anaerobic transcription factor FNR is the major activator of yfiD expression, but that the extent of FNR-mediated activation can be modulated by the indirect oxygen sensor ArcA and the pyruvate sensor PdhR. Moreover, in accord with the previous proteomic analysis, expression from a yfiD::lac reporter fusion and the intracellular content of YfiD was found to be greatest during growth at low pH for both aerobic and anaerobic cultures. Therefore, from the data presented here, the model describing the regulation of yfiD expression in E. coli can be adapted to include three new features (Fig. 6). The first is competition between FNR (acting negatively) and ArcA (acting positively) for occupation of the -93·5 region of the promoter (FNR II; Fig. 6). This competition is consistent with the observation that yfiD expression is maximal under microaerobic conditions (Marshall et al., 2001 ). Thus, aerobically, FNR is largely inactive, neither FNR site (FNR I or FNR II; Fig. 6) within the yfiD promoter is occupied and yfiD expression is low. Microaerobically, a low level of active FNR appears in the bacteria, and the high affinity activating FNR site (FNR I) becomes occupied, allowing yfiD expression. Under these conditions there is insufficient active FNR to occupy the low affinity upstream repressing FNR site (FNR II), and the data presented here suggests that ArcA binding to this region ensures that this is so, thereby facilitating maximal yfiD expression. Because the level of yfiD expression observed when ArcA is present in multiple copies is similar to that reported for expression from a modified yfiD promoter that lacks the FNR II site (Green et al., 1998 ), it is most likely that ArcA acts as an anti-repressor, minimizing FNR occupation of the FNR II site during microaerobic growth. Anaerobically, the intracellular concentration of active FNR becomes sufficient to allow occupation of the FNRII site by displacing ArcA, and thereby inhibit the transcription activation mediated by the FNR dimer bound at FNR I. The next phase of investigation should consider the relative binding affinities of FNR and ArcA for the -93 region of the yfiD promoter. Such studies will have to be done in vitro because mutagenesis of either type of upstream binding site would inevitably affect the properties of the other because of the large sequence overlap. The second feature is the PdhR-mediated repression of yfiD indicating a link to pyruvate/central metabolism. Previous reports have indicated that intracellular YfiD levels are enhanced by growth in the presence of pyruvate (Blankenhorn et al., 1999 ), which may now be explained (at least in part) by the relief of PdhR-mediated repression. The third feature is that yfiD expression is enhanced during growth at acid pH. It is well known that the effects of pH on gene expression overlap with other environmental factors and particularly with oxygen availability (Cotter et al., 1990 ; Meng & Bennett, 1992 ; Neely et al., 1994 ; Slonczewski & Foster, 1996 ; Blankenhorn et al., 1999 ) and it now appears that yfiD expression responds to both oxygen starvation and acidic pH via FNR, ArcA and an as yet unknown acid-responsive component.



View larger version (9K):
[in this window]
[in a new window]
 
Fig. 6. Regulation of the yfiD gene of E. coli showing the effects of global regulators on the activity of the yfiD promoter. The pyruvate responsive regulator PdhR acts as a minor repressor of yfiD expression when the intracellular pyruvate pool is low. The anaerobic regulators FNR and ArcA combine to allow maximum YfiD expression during anaerobic growth. Aerobically, FNR is largely inactive, neither FNR site is occupied and yfiD expression is low. Microaerobically, active FNR begins to appear in the bacterial cytoplasm and the high affinity activating FNR site (FNR I) becomes occupied, allowing yfiD expression. Under these conditions there is insufficient active FNR to occupy the upstream repressing FNR site (FNR II) and furthermore, ArcA binding to overlapping sites (solid bars) within FNR II ensures that this is so. Anaerobically, the intracellular concentration of active FNR is now sufficient to bind at the FNR II site by displacing ArcA and thereby inhibit the transcription activation mediated by the FNR dimer bound at FNR I. At low pH, an as yet uncharacterized acid-responsive regulator (indicated by ? and broken arrows) enhances yfiD expression and intracellular YfiD content in acid cultures in the presence and absence of FNR. + indicates a positive effect; - a negative effect on yfiD expression.

 
The pattern of regulation observed for yfiD has some features in common with those of the pflB gene, which also encodes a radical-containing protein, PFL. The pflB gene is transcribed from multiple promoters, and promoters 6 and 7 are, like the yfiD promoter, positively regulated in response to anaerobiosis by both FNR and ArcA (Sawers, 1993 ). The pflB gene is part of an operon consisting of focApflB and it has been reported that expression of pflB is positively regulated by intracellular pyruvate (Sawers & Bock, 1988 ). However, analysis of the pflB locus revealed no good matches to the PdhR binding site close to promoter 6 or promoter 7, and so they are unlikely to be directly influenced by intracellular pyruvate pools via PdhR. A potential PdhR site was found within the pflB coding region, some 1787 bp upstream of the PFL activase gene (pflA) transcript start, and it is thus unlikely that this potential PdhR site is transcriptionally relevant. However, from the PdhR-mediated repression of yfiD described here, the report that growth in the presence of pyruvate increases the intracellular content of YfiD (Blankenhorn et al., 1999 ), and the pyruvate-mediated induction of pflB (Sawers & Bock, 1988 ), it appears that yfiD and pflB could be co-ordinately regulated in response to both pyruvate and anaerobiosis. It will be of interest to determine whether pflB expression is also acid responsive. An indication that this may be so was provided by the presence of a cross-reacting species of Mr~85000 (consistent with the size of PFL) in our Western blots, the intensity of which increased as intracellular pH decreased (not shown).

The overlapping pattern of regulation of yfiD and pflB has acquired new significance because during the re viewing of this work it has been shown that YfiD can act as a ‘spare part’ by providing an active glycyl radical domain to restore activity to oxygenolytically cleaved PFL (Wagner et al., 2001 ). PFL cleaves pyruvate to form acetyl-CoA (which is subsequently converted to acetate) and formate (which is converted to carbon dioxide and hydrogen) during anaerobic growth. Thus, it is suggested that YfiD may act to maintain/restore PFL activity to bacteria that have been exposed to oxygen (Wagner et al., 2001 ). Such a role is consistent with FNR-mediated regulation tuned for maximum yfiD expression in microaerobic conditions (Marshall et al., 2001 ), and with PdhR-mediated repression of yfiD expression in response to low intracellular pyruvate. Moreover, the chemostat studies with the yfiD mutant provide further support for the ‘spare part’ hypothesis, which predicts that in the absence of YfiD under microaerobic conditions, carbon flux through PFL would be reduced due to an inability to repair oxygenolytically cleaved PFL. Reduced flux through PFL could lead to increased excretion of pyruvate, succinate and lactate, but not acetate, ethanol or formate. The data presented here indicate that a lesion in yfiD leads to excretion of increased levels of pyruvate (16-fold greater than for the parent strain), succinate (>15-fold higher) and lactate (>140-fold higher) whereas the levels of ethanol and acetate were similar for both the parent and yfiD mutant strains (Table 3). All these features are consistent with YfiD acting to maintain high levels of PFL activity during microaerobic growth. The sevenfold increase in formate excretion for the yfiD mutant cultures does not at first sight appear to be consistent with this role. However, because the growth medium contains limited amounts of oxygen and no other exogenous electron acceptors, it could be that the reduced flux through PFL, caused by the inability to rescue the activity of oxygen-damaged PFL because of the absence of YfiD, has not allowed sufficient formate to accumulate to activate the FhlA-regulated formate hydrogen-lyase system in the mutant cultures. Previous studies have shown that maximal anaerobic expression of hycB (a component of formate hydrogen-lyase) at pH 7·0 in vivo requires the presence of 30 mM formate in the culture medium (Rossmann et al., 1991 ). Moreover, 10 mM formate has been shown to be necessary for maximum activity of FhlA in vitro (Hopper et al., 1994 ). Although these studies are not directly comparable to those described here, they do indicate that substantial formate accumulation is required for the optimal induction of formate hydrogen-lyase. Consequently, if formate hydrogen-lyase is not induced in the yfiD mutant strain, due to subinductive intracellular formate concentrations resulting from reduced flux through PFL, then formate would not be disproportionated to carbon dioxide and water. This could lead to the detection of increased levels of formate in the yfiD mutant cultures compared to the parental cultures, where, because the carbon flux through PFL is uninhibited, sufficient formate is produced to induce a functional formate hydrogen-lyase. Other possibilities that cannot be discounted are that YfiD may contribute directly or indirectly to the metabolism of formate, perhaps by reversing the PFL reaction, or that the formate accumulating in the yfiD mutant cultures arises from a source other than PFL. Further work will be needed to resolve these questions.

Consistent with the proposal that YfiD acts as a radical domain for damaged PFL and possibly other radical proteins in E. coli, the YfiD radical shares many features with that of PFL. The location of the YfiD radical was presumed to be G102 by analogy with PFL (Wagner et al., 1992 ); and this has recently been shown to be the case (Wagner et al., 2001 ). Here it is shown that the YfiD radical is inserted by the PFL activase, PflA. Therefore, yfiD and pflB not only share similar regulation at the transcriptional level, but the activation, and thus the activities, of the two proteins are also linked. In PFL the glycine radical acts as a ‘radical store’ when the protein is in the resting state (Becker et al., 1999 ) and it would now seem that a similar function is fulfilled by G102 in YfiD (Wagner et al., 2001 ). The observation that the radical content of YfiD is increased at low pH suggests that activated PFL is more easily oxygenolytically cleaved under such conditions and consequently more YfiD is activated to restore activity to the damaged PFL protein. Whether the YfiD protein can, under certain conditions, have activity per se or whether its function is only to repair oxidatively damaged PFL is as yet unknown. The reaction mechanism of PFL requires an amino acid triad consisting of G734, C418 and C419. In the resting state the radical resides at G734 but during initiation of the PFL reaction it is transferred to C419 and then to C418 (Becker et al., 1999 ). This sequence can be reversed to return the protein to the resting state, allowing multiple turnovers without regeneration of the glycyl radical. In contrast to PFL, the YfiD protein has only one cysteine residue (C31) and thus if it has independent activity it must use a different catalytic mechanism to that of PFL. Other glycyl radical enzymes are known that lack the adjacent cysteine motif of PFL (Sawers & Watson, 1998 ; Selmer & Andrei, 2001 ). The failure to observe increased oxygenolytic cleavage of YfiD from cultures unable to express the PFL deactivase AdhE (a multifunctional alcohol dehydrogenase), which acts to quench the PFL radical and thereby prevent oxidative cleavage occurring is consistent with the proposed role of YfiD, because YfiD would need to be activated under conditions that could damage PFL. Thus, the work reported here describes strong links between the regulation of expression and activity of YfiD and PFL, in accord with the proposed role of YfiD as a radical-containing module that can be plugged into oxidatively damaged PFL to restore activity to this key enzyme.


   ACKNOWLEDGEMENTS
 
We would like to thank J. R. Guest for many useful discussions, S. Thorpe for mass spectrometry and A. Cox for the metal ion analyses. This work was funded by the Biotechnology and Biological Sciences Research Council project grant P08299.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Becker, A., FritzWolf, K., Kabsch, W., Knappe, J., Schultz, S. & Wagner, A. F. V. (1999). Structure and mechanism of the glycyl radical enzyme pyruvate formate-lyase. Nat Struct Biol 6, 969-975.[Medline]

Bergmeyer, H. U. & Bernt, E. (1968). Determination of D-glucose with glucose oxidase and peroxidase. In Methods in Enzymatic Analysis , pp. 1205-1215. Edited by H. U. Bergmeyer. New York & London:Academic Press.

Blankenhorn, D., Phillips, J. & Slonczewski, J. (1999). Acid- and base-induced proteins during aerobic and anaerobic growth of Escherichia coli revealed by two-dimensional gel electrophoresis. J Bacteriol 181, 2209-2216.[Abstract/Free Full Text]

Cotter, P. A., Chepuri, R. B., Gennis, R. B. & Gunsalus, R. P. (1990). Cytochrome o (cyoABCDE) and d (cydAB) oxidase gene expression in Escherichia coli is regulated by oxygen, pH and the fnr gene product. J Bacteriol 172, 6333-6338.[Medline]

Degraeve, P., Delorme, P. & Lemay, P. (1996). Pressure-induced inactivation of E. coli ß-galactosidase: influence of pH and temperature. Biochim Biophys Acta 1292, 61-68.[Medline]

Freestone, P., Grant, S., Trinei, M., Onoda, T. & Norris, V. (1998). Protein phosphorylation in Escherichia coli L-form NC-7. Microbiology 144, 3289-3295.[Abstract]

Green, J. & Baldwin, M. L. (1997). HlyX, the FNR homologue of Actinobacillus pleuropneumoniae, is a [4Fe 4S]-containing oxygen-responsive transcription regulator that anaerobically activates FNR-dependent class I promoters via an enhanced AR1-contact. Mol Microbiol 24, 593-605.[Medline]

Green, J. & Marshall, F. A. (1999). Identification of a surface of FNR overlapping AR1 that is required for repression of gene expression. J Biol Chem 274, 10244-10248.[Abstract/Free Full Text]

Green, J., Baldwin, M. L. & Richardson, J. (1998). Downregulation of Escherichia coli yfiD expression by FNR occupying a site at -93·5 involves the AR1-containing surface of FNR. Mol Microbiol 29, 1113-1123.[Medline]

Guan, K. & Dixon, J. E. (1991). Eukaryotic proteins expressed in Escherichia coli: an improved thrombin cleavage and purification procedure of fusion proteins with glutathione S-transferase. Anal Biochem 192, 262-267.[Medline]

Hamilton, C. M., Aldea, M., Washburn, B. K., Babitzke, P. & Kushner, S. R. (1989). New method for generating deletions and gene replacements in Escherichia coli. J Bacteriol 171, 4617-4622.[Medline]

Han, M.-E., Yoon, S. S. & Lee, S. Y. (2001). Proteome analysis of metabolically engineered Escherichia coli producing poly(3-hydroxybutyrate). J Bacteriol 183, 301-308.[Abstract/Free Full Text]

Heßlinger, C., Fairhurst, S. A. & Sawers, G. (1998). Novel keto acid formate-lyase and propionate kinase enzymes are components of an anaerobic pathway in Escherichia coli that degrades L-threonine to propionate. Mol Microbiol 27, 477-492.[Medline]

Hopper, S., Babst, M., Schlensog, V., Fischer, H. H., Hennecke, H. & Bock, A. (1994). Regulated expression in vitro of genes coding for formate hydrogenlyase components of Escherichia coli. J Biol Chem 269, 19597-19604.[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.[Medline]

Kessler, D., Herth, W. & Knappe, J. (1992). Ultrastructure and pyruvate formate-lyase radical quenching property of the multimeric AdhE protein of Escherichia coli. J Biol Chem 267, 18073-18079.[Abstract/Free Full Text]

Kiley, P. J. & Beinert, H. (1999). Oxygen sensing by the global regulator FNR: the role of the iron-sulfur cluster. FEMS Microbiol Rev 22, 341-352.

Lazzazera, B. A., Beinert, H., Khoroshilova, N., Kennedy, M. C. & Kiley, P. J. (1996). DNA-binding and dimerization of the Fe-S containing FNR protein Escherichia coli are regulated by oxygen. J Biol Chem 271, 2762-2768.[Abstract/Free Full Text]

Marshall, F. A., Messenger, S. L., Wyborn, N. R., Guest, J. R., Wing, H. J., Busby, S. J. W. & Green, J. (2001). A novel promoter architecture for microaerobic activation by the anaerobic transcription factor FNR. Mol Microbiol 39, 747-753.[Medline]

Meng, S.-Y. & Bennett, G. N. (1992). Nucleotide sequence of the Escherichia coli cad operon: a system for neutralization of low extracellular pH. J Bacteriol 174, 2659-2669.[Abstract]

Miller, J. H. (1972). Assay of ß-galactosidase. In Experiments in Molecular Genetics, pp. 352–355. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Neely, M. N., Dell, C. L. & Olson, E. R. (1994). Roles of LysP and CadC in mediating the lysine requirement for acid induction of the Escherichia coli cad operon. J Bacteriol 176, 3278-3285.[Abstract]

Perrot, F., Hebraud, M., Charlionet, R., Junter, G. A. & Jouenne, T. (2000). Protein patterns of gel-entrapped Escherichia coli cells differ from those of free floating organisms. Electrophoresis 21, 645-653.[Medline]

Popescu, C. V., Bates, D. M., Beinert, H., Munck, E. & Kiley, P. J. (1998). Mössbauer spectroscopy as a tool for the study of activation/inactivation of the transcription regulator FNR in whole cells of Escherichia coli. Proc Natl Acad Sci USA 95, 13431-13435.[Abstract/Free Full Text]

Quail, M. A. & Guest, J. R. (1995). Purification, characterization and mode of action of PdhR, the transcriptional repressor of the pdhR-aceEF-lpd operon of Escherichia coli. Mol Microbiol 15, 519-529.[Medline]

Rossmann, R., Sawers, G. & Bock, A. (1991). Mechanism of regulation of the formate-hydrogenlyase pathway by oxygen, nitrate, and pH: definition of the formate regulon. Mol Microbiol 5, 2807-2814.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Press.

Sawers, G. (1993). Specific transcriptional requirements for positive regulation of the anaerobically inducible pflB operon by ArcA and FNR. Mol Microbiol 10, 737-747.[Medline]

Sawers, G. & Bock, A. (1988). Anaerobic regulation of pyruvate formate-lyase from Escherichia coli K-12. J Bacteriol 170, 5330-5336.[Medline]

Sawers, G. & Watson, G. (1998). A glycyl radical solution: oxygen-dependent interconversion of pyruvate formate-lyase. Mol Microbiol 29, 945-954.[Medline]

Selmer, T. & Andrei, P. I. (2001). p-Hydroxyphenylacetate decarboxylase from Clostridium difficile. Eur J Biochem 268, 1363-1372.[Abstract/Free Full Text]

Slonczewski, J. L. & Foster, J. W. (1996). pH-regulated genes and survival at extreme pH. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 1539–1549. Edited by F. C. Niedhardt and others. Washington, DC: American Society for Microbiology.

Vishniac, W. & Santer, M. (1957). The Thiobacilli. Bacteriol Rev 21, 195-209.

Wagner, A. F. V., Frey, M., Neugebauer, F. A., Schafer, W. & Knappe, J. (1992). The free radical in pyruvate formate-lyase is located on glycine-734. Proc Natl Acad Sci USA 89, 996-1000.[Abstract]

Wagner, A. F. V., Schultz, S., Bomke, J., Pils, T., Lehmann, W. D. & Knappe, J. (2001). YfiD of Escherichia coli and Y061 of bacteriophage T4 as autonomous glycyl radical cofactors reconstituting the catalytic center of oxygen-fragmented pyruvate formate-lyase. Biochem Biophys Res Commun 285, 456-462.[Medline]

Wing, H. J., Williams, S. M. & Busby, S. J. W. (1995). Spacing requirements for transcription activation by Escherichia coli. J Bacteriol 177, 6704-6710.[Abstract]

Received 7 September 2001; revised 20 November 2001; accepted 27 November 2001.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Wyborn, N. R.
Articles by Green, J.
Articles citing this Article
PubMed
PubMed Citation
Articles by Wyborn, N. R.
Articles by Green, J.
Agricola
Articles by Wyborn, N. R.
Articles by Green, J.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2002 Society for General Microbiology.