N-terminal truncations in the FhlA protein result in formate- and MoeA-independent expression of the hyc (formate hydrogenlyase) operon of Escherichia coli

William T. Selfa,1, Adnan Hasona1 and K. T. Shanmugam1

Department of Microbiology and Cell Science, Box 110700, University of Florida, Gainesville, FL 32611, USA1

Author for correspondence: K. T. Shanmugam. Tel: +1 352 392 2490. Fax: +1 352 392 5922. e-mail: shan{at}ufl.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
 
The formate hydrogenlyase complex of Escherichia coli catalyses the cleavage of formate to CO2 and H2 and consists of a molybdoenzyme formate dehydrogenase-H, hydrogenase 3 and intermediate electron carriers. The structural genes of this enzyme complex are activated by the FhlA protein in the presence of both formate and molybdate; ModE-Mo serves as a secondary activator. Mutational analysis of the FhlA protein established that the unique N-terminal region of this protein was responsible for formate- and molybdenum-dependent transcriptional control of the hyc operon. Analysis of the N-terminal sequence of the FhlA protein revealed a unique motif (amino acids 7–37), which is also found in ATPases associated with several members of the ABC-type transporter family. A deletion derivative of FhlA lacking these amino acids (FhlA9-2) failed to activate the hyc operon in vivo, although the FhlA9-2 did bind to hyc promoter DNA in vitro. The ATPase activity of the FhlA9-2–DNA–formate complex was at least three times higher than that of the native protein–DNA–formate complex, and this degree of activity was achieved at a lower formate level. Extending the deletion to amino acid 117 (FhlA167) not only reversed the FhlA- phenotype of FhlA9-2, but also led to both molybdenum- and formate-independence. Deleting the entire N-terminal domain (between amino acids 5 and 374 of the 692 amino acid protein) also led to an effector-independent transcriptional activator (FhlA165), which had a twofold higher level of hyc operon expression than the native protein. Both FhlA165 and FhlA167 still required ModE-Mo as a secondary activator for an optimal level of hyclac expression. The FhlA165 protein also had a twofold higher affinity to hyc promoter DNA than the native FhlA protein, while the FhlA167 protein had a significantly lower affinity for hyc promoter DNA in vitro. Although the ATPase activity of the native protein was increased by formate, the ATPase activity of neither FhlA165 or FhlA167 responded to formate. Removal of the first 117 amino acids of the FhlA protein appears to result in a constitutive, effector-independent activation of transcription of the genes encoding the components of the formate hydrogenlyase complex. The sequence similarity to ABC-ATPases, combined with the properties of the FhlA deletion proteins, led to the proposal that the N-terminal region of the native FhlA protein interacts with formate transport proteins, both as a formate transport facilitator and as a cytoplasmic acceptor.

Keywords: FhlA mutations, molybdenum, operon regulation

Abbreviations: FDH-H, formate dehydrogenase isoenzyme; FHL, formate hydrogenlyase enzyme complex; Phyc, hyc operon promoter

a Present address: NHLBI, NIH, Bethesda, MD, USA.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
 
Escherichia coli catalyses the cleavage of intracellular formate to H2 and CO2 via the formate hydrogenlyase enzyme complex (FHL) (reviewed by Böck & Sawers, 1996 ). This complex includes a formate dehydrogenase isoenzyme (FDH-H; encoded by the fdhF operon), a hydrogenase isoenzyme (HYD3; encoded by the hyc operon), and intermediate electron carriers (also encoded by the hyc operon). FHL is only produced under anaerobic conditions and in the absence of alternative terminal electron acceptors, such as nitrate. Reduced levels of expression of the hyc operon under these conditions have been attributed to a reduced pool of intracellular formate (Rossmann et al., 1991 ). This lack of available formate affects the transcriptional activator FhlA, as it requires formate as an effector molecule (Maupin & Shanmugam, 1990 ; Schlensog et al., 1994 ).

Molybdenum is required for FHL activity and is present in the FDH-H protein as molybdopterin guanine dinucleotide (Boyington et al., 1997 ). Molybdenum is transported in the form of molybdate by a high-affinity transporter encoded by the modABC operon (reviewed by Grunden & Shanmugam, 1997 ). Previous studies have shown that the expression of both the hyc and fdhF operons also requires molybdenum for optimal transcription (Rosentel et al., 1995 ; Hasona et al., 1998b ). ModE protein, originally characterized as a molybdate-dependent repressor of the modABC operon (Grunden et al., 1996 ), was also identified as a secondary transcriptional activator of the hyc operon (Self et al., 1999 ). The FhlA protein has been proposed to serve as a molybdenum sensor (Hasona et al., 1998b ), possibly interacting with the product of molybdate activation by the MoeA protein (Hasona et al., 1998a ).

The FhlA protein has three distinguishable domains based, in part, on sequence similarity with other transcriptional activators, such as NtrC from E. coli (Nixon et al., 1986 ) and NifA from Klebsiella pneumoniae (Drummond et al., 1986 ). The C-terminal region contains a helix–turn–helix motif which constitutes a DNA-binding domain; this was confirmed by the DNA-binding activity of a C-terminal fragment of FhlA containing the amino acids 379–692 (Self, 1998 ; Leonhartsberger et al., 2000 ). The central region of FhlA is conserved among all members of the NtrC/NifA family of activators and contains a nucleotide-binding motif (Weiss et al., 1991 ): this region is probably responsible for formate- and DNA-dependent ATP hydrolysis and interaction with the {sigma}54 form of RNA polymerase in the initiation of transcription (Hopper & Böck, 1995 ; Korsa & Böck, 1997 ). The N-terminal segment of the FhlA protein appears to contain a region which modulates the transcriptional activity of the C-terminal segment. In many proteins of the NtrC family, the unique N-terminal region (of varying length) contains a receiver domain which must be phosphorylated by a cognate sensor-protein for activity (Keener & Kustu, 1988 ). Some members of this family of activators are not known to be phosphorylated, and are presumed to interact with appropriate effector molecules directly.

The N-terminal region of FhlA is one of the largest (380 aa), when compared to other response regulators of the NtrC family. The unique N-terminal region of the FhlA protein displays no significant sequence similarity to any known protein, and is not believed to be phosphorylated (Böck & Sawers, 1996 ). Point mutations in the N-terminal region of FhlA lead to partial independence from formate (Korsa & Böck, 1997 ) and molybdenum (Self & Shanmugam, 2000 ) in the activation of the hyc operon. Based on these studies, it has been proposed that the N-terminal region of FhlA binds the effector molecule formate and may also interact with the product of the MoeA protein (activated molybdenum).

Removal of the first 378 amino acids of the 692 aa FhlA protein led to a formate-independent transcriptional activator (Self, 1998 ; Leonhartsberger et al., 2000 ). The N-terminal half of the protein was also shown to inhibit the ATPase activity of the C-terminal half of the protein and to inhibit the formate-dependent ATPase activity of the native protein. Since the FhlA protein has been proposed to interact with the MoeA product for optimum transcription of the hyc operon, and mutations in the unique N-terminal domain of FhlA lead to molybdenum-independence, it is possible that removing the N-terminal domain would also impart molybdenum-independence to the truncated FhlA protein. It is also not known which part of the nearly 400 aa N-terminal segment of the protein is essential for formate- and molybdenum-dependence for hyc activation. Therefore, several deletion derivatives extending to varying lengths of the N-terminal domain of FhlA were constructed to identify the critical region(s) responsible for effector-independence, and the results of this study are reported here.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
 
Bacterial strains.
The bacterial strains, phages and plasmids used in this study are detailed in Table 1. All strains are derivatives of Escherichia coli K-12.


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

 
Media and growth conditions.
Media used for bacterial growth were as described by Rosentel et al. (1995) . L-broth (LB; Miller, 1972 ) was supplemented with glucose (0·3%), sodium formate (30 mM), sodium molybdate (1 mM) or sodium nitrate (30 mM). Antibiotics were added to the media, as required, at the following concentrations: 100 µg ampicillin ml-1; 30 µg tetracycline ml-1; 50 µg chloramphenicol ml-1 (plates), 10 µg chloramphenicol ml-1 (broth); 50 µg kanamycin ml-1.

Transduction with phages P1 and {lambda} was performed as described by Grunden et al. (1996) . Genetic and molecular experiments were performed essentially as previously described (Self et al., 1999 ). Plasmids pWS2, pWS164 and pWS165 are derivatives of the low-copy vector pACYC184. Plasmid pWS167 is a derivative of pT7-7. For experiments involving pcnB mutant (Lopilato et al., 1986 ), the appropriate fhlA+ and fhlA165 DNAs were transferred to plasmid vector pBR322 in the construction of pWS3 and pWS165-2, respectively. Plasmid pWS165-2 also carries the gene for spectinomycin resistance from pSJS1240 (Kim et al., 1998 ).

Isolation of N-terminal deletion fhlA alleles.
Internal deletion mutations within the fhlA gene were isolated by Bal31 nuclease treatment of pWS2 (Self & Shanmugam, 2000 ) after HpaI restriction endonuclease hydrolysis. HpaI hydrolyses pWS2 approximately in the centre of the DNA encoding the unique N-terminal region of the FhlA protein. The reactions were terminated by the addition of ethanol (70% final concn) after 5, 10 and 15 min of nuclease treatment. The DNA was precipitated and subsequently ligated with T4 DNA ligase. The ligation mixture was transformed into strain SE1174 (fhlA::Tn10), and colonies expressing FDH-H activity were selected by the dye-overlay procedure as described by Lee et al. (1985) . Plasmids from selected clones were isolated and the size of the deletion was determined using restriction endonucleases. Three plasmids with a deletion of >=1 kb were selected for further analysis. The extent of the deletion in each mutant fhlA allele was determined by sequencing the DNA using the Sanger dideoxy sequencing method (Sanger et al., 1977 ; Sequenase version 2·0 DNA sequencing kit). This procedure yielded fhlA alleles fhlA164, fhlA165 and fhlA166 (plasmids pWS164, pWS165 and pWS166, respectively; Fig. 1). fhlA167 was isolated after removing the fhlA DNA as a PflMI–SmaI fragment and cloning it into a SmaI-digested vector, pT7-7. This construct, pWS41, carries a deletion of the first 117 aa of the N-terminus, and an extra 7 aa in the N-terminus derived from the vector DNA sequence (MARIRAL) which also provided the translation start site. Plasmid pWS167 was constructed by inserting a chloramphenicol-resistance gene cartridge into the ScaI site of pWS41. The fhlA167 allele DNA was sequenced to confirm the deletion. DNA sequences were analysed using GENEPRO software (Riverside Scientific).



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 1. Extent of the N-terminal deletions within the FhlA protein. The solid lines represent the amino acid sequence and the dashed lines correspond to the deletions in the FhlA protein sequence. The ß-galactosidase activity [nmol (mg cell protein)-1 min-1] represents the expression of hyclac in the presence of the various FhlA alleles in a {Delta}(srlfhlA) strain carrying {lambda}WS1, grown in L-broth containing glucose (0·3%) and sodium formate (30 mM) under anaerobic conditions. The solid boxes represent the motif for the ATP-binding domain; the hatched boxes represent the helix–turn–helix motif (DNA binding domain).

 
For the construction of pWS9-2, plasmid pWS9 was modified by introducing a chloramphenicol-resistance gene cartridge from plasmid pBR325 into the SspI–DraI sites, to generate pWS9-1, followed by PCR-mediated deletion of the fhlA DNA corresponding to amino acids 7–37 of the FhlA protein. The fhlA gene on both sides of the deletion was amplified using two primer pairs (pair one, 5'-AAATGTCATATACACCGATGGTAAAGCGTTCTGCGCTCGCCG-3' and 5'-GACTTCGCTGACCAGTTCGTCCATA-3'; pair two, 5'-CGAGCGCAGAACGCTTTACCATCGGTGTATATGACATTTTGTTC-3' and 5'-TGTGGATAACCGTATTACCGCCT-3'). The PCR products were ligated together to create the deletion derivative (plasmid pWS9-2), and the PCR-amplified region was sequenced to confirm that the fhlA gene in plasmid pWS9-2 did not carry any additional mutations, besides the intended deletion.

Purification of the FhlA165 protein.
Purification of the FhlA165 protein was essentially as described for the native FhlA protein by Schlensog et al. (1994) . Briefly, pWS16, which carries the fhlA165 allele under the control of phage T7 promoter, was constructed by ligating a 1·4 kb BstBI–ClaI fragment from pWS165 into the ClaI site in the expression vector pT7-7 (Tabor & Richardson, 1985 ). Fresh transformants of BL21(DE3) with pWS16 were grown in 1 litre of LB with ampicillin in a 2·8 litre Fernbach flask at 37 °C with shaking (225 r.p.m.). When the culture reached OD420 0·7 (Spectronic 710 spectrophotometer), IPTG was added (0·34 mM final concn) and the culture was incubated at 37 °C with shaking (225 r.p.m.) for an additional 3 h. Cells were harvested and washed once with Tris buffer (50 mM Tris, pH 8·0; 0·5 mM EDTA; 1 mM benzamidine; 1 mM DTT). The cells were then resuspended in 12 ml Tris buffer and broken by passage through a French pressure cell [20000 p.s.i. (138 MPa)]. All subsequent steps were carried out at 4 °C. The crude extract was centrifuged at 30000 g for 30 min, and the supernatant (14 ml) was further clarified by centrifugation for 2 h at 120000 g. Ammonium sulfate (37% saturation) was added to the supernatant, and the proteins were precipitated for 30 min. The sample was then centrifuged for 30 min at 30000 g, and the pellet was dissolved in 1·2 ml Tris buffer. This sample was dialysed overnight in 1 litre Tris buffer. Unlike the wild-type FhlA protein, which precipitated out of solution at this step (Schlensog et al., 1994 ), the FhlA165 remained soluble. This dialysed sample was analysed by chromatography, using a 40 ml heparin-agarose affinity column (Pharmacia Biotech). The column matrix was washed with Tris buffer and the proteins were eluted with a linear gradient of KCl (0–0·5 M). FhlA165-containing fractions, as determined by SDS-PAGE analysis, were pooled and dialysed overnight in Tris buffer. Further purification of FhlA165 was achieved using a 12 ml Q-Sepharose column (Pharmacia) and a linear KCl gradient (0–0·4 M). The FhlA165 protein eluted at approximately 0·3 M KCl, and was determined to be pure based on SDS-PAGE analysis. Fractions containing the FhlA165 protein were pooled and dialysed overnight in Tris buffer (25 mM, pH 7·5) with 1 mM EDTA and 50% glycerol (v/v). The FhlA165 protein was stored at -20 °C.

FhlA167 and FhlA9-2 proteins were purified essentially as described for FhlA165. However, the two proteins obtained after the Q-Sepharose step were subjected to further chromatography on a Q-Sepharose column with sodium citrate as the eluent. Both proteins eluted at a citrate concentration of 0·15 M. The pure proteins were dialysed against Tris buffer (50 mM, pH 7·5) and stored on ice. Freshly prepared proteins were used immediately.

Purification of native FhlA protein.
The purification procedure similar to the one used for FhlA165 protein led to the characteristic precipitation of the native protein during dialysis after the ammonium sulfate precipitation step (Schlensog et al., 1994 ). The FhlA protein required high salt (0·2 M KCl) to remain in solution; thus, it may not have been present in the native form. In addition, most of the FhlA protein produced upon expression from a T7 promoter was present in inclusion bodies. Therefore, the conditions used for expression of fhlA+ and purification of the native FhlA were modified, and the method is described below.

The fhlA+ gene was cloned into the expression vector pT7-7, by removing a 2·4 kb SpeI–ClaI fragment from plasmid pSE133 (Sankar et al., 1988 ) and ligating this DNA to a 2·4 kb XbaI–ClaI fragment from pT7-7 (Tabor & Richardson, 1985 ). The resultant plasmid, pWS9, carried the fhlA gene under the control of phage T7 gene 10 promoter. Fresh BL21(DE3) transformants of pWS9 were inoculated into 1 litre LB containing ampicillin, in a 2·8 litre Fernbach flask. The culture was grown with shaking (225 r.p.m.) at 37 °C until an OD420 of 0·7 was reached. IPTG was added (0·5 mM final concn) to the culture, and incubation was continued at 23 °C with shaking (225 r.p.m.) for an additional 6 h. Incubating the culture at this temperature allowed a larger fraction of the newly synthesized FhlA to stay in solution. Cells were harvested at 4 °C and washed once with Tris buffer (50 mM Tris, pH 8·0; 0·5 mM EDTA; 1 mM benzamidine; 1 mM DTT). All subsequent operations were carried out at 4 °C. Cells were resuspended in 25 ml Tris buffer and broken by passage through a French pressure cell at 20000 p.s.i. (138 MPa). The crude extract was centrifuged at 30000 g for 2 h and the supernatant fraction was loaded onto Q-Sepharose column (30 cmx2·5 cm; Pharmacia). The column was then washed with Tris buffer and the proteins were eluted with a linear gradient of NaCl (0–0·4 M): the FhlA protein eluted at approximately 0·37 M NaCl. FhlA-containing fractions, as determined by SDS-PAGE analysis, were pooled and dialysed overnight in Tris buffer. This protein mixture was loaded onto a Hi-trap heparin column (Pharmacia) and eluted using a linear NaCl gradient (0–0·4 M): the FhlA protein eluted at approximately 0·3 M NaCl. FhlA-containing fractions were pooled and dialysed overnight in HEPES buffer (30 mM HEPES, pH 7·0 with 0·5 mM EDTA, 1 mM benzamidine and 1 mM DTT). The FhlA protein was further purified using a 1 ml High S column (cation exchange, Bio-Rad) equilibrated with HEPES buffer and eluted with a linear gradient of 0–0·5 M NaCl. The FhlA protein eluted at approximately 0·25 M NaCl. Fractions containing FhlA protein were pooled and dialysed overnight in HEPES buffer. The final purification step required chromatography on a 1 ml High Q column (anion exchange, Bio-Rad) equilibrated with HEPES buffer and elution with a linear gradient of sodium citrate (0–0·3 M). The pure FhlA protein, as judged by SDS-PAGE analysis, was dialysed in storage buffer (25 mM HEPES, pH 7·0; 0·8 mM DTT; 10% glycerol). FhlA protein was quickly frozen in liquid nitrogen and stored at -70 °C, in aliquots.

Enzyme assay.
FHL activities [expressed as nmol (mg cell protein)-1 min-1] of late-exponential phase cultures were determined as described by Lee et al. (1985) . All experiments were carried out with whole cells to minimize O2 inactivation of the enzyme. ß-Galactosidase activities [expressed as nmol (mg cell protein)-1 min-1] were determined in late-exponential- to early-stationary-phase cultures as described by Rosentel et al. (1995) .

ATPase activity.
ATPase activity associated with the FhlA protein was determined as described by Korsa & Böck (1997) , and expressed as pmol ATP hydrolysed (pmol FhlA)-1.

DNA electrophoretic mobility-shift experiments.
DNA mobility-shift experiments were performed essentially as described by Grunden et al. (1996) . Briefly, DNA containing the hychyp intergenic region was PCR-amplified from plasmid pFGH100, using two primers: 5'-CCGGATCCGTTATTTCGAGCATATC-3' and 5'-CCCTGCAGTTTAAGCTAAAGATGAA-3'. This amplified DNA fragment, after digestion with PstI and BamHI, was cloned into plasmid pUC19, also digested with PstI and BamHI. The resulting plasmid, pWS40, was the source of a 154 bp PstI–BamHI fragment used in the electrophoretic mobility-shift experiments. The 154 bp DNA fragment, after isolation from a 10–30% sucrose gradient, was labelled with 32P-dNTPs by filling in the recessed 3' end of the BamHI site with the DNA polymerase Klenow fragment. FhlA or FhlA165 protein was pre-incubated in binding buffer (10 mM Tris, pH 7·5; 1 mM EDTA; 1 mM DTT; 1 mM ATP) for 5 min at 4 °C before the labelled DNA was added. Binding reaction mixtures (total volume 20 µl) were incubated for 20 min at 37 °C. Polyacrylamide gels (5%) were pre-run (100 V, 60 min, room temperature), with circulation of TBE buffer (Tris, 89 mM, pH 8·3; borate, 89 mM; EDTA, 2·5 mM). Samples were loaded and the gels were run for a further 60 min (100 V), in TBE buffer. Buffer circulation re-started after the samples had been loaded. The gels were then dried and exposed to pre-flashed X-ray film. The band intensity was determined using a phosphorimager (Molecular Dynamics).

Materials.
Biochemicals were purchased from Sigma. Other organic and inorganic chemicals were from Fisher Scientific and were of analytical grade. Restriction endonucleases and DNA-modifying enzymes were purchased from Promega and New England Biolabs.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
 
Isolation of deletions within the N-terminal domain of FhlA
A number of deletion derivatives of fhlA which allowed the production of dihydrogen by an fhlA mutant strain, E. coli SE1174 (fhlA::Tn10) were isolated. After analysis of the size of the fhlA gene in these plasmids and screening for the ability to support FHL production, four fhlA mutations with large deletions within the unique N-terminal domain of FhlA were selected for further analysis. Based on DNA sequence data, all four alleles (fhlA164, {Delta}26–370, L371V; fhlA165, {Delta}5–374; fhlA166, {Delta}36–314; fhlA167, {Delta}1–117) carried large deletions in the unique N-terminal domain of FhlA. Although fhlA164 and fhlA166 supported production of FHL activity in a fhlA mutant, the level of hyc–lac expression in the presence of these two alleles was only about 25% of the values obtained with native fhlA [approximately 350 units ß-galactosidase activity, compared to approximately 1800 units activity for the strain with native FhlA]. The alleles fhlA165 and fhlA167, which supported the highest level of ß-galactosidase production (Fig. 1), were studied in detail.

FhlA165 and FhlA167 are formate and molybdate independent
In the presence of native FhlA protein (pWS2), the optimum level of hyclac expression required both molybdate and formate (Table 2). Most notably, in the mod pfl double mutant, strain WS118(pWS2), hyclac expression was very low (140 units ß-galactosidase activity) in the absence of molybdate and formate. The addition of formate to the growth medium increased the level of hyclac expression to approximately 50% of the maximal level (840 units); supplementing the medium with molybdate alone increased the level of expression by approximately threefold (470 units). A level of hyclac expression equivalent to that of wild-type E. coli was attained only when both formate and molybdate were present in the growth medium of the mod pfl double mutant.


View this table:
[in this window]
[in a new window]
 
Table 2. Transcription of PhyclacZ by FhlA165 and FhlA167 is formate- and molybdate-independent

 
In contrast, activation of hyclac expression by FhlA165 and FhlA167 was no longer dependent on formate and molybdate: this was demonstrated by the high level of ß-galactosidase activity present in all three strains carrying pWS165 or pWS167 (Table 2). Although both deletion derivatives led to molybdenum- and formate-independence, only FhlA165 supported a higher level of hyclac expression than the wild-type (4000 and 2000 units ß-galactosidase activity, respectively; Table 2). The higher than expected level of expression in the presence of FhlA165 suggests that the N-terminal half of the FhlA protein negatively regulates the activation of the hyc operon by the native FhlA protein, even in the presence of the effectors formate and molybdate. Removal of the N-terminal half of FhlA165 allowed the C-terminal DNA-binding segment to activate the hyc operon maximally; this was in agreement with a previous report by Leonhartsberger et al. (2000) .

The FhlA167 mutant carried a deletion of amino acids 1–117, while the deletion in FhlA165 was of amino acids 5–374. These results show that removal of the first 117 aa is sufficient to produce an effector-independent transcriptional activator, but that this is not sufficient to overcome the inhibitory effect of the N-terminal domain of FhlA on the transcription of the hyc operon by the C-terminal domain. The critical structure responsible for modulating the activity of the C-terminal DNA-binding domain appears to reside between amino acids 118 and 374.

The product of the first gene of the hyc operon, hycA, appears to act by negatively modulating the level of hyc operon expression (Sauter et al., 1992 ). The nature of this apparent repression is yet to be determined. Even in the presence of FhlA165 protein as activator, hyclac expression was reduced by HycA by the same level as seen with the native FhlA protein (data not shown). These results suggest that the effect of HycA on hyc operon expression is independent of formate and molybdate.

FhlA165 and FhlA167 activate transcription of PhyclacZ in a modE moeA double mutant
Both ModE and MoeA proteins are required for the molybdate-dependent transcription of the hyc operon (Hasona et al., 1998b ). ModE was shown to be a secondary transcriptional activator and directly interacts with hyc promoter (Phyc) DNA (Self et al., 1999 ). The role of MoeA in this regulation is not known but it has been proposed that the FhlA protein binds the catalytic product of MoeA (activated molybdenum; Hasona et al., 1998a ) and modulates its activity (Self & Shanmugam, 2000 ). Since FhlA165 and FhlA167 activated transcription of hyclacZ even in the absence of molybdate, the need for ModE and MoeA proteins for the expression of hyclac with these deletion derivatives was determined.

A modE moeA double mutant containing the native FhlA protein, strain WS198(pWS2), produced a measurable, but low, level of ß-galactosidase activity (approximately 180 units) when cultured in L-broth containing glucose (Table 3); the addition of molybdate to the medium had no effect on hyclac expression. In contrast, the same double mutant containing the FhlA165 protein, E. coli strain WS198(pWS165), produced an approximately 15-fold more ß-galactosidase activity (2600 units): this level of activity is also approximately 2·2-fold higher than the activity of a mod+ strain containing the native FhlA, strain WS127(pWS2). The FhlA167 protein also overcame the requirement of ModE and MoeA proteins for hyc operon expression, and the level of expression of hyclac was similar to that of the mod+ strain with native FhlA (approximately 1500 units). As expected, addition of molybdate had no significant effect on the expression of the hyc operon when FhlA165 or FhlA167 was used as the activator. Although hyclac expression in the modE moeA mutant was higher in the presence of FhlA165 or FhlA167, the level of ß-galactosidase activity produced was only approximately 65–75% of the value obtained with the mod+ strain with these mutated forms of the FhlA proteins [WS127(pWS165); WS127(pWS167)]. This lower than expected value was apparently a consequence of the absence of ModE protein, a required secondary activator, since a mutation in modE alone accounted for this difference. In the presence of either FhlA165 or FhlA167 the moeA mutation had no effect on hyclac expression. These results show that the deletions in FhlA165 and FhlA167 overcame the need for the MoeA-catalysed product for hyc operon expression, but not that for ModE-Mo, a required secondary activator of the hyc operon.


View this table:
[in this window]
[in a new window]
 
Table 3. Expression of PhyclacZ by FhlA165 and FhlA167 requires ModE-Mo

 
In a separate study, expression of hyclac was found to be limited by the concentration of FhlA in the cell when the protein was produced from a chromosomal copy of fhlA (J. A. Maupin & K. T. Shanmugam, unpublished data). Under the present experimental conditions, the amount of ß-galactosidase activity produced by a culture carrying {lambda}WS1 (hyclac) and a chromosomal copy of fhlA+ was approximately 600 units. This value did not change significantly when the fhlA+ gene was introduced from a plasmid (pWS3; vector pBR322) in a pcnB mutant background which limits the copy-number of the plasmid (Lopilato et al., 1986 ). In a pcnB+ strain, the level of ß-galactosidase activity produced by a culture carrying plasmid pWS2 (vector pACYC184), which contains fhlA+, was approximately 1300 units, and in the presence of plasmid pWS3 (vector pBR322) the level of ß-galactosidase activity produced from hyclac increased further to approximately 3000 units. This observed copy-number effect is not related to the proposed antagonistic effect of HycA on hyclac expression (Sauter et al., 1992 ), as the strain used in these experiments carries a deletion of the entire srlfhlA region. In the experiments described above (Tables 2 and 3), the fhlA gene was present in a pBR322-based plasmid and it is possible that the observed independence from MoeA could be a result of overexpression of the fhlA mutant allele. However, it should be noted that in a modE background, the moeA mutation severely limited hyclac expression even in the presence of multiple copies of the fhlA gene. To rule out the effect of multiple copies of fhlA, hyclac expression was investigated in a pcnB mutant background with the fhlA165 allele.

When the overall level of hyclac expression was reduced to approximately 600–700 units ß-galactosidase activity (in a pcnB mutant, strain WS244) in the presence of native FhlA, the requirement for ModE-Mo and MoeA for hyclac expression could be readily detected. In the presence of FhlA165, the modE mutation reduced the level of hyclac expression by about 33%. Even in a modE moeA double mutant, the level of hyclac expression catalysed by FhlA165 was only reduced by about 30% compared to the corresponding isogenic moeA mutant, strain WS246. However, with FhlA165, a moeA mutation alone had a positive effect on hyclac expression (1·4-fold higher than with the native FhlA protein) in this pcnB mutant, in contrast to a negative effect of the moeA mutation (27% lower) in the presence of FhlA+. This positive effect of MoeA in the presence of the FhlA165 protein could be a result of a large pool of ModE-Mo in the cell, since molybdate is not processed in the moeA mutant combined with the MoeA-independence of the FhlA165 protein. The large pool of ModE-Mo could increase the level of expression of hyclac in the presence of FhlA165, but could not overcome the MoeA requirement for optimal activation by native FhlA. These results further demonstrate that transcription of the hyc operon by the N-terminal deletion derivative of FhlA is independent of the MoeA protein.

Activation of PhyclacZ by FhlA165 and FhlA167 is O2- and nitrate-independent
Böck et al. previously proposed that the lack of hyc operon expression during conditions of aerobic (O2) or anaerobic respiration (nitrate) was due to a lack of sufficient levels of formate to interact with FhlA (Rossmann et al., 1991 ). The formate-independence of hyclacZ expression in the presence of FhlA165 and FhlA167 should lead to the production of significant ß-galactosidase activity even under aerobic growth conditions.

Strain WS127(pWS2) with wild-type fhlA+ did not produce ß-galactosidase activity when grown under aerobic growth conditions (Table 4). Addition of formate to the medium increased the level of ß-galactosidase activity only to approximately 200 units and this represents 21% of the anaerobic level (Table 4). Exogenously added formate appears to be unable to support hyc expression to an optimum level when the cells are cultured under aerobic conditions. Even in the presence of O2, PhyclacZ was activated by FhlA165 protein to a level higher than that of the wild-type strain grown anaerobically in L-broth containing glucose (1600 and 1200 units ß-galactosidase activity, respectively). However, the level of ß-galactosidase activity produced by WS127(pWS165) was still only about 50% of the same strain grown under anaerobic conditions. Including formate in the aerobic growth medium did not increase the level of PhyclacZ expression in the presence of FhlA165. Since FhlA165 is formate-independent for the transcription of the hyc operon, this O2 effect could be due to a reduction in the expression of the fhlA gene. Previous work from our laboratory showed that fhlAlacZ expression under aerobic growth conditions is only approximately 15% of the values obtained with anaerobic cultures (J. A. Maupin & K. T. Shanmugam, unpublished data). Recently, the oxyS small RNA species was shown to play a role in the regulation of fhlA, and this could account for O2-dependent control of the fhlA and fhlA165 genes (Altuvia et al., 1998 ) which indirectly controls the level of FHL in the cell. However, transducing an oxyS deletion mutation into strain WS127 did not affect the level of expression of hyclac or fhlAlac in the presence of O2 (data not shown). There appears to be an as yet unidentified O2-dependent control of the hyc operon in E. coli. Although both the native FhlA and FhlA165 are produced to the same level in the cell, it is possible that there is a difference in the affinity of the two proteins for the Phyc which may account for the difference in the level of expression of hyclac in the aerobic cell.


View this table:
[in this window]
[in a new window]
 
Table 4. Effect of O2 and nitrate on the FhlA165- and FhlA167-mediated expression of PhyclacZ

 
As formate is also an electron donor for nitrate reduction, the presence of nitrate in vivo is expected to decrease the formate pool and consequently reduce the expression of the hyc operon. As expected, hyclacZ was expressed at a very low level in the presence of nitrate when native FhlA served as the activator (Table 4). Supplementing the medium with formate restored hyclacZ expression. However, with FhlA165 as the activator, nitrate had no detectable effect on the expression of the hyc operon, supporting the model that formate is not a required effector for FhlA165.

It is interesting to note that although FhlA167 supported effector-independent expression of hyclac when the cells were grown anaerobically, in an aerobic cell this allele supported the production of only approximately 200 units ß-galactosidase activity (approximately 15% of the anaerobic level). The addition of formate to the medium had a minimal effect on FhlA167-mediated hyc expression in the aerobic cell: this is in agreement with its formate independence. Addition of nitrate to the anaerobic culture also reduced the hyclac expression in the presence of FhlA167, but the nitrate effect was minimal compared to that of O2. This difference in the response of FhlA165 and FhlA167 to terminal electron acceptors suggests that the amino acids between 117 and 374 play a significant role in sensing the redox properties of the environment. Alternatively, the affinity of the two proteins for the Phyc may be different.

A sequence in the N-terminal region of FhlA is similar to a conserved region in ATPases of ABC transporters
As the first 117 amino acids of FhlA had been shown to be responsible for effector-mediated control of the hyc operon, this region of the FhlA sequence was compared with the sequences of other known E. coli proteins. This comparison revealed significant similarity between the N-terminal amino acid sequence of FhlA and sequences found in the energizer proteins of ABC transporters (Fig. 2). The C-terminal end of the presented ABC protein sequences corresponds to the ‘Walker’ B site, and the N-terminal end represents the unique ABC signature sequence (Linton & Higgins, 1998 ). Upstream of this ATPase-like sequence resides a helical domain which interacts with the membrane component of the ABC transport system (Hunke et al., 2000 ). Although the critical aspartate and proline amino acids, present in all ‘Walker’ B sites of the ABC proteins and the unique ABC signature sequence, were not present in the FhlA protein 18 of the 31 amino acids within this sequence (between amino acids 7 and 37) were similar or identical to a derived consensus sequence of the ABC proteins (Fig. 2). The observed sequence similarity between the FhlA and ABC-ATPases raised interesting questions about the role of this N-terminal region of the FhlA protein. Does this region of FhlA along with the central domain of the protein contribute to formate-dependent ATP hydrolysis? However, it should be noted that the N-terminal domain of FhlA by itself has no detectable ATPase activity (Leonhartsberger et al., 2000 ).



View larger version (36K):
[in this window]
[in a new window]
 
Fig. 2. Alignment of the N-terminal region of the FhlA protein with ATPases from various E. coli ABC transport systems. Amino acids 7–37 of the FhlA N-terminal region are shown on the top line. Below are proteins which all act as energizer proteins (cytoplasmic ATPases) in ABC-type transport systems within E. coli. The bottom line represents a consensus sequence derived from analysis of all nine proteins. References for each of the proteins are as follows: FhlA (Maupin & Shanmugam, 1990 ; Schlensog & Böck, 1990 ); PotA (Kashiwagi et al., 1993 ); ModF (Grunden et al., 1996 ); MalK (Shuman & Silhavy, 1981 ); XylG (Rosenfeld et al., 1984 ); LivG (Adams et al., 1990 ); HisP (Ames & Lever, 1970 ); ModC (Maupin-Furlow et al., 1995 ); PotG (Pistocchi et al., 1993 ).

 
To evaluate the significance of the N-terminal region of the FhlA protein, the DNA sequence corresponding to the amino acids 7–37 was deleted from the fhlA gene, and the resulting fhlA9-2 allele was used in hyclac expression studies (Fig. 1). The fhlA9-2 allele failed to support hyc expression or production of FHL activity, even when the medium was supplemented with formate at levels as high as 30 mM and/or 1 mM molybdate. These results show that this segment of the FhlA protein is critical for in vivo activity of the FhlA protein. It is possible that the FhlA9-2 protein is still formate-dependent for its activity and the inability to produce a FhlA9-2–formate complex in vivo apparently negated its activity. Removal of an additional 80 amino acids (to amino acid 117) completely reversed this defect, as FhlA167 is formate-independent (Fig. 1, Table 2).

Production of active FHL is independent of formate in the presence of FhlA165
Since transcription of the hyc operon was at its highest in the presence of FhlA165, the need for formate in the conversion of apo-FDH-H to active FDH-H was determined. Strain SE2007, which carries a mutation in the chromosomal fhlA gene, produced approximately 125 units FHL activity with plasmid pWS165. This value was slightly higher than the 110 units FHL activity produced by strain SE2007(pWS2) with wild-type fhlA. However in a fhlA/pfl double mutant, only the strain carrying fhlA165 produced FHL activity (approximately 90 units), again confirming the formate-independence for hyc and fdhF expression in the presence of FhlA165. These results further demonstrate that formate is only required for transcriptional activation of fdhF and hyc, and not for maturation and activation of the apoproteins to activate the FHL.

FhlA165 has a higher affinity for hyc DNA
The higher level of hyclac expression when activated by FhlA165 (Table 2) prompted investigation into the affinity of FhlA165 protein for the target DNA. Wild-type FhlA (250 nM) shifted the electrophoretic mobility of the Phyc DNA and completely shifted the DNA at a concentration of 1 µM FhlA (Fig. 3). In contrast, the FhlA165 protein shifted the mobility of the DNA at a concentration of 100 nM and completely shifted the DNA at 500 nM FhlA165, suggesting that FhlA165 had a higher affinity for the target DNA sequence. Based on phosphorimager analysis of the DNA electrophoretic mobility shift presented in Fig. 3, the apparent Kd for the binding of FhlA165 to DNA was calculated to be 125 nM, while the value for the native FhlA protein was 270 nM. In similar mobility shift experiments, including formate in the binding reaction had no effect on the apparent Kd of the interaction between DNA and protein (data not shown). The higher affinity of FhlA165 for the Phyc DNA could explain the higher level of hyc operon expression seen in the presence of the FhlA165 protein (Tables 2 and 4), compared to the level of expression when activated by native FhlA. However, it could not be ruled out that FhlA165 had an increased ability to activate transcription.



View larger version (89K):
[in this window]
[in a new window]
 
Fig. 3. DNA electrophoretic mobility-shift experiments with purified FhlA and mutated FhlA proteins with Phyc DNA. All reactions contained 50 fmol of 154 bp hyc upstream DNA, and (A) FhlA, (B) FhlA165, (C) FhlA9-2 and (D) FhlA167. The concentration of each of the FhlA proteins added is given in µM below each image. Arrows indicate the DNA–protein complex.

 
FhlA167 had a very low affinity for Phyc DNA (Fig. 3D), although this allele activated hyclac to the same level as the native protein, in vivo. The addition of formate to the binding reaction or to the electrophoresis conditions had no effect on the FhlA167–Phyc DNA interaction. In contrast, the FhlA9-2 allele, which did not support hyc operon expression, did bind to the Phyc DNA in vitro and shifted the electrophoretic mobility of the Phyc DNA (Fig. 3C). This change in the electrophoretic mobility of the Phyc DNA saturated at a protein concentration of approximately 100 nM FhlA9-2, and at this concentration only a small fraction of the total DNA was protein-bound. At a protein concentration of 150 nM, most of the DNA was found in larger DNA–protein complexes.

ATPase activity of FhlA mutant derivatives
Members of the NtrC/NifA family of proteins including FhlA have a promoter DNA-dependent ATPase activity (Weiss et al., 1991 ; Hopper & Böck, 1995 ). This activity is associated with the formation of the DNA-open complex, a prelude to the initiation of transcription (Wedel & Kustu, 1995 ). Since the FhlA protein alleles had differing in vivo activities and promoter DNA-binding characteristics, the Phyc-dependent ATPase activity of the various FhlA alleles was determined.

Purified native FhlA protein had a basal ATPase activity of 250 units (Fig. 4). This activity increased to 320 units in the presence of formate, and the optimum concentration of formate was between 10 and 25 mM. At 50 mM formate, the ATPase activity of FhlA protein decreased. In the presence of Phyc DNA the ATPase activity of the FhlA protein increased by over 200 pmol ATP hydrolysed (pmol FhlA)-1 (approximately 25 mM formate present) (Fig. 4a). This characteristic formate- and DNA-dependent increase in ATPase activity of the FhlA protein is similar to that reported by Hopper & Böck (1995) .



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 4. ATPase activity of native FhlA protein and mutant forms of the FhlA protein in the presence of formate and Phyc DNA. ATPase activities are given in relative units [pmol ATP hydrolysed (pmol FhlA)-1]. DNA (plasmid pFGH100) was added to give a final concentration of 0·1 µM. (a) Native FhlA protein and FhlA165. {bullet}, FhlA (final concn 1 µM) and DNA; {circ}, FhlA only (final concn 1 µM); {blacksquare}, FhlA165 (final concn 2 µM) and DNA; {square}, FhlA165 only (final concn 2 µM). (b) FhlA9-2 and FhlA167. {bullet}, FhlA9-2 (final concn 0·13 µM) and DNA; {circ}, FhlA9-2 only (final concn 0·13 µM); {blacksquare}, FhlA167 (final concn 0·13 µM) and DNA; {square}, FhlA167 only (final concn 0·13 µM).

 
The basal ATPase activity of the FhlA165 protein was only 60 pmol ATP hydrolysed (pmol FhlA)-1 (Fig. 4), and as expected this activity was not increased by formate. In the presence of Phyc DNA, the level of ATPase activity also increased by approximately 200 pmol ATP hydrolysed (pmol FhlA)-1. In contrast to the native protein, which responded positively to lower concentrations of formate, the FhlA165 reacted negatively to formate in the ATPase assay. The negative response to formate seen in vitro could also account for the formate-dependent decrease in hyclacZ activation in the mod mutant, strain WS113(pWS165) (Table 2). The similar response of ATPase activity of the native FhlA and FhlA165 proteins to DNA suggested that these proteins, once bound to the hyc upstream DNA, activated transcription at the same level, and the observed difference in the hyc operon expression could be attributed to the differences in the affinity of the proteins to the Phyc DNA.

In these experiments the native FhlA protein had a higher than expected basal ATPase activity. This is in contrast to the ATPase activity of FhlA protein reported by Hopper & Böck (1995) . This difference could be due to variations in the protein purification methods employed in these two studies. The absence of formate-dependent ATPase activity in FhlA165 suggested that the deletion of the unique N-terminal domain of FhlA removed a previously unknown ATPase activity, which may be specific for formate binding.

The FhlA167 protein had an ATPase activity of approximately 300 units which was not altered either by formate or by Phyc DNA. This is in agreement with its lower affinity to Phyc DNA (Fig. 4b). The FhlA9-2 protein also produced approximately 300 pmol ATP hydrolysed (pmol FhlA)-1 ATPase activity which was not influenced by the presence of formate. However, the ATPase activity of this protein increased by more than threefold in the presence of Phyc DNA. Addition of both Phyc DNA and formate to the reaction further increased the ATPase activity of the FhlA9-2 protein by twofold. In the presence of 10 mM formate and Phyc DNA the ATPase activity was over 1500 pmol ATP hydrolysed (pmol FhlA)-1, the highest among the four FhlA alleles tested. Increasing the formate concentration beyond 10 mM inhibited the DNA-dependent ATPase activity. The ability of FhlA9-2 to bind to Phyc DNA in vitro and to hydrolyse ATP in a DNA-dependent manner shows that the FhlA9-2 protein is active in vitro. The formate-dependent increase in ATPase activity of this allele is similar to that of the native FhlA protein and is in contrast to that of FhlA165 and FhlA167. These results suggest that the FhlA9-2 protein requires formate for activity and the FhlA9-2–formate complex is not formed in vivo because of the deletion of the amino acids 7–37.

The inability of the FhlA9-2 protein to activate the expression of the hyc operon in vivo combined with the higher than expected activity in vitro raises the possibility that the fhlA9-2 is not transcribed in vivo. To test this, the upstream region of the fhlA gene in both pWS9 and pWS9-2 was sequenced. The two sequences were identical for approximately 500 bases upstream of the translation start site. In a separate experiment, total RNA was isolated from strains carrying either pWS9-2 or its parent pWS9 grown under anaerobic conditions in L-broth containing glucose (0·3%). A cDNA corresponding to the fhlA gene was synthesized from both samples using Moloney Murine Leukaemia virus reverse transcriptase (Superscript II; Life Technologies). The cDNA was quantitated by real-time PCR using SYBR green (iCycler; Bio-Rad). The results of these experiments showed that the level of cDNA produced from the total RNA from strains with plasmid pWS9 or pWS9-2 was comparable, indicating that the fhlA9-2 allele was also transcribed in E. coli. The inability to activate the hyc operon is apparently due to the deletion of amino acids in the protein.


   CONCLUSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
 
Results from this study show that removal of the first 117 amino acids from the FhlA protein (FhlA167) leads to effector-independence for expression of the hyc operon. Extending the deletion to amino acid 374 (FhlA165) increases the ability of the protein to bind Phyc DNA in vitro, enhances the level of hyc operon expression in vivo, and also overcomes the redox-dependent control of hyc operon expression. The amino acids between 117 and 374 could play a significant role in modulating the level of hyc operon transcription via the central ATP-binding domain and the C-terminal domain (amino acids 375–692 of the FhlA protein). Although the need for MoeA for expression of the hyc operon is overcome by either of the two deletion derivatives FhlA165 and FhlA167, ModE-Mo is still required as a secondary activator for the optimum expression of the hyc operon. These results further support the proposal that the MoeA-catalytic product interacts with the native FhlA protein for optimum expression of the hyc operon. Removing the sequence of amino acids with similarity to the sequences found in the ATPase components of ABC-type transport systems eliminates the in vivo activity, but not the formate and Phyc DNA dependent in vitro ATPase activity. This suggests that in vivo the FhlA9-2 protein is unable to produce a formate-bound complex. The FhlA protein could accept formate directly from the transporter, and the formate-dependent ATPase activity may signify this interaction. Direct transfer of formate to the FhlA protein would also minimize formate accumulation in the cytoplasm. The amino acids 7–37 may play a role in transferring formate from the transporter to the appropriate location in the FhlA protein. The sequence similarity with the ABC-ATPases, combined with the absence of formate-dependent ATPase activity in FhlA165 and FhlA167, suggests that the native FhlA protein contains two ATPase activities: one associated with DNA-binding and transcription activation and the other related to the interaction with formate.


   ACKNOWLEDGEMENTS
 
We thank G. Storz for providing oxyS mutants of E. coli. This work was supported in part by funds from the National Institutes of Health (GM48667) and the Florida Agricultural Experiment Station.

Florida Agricultural Experiment Station Journal Series No. R-08430.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES
 
Adams, M. D., Wagner, L. M., Graddis, T. J., Landick, R., Antonucci, T. K., Gibson, A. L. & Oxender, D. L. (1990). Nucleotide sequence and genetic characterization reveal six essential genes for the LIV-I and LS transport systems of Escherichia coli. J Biol Chem 265, 11436-11443.[Abstract/Free Full Text]

Altuvia, S., Zhang, A., Argaman, L., Tiwari, A. & Storz, G. (1998). The Escherichia coli OxyS regulatory RNA represses fhlA translation by blocking ribosome binding. EMBO J 17, 6069-6075.[Abstract/Free Full Text]

Ames, G. F. & Lever, J. (1970). Components of histidine transport: histidine-binding proteins and HisP protein. Proc Natl Acad Sci USA 66, 1096-1103.[Abstract]

Böck, A. & Sawers, G. (1996). Fermentation. In Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd edn, pp. 262–282. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Boyington, J. C., Gladyshev, V. N., Khangulov, S. V., Stadtman, T. C. & Sun, P. D. (1997). Crystal structure of formate dehydrogenase H: catalysis involving Mo, molybdopterin, selenocysteine, and an Fe4S4 cluster. Science 275, 1305-1308.[Abstract/Free Full Text]

Drummond, M., Whitty, P. & Wootton, J. (1986). Sequence and domain relationships of ntrC and nifA from Klebsiella pneumoniae: homologies to other regulatory proteins. EMBO J 5, 441-447.[Abstract]

Grunden, A. M. & Shanmugam, K. T. (1997). Molybdate transport and regulation in bacteria. Arch Microbiol 168, 345-354.[Medline]

Grunden, A. M., Ray, R. M., Rosentel, J. K., Healy, F. G. & Shanmugam, K. T. (1996). Repression of the Escherichia coli modABCD (molybdate transport) operon by ModE. J Bacteriol 178, 735-744.[Abstract]

Hasona, A., Ray, R. M. & Shanmugam, K. T. (1998a). Physiological and genetic analyses leading to identification of a biochemical role for the moeA (molybdate metabolism) gene product in Escherichia coli. J Bacteriol 180, 1466-1472.[Abstract/Free Full Text]

Hasona, A., Self, W. T., Ray, R. M. & Shanmugam, K. T. (1998b). Molybdate-dependent transcription of hyc and nar operons of Escherichia coli requires MoeA protein and ModE-molybdate. FEMS Microbiol Lett 169, 111-116.[Medline]

Hopper, S. & Böck, A. (1995). Effector-mediated stimulation of ATPase activity by the sigma 54-dependent transcriptional activator FHLA from Escherichia coli. J Bacteriol 177, 2798-2803.[Abstract]

Hunke, S., Mourez, M., Jehanno, M., Dassa, E. & Schneider, E. (2000). ATP modulates subunit–subunit interactions in an ATP-binding cassette transporter (MalFGK2) determined by site-directed chemical cross-linking. J Biol Chem 275, 15526-15534.[Abstract/Free Full Text]

Kashiwagi, K., Miyamoto, S., Nukui, E., Kobayashi, H. & Igarashi, K. (1993). Functions of potA and potD proteins in spermidine-preferential uptake system in Escherichia coli. J Biol Chem 268, 19358-19363.[Abstract/Free Full Text]

Keener, J. & Kustu, S. (1988). Protein kinase and phosphoprotein phosphatase activities of nitrogen regulatory proteins NTRB and NTRC of enteric bacteria: roles of the conserved amino-terminal domain of NTRC. Proc Natl Acad Sci USA 85, 4976-4980.[Abstract]

Kim, R., Sandler, S. J., Goldman, S., Yokota, H., Clark, A. J. & Kim, S. H. (1998). Overexpression of archael proteins in E. coli. Biotechnol Lett 20, 207-210.

Korsa, I. & Böck, A. (1997). Characterization of fhlA mutations resulting in ligand-independent transcriptional activation and ATP hydrolysis. J Bacteriol 179, 41-45.[Abstract]

Lee, J. H., Patel, P., Sankar, P. & Shanmugam, K. T. (1985). Isolation and characterization of mutant strains of Escherichia coli altered in H2 metabolism. J Bacteriol 162, 344-352.[Medline]

Leonhartsberger, S., Ehrenreich, A. & Böck, A. (2000). Analysis of the domain structure and the DNA binding site of the transcriptional activator FhlA. Eur J Biochem 267, 3672-3684.[Abstract/Free Full Text]

Linton, K. J. & Higgins, C. F. (1998). The Escherichia coli ATP-binding cassette (ABC) proteins. Mol Microbiol 28, 5-13.[Medline]

Lopilato, J., Bortner, S. & Beckwith, J. (1986). Mutations in a new chromosomal gene of Escherichia coli K-12, pcnB, reduce plasmid copy number of pBR322 and its derivatives. Mol Gen Genet 205, 285-290.[Medline]

Maupin, J. A. & Shanmugam, K. T. (1990). Genetic regulation of formate hydrogenlyase of Escherichia coli: role of the fhlA gene product as a transcriptional activator for a new regulatory gene, fhlB. J Bacteriol 172, 4798-4806.[Medline]

Maupin-Furlow, J. A., Rosentel, J. K., Lee, J. H., Deppenmeier, U., Gunsalus, R. P. & Shanmugam, K. T. (1995). Genetic analysis of the modABCD (molybdate transport) operon of Escherichia coli. J Bacteriol 177, 4851-4856.[Abstract]

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

Nixon, B. T., Ronson, C. W. & Ausubel, F. M. (1986). Two-component regulatory systems responsive to environmental stimuli share strongly conserved domains with the nitrogen assimilation regulatory genes ntrB and ntrC. Proc Natl Acad Sci USA 83, 7850-7854.[Abstract]

Pistocchi, R., Kashiwagi, K., Miyamoto, S., Nukui, E., Sadakata, Y., Kobayashi, H. & Igarashi, K. (1993). Characteristics of the operon for a putrescine transport system that maps at 19 minutes on the Escherichia coli chromosome. J Biol Chem 268, 146-152.[Abstract/Free Full Text]

Rosenfeld, S. A., Stevis, P. E. & Ho, N. W. (1984). Cloning and characterization of the xyl genes from Escherichia coli. Mol Gen Genet 194, 410-415.[Medline]

Rosentel, J. K., Healy, F., Maupin-Furlow, J. A., Lee, J. H. & Shanmugam, K. T. (1995). Molybdate and regulation of mod (molybdate transport), fdhF, and hyc (formate hydrogenlyase) operons in Escherichia coli. J Bacteriol 177, 4857-4864.[Abstract]

Rossmann, R., Sawers, G. & Böck, 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]

Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74, 5463-5467.[Abstract]

Sankar, P., Lee, J. H. & Shanmugam, K. T. (1988). Gene-product relationships of fhlA and fdv genes of Escherichia coli. J Bacteriol 170, 5440-5445.[Medline]

Sauter, M., Böhm, R. & Böck, A. (1992). Mutational analysis of the operon (hyc) determining hydrogenase 3 formation in Escherichia coli. Mol Microbiol 6, 1523-1532.[Medline]

Schlensog, V. & Böck, A. (1990). Identification and sequence analysis of the gene encoding the transcriptional activator of the formate hydrogenlyase system of Escherichia coli. Mol Microbiol 4, 1319-1327.[Medline]

Schlensog, V., Lutz, S. & Böck, A. (1994). Purification and DNA-binding properties of FHLA, the transcriptional activator of the formate hydrogenlyase system from Escherichia coli. J Biol Chem 269, 19590-19596.[Abstract/Free Full Text]

Self, W. T. (1998). Genetic and biochemical analysis of the molybdate-dependent expression of hyc operon in Escherichia coli. PhD thesis, University of Florida, USA.

Self, W. T. & Shanmugam, K. T. (2000). Isolation and characterization of mutant FhlA proteins which activate transcription of hyc operon (formate hydrogenlyase) of Escherichia coli in the absence of molybdate. FEMS Microbiol Lett 184, 47-52.[Medline]

Self, W. T., Grunden, A. M., Hasona, A. & Shanmugam, K. T. (1999). Transcriptional regulation of molybdoenzyme synthesis in Escherichia coli in response to molybdenum: ModE-molybdate, a repressor of the modABCD (molybdate transport) operon is a secondary transcriptional activator for the hyc and nar operons. Microbiology 145, 41-55.[Abstract]

Shuman, H. A. & Silhavy, T. J. (1981). Identification of the malK gene product. A peripheral membrane component of the Escherichia coli maltose transport system. J Biol Chem 256, 560-562.[Abstract/Free Full Text]

Tabor, S. & Richardson, C. C. (1985). A bacteriophage T7 RNA polymerase/promoter system for controlled exclusive expression of specific genes. Proc Natl Acad Sci USA 82, 1074-1078.[Abstract]

Wedel, A. & Kustu, S. (1995). The bacterial enhancer-binding protein NTRC is a molecular machine: ATP hydrolysis is coupled to transcriptional activation. Genes Dev 9, 2042-2052.[Abstract]

Weiss, D. S., Batut, J., Klose, K. E., Keener, J. & Kustu, S. (1991). The phosphorylated form of the enhancer-binding protein NTRC has an ATPase activity that is essential for activation of transcription. Cell 67, 155-167.[Medline]

Received 26 June 2001; revised 14 August 2001; accepted 20 August 2001.