HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 3, 1749-1754, January 18, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/3/1749    most recent M106192200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Scott, C. Articles by Green, J. Search for Related Content PubMed PubMed Citation Articles by Scott, C. Articles by Green, J. Social Bookmarking                      What's this?

Miscoordination of the Iron-Sulfur Clusters of the Anaerobic Transcription Factor, FNR, Allows Simple Repression but Not Activation^*

Colin Scott and Jeffrey Green

From the Krebs Institute for Biomolecular Research, Department of Molecular Biology and Biotechnology, University of Sheffield, Western Bank, Sheffield S10 2TN, United Kingdom

Received for publication, July 3, 2001, and in revised form, October 9, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The FNR protein of Escherichia coli regulates target genes in response to anaerobiosis. Environmental oxygen is sensed by the acquisition of oxygen-labile [4Fe-4S] clusters that promote dimerization, DNA binding, and productive interactions with RNA polymerase. Three N-terminal cysteine residues (Cys²⁰, Cys²³, and Cys²⁹) and Cys¹²² act as ligands for the FNR iron sulfur clusters. An FNR variant, FNR-C20S, that retains only trace activity in vivo can acquire [4Fe-4S] clusters in vitro that enhance site-specific DNA binding. Second site substitutions in activating regions AR1, AR2, and AR3 restore in vivo activity to FNR-C20S, suggesting that the impairment in FNR-C20S activity is due to a failure to communicate with RNA polymerase effectively. Here we show that FNR-C20S can repress a simple FNR-regulated promoter in vivo and that it can form productive heterodimers with an FNR variant with altered DNA binding specificity, FNR-E209V. Transcription studies with FNR-E209V·FNR-C20S heterodimers indicate that the presence of a miscoordinated iron-sulfur cluster (FNR-C20S) in the downstream (but not the upstream) subunit of the FNR dimer impairs activation from a class II promoter and that this impairment can be overcome by amino acid substitutions known to unmask AR2 or improve AR3 in the affected subunit.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The transition from aerobic to anaerobic respiration in the facultative anaerobe Escherichia coli is, in part, governed by the global transcription factor FNR (the regulator of fumarate and nitrate reduction) (1, 2). FNR is a one-component sensor/regulator protein that senses changes in cytoplasmic oxygen concentration directly via an oxygen-labile [4Fe-4S] cluster. Each FNR monomer can acquire a single [4Fe-4S] cluster, which is coordinated by three N-terminal cysteine residues (Cys²⁰, Cys²³, and Cys²⁹, but not Cys¹⁶) and a fourth cysteine residue (Cys¹²²) (3, 4). Upon incorporation of [4Fe-4S] clusters, the FNR monomers dimerize; the resultant dimeric FNR can bind DNA and contact RNA polymerase (RNAP)¹ holoenzyme to produce a productive complex at FNR-dependent promoters (5-8). Activation of FNR-dependent promoters by [4Fe-4S] cluster-containing, dimeric FNR requires appropriate contacts between FNR and RNAP. Two surface-exposed activating regions (AR1 and AR3) that are involved in FNR-RNAP communication have been characterized (7, 9, 10). At class I promoters (where the FNR site is centered at 61.5 bp or further from the transcription initiation site), AR1 of the downstream subunit of the FNR dimer contacts the C-terminal domain of the -subunit of RNAP (10, 11). At class II promoters (where the FNR site is centered at or about 41.5 bp from the transcription initiation site), AR3 makes the dominant contact with RNAP via the ⁷⁰-subunit, and now AR1 of the upstream FNR subunit contacts the C-terminal domain of the -subunit (7). The cAMP receptor protein (CRP) of E. coli, a structural homologue of FNR, also uses AR1 to activate transcription at class I promoters but has a different strategy at class II promoters, where it uses a different activating region (AR2) to contact the N-terminal domain of the -subunit of RNAP instead of the ⁷⁰-AR3 contact formed by FNR (12). Although apparently not normally active in FNR, AR2 activity can be generated by screening for second site mutations that compensate for an impaired AR1 (11).

Aerobic inactivation of FNR occurs when the iron-sulfur cluster reacts with molecular oxygen. The [4Fe-4S] cluster is disassembled in response to oxygen, ultimately yielding inactive, monomeric apo-FNR (13, 14).

Substitution of any of the four cysteine ligands of the iron-sulfur cluster has profound effects upon the in vivo activity of FNR. Altering Cys¹²² abolishes the ability of FNR to activate transcription from model FNR-dependent promoters in vivo (3, 4). On the other hand, replacing Cys²⁰, Cys²³, or Cys²⁹ generates FNR proteins with impaired but detectable in vivo activity (3, 4). Recent studies have focused on determining the cause of the residual activity of the FNR proteins with N-terminal cysteine substitutions, particularly the FNR-C20S variant. Anaerobically purified FNR-C20S, in contrast to wild-type FNR, did not contain [4Fe-4S] clusters, suggesting that FNR-C20S cannot incorporate a [4Fe-4S] cluster (2). However, FNR-C20S protein purified under aerobic conditions has been shown to have a similar iron content to wild-type FNR prepared in the same way, suggesting that both proteins contain degraded iron-sulfur clusters (15). Additionally, in vitro analysis has revealed that aerobically purified FNR-C20S can bind DNA in footprinting reactions (15), albeit with lower affinity than aerobically purified, unaltered FNR. Isolated FNR-C20S can also acquire a [4Fe-4S] cluster in vitro with properties similar to that wild-type FNR, and the formation of the cluster drives FNR-C20S dimerization as judged by enhanced site-specific DNA binding (16). Moreover, second site substitutions in the activating regions of FNR-C20S (AR1, AR2, or AR3) can correct for the C20S substitution, allowing in vivo activation of expression from FNR-dependent promoters. These observations suggest that FNR-C20S can assemble oxygen-labile [4Fe-4S] clusters in vivo and that the presence of the miscoordinated iron-sulfur clusters promotes dimerization and DNA binding but not the conformational changes required for effective contact with RNAP and transcription regulation (16). Here evidence is presented to show that FNR-C20S can acquire an oxygen-sensitive cluster in vivo and that this is sufficient to mediate anaerobic repression of a simple FNR-regulated promoter. The FNR-C20S variant also formed heterodimers with FNR-E209V, further indicating that the miscoordinated iron-sulfur cluster of FNR-C20S can promote dimerization. These heterodimers activated transcription from an FNR-dependent class II promoter, but only when FNR-C20S was the upstream subunit of the FNR dimer, indicating that the impaired activity of FNR-C20S in vivo is probably due to a failure to make productive interactions with RNAP. This was supported by the finding that amino acid substitutions in AR3 or those that unmask AR2 suppress the effects of the orientation of the FNR-C20S·FNR-E209V heterodimers at class II promoters. Finally, studies with FNR29, which lacks all the N-terminal cysteine residues and thus cannot form iron-sulfur clusters, indicated that two iron-sulfur clusters, one per subunit are required for FNR dimerization. Therefore, it is concluded that the model describing the aerobic-anaerobic FNR switch can be refined such that iron-sulfur cluster assembly by two FNR monomers permits dimerization and enhanced DNA binding but that only correctly liganded iron-sulfur clusters can cause the necessary conformational changes to establish effective contacts with RNAP and subsequent transcription activation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Strain Construction and Growth-- The E. coli strains used were based upon JRG1728 (fnr lac) (17). In the in vivo repression experiments, JRG1728 was transformed with a pRW50-based reporter plasmid carrying the lac operon under the control of the FFgal4 promoter (18) and pBR322 derivatives encoding either FNR (pGS196), FNR29 (pGS198), or FNR-C20S (pGS197) (3, 17). For the oriented heterodimer study, JRG1728 was transformed with pRW50-based plasmids containing the lac operon under the control of the FFmelR, YYmelR, FYmelR, or YFmelR class II (regulator site centered at 41.5) semi-synthetic promoters (19). The resultant reporter strains were then transformed with a kanamycin-resistant pLG339 derivative encoding the DNA recognition variant FNR-E209V (19). The strains produced were then transformed with pBR322 derivatives encoding FNR (pGS196), FNR29 (pGS198), FNR-C20S (pGS197), FNR-C20S/E47K (pGS1341a), or FNR-C20S/K60R (pGS1543), (3, 16, 17).

Cultures were grown in Lennox broth (20) with the appropriate antibiotics added. For the -galactosidase assays, anaerobic cultures were grown in Lennox broth supplemented with 0.5% (w/v) glucose, using anaerobic jars and anaerobic gas generating kits (Oxoid). Aerobic cultures were grown on the same medium (5 ml) in 250-ml conical flasks shaken at 250 rpm. The media were supplemented with tetracycline (35 µg ml¹), kanamycin (25 µg ml¹), and ampicillin (100 µg ml¹) as appropriate.

-Galactosidase Assay-- -Galactosidase assays were in accordance with the method of Miller (21), and the strains used were grown from an inoculum of 1:200 until an A₆₀₀ of 0.3-0.6 was reached before -galactosidase activity was estimated.

Quantitative Reverse Transcriptase-PCR-- Total RNA was prepared using a Qiagen RNeasy mini kit in accordance with the manufacturer's instructions. Total RNA was extracted from cultures of E. coli strain JRG1728 (fnr lac) containing pGS196 (encoding wild-type FNR) and the FFmelR lac reporter plasmid grown either aerobically or anaerobically to an A₆₀₀ of 0.3-0.4. The concentration of total RNA isolated was determined using a UNICAM UV4 UV-visible spectrophotometer. First strand cDNA synthesis was performed using a Qiagen Omniscript Reverse Transcriptase kit in accordance with the manufacturer's instructions using 100 nmol of total RNA/µl of reaction volume. The cDNA was then used as the template for PCR using appropriate primers for lacZ and tatE. A PCR master mix was prepared for each condition and then aliquoted into 20 µl of reaction volumes. The products were labeled by including [³²P]dGTP in the reaction mix. One reaction aliquot was removed from the thermocycler (Hybaid Omn-E) after five cycles (melting, 95 °C, 30 s; annealing, 62 °C, 45 s; extension, 72 °C, 60 s) and after every subsequent cycle (total cycles, 15). The samples were resolved on Tris-borate EDTA buffered 20% polyacrylamide gels and detected by autoradiography. The relative quantities of each product were determined by scanning (Sharp JX-330) the autoradiograph and subsequent quantitative analysis using ImageMaster software (Amersham Biosciences, Inc.).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

FNR-C20S Can Repress Gene Expression in Vivo-- FNR-C20S has been shown to incorporate an iron-sulfur cluster in vitro that enhances binding of FNR-C20S dimers to DNA containing an FNR site. However, in vivo, FNR-C20S has a much reduced ability to activate expression from FNR-dependent promoters (FFmelR, dms, and frdA), nor can FNR-C20S repress transcription from the FNR-repressed cyoA and ndh promoters as effectively as the wild-type protein (3, 4). This could suggest that the [4Fe-4S]-containing form of FNR-C20S is not generated in vivo, resulting in diminished activity. However, activation by FNR requires interaction between FNR and RNAP, and so defects in in vivo FNR-C20S-dependent transcription activation may not be indicative of the absence of DNA binding by FNR-C20S. Moreover, FNR-dependent repression at many natural promoters has proved to be a complex process (1), so again the action of FNR proteins at these promoters may not be wholly representative of their ability to bind DNA per se. Fortunately, a simple, synthetic, FNR-repressed lacZ reporter is available (the FFgal4 promoter) in which transcriptional activity is blocked by the binding of active FNR close to the 10 element (18).

The activity of the FFgal4 promoter in a strain containing FNR was repressed in an oxygen-dependent manner (Table I), confirming that this promoter is repressed in the presence of anaerobic, active FNR. As would be expected, the FNR variant, FNR29, that is incapable of incorporating an iron-sulfur cluster because it lacks all four N-terminal cysteines, could not repress transcription from the FFgal4 promoter. However, when FNR-C20S was present a 5.7-fold anaerobic repression of FFgal4, compared with 13-fold with wild-type FNR, was observed. Therefore, it seems likely that FNR-C20S can bind the FNR site of the FFgal4 promoter in vivo in an oxygen-responsive manner and repress transcriptional activity by promoter occlusion. Because FNR29 cannot repress transcription from this simple promoter, it may be concluded that incorporation of iron-sulfur clusters is required for this activity and thus that FNR-C20S can acquire iron-sulfur clusters in vivo.

View this table: [in this window] [in a new window]   Table I Transcriptional repression of FFgal4 by FNR and FNR-C20S Each strain tested contained a pRW50 derivative in which the expression of the lac operon was driven from the FFgal4 promoter. Each strain also carried a pBR322 derivative encoding wild-type FNR, FNR-C20S, or FNR29. The mean -galactosidase activity of each strain was determined using the method of Miller (21) from at least three independent cultures. The standard deviations are shown in parentheses.

FNR·FNR-C20S Heterodimers Can Activate Gene Expression in Vivo but in an Orientation-dependent Manner-- Active, dimeric FNR binds DNA in a site-specific manner, targeting a 14-bp imperfect palindrome (TTGATNNNNATCAA). In the simplest case of FNR-dependent transcription activation, FNR binds at this site centered 41.5 base pairs upstream of the transcript start. However, although many natural FNR-dependent promoters have such an architecture, their activities are also influenced by other transcription factors responding to different environmental stimuli (such as the availability of alternative electron acceptors). The semi-synthetic FFmelR promoter (with an FNR consensus site TTGATCTAGAATCAA centered at 41.5) was created to provide a simple model FNR-regulated promoter and has subsequently been modified to facilitate the study of FNR-dependent transcription activation (19). Mutagenesis of the DNA-binding helix of FNR at position 209 (E209V) changes the binding specificity of the regulator from TTGAT (F) to TTAAT (Y) (19). This altered FNR-binding site has been engineered into the simple class II FNR-dependent FFmelR promoter (described above), creating the FNR-E209V-activated YYmelR promoter by replacing the consensus FNR (FF) site with the YY site (TTAATCTAGAATTAA). The wild-type half-site (F) can also be fused with the variant FNR half-site (Y), creating a promoter with either FY or YF regulatory sites. These sites confer specificity for FNR·FNR-E209V hybrid dimers, and the orientation of the dimer relative to the transcriptional machinery is determined by the orientation of the DNA-binding half-sites in the hybrid promoter (Fig. 1). Thus, this series of lacZ reporters possess variations on the FNR consensus within the context of the same promoter sequence. This system can be adapted to probe the effect of introducing one FNR-C20S iron-sulfur cluster and one wild-type iron-sulfur cluster into an FNR dimer. By co-expressing FNR-E209V, which recognizes the Y half-site and has a normal iron-sulfur cluster, with FNR-C20S, which recognizes the F half-site, the impact of the miscoordinated FNR-C20S cluster can be assessed at these simple model FNR-activated promoters. The plasmids expressing the FNR proteins are selected such that less FNR-E209V is present in the bacteria than FNR-C20S. Thus, most FNR-E209V should be in the form of FNR-E209V·FNR-C20S heterodimers. Therefore, a strain of E. coli lacking chromosomally encoded FNR (JRG1728) was transformed a pLG339 derivative encoding FNR-E209V and with the FFmelR, FYmelR, YFmelR, or YYmelR reporter plasmids. The four reporter strains were transformed with a pBR322 derivative encoding either FNR, FNR-C20S, or FNR29. Using this system the presence of FNR-C20S·FNR-E209V or FNR-C20S·FNR29 heterodimers would be exposed either by activation from one or both hybrid promoters (FY and YF) or by a reduction in FNR-E209V-mediated activation of the YYmelR promoter (Table II).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Schematic representation of FNR homodimers, FNR-E209V homodimers, and FNR·FNR-E209V heterodimers interacting at their cognate promoters. The FNR [4Fe-4S] subunits are represented as ovals containing squares (indicating a [4Fe-4S] cluster) with either a white base, indicating a wild-type DNA recognition helix, or a black base, indicating an FNR-E209V DNA recognition helix. The appropriate homodimers and heterodimers are shown interacting with the FFmelR (a), YFmelR (b), FYmelR (c), or YYmelR promoters (d), in which the wild-type (F) FNR half-sites (TTGAT) are shown as white blocks and the variant (Y) FNR half-sites (TTAAT) are shown as black blocks. The orientation of the FNR-binding sites are indicated relative to the transcription initiation site (shown as a black arrow).

View this table: [in this window] [in a new window]   Table II Transcription activation by oriented FNR heterodimers Each strain tested expressed FNR-E209V and a pRW50 derivative in which the expression of the lac operon was driven from the FFmelR, YYmelR, FYmelR, or YFmelR promoters. Each strain also carried a pBR322 derivative encoding wild-type FNR, FNR-C20S, or FNR29. The mean -galactosidase activity of each strain was determined using the method of Miller (21) from at least three independent cultures. The standard deviations are shown in parentheses.

Analysis of the strains containing the FFmelR reporter plasmid showed that a significant level of activation was only observed in the presence of wild-type FNR, suggesting that neither FNR-C20S or FNR29 homodimers nor FNR-E209V·FNR-C20S or FNR-E209V·FNR29 heterodimers can activate expression from this promoter, whereas wild-type FNR homodimers can.

This type of analysis, in which -galactosidase activity is used to report promoter activity, is a frequently used, simple, and convenient approach. However, a number of other factors not directly related to transcription (such as mRNA turnover, protein assembly, and protein turnover) could influence the measured output. Therefore, it was deemed prudent to obtain an independent measure of transcription to ensure that the effects observed in the lacZ reporter experiments are due to changes in transcription. Such a correlation between -galactosidase activity and lacZ transcript was sought by analyzing the amount of transcript produced from the FNR-dependent FFmelR-reporter plasmid in JRG1728 (fnr lac) expressing fnr from pGS196 under aerobic and anaerobic conditions using quantitative reverse transcriptase-PCR (see "Experimental Procedures"). The amount of FNR-dependent lacZ mRNA produced was compared with the amount of mRNA produced from the chromosomal tatE gene, which has been shown to be transcribed at the same rate in the presence or absence of oxygen (22). The finding that the ratio of tatE transcript in aerobic compared with anaerobic cultures was in the range 1.2-1.3 confirmed the previous reports (22) and indicated that this transcript is an appropriate internal control for measuring changes in the lacZ transcript. During aerobic growth the ratio of lacZ to tatE mRNA was 0.3-0.4, whereas anaerobic growth the ratio was 2.2-2.5. This represents up to 8.3-fold induction in FNR-dependent anaerobic induction of this transcript. Measurement of -galactosidase activity under the same conditions revealed a 12-fold increase when anaerobic cultures were compared with aerobic cultures. Thus, these data suggest that -galactosidase activity is representative of transcriptional activity from the pRW50-based FNR-dependent lacZ fusion reporters used here.

The YYmelR reporter was most active when FNR-E209V was combined with FNR29 and displayed reduced activity when either wild-type FNR or FNR-C20S was present. Because FNR-E209V-driven expression from YYmelR was high in the presence of FNR29, it is suggested that iron-sulfur cluster incorporation into both FNR subunits is essential for dimerization. The reduced level of FNR-E209V-driven expression from YYmelR in the presence of FNR and FNR-C20S indicates that both wild-type FNR and FNR-C20S can dimerize with FNR-E209V but that these heterodimers fail to recognize the YYmelR promoter and therefore fail to activate transcription. Furthermore, the formation of heterodimers reduces the pool of available FNR-E209V homodimer, thereby reducing expression from the YYmelR promoter relative to that observed when FNR29 is co-expressed with FNR-E209V. The extent of the reduction in activity from the YYmelR promoter was not as great for FNR-C20S-expressing strains compared with that observed for those containing wild-type FNR (a 6.7-fold difference between wild-type FNR and FNR29 compared with a 2.3-fold difference between FNR-C20S and FNR29). This suggests that although FNR-C20S is able to form dimers, its ability to do so is impaired relative to the wild-type FNR or that FNR-C20S dimers are efficiently formed but do not bind DNA as well as the unaltered FNR protein.

At promoters that require the action of heterodimers (FY and YF) the presence of FNR29 fails to lead to activation from the hybrid-binding sites, again indicating that the presence of a cluster in each FNR monomer may be required for dimerization. The presence of the wild-type FNR allowed transcription to occur, demonstrating that the wild-type FNR and FNR-E209V can form an active dimer. The ability of FNR-E209V·FNR-C20S heterodimers to mediate transcription activation at the FYmelR and YFmelR hybrid promoters was found to be dependent upon the orientation of the two subunits of the hybrid dimer. Activation of transcription from the FYmelR promoter was found to be similar for heterodimers of FNR-E209V with both FNR or FNR-C20S, but at the YFmelR promoter the observed level of transcription for FNR-E209V heterodimers with FNR-C20S was closer to that found for FNR-E209V·FNR29 heterodimers. These observations are in accord with the hypothesis that the iron-sulfur clusters acquired by FNR-C20S promote dimerization and DNA binding but fail to align the activating regions of FNR with their cognate sites in RNAP (Fig. 2). Thus, at the FYmelR promoter the correctly configured AR3 of the FNR-E209V subunit is juxtaposed to the ⁷⁰ subunit of RNAP allowing the strong activating contact to be made. It has been shown previously that class I activation by the FNR-C20S homodimer is less affected by the C20S induced iron-cluster defect than class II activation, and thus AR1 is probably less affected than AR3 (16). Therefore, it is the slightly impaired AR1 of the FNR-C20S monomer that is adjacent to the C-terminal domain of the RNAP  subunit at the FYmelR promoter, allowing a partial contact to be made. Thus, the FYmelR promoter can be activated by the FNR-E209V·FNR-C20S heterodimer. Conversely, at the YFmelR promoter, the FNR-C20S subunit would be required to make the essential AR3 contact with RNAP but cannot, presumably because its AR3 is incorrectly configured as a consequence of the miscoordinated iron-sulfur cluster. Thus, activation from the YFmelR promoter cannot be mediated by the FNR-C20S·FNR-E209V heterodimer.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Model of the orientation-dependent interaction between the FNR-C20S·FNR-E209V heterodimers and RNAP at the YFmelR and FYmelR promoters. The wild-type [4Fe-4S] clusters (squares) of the FNR-E209V subunits recognize the Y half-site (black base). The miscoordinated [4Fe-4S] clusters (parallelograms) of the FNR-C20S subunits recognize the F half-site (white base). The heterodimers are shown interacting at the YFmelR (a) or FYmelR promoters (b). The orientation of the FNR-binding sites is indicated relative to the transcription initiation site (+1), and the relative level of transcription is indicated by the width of the arrow. The C-terminal domain of the  subunit () and the 70 subunit () of RNAP are indicated, (the ' subunits and the  subunit N-terminal domain are omitted for clarity). The activating regions AR1 (diamond) and AR3 (asterisk) that contact RNAP are shown as a functional contact in white and an impaired contact in black.

Transcriptional Activation by the FNR-C20S Variant Can Be Restored by Second Site Mutations in the RNA Polymerase Contacts AR2 and AR3-- It has been shown that the inability of the FNR-C20S homodimer to activate transcription can be compensated for by the introduction of second site mutations that augment AR1 and AR3 or by unmasking the normally silent AR2 (16). However, the possibility that these second site substitutions affect a signal transduction step other than communication with RNAP (such as improving DNA binding affinity) has not been excluded. Thus, the orientation dependence of FNR-E209V·FNR-C20S heterodimers, shown in Fig. 2, should be abolished when FNR-C20S derivatives carrying compensatory mutations in AR2 or AR3 (e.g. FNR-C20S/E47K and FNR-C20S/K60R, respectively) are used in the place of FNR-C20S.

The data (Table III) show that both FNR-C20S/E47K and FNR-C20S/K60R are more active than FNR-C20S in the strain containing the FFmelR reporter, which confirms that the second-site mutations do restore some activity to FNR-C20S. Because both FNR-C20S/E47K and FNR-C20S/K60R could interfere with transcriptional activation by the FNR-E209V homodimer at the YYmelR promoter, it seems likely that FNR-C20S/E47K and FNR-C20S/K60R can dimerize efficiently with FNR-E209V. At the FYmelR and YFmelR promoters FNR-C20S/E47K·FNR-E209V and FNR-C20S/K60R·FNR-E209V heterodimers behave like the FNR·FNR-E209V heterodimer, activating transcription from both promoters. This demonstrates that by adjusting RNAP contacts through the modified AR2 or AR3, communication with RNAP was re-established, and productive complexes could then be formed allowing transcription from the YFmelR promoter.

View this table: [in this window] [in a new window]   Table III Transcription activation by oriented FNR heterodimers carrying amino acid substitutions in AR2 or AR3 Each strain tested expressed FNR-E209V and harbored a pRW50 derivative in which the expression of the lac operon was driven from the FFmelR, YYmelR, FYmelR, or YFmelR semi-synthetic promoters. Each strain also carried a pBR322 derivative encoding either wild-type FNR, FNR-C20S/E47K, or FNR-C20S/K60R. The mean -galactosidase activity of each strain was determined using the method of Miller (21) from at least three independent cultures. The standard deviations are shown in parentheses.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Signal Perception by FNR and FNR-C20S-- The purpose of this work was to investigate the effects of a miscoordinated iron-sulfur cluster on the ability of FNR to regulate transcription. It had been previously shown that although FNR-C20S had only trace activity in vivo, it was able to acquire an oxygen-labile iron-sulfur cluster that enhanced DNA binding in vitro (3, 4, 16). The finding that amino acid substitutions in the activating regions of FNR-C20S could partially rescue activity in vivo suggested that the inactivity of FNR-C20S in vivo is due to a lesion in signal transduction downstream of cluster acquisition and DNA binding (16). The evidence presented here supports this contention by demonstrating that FNR-C20S can regulate the activity of a simple FNR-repressed promoter in an oxygen-dependent manner in vivo. The simplest explanation for this observation is that FNR-C20S can dimerize and bind DNA in response to oxygen-starvation through the assembly and disassembly of a [4Fe-4S] cluster. Conversely, FNR29 cannot regulate the same simple FNR-repressed promoter, presumably because it cannot form an iron-sulfur cluster. This is consistent with the in vitro data suggesting that the acquisition of a [4Fe-4S] cluster by FNR-C20S is sufficient to cause dimerization and DNA binding (16).

Like wild-type FNR, FNR-C20S was found to form functional heterodimers with FNR-E209V, again indicating that FNR-C20S is able to dimerize in vivo. Both wild-type FNR and FNR-C20S can interfere with FNR-E209V-dependent activation from the YYmelR promoter, but FNR-C20S was less able to so than wild-type FNR. It is possible that the possession of a miscoordinated [4Fe-4S] cluster slightly impairs the dimerization of FNR-C20S. Because FNR29 cannot form active heterodimers with FNR-E209V nor interfere with FNR-E209V-dependent regulation, it seems likely that FNR29 cannot form heterodimers. This suggests that to form a dimer both, and not just one, of the FNR subunits must contain an iron-sulfur cluster, although one, or both, of these may be miscoordinated as in the case of FNR-C20S. The conclusion that both subunits of an FNR dimer must contain a cluster to permit dimerization and transcriptional activity is in accord with previous studies (23). In contrast the structurally homologous CRP of E. coli has been demonstrated to be most active when one of the available allosteric sites of the CRP dimer remains unoccupied (a CRP₂·cAMP₁ complex) rather than a CRP₂·cAMP₂ complex in which one cAMP molecule is bound per subunit of CRP (24).

Signal Transduction-- Although incorporation of a miscoordinated [4Fe-4S] cluster has only a small effect on FNR dimerization (16), it does dramatically influence the effectiveness of subsequent processes vital to transcription activation, crucially including configuring AR3 such that it can make productive contacts with RNAP. Moreover, although FNR-C20S can regulate the simple FNR-repressed FFgal4 promoter, it is unable to repress transcription from more complex promoters to the same extent as the wild-type FNR (3, 4). This may suggest that repression by FNR requires the correct spatial organization of repression-specific regions of amino acids. Conformational changes required for repression may, as those required for activation, be dependent upon the incorporation of a properly coordinated iron-sulfur cluster. The majority of characterized FNR-repressed promoters contain multiple FNR-binding sites (for example, see Refs. 25-27), and recent evidence suggests that FNR·FNR contacts between tandemly bound FNR dimers, but not FNR·RNAP contacts, may be involved in repression of transcription from these promoters (25, 28). Thus, the correct coordination of the [4Fe-4S] cluster may be required for both FNR·RNAP communication, required in FNR-mediated activation, and FNR·FNR contacts, required for FNR-mediated repression at some promoters.

Because the ligands of the FNR iron-sulfur cluster appear to have such a pivotal role in positioning the activating regions, it is interesting to note that many FNR homologues possess N-terminal cysteine signatures different to that of E. coli FNR (for example FnrP from Paracoccus denitrificans, AadR from Rhodopseudomonas palustris, and the FnrN proteins from Rhizobium leguminosarum; Refs. 29-31). It is tempting to speculate that different arrangements of ligands found in FNR proteins have evolved to allow optimal alignment of their activating regions with their cognate RNAP.

Conclusion-- Taking account of the evidence presented here, the model describing the aerobic-anaerobic FNR switch can be refined such that iron-sulfur cluster assembly into two FNR monomers permits their dimerization and enhances site-specific DNA binding. However, correctly engaged iron-sulfur cluster ligands are not essential for these processes, but a correctly engaged Cys²⁰ ligand is essential to establish productive communication with RNAP (via the correct arrangement of activating regions, and possibly repressing regions). Thus, the FNR iron-sulfur cluster is not only the sensor of oxygen and the promoter of dimerization and site-specific DNA binding, but its configuration is fundamental in transmitting the oxygen starvation signal to RNAP and thus plays an important structural role in FNR.

	ACKNOWLEDGEMENT

We thank S. J. W. Busby for the kind gift of the FFgal4, YYmelR, YFmelR, and FYmelR reporter plasmids and for many useful discussions.

	FOOTNOTES

^* This work was supported by Biotechnology and Biological Sciences Research Council Grant PRS12148.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Molecular Biology and Biotechnology, Krebs Institute for Biomolecular Research, University of Sheffield, Firth Court, Western Bank, Sheffield S10 2TN, UK. Tel.: 44-114-222-4403; Fax: 44-114-272-8697; E-mail: jeff. green@sheffield.ac.uk.

Published, JBC Papers in Press, November 8, 2001, DOI 10.1074/jbc.M106192200

	ABBREVIATIONS

The abbreviations used are: RNAP, RNA polymerase; CRP, cAMP receptor protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				G. P. Roberts, H. Youn, and R. L. Kerby CO-Sensing Mechanisms Microbiol. Mol. Biol. Rev., September 1, 2004; 68(3): 453 - 473. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/3/1749    most recent M106192200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Scott, C. Articles by Green, J. Search for Related Content PubMed PubMed Citation Articles by Scott, C. Articles by Green, J. Social Bookmarking                      What's this?