![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,
Inna Grinberg,2,
Batia Gorovitz-Harris,2
Hanna Lundström,1
Ilya Borovok,2
Yair Aharonowitz,2
Britt-Marie Sjöberg,1* and
Gerald Cohen2*
Department of Molecular Biology and Functional Genomics, Stockholm University, SE-10691 Stockholm, Sweden,1 Department of Molecular Microbiology and Biotechnology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv, 69978, Israel2
Received 25 March 2007/ Accepted 1 May 2007
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
While eukaryotes in general employ just the class I RNR, many bacteria possess two or even all three RNR classes, the genes of which are typically organized in operons (45) (for a comprehensive listing of RNRs in bacteria, see http://rnrdb.molbio.su.se). Escherichia coli, Salmonella enterica serovar Typhimurium, and many enterobacteria contain genes encoding three RNRs, class Ia and class Ib RNRs (subdivisions of the class I RNR) and the class III RNR. In E. coli, the class Ia RNR operon contains nrdA and nrdB genes that code for the NrdA catalytic and the NrdB radical-generating subunits, respectively; the class Ib RNR operon consists of four genes, nrdE and nrdF code for the corresponding subunits NrdE and NrdF, respectively, while nrdH codes for NrdH, an 
9-kDa protein that functions as a specific electron donor, and nrdI codes for NrdI, an 15-kDa protein whose function is still unknown. The NrdAB and NrdEF RNRs have limited sequence identity but share many catalytic properties (27). Both require oxygen for generation of a tyrosyl radical stabilized by an iron center, which transfers the radical to an active-site cysteine of NrdA or NrdE. Class III RNRs are encoded by nrdD, which occurs in an operon containing nrdG, coding for a specific activase that uses S-adenosylmethionine to create a stable oxygen-sensitive glycyl radical close to the active site of NrdD. Although considerable information exists regarding the biochemical, structural, and allosteric properties of RNRs (34), our knowledge of the molecular genetic mechanisms by which their differential expression is regulated is more limited.
Aerobic growth of E. coli depends on a functional class Ia RNR (23). The aerobic class Ib RNR is not essential for growth, is ordinarily very poorly expressed, and cannot take the place of the class Ia RNR (23). However, when additional copies of the class Ib nrdHIEF genes are present on a low-copy-number plasmid or an additional copy is inserted into the chromosome, class Ib RNR can support the growth of an nrdA temperature-sensitive mutant at the restrictive temperature (24) or an nrdAB deletion mutant (16).
The E. coli class Ia nrdAB RNR genes constitute a tightly regulated transcription unit (19). The operon contains a third nonessential gene, annotated as yfaE, that is predicted to encode a ferredoxin-related protein (3). Under aerobic conditions, the DnaA, Fis, and IciA proteins bind to specific sites in the promoter region to regulate nrdAB expression, while several cis-acting sites and an AT-rich region are important in coupling transcription, in a complex way, to the cell cycle (18, 22). Herrick and Sclavi (21) have recently argued in a comprehensive review that an unknown protein is responsible for controlling cell cycle-dependent nrdAB expression. They propose that NrdR, a transcription factor first identified in streptomycetes (6), is a likely candidate to constitute the regulatory link between nrdAB expression and nucleotide pool size. The less well-studied E. coli class Ib nrdHIEF RNR genes form a tight transcription unit in which all four genes are coordinately transcribed from a single promoter (23, 33). Expression of the two sets of RNR genes is dependent on environmental conditions. Under standard laboratory conditions, transcription of E. coli nrdAB is much greater than that of nrdHIEF. Both nrdAB and nrdHIEF expression are elevated by DNA damage and by inhibitors of DNA synthesis (15, 23, 33). Hydroxyurea, a free-radical scavenger that inhibits class I RNR enzyme activity (2, 41), induces expression of both nrdAB and nrdHIEF operons (33). In contrast, nrdHIEF, but not nrdAB, is highly expressed in response to exposure to oxidative-stress agents (33, 36). Also, nrdHIEF, but not nrdAB, expression is strongly enhanced in mutants defective in catalase, superoxide dismutase, and alkyl hydroperoxide reductase, which eliminate superoxides and peroxides (33). Furthermore, nrdHIEF transcription is highly upregulated in mutants lacking a functional thioredoxin Trx1 or glutaredoxin Grx1, the major reductants of class Ia RNRs (33). Monje-Casas et al. have suggested that the physiological role of the class Ib RNR is primarily a response to oxidative stress, which causes DNA damage (33). However, the mechanism that triggers nrdHIEF expression is distinct from either of the known global oxidative-stress regulators, OxyR and SoxRS (43). In addition, the presence of a FUR box in the nrdHIEF promoter region suggests that iron metabolism might regulate the transcription of class Ib RNR (31, 48).
Anaerobic growth of E. coli requires a functional class III RNR (14). When E. coli is shifted from aerobic to microaerophilic or anaerobic growth conditions, nrdAB expression is downregulated with concomitant upregulation of nrdDG expression. Induction of nrdDG transcription is dependent on the FNR system but is independent of the ArcAB two-component system (8). In an nrdDG mutant, nrdAB expression is upregulated. NrdAB can allow partial growth in the absence of the class III RNR system when traces of oxygen are present, whereas NrdEF cannot, even when the nrdEF genes are overexpressed (14). However, the cells grow poorly, have abnormal morphology, and exhibit extensive filamentation. Moreover, deletion of nrdAB and nrdDG abolishes growth in strict anaerobiosis. Similar findings were reported for Lactococcus lactis, which contains class Ib and class III RNRs, an nrdD mutant of which was able to grow well under microaerophilic conditions (26). However, when strict anaerobic conditions were applied, growth was dependent on a functional NrdDG (46). These results suggest that while NrdAB and NrdEF both require oxygen, their catalytic activities differ under conditions of limiting oxygen concentration. In contrast, in Bacillus subtilis, the class Ib RNR, the only RNR identified in the organism, is essential for both aerobic and anaerobic growth (20).
The recent discovery in Streptomyces of a novel regulatory protein, NrdR, that controls RNR gene expression prompted the studies reported here. Streptomyces coelicolor contains genes encoding class Ia and class II RNRs that are differentially expressed during vegetative growth (7). Either RNR is sufficient for aerobic growth (6). The S. coelicolor NrdR protein is a zinc finger/ATP cone (1) transcriptional regulatory protein (COG1327). Abolishment of NrdR function by a gene knockout resulted in a dramatic increase in transcription of both sets of RNR genes (6). Further studies revealed that NrdR binds to tandem repeat sequences, called NrdR boxes, located in or near to the promoter regions of the class Ia and class II RNR operons (17). NrdR boxes are widespread in diverse bacterial genomes and are almost invariably located in the regulatory regions of different RNR genes (38). In this paper, we explore the role of the E. coli homolog of NrdR in regulating the transcription of RNR genes. Although NrdR is thought to regulate the expression of all RNR operons, here, we focus primarily on its effect on control of the aerobic class Ia and class Ib RNR sets of genes.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
– ilvG rfb-50 rph-1] is referred to as the wild type. E. coli strain XL1 {recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F– proAB lacIq lacZ 
M15 Tn10 (Tcr)]} was used for plasmid constructions; BL21(DE3) [F– ompT hsdSB(rB– mB–) gal dcm (DE3)] was used for protein overexpression. Culture media and DNA manipulations. E. coli strains were grown in Luria-Bertani (LB) medium and supplemented with kanamycin or ampicillin (50 µg/ml and 100 µg/ml, respectively) when appropriate. Plasmid DNA was isolated using the High Pure Plasmid Isolation Kit (Roche). DNA restriction digestions and ligations were carried out according to the manufacturer's instructions. DNA linear fragments were isolated from a 0.9% agarose (Sigma) gel using the QIAquick Gel Extraction Kit (QIAGEN). DNA manipulations were as described previously (39). Electroporation was performed with a Gene Pulsar II apparatus (Bio-Rad Laboratories) according to the manufacturer's instructions.
Cloning of the nrdR gene. The E. coli nrdR gene was amplified by PCR from MG1655 genomic DNA as described previously (17) using nrdR-For (5'-CATATGCATTGCCCATTCTGTTTCGC) and nrdR-Rev (5'-CTCGAGGTCCTCCAGGCGCGCGATC) as the forward and reverse primers, containing NdeI and XhoI restriction sites, respectively. The PCR-amplified fragment was eluted from an agarose gel, ligated to pGEM-T Easy vector (Promega), and introduced into E. coli BL21 by electroporation. Positive transformants were detected by blue-white screening and screened by colony PCR. DNA inserts were sequenced to verify their correctness. The pGEM construct was digested with the restriction endonucleases NdeI and XhoI (Fermentas) and electrophoresed, and the small fragment containing the nrdR gene was eluted. The expression vector pET30a(+) (Novagen) was digested with NdeI and XhoI, electrophoresed, and eluted. T4 DNA ligase (Takara) was used to join the two DNA fragments to obtain pET30a(+)::nrdR, which was introduced into XL1-Blue by electroporation, and transformants were selected for kanamycin resistance. The nrdR DNA insert and the adjacent DNA regions were sequenced to verify the correctness of the construct. The E. coli nrdR recombinant protein was expressed with a hexahistidine tag at its C terminus.
Protein overexpression. An overnight culture of E. coli BL21(DE3)/pET30a(+)::nrdR was diluted in LB medium containing kanamycin (50 µg/ml) to an absorbance of 0.1 at 600 nm and shaken vigorously at 37°C. At an absorbance at 600 nm of 0.6, isopropyl-ß-D-thiogalactopyranoside (IPTG) (Sigma) was added to a final concentration of 0.4 mM. The cells were incubated for 22 h at 22°C and harvested by centrifugation at 4,000 x g for 20 min at 4°C. The supernatant was discarded, and the cell pellet was stored at –70°C.
Protein purification. Frozen cells were thawed and suspended in sonication buffer: 50 mM Tris-HCl (pH 8.3), 300 mM NaCl, 5 mM imidazole. Phenylmethylsulfonyl fluoride was added to the cell suspension to 1 mM, and the mixture was sonicated in an ultrasonic processor (Misonics) until it was clear. The sonicate was centrifuged at 10,000 x g for 45 min at 4°C. The supernatant was loaded on a 5-ml high-capacity Ni2+-CAM resin column (Sigma) equilibrated with the sonication buffer, and the column was washed first with 40 ml sonication buffer, followed by 20 ml buffer containing 15 mM imidazole. Protein was eluted with buffer containing 250 mM imidazole. Protein samples after Ni affinity purification were dialyzed against 50 mM Tris-HCl (pH 8.5), 300 mM NaCl, 5 mM dithiothreitol, 50 µM ZnCl2, 20% glycerol. Recovery of recombinant protein was monitored by the Bradford assay (Bio-Rad) with bovine serum albumin as the standard (9). Purified recombinant proteins were stored at –70°C.
Gel electrophoretic mobility shift assay. Gel electrophoretic mobility shift assays were performed with DNA probes containing the nrdAB, nrdHIEF, and nrdDG promoter regions as described previously (17). Probes were generated by PCR and purified using a PCR purification kit (Roche). They are denoted nrdA (185 bp), nrdH (180 bp), and nrdD (177 bp), respectively, and were designed so that the two NrdR boxes were positioned in the middle of the probes. A control rib probe (190 bp) containing the promoter region of the rib operon (which includes nrdR) served as a negative control. Mutant nrdA probes, in which the sequence of the NrdR box 1 (upstream box) was changed from 5'-TCACACTATCTTGCAG to 5'-TgAgACataCaTcCAG or the sequence of the NrdR box 2 (downstream box) was changed from 5'-CCCCTATATATAGTGT to 5'-CgCgTAataAaAcTGT (the changed bases are shown in lowercase letters), were created by an overlap PCR procedure as previously described (17). The changes were made from a comparison of the two 16-bp NrdR box sequences; seven conserved bases were changed to the complementary bases. The primers used in generating wild-type and mutant nrd probes are listed in the supplemental material.
DNA fragments were labeled at the 3' end with digoxigenin (DIG)-dUTP using the Terminal Transferase kit (Roche, Mannheim, Germany). Binding reactions were carried out in a final volume of 20 µl containing labeled DNA (3 fmol), binding buffer (20 mM Tris-HCl, pH 9.5, 5% [vol/vol] glycerol, 1 mM MgCl2, 40 mM KCl, 1 mM dithiothreitol), purified recombinant NrdR (1 to 15 µg protein), 1 µg salmon sperm DNA, and 0.1 µg bovine serum albumin. Following 30 min of incubation at 37°C, the reaction products were separated on a native 6% polyacrylamide gel in 0.5x Tris-borate buffer, pH 9.5. The gel was contact blotted onto a Hybond N+ membrane (Amersham Biosciences). Chemiluminescence detection was performed according to the manufacturer's instructions (Roche, Mannheim, Germany). The membrane was exposed to X-ray film (FUJI) for 15 to 40 min at 37°C.
Construction of E. coli nrdR mutants.
The PCR targeting method of Datsenko and Wanner (12) was used to create deletions in nrdR. A 1,371-bp DNA fragment containing the nrdR gene with flanking upstream (463-bp) and downstream (484-bp) regions was amplified by PCR from genomic DNA with the forward and reverse primers upR1 (5'-CGGGCAAGGGATCATTCGAC) and downR1 (5'-GTGCGCATCCGCTTGAGAAAGC). The PCR product was cloned into pGEM-T Easy vector to give pGEMT::nrdR. The plasmid was digested with HpaI and Bsp119I restriction endonucleases (Fermentas) and ligated to a PCR-amplified 
2 promoterless and terminatorless kanamycin (aph-3) resistance cassette (35) digested with HpaI and Bsp119I enzymes to give pGEMT::nrdR1. The PCR product obtained after amplification of pGEMT::nrdR1 with the upR1 and downR1 primers was electroporated into E. coli MG1655 containing plasmid pKD46 (12). Transformants were incubated at 30°C for 2 h in SOC medium (39) supplemented with 10 mM arabinose, and kanamycin-resistant colonies were selected on LB agar plates supplemented with kanamycin. Curing of pKD46 plasmid was effected by overnight incubation at 37°C. The resulting strain was designated MG1655nrdR1. The nrdR1 deletion was verified by PCR and sequencing. Figure 1A shows the positions of the HpaI and Bsp119I restriction sites used to create the 424-bp nrdR1 deletion in which codons 1 to 134 were replaced by the kanamycin cassette. An nrdR deletion mutant lacking the major portion of the ATP cone domain was constructed in a similar way. Plasmid pGEMT::nrdR was digested with AjiI and Bsp119I restriction endonucleases and ligated with the HpaI- and Bsp119I-digested kanamycin resistance cassette to give pGEMT::nrdR2. Transformants were processed as described above to yield MG1655 nrdR2. The nrdR2 deletion was verified by PCR and sequencing. Figure 1A shows the positions of the AjiI and Bsp119I restriction sites used to create the 190-bp nrdR2 deletion, in which codons 57 to 134 were replaced by the kanamycin cassette.
|
nrdR1. Genomic-DNA extraction. Genomic DNA was extracted using the Genomic DNA Extraction Kit (QIAGEN) according to the manufacturer's recommendations.
RNA extraction and reverse transcription (RT)-PCRs. For total-RNA preparation, E. coli cultures taken at different optical densities at 550 nm (OD550) were mixed directly with RNA Protect Bacteria Reagent (QIAGEN). RNA was extracted using the RNeasy Mini Kit (QIAGEN) according to the manufacturer's recommendations, including the on-column DNase I treatment. To remove further DNA contamination, the eluted RNA was treated with 2 U RNase-free TURBO DNase (Ambion) and kept at –80°C. The amount of RNA was determined from its 260-nm absorption using a Smart SpecTM spectrophotometer (Bio-Rad).
Conversion of RNA to cDNA was performed using a ThermoScript RT-PCR system from Invitrogen. For all RT-PCRs, 1 µg RNA was used and 1 pmol primer (lw; see Table S1 in the supplemental material) was added, together with 2 µl of a 20-pmol deoxynucleoside triphosphate (dNTP) mixture and diethylpyrocarbonate-treated water to a final volume of 12 µl. The RT primers are listed in Table S1 in the supplemental material. The mixture was then incubated at 65°C for 5 min and then transferred to room temperature. To continue the RT reaction, a master mixture was prepared according to the manufacturer's protocol (Invitrogen), added to the RNA tube, and incubated at 50°C for 1 h, followed by a 5-min inactivation step at 85°C in a Robocycler 96 (Stratagene). cDNA was kept at –80°C.
Real-time PCR.
Real-time PCR was performed in a 20-µl reaction volume with 1 µl cDNA, TaqMan Universal PCR Master Mix, and TaqMan primers and probes according to the manufacturer's protocol (Applied Biosystems). The detections were performed in a model 7000 ABI PRISM sequence detection system from Applied Biosystems. The primers for real-time PCR probes are listed in Table S1 in the supplemental material. Standard curves for each nrd gene were used to analyze the amount of RNA in each sample and in the gapA internal control. The amount of target, normalized to the endogenous reference and relative to the calibrator, is given by 2–CT. In all cases, the values are the mean of three experiments, and the error was below 4%.
Other methods. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed as described by Laemmli (30). Nucleotide sequencing was determined using an AB1 Prism 3100 genetic analyzer (Applied Biosystems) and the Big Dye terminator cycle sequencing kit (Applied Biosystems) as recommended by the manufacturer.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To establish that nrdR is expressed during growth, total RNA was isolated from cells at the early exponential phase (OD550 = 0.5) and early stationary phase (OD550 = 1.5), reverse transcribed with primers specific for nrdR, and analyzed by semiquantitative RT-PCR. Figure 1C shows that nrdR is transcribed at similar levels during growth. To determine whether E. coli NrdR regulates expression of the different nrd genes, we initially used the Targetron Gene Knockout System (Sigma-Aldrich) to disrupt the nrdR gene by insertion of a group II intron conferring kanamycin resistance. The insertion mutation caused a pronounced drop in the growth rate compared to the wild type, presumably due to a strong polar effect on transcription of the downstream ribD, ribH, nusB, thiL, and pgpA genes (data not shown). We therefore created deletions in nrdR in which the zinc finger and ATP cone domains (MG1655nrdR1) or the ATP cone domain alone (MG1655nrdR2) were replaced by a kanamycin cassette without altering the transcription of the downstream nrdR genes (see Material and Methods). Both nrdR deletion-substitution mutants were indistinguishable in their growth on LB broth and minimal medium from the parent strain (data not shown).
To assess the effect of NrdR on the transcription of nrdAB and nrdHIEF genes, quantitative measurements of specific RNA levels were made in the wild type and in nrdR mutants using real-time RT-PCR. The nrdA and nrdE genes served as the respective target sequences for the aerobic RNR genes. In parallel, measurements were made of RNA levels of the class III nrdDG RNR genes using nrdD as the target sequence, even though NrdD is functional only in anaerobic growth. Table 1 shows the amounts of nrdA, nrdE, and nrdD RNAs present in wild-type and mutant strains determined at what are conveniently defined as the early (OD550 = 0.4), mid-exponential (OD550 = 0.8), late exponential (OD550 = 1.3), and early stationary (OD550 = 1.6) phases of growth. In MG1655 (wild type), the amount of nrdA RNA decreased three- to fourfold as the culture progressed from the early exponential to the early stationary phase (Table 1). The fall in nrdA RNA levels presumably reflects the decrease in need for new enzyme synthesis as cells enter the later phases of growth. In the case of nrdE, the decrease in RNA levels in MG1655 during growth is much greater. The amount of nrdE RNA at the early exponential phase is some 30-fold less than that of nrdA RNA at the same growth phase. At later growth phases, the amount of nrdE drops dramatically to several hundredfold less than that of nrdA at the same growth phases. In the case of nrdD, the changes in RNA levels during growth are modest and are less than a twofold increase. Interestingly, the amounts of nrdD RNA relative to nrdA RNA at the early exponential and early stationary phases increased by fivefold. Comparison of the actual amounts of nrdA, nrdE, and nrdD RNAs in MG1655 (Table 1) shows clearly that nrdAB transcripts were far more abundant than nrdHIEF and nrdDG transcripts at all times during growth. Measurements of the amount of nrdR RNA in MG1655 showed a twofold overall drop in level during the growth cycle (Table 1).
|
nrdR1 and MG1655nrdR2, are the ratios of the amounts of specific RNAs made in these strains to that made in MG1655 (wild type) at the same growth phase. The induction values for MG1655 at each growth phase were arbitrarily set as 1.0. In the MG1655nrdR1mutant (which is principally an nrdR null mutant), the level of nrdA RNA changed little during growth. At the early exponential phase, the nrdA induction factor (the nrdA change compared to that of the wild type at the same growth phase) was about 1.5-fold more than that of the wild type at the same growth phase. Subsequently, the induction factor increased and reached about fourfold higher at the early stationary phase. Similar increases occurred in the nrdA induction factor in the MG1655nrdR2 mutant (which lacks only the ATP cone domain). The induction factor was 0.75 at the early exponential phase and increased to some eightfold at the early stationary phase.
|
nrdR1, the nrdE induction value (the nrdE change compared to that of the wild type at the same growth phase) at the early exponential phase was some 4.6-fold more than that of the wild type and markedly increased during growth. It was some 22-fold higher at the mid-exponential, 50-fold higher at the late exponential, and 26-fold higher at the early stationary phases of growth. In MG1655nrdR2, the changes were even more striking. The nrdE induction value was about 7-fold higher than that in the wild type at the early exponential growth phase, 178-fold higher at the mid-exponential phase, and 70- to 75-fold higher at the late exponential and early stationary phases of growth. To confirm that induction of nrdE expression in the nrdR1 deletion strain is a consequence of abrogation of NrdR function, we introduced a pACYC184 low-copy-number plasmid bearing the intact nrdR gene and its regulatory region into MG1655nrdR1. Semiquantitative RT-PCR analysis showed that the plasmid expresses nrdR and causes down regulation of nrdE to the level present in the wild type (data not shown). The impressive increases in nrdE expression in the two nrdR mutant strains over that of the wild type from the mid-exponential stage of growth reflect in part the very low nrdE RNA levels present in the wild type at these stages. As can be seen from Table 1 (and the supplemental material), which reports absolute amounts of nrdA and nrdE RNAs, the class Ia RNR operon is clearly the dominant RNR operon expressed throughout aerobic growth, and the actual amounts of nrdE RNA are 5- to 50-fold less than those of nrdA throughout the growth cycle.
We also monitored the levels of nrdD transcripts, coding for the NrdD anaerobic RNR, in wild-type and mutant strains. Induction levels (the nrdD change compared to that of the wild type at the same growth phase) in MG1655nrdR1 and MG1655nrdR2 were about seven- and fourfold higher at the early exponential phase, respectively. At the later growth phases, the induction values in MG1655nrdR1 were 4- to 6-fold higher, and they were 5- to 10-fold higher in MG1655nrdR2. However, the actual amounts of nrdD RNA were 4- to 10-fold less than those of nrdA RNA (see the supplemental material).
NrdR binds to the promoter regions of all three E. coli RNR operons. On the basis of previous studies that showed that NrdR regulates the transcription of the S. coelicolor class Ia and class II RNR operons and binds to their promoter regions, we supposed that NrdR might control the expression of the E. coli RNR genes in a similar fashion. NrdR is a zinc finger/ATP cone protein that binds to tandem 16-bp repeat sequences, termed NrdR boxes, located almost invariably in front of RNR genes (17, 38). The 5' upstream regions of all three E. coli RNR operons contain two tandem NrdR boxes in their promoter regions (38). Figure 3 shows an alignment of the E. coli and S. enterica serovar Typhimurium nrdAB, nrdHIEF, and nrdDG promoter regions. The NrdR boxes and the positions of the –10 and –35 promoter elements and transcription start sites, where known, are indicated. The alignment reveals that (i) the positions of the NrdR boxes in the nrdAB, nrdHIEF, and nrdDG regulatory regions differ with respect to the positions of their promoter elements; (ii) the sequences of NrdR boxes belonging to the same class of RNR are significantly more alike than the sequences of NrdR boxes from different RNRs of the same organism; and (iii) the proximal NrdR box 2 sequences are more conserved than the distal NrdR box 1 sequences.
|
180-bp DNA probes that spanned the two NrdR boxes. The probes, denoted nrdA (185 bp), nrdH (180 bp), and nrdD (177 bp), were designed so that the two boxes were positioned in the central portion of the probe. Figure 4A shows that NrdR binds to all three probes. The relatively large amounts of protein used in the gel shift experiments and the appearance of multiple DNA-NrdR complexes probably reflect the tendency of NrdR to aggregate into oligomeric forms (17). NrdR did not bind to a control DNA probe consisting of the upstream regulatory region of the rib operon at any of the NrdR concentrations used.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
One possible mechanism to account for differential expression of class I RNR genes assumes that NrdR repression depends on the position of the NrdR boxes with respect to the nrdAB and nrdHIEF promoters. Inspection of the regulatory regions of the E. coli nrdAB and nrdHIEF (and nrdDG) operons shows that they contain tandem repeat sequences that essentially conform to the consensus NrdR box motif, acaCwAtATaTwGtgt (38). The two 16-bp repeats are separated by 15 bp, and a similar organization occurs in the corresponding regulatory regions of the S. enterica serovar Typhimurium RNR operons. One of the sequences (NrdR box 2) was first noted in a comparison of the E. coli and S. enterica serovar Typhimurium promoter regions (25). It was also identified in E. coli as a conserved motif, termed Even 35, by phylogenetic footprinting of orthologous nrdA and nrdD genes in closely related species (37). The positions of the NrdR boxes with respect to the nrdAB and nrdHIEF promoter elements differ (Fig. 3). In the nrdAB promoter region, the two NrdR boxes are located downstream of the –35 promoter recognition element, with NrdR box 1 overlapping the –10 promoter element and the transcription start site. In the nrdHIEF promoter region, the two NrdR boxes bracket the –35 promoter recognition element, with NrdR box 2 overlapping the –10 promoter element. A similar arrangement occurs in the promoter regions of the S. enterica serovar Typhimurium RNR operons. We suppose that NrdR may be more effective in inhibiting nrdHIEF transcription than nrdAB because it more likely blocks the binding of RNA polymerase to the nrdHIEF promoter than to the nrdAB promoter. This view is consistent with the finding that NrdR box 2, but not NrdR box 1, is necessary for NrdR binding (Fig. 4B) and that the position of NrdR box 2 in nrdHIEF overlaps with the –10 promoter element, whereas in nrdAB it is located 16 bp downstream of the –10 element (Fig. 3). A similar mechanism was proposed to account for the differential effect of NrdR on transcription of the S. coelicolor nrdAB and nrdRJ genes, where the NrdR boxes either overlap with the promoter elements or are located just upstream, respectively (17).
Alternatively, differential regulation could be a consequence of differences in the sequences of the NrdR box motifs. Comparison of the sequences of the E. coli and S. enterica serovar Typhimurium nrdAB NrdR boxes reveals high conservation; NrdR box 1 shares 9/16 identities, while NrdR box 2 is identical in all but one of the 16 bases (Fig. 3) A similar picture emerges from a comparison of the E. coli and S. enterica serovar Typhimurium nrdHIEF and nrdDG NrdR box sequences. In contrast, sequence conservation between NrdR boxes belonging to nrdAB, nrdHIEF, and nrdDG promoter regions is less pronounced. The greater conservation in box 2 suggests that it may be the primary site for NrdR binding (all three probes showed similar NrdR binding profiles). Indeed, preliminary studies in which the two nrdAB NrdR box sequences were mutated showed that box 2 is critical for binding (Fig. 4B). Although NrdR was found to bind in vitro to nrdAB and nrdHIEF (and nrdDG) promoter probes (Fig. 4A), differences in the positions and extents of binding could have a significant influence on promoter activity. Possibly, differences in both the positions and sequences of NrdR boxes may account, at least in part, for differential effects on transcription of nrdAB and nrdHIEF. Further studies will be required to determine the correctness of these views. Nevertheless, it is of interest that deletions in the region immediately downstream of the nrdAB transcription start site were reported to cause a 10-fold increase in expression (47). We note that the deletion removes the downstream NrdR box 2, supporting the view that NrdR controls transcription via its interaction with NrdR box motifs.
The effect of NrdR on transcription of the class III anaerobic RNR genes in aerobic growth is unexpected. In the wild type, the level of nrdD transcription did not change appreciably during the growth cycle. In the nrdR mutants, nrdD transcription increased by four- to eightfold throughout growth. Expression of the nrdDG genes occurs under oxygen-limiting conditions, during anaerobiosis, and, to a lesser extent, in the deceleration phase of aerobically growing cultures and is completely dependent on the global transcription regulator FNR, but not on the ArcA-ArcB two-component system (8). The results presented here suggest that induction of nrdDG expression is, in part, controlled by NrdR. Sun et al. (44) identified an FNR binding sequence in the upstream region of nrdD, and Boston and Atlung (8) noted a second site; one site is centered at position –65 relative to the transcription start site, and the second overlaps the –35 promoter sequence. The latter FNR site also overlaps with the upstream NrdR box 1 sequence. Hence, the E. coli class III nrdDG operon appears to be under positive regulation by FNR and negative regulation by NrdR.
Numerous environmental factors are known to influence expression of the E. coli class Ia and class Ib RNR aerobic genes, including oxygen tension, inhibition of DNA synthesis, DNA-damaging agents, nutritional status, and oxidative- and thiol-redox stress agents (21). The coordination of DNA replication with cell growth involves multiple levels of regulation (reviewed in reference 21). Gon and coworkers have recently proposed a model in which the nucleotide bound state of E. coli DnaA regulates transcription of the nrdAB genes (16). In addition to its role in initiating DNA replication, DnaA is a transcriptional regulator that controls positively or negatively the expression of nrdAB by binding to DnaA boxes immediately upstream of the nrdAB operon (32, 42). In their model, ATP-DnaA acts as a repressor of nrdAB transcription; after initiation of DNA replication, ATP DnaA is converted to ADP DnaA, resulting in increased nrdAB expression and thereby coordinating DNA replication and RNR synthesis during the cell cycle. Class Ib nrdEF genes would not be subject to this regulatory mode, since the promoter region lacks a DnaA binding site. Gon and coworkers have proposed an additional control mechanism to regulate NrdAB activity in which an excess of dNTPs leads to binding of dNTP by DnaA and repression of nrdAB expression (16). This is consistent with findings that DnaA binds dATP (40). Thus, high dNTP levels may negatively control expression of the class Ia nrdAB genes by transcriptional repression and, because NrdA contains a regulatory site that binds dATP, negatively control enzyme activity by allosteric regulation. Class Ib RNRs lack this allosteric regulatory site. Hence, class Ia RNRs would provide protection against possible harmful effects of overproduction of dNTPs by negative feedback, while class Ib RNRs may serve to enable synthesis of elevated levels of dNTPs under conditions where higher levels are needed, such as in DNA repair.
In this communication, we have described a new mechanism for controlling expression of the E. coli nrdAB and nrdHIEF genes during aerobic growth via binding of the transcriptional repressor NrdR to specific motifs in the promoter regions. The physiological role of the nrdHIEF operon in E. coli is not clear, since under aerobic conditions, NrdAB is the major source of the class I RNR and NrdEF is made in insufficient amounts to be able to replace it for growth. Nevertheless, it was reported that an extra chromosomal copy of nrdHIEF (or low-copy-number nrdHIEF plasmids) suffice to enable growth of an nrdA mutant, with the extra copy dramatically increasing the nrdHIEF level (16, 23). Recent observations suggest a possible new function for class Ib RNR under iron limitation. A Fur box was described in the region immediately upstream of nrdHIEF (Fig. 3) (48). Global transcription analysis of an E. coli fur mutant showed that nrdHIEF was partially derepressed (31). These observations imply the existence of a finely tuned regulatory switch that controls nrdHIEF expression. This is also evident in the marked changes in nrdHIEF expression found in the wild type during growth and other reports of the dependence of nrdHIEF expression on the growth phase and growth medium (33). The studies reported here suggest that the cellular level of NrdR is ordinarily sufficient to repress nrdHIEF, but not nrdAB, expression but is insufficient to repress additional nrdHIEF copies. This view is consistent with the finding that abolishing NrdR function caused drastic increases in nrdHIEF expression. These results suggest that induction of the E. coli nrdHIEF operon is likely to be physiologically important under conditions in which nrdAB expression is inactive or severely limited and possibly under conditions, such as in infection, when iron availability may be important.
NrdR, like DnaA, binds ATP and dATP and controls the transcription of RNR genes by binding to specific target sites in the promoter regions. Elsewhere, it was shown in S. coelicolor that NrdR binds to these motifs via the N-terminal zinc finger domain and that efficient NrdR binding is dependent on the presence of an intact ATP cone domain that binds ATP/dATP (17). Plausibly, the intracellular level of ATP/dATP modulates NrdR binding and enables differential regulation of nrdAB and nrdHIEF transcription. As noted above, dATP also binds DnaA, and the nrdAB operon contains two tandem DnaA boxes proximal to NrdR box 1. Thus, NrdR and DnaA may both function, in concert or otherwise, as dNTP-dependent repressors of RNR expression. Finally, we note that because ATP/dATP allosterically regulate the class Ia RNR, but not the class Ib RNR, enzyme activity, ATP/dATP may play a dual role in controlling transcription, as well as enzyme activity.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Published ahead of print on 11 May 2007.
Supplemental material for this article may be found at http://jb.asm.org/.
Present address: Institute for Bioengineering of Catalonia (IBEC) and Microbiology Department Edifici Anexe—Laboratory 4, Biology Faculty, Avinguda Diagonal 645, 08028 Barcelona, Spain.
E.T. and I.G. contributed equally to this study.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
