INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Proteus mirabilis infects the urinary tract of humans and is most commonly responsible for causing disease in individuals with structural abnormalities of the urinary tract or in patients who undergo long-term catheterization (16). Cystitis, acute pyelonephritis, and urinary stone formation are all possible consequences of P. mirabilis infection (17).

P. mirabilis produces a urea-inducible urease, a high-molecular-weight, multimeric, cytoplasmic nickel metalloenzyme. Urease catalyzes the hydrolysis of urea to ammonia and carbon dioxide (18). During the course of infection, the production of ammonia by urea hydrolysis raises the pH in the local environment, subsequently precipitating polyvalent ions that are normally soluble in urine. The result is the formation of urinary stones. The elevated pH also creates an environment that is more favorable for growth of this species (4). Increased ammonia production can also lead to acute inflammation with possible tissue necrosis (18).

The P. mirabilis urease gene cluster is found in single copy on the chromosome and consists of eight contiguous genes, ureRDABCEFG (12, 19, 24). The ureA (UreA, 11 kDa), ureB (UreB, 12 kDa), and ureC (UreC, 61 kDa) genes encode the structural polypeptides required for the assembly of a catalytically inactive urease apoenzyme (18). The accessory genes, ureD (UreD, 31 kDa), ureE (UreE, 18 kDa), ureF (UreF, 23 kDa), and ureG (UreG, 22 kDa), encode proteins required for insertion of nickel ions into the metalloenzyme resulting in catalytically active urease (18). The urease gene cluster is regulated by the gene product of ureR (UreR, 33 kDa).

P. mirabilis UreR and the plasmid-encoded UreR found in Escherichia coli are positive transcriptional activators of the urease genes. The two proteins share 70% amino acid identity (6) and are functionally interchangeable in the activation of transcription from the ureR (p_ureR) and ureD (p_ureD) promoters in both the P. mirabilis and plasmid-encoded urease gene clusters (6). The UreR binding sites of both promoters have the consensus sequence T(A/G)(T/C)(A/T)(T/G)(C/T)T(A/T)(T/A)ATTG (25). Both UreR proteins have been shown to activate transcription from p_ureD in the presence of urea (11, 6). In addition, UreR regulates its own transcription in the presence of urea from p_ureR in the direction opposite the rest of the gene cluster (6). In the absence of urea induction, H-NS represses ureR expression (3). Because UreR activates transcription in a urea-inducible manner, it is hypothesized that UreR binds urea; however, this has not been directly demonstrated.

UreR is a member of the AraC family of transcriptional regulators and contains a putative helix-turn-helix in addition to an AraC signature sequence (5, 19). The AraC signature sequence, found within all AraC family members, is a second helix-turn-helix that is hypothesized to also bind DNA (7). Moreover, UreR also contains three conserved leucine residues (Leu147, Leu148, and Leu158) in the same relative location with the same spatial distance relative to each other as in AraC (Leu150, Leu151, and Leu161). These leucine residues are critical for AraC dimerization (23), and we therefore also hypothesize that UreR dimerizes via this mechanism. In the presence of arabinose, AraC uses these three critical leucines for dimerization via an antiparallel coiled-coil in a "knobs-into-holes" manner, as elucidated by X-ray crystallographic studies (23). This coiled-coil is also the primary dimerization face in the absence of arabinose, shown by both size exclusion chromatography and sedimentation velocity analytical ultracentrifugation of an AraC mutant with mutations in Leu150, Leu151, Asn154, and Leu161 (15). A secondary dimerization face in the  barrel of AraC is evident; however, it does not appear to represent the primary means of dimer interaction (15).

AraC contains two separate and independent domains, each with a distinct function, namely, dimerization and DNA binding; UreR is predicted to have similar domains with similar functions. Previously, chimeric proteins containing the two domains of AraC to characterize each of the domain's functions were synthesized (2). The predicted AraC DNA-binding domain was fused to C/EBP, a known eukaryotic transcriptional activator that dimerizes via a leucine zipper. The C/EBP-AraC fusion was found to bind to p_BAD and activate transcription (2). The hypothesized AraC dimerization domain was fused to the LexA DNA-binding domain. This fusion was predicted to mimic full-length LexA and demonstrated the need for dimerization in order to repress transcription of genes normally turned off by LexA. A chromosomal transcriptional fusion of p_sulA to lacZ was repressed in the presence of the AraC-LexA fusion protein. This strategy was used to identify both domains of AraC (2).

In this study, we constructed fusion proteins to identify putative domains of UreR and assign dimerization and DNA-binding functions to each of the domains as well as identifying key amino acid residues involved in dimerization and urea induction.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and enzymes. All enzymes were purchased from Life Technologies (Rockville, Md.) or New England Biolabs (Beverly, Mass.). All chemicals were purchased from Sigma-Aldrich (St. Louis, Mo.) unless otherwise noted. The DIG (digoxigenin) gel shift kit was obtained from Amersham Pharmacia Biotech (Piscataway, N.J.).

Bacterial strains and growth conditions. Bacterial strains used in this study are listed in Table 1. Bacteria were grown either in Luria-Bertani (LB) broth at 37°C with aeration in a shaking incubator (200 rpm) or on LB plates containing 1.5% agar at 37°C. Plates and media were supplemented with the antibiotic chloramphenicol (10 µg/ml), tetracycline (7.5 µg/ml), ampicillin (100 µg/ml), or kanamycin (50 µg/ml).

PCR amplification of DNA used to make fusion proteins. PCR primers are listed in Table 2. The amplification protocol for PCRs was as follows: denaturation, 94°C, 3 min; annealing, 50 to 55°C, 45 s; elongation, 72°C, 1 min; for 30 cycles.

Cloning of PCR products. PCR products were ligated into pCR-BluntII-TOPO (Invitrogen, Carlsbad, Calif.). Inserts were excised using the appropriate restriction enzymes (either NcoI-BamHI, BamHI-XhoI, NcoI-XhoI, or NcoI-HindIII) and separated by agarose gel electrophoresis. PCR amplification products for the inserts encoding C/EBP_302-350-Myc-His and LexA_1-87-Myc-His were digested with restriction enzymes; Table 2 lists inserts and restriction enzymes used. All constructs use an NcoI site to ligate into the vector at the gene sequence encoding the start codon. A BamHI site is at the junction of the gene sequences encoding the domains in both C/EBP_302-350-UreR_164-293 and UreR_1-182-LexA_1-87. All constructs except the gene sequences encoding UreR-Myc-His, UreR_164-293-Myc-His, C/EBP_302-350-Myc-His, LexA_1-87-Myc-His, L147A-L148-Myc-His, and L147A-L148A-L158A-Myc-His, which use a 3' HindIII site, contain a XhoI site at the 3' end for cloning into the expression vector. Inserts were purified using a Qiaquick gel extraction kit (Qiagen) and ligated into either pBAD/MHA or pSE380 (Invitrogen). Plasmids were introduced into the corresponding laboratory strain by CaCl₂ transformation (21). The LexA_1-87-Myc-His gene sequence was ligated into pBAD/MHA to take advantage of the Myc-His epitopes. After transformation, the pBAD vector containing the gene sequence encoding LexA_1-87-Myc-His was purified using a Qiagen miniprep kit. The plasmid was then used in a PCR with primers to amplify the gene sequence for LexA_1-87-Myc-His. The PCR product was then cut with NcoI and XhoI and ligated into the NcoI and XhoI sites in pSE380.

Construction of p_ureD-lacZ reporter plasmid. A low-copy-number ureD-lacZ fusion reporter plasmid compatible with pBAD-Myc-His was constructed. Primers MOB906 and MOB915 were used to PCR amplify an approximately 4.3 kb DNA fragment from plasmid pR10 ureD-lacZ (11) under the following conditions: 95°C denaturation, 54°C annealing, and 72°C elongation for 30 cycles, using Vent DNA polymerase in the presence of 1 mM MgSO₄. The PCR product was gel purified and ligated to PCR-Blunt (Invitrogen), forming pCC026. pCC026 was digested with EcoRI, and the 4.3-kb DNA fragment encoding ureD-lacZ was ligated to an EcoRI-digested derivative of pACYC184 that had previously been cut with PvuII and religated (thus, it does not encode a functional chloramphenicol acetyltransferase). The resulting recombinant plasmid, pCC042, carries a tetracycline resistance marker and encodes a ureD-lacZ transcriptional fusion that is activated in the presence of UreR and urea.

Sequencing. Both strands of plasmid constructs were sequenced across each junction and throughout the insert. Sequencing was done by the Biopolymer Laboratory at the University of Maryland, Baltimore.

-Galactosidase expression assays. Fresh medium was inoculated with a single colony from LB-agar plates containing the appropriate antibiotics and cultured at 37°C overnight. Overnight cultures were used to inoculate fresh medium. Cultures were monitored until they reached an optical density at 600 nm (OD₆₀₀) of ~0.4 to 0.6, at which time inducer (isopropyl--D-thiogalactopyranoside [IPTG], arabinose, or urea) was added, and cultures were incubated for an additional hour. Cultures were placed on ice, and the OD₆₀₀ was measured. Chloroform (100 µl) and 0.1% sodium dodecyl sulfate (SDS; 50 µl) were added, and cultures were vortexed. The suspension of permeabilized cells (10 µl) was added to 990 µl of Z buffer and 200 µl of o-nitrophenyl--D-galactopyranoside (4 mg/ml in H₂O). Timed reactions were stopped with 500 µl of 1 M Na₂CO₃. OD₄₂₀ and OD₅₅₀ measurements were recorded, and Miller units were calculated (20). All constructs were assayed in three or more independent experiments.

Western blot analysis. Overnight cultures were used to inoculate 4 ml of fresh LB medium containing the appropriate antibiotics and allowed to grow to mid-exponential phase. Cultures were induced with the appropriate inducer and incubated for the indicated time. Cultures were placed on ice. OD₆₀₀ was determined, and all cultures were adjusted to the same reading. Bacteria were harvested from 1 ml of culture by centrifugation and resuspended in equivalent amounts of Laemmli sample buffer. Samples were boiled for 5 min and placed on ice. The sample volume listed for each experiment was loaded onto a 3.75% stacking and either a 12.5 or 15% SDS-polyacrylamide gel by the method of Laemmli (13). SDS running buffer was used for electrophoresis. Gels were transferred onto Immobilon P membranes in a transfer chamber containing transfer buffer. Transfer occurred overnight at ~12 V at 4°C. Membranes were blocked in 5% dry milk in 0.1% TTBS (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% Tween 20) for 1 h. Primary antibodies were added to 0.1% TTBS (anti-LexA, anti-Myc and anti-His, diluted 1:5,000), and membranes were exposed to antibodies for 2 h at room temperature. Membranes were washed in 0.1% TTBS three times for 15 min each. Secondary antibodies (anti-mouse immunoglobulin G coupled to alkaline phosphatase detected anti-Myc and anti-His antibodies; anti-rabbit immunoglobulin G coupled to alkaline phosphatase detected anti-LexA antibodies) were placed in 0.1% TTBS (dilution 1:2,000), and membranes were incubated in the secondary antibodies for 1 h. Membranes were washed three times for 15 min each in 0.1% TTBS and developed with nitroblue tetrazolium-5-bromo-4-chloro-3-indolylphosphate dissolved in H₂O (according to the Western blot protocol as described in reference 1).

Gel shift analysis. Gel shift experiments used either a 60-bp (DNA-binding domain study) or a 64-bp (dimerization study) double-stranded oligonucleotide that was synthesized from smaller overlapping oligonucleotide fragments that were allowed to anneal and extend using Vent polymerase. One microliter of each reaction was run on a gel for quantification purposes. Both double-stranded oligonucleotides (3.85 pmol of each) were labeled by the DIG gel shift (Amersham Pharmacia) protocol. Each double-stranded oligonucleotide was diluted to 30 fmol for the binding study.

Purification of UreR derivatives. E. coli Top10 transformed with pCP016 or pCP088 was grown in Luria broth at 37°C. Expression of UreR-Myc-His₆ was induced with 0.2% arabinose when an OD₆₀₀ of ~0.6 was reached. Following 3 h of induction, cells were collected by centrifugation and lysed by two passages through a French pressure cell (18,000 lb/in²). Single-step purification of UreR-Myc-His₆ was performed by nickel-chelating nitrilotriacetic acid affinity chromatography. A single polypeptide of approximately 33 kDa was eluted from the column with 250 mM imidazole as seen on a Coomassie blue-stained 12% SDS-polyacrylamide gel (data not shown). In a similar experiment, purified protein was electrophoresed, transferred to a nitrocellulose membrane, and reacted with rabbit antiserum specific for the Myc epitope. Western blot anlysis showed that arabinose-induced E. coli Top10 transformed with either pCP016 or pCP088 produced a single species of 33 kDa, consistent with the predicted size for UreR-Myc-His₆ (see Fig. 3). The band was absent from the vector control under identical conditions.

Gel filtration chromatography. Gel filtration chromatography was performed at room temperature on a Sephadex G-75 column (1 by 35 cm) equilibrated with running buffer (50 mM phosphate [pH 7.5], 150 mM NaCl) at a flow rate of 0.5 ml/min. Molecular weight standards bovine serum albumin (68 kDa) and carbonic anhydrase (34.5 kDa) were used to calibrate the column. Purified wild-type and mutant UreR-Myc-His₆ proteins (100 µl of 0.1 mg/ml) were injected onto the column, and 0.5-ml fractions were collected. The protein elution profile was monitored by absorbance at 280 nm. An aliquot (100 µl) of each fraction was transferred to an Immobilon P membrane, and immunoblot analysis performed with anti-Myc antibodies.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Localization of the UreR DNA-binding domain. To localize the UreR DNA-binding domain, a protein chimera was constructed by fusing the gene sequence encoding the C-terminal half of UreR to the gene sequence encoding the leucine zipper dimerization domain of C/EBP (Fig. 1; primers are listed in Table 2). All translational fusion constructs were cloned into the E. coli arabinose-inducible expression vector pBAD/Myc-His A (Invitrogen) (Table 1). A p_ureD-lacZ plasmid reporter was constructed in pACYC184 to assay specific induction of the urease gene cluster, using -galactosidase activity as the readout. Both reporter plasmid and pBAD constructs with and without insert were transformed into E. coli Top10.

View larger version (26K): [in this window] [in a new window]   FIG. 1.   Schematic of fusion constructs. The domain structure of each of the chimeric proteins is shown; the amino acid residue number is labeled at each boundary. A BamHI restriction site (coding for a Gly-Ser amino acid linker (GS)) was inserted between gene sequences encoding each domain. Key leucine residues and helix-turn-helix motifs (H-T-H) are indicated. Sig., signature.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   -Galactosidase assays using pureD-lacZ reporter construct. -Galactosidase activity was measured in E. coli Top10 transformed with the pureD-lacZ reporter construct and the chimeric fusion constructs. Cultures (4 ml) were grown in LB medium to mid-exponential phase and induced with 0.02% arabinose with or without 100 mM urea for 1 h. -Galactosidase activity is expressed in Miller units (18). All constructs were assayed in duplicate in at least three different experiments.

View larger version (49K): [in this window] [in a new window]   FIG. 3.   Western blot analysis of fusion constructs. E. coli Top10 (A, B, C, and F) or E. coli JL1436 (D and E), transformed with vector or clones expressing UreR or chimeric proteins, was cultured in LB medium to mid-exponential phase and induced with 2% arabinose for 4 h (unless otherwise noted). Samples (10 µl) of each suspension were boiled for 5 min in gel sample buffer and electrophoresed on an SDS-polyacrylamide gel. Membranes were incubated with antiserum or monoclonal antibodies and developed with secondary antibodies conjugated to alkaline phosphotase (Sigma). (A) pBAD, UreR, UreR-Myc-His; 12.5% SDS-polyacrylamide gel. Membranes were reacted with mouse anti-His antibody (diluted 1:1,000) (Invitrogen). (B) pBAD, UreR, UreR164-293, or UreR164-293-Myc-His; 15% SDS-polyacrylamide gel. Membranes were reacted with mouse anti-His antibody (diluted 1:1,000) (Invitrogen). (C) pBAD, UreR, C/EBP302-350, C/EBP302-350-Myc-His; 15% SDS-polyacrylamide gel. Membranes were reacted with mouse anti-His antibody (diluted 1:1,000) (Invitrogen). (D) LexA, LexA1-87, LexA1-87-Myc-His; induced with 2 mM IPTG. Samples (2 µl) of each were electrophoresed on a 15% SDS-polyacrylamide gel. Membranes were hybridized with rabbit polyclonal anti-LexA antibody (diluted 1:5,000) (Invitrogen). (E) LexA, LexA1-87, UreR1-182-LexA1-87; induced with 2 mM IPTG. Samples (1.25 µl) of each, including strain only, were electrophoresed on a 12.5% SDS-polyacrylamide gel. Membranes were reacted with rabbit polyclonal anti-LexA antibody (diluted 1:5,000) (Invitrogen). (F) pBAD, UreR, UreR-Myc-His, UreR L147A-L148A, UreR L147A-L148A-Myc-His, UreR L147A-L148A-L158A, and UreR L147A-L148A-L158A-Myc-His; induced with 2% arabinose for 4 h. Samples (10 µl) of each were electrophoresed on a 12.5% SDS-polyacrylamide gel. Membranes were hybridized with mouse anti-Myc antibody (diluted 1:5,000) (Invitrogen).

DNA binding by UreR constructs assessed by gel shift. Gel shift experiments were performed to verify that C/EBP_302-350-UreR_164-293 used in the -galactosidase assays could bind to p_ureD. A 60-bp double-stranded oligonucleotide that contained the DNA sequence for p_ureD (66 to 6 upstream of the transcriptional start of ureD) was synthesized and labeled with DIG-11-ddUTP by terminal transferase.

View larger version (66K): [in this window] [in a new window]   FIG. 4.   Gel mobility shift assay for interaction of UreR with the P. mirabilis pureD. A 60-bp double-stranded oligonucleotide was synthesized based on the sequence of pureD. Whole-cell extracts of E. coli Top10 transformed with either pBAD or vector expressing C/EBP302-350-UreR164-293 were obtained by inducing a 4-ml culture in mid-exponential phase with 2% arabinose for 4 h. Whole-cell extracts of E. coli Top10 expressing UreR and UreR-Myc-His were obtained by inducing a 4-ml culture in mid-exponential phase with 2% arabinose and 100 mM urea for 4 h. Bacterial suspensions were adjusted to the same OD, and 3 ml of each culture was harvested by centrifugation (10,000 × g, 5 min, 4°C). Bacteria were resuspended in 1 ml of TEN buffer and lysed in a French press (18,000 1b/in2). Binding reactions with the 60-bp pureD dsDNA fragment were carried out in the presence of 100 mM urea with various amounts of the extracts.

Localization of the UreR dimerization domain. Protein chimeras were constructed to localize the UreR dimerization domain. The gene sequence encoding the N-terminal half (amino acids 1 to 182) of UreR was fused to the gene sequence encoding the LexA DNA-binding domain (amino acids 1 to 87), forming the UreR_1-182-LexA_1-87 fusion protein (Fig. 1; primers are listed in Table 2). All fusion constructs used in these experiments were cloned into pSE380 (Invitrogen) (Table 1) under the control of an IPTG-inducible promoter. The reporter strain, JL1436 (Table 1), originally described by Bustos and Schleif (2), consisted of a chromosomal transcriptional fusion of p_sulA with lacZ placed downstream; the sulA gene is repressed by dimerized LexA (2).

View larger version (16K): [in this window] [in a new window]   FIG. 5.   -Galactosidase reporter activity of the psulA-lacZ chromosomal fusion reporter construct in strain JL1436. E. coli JL1436 (carries a single copy of a psulA-lacZ chromosomal fusion) transformed with the chimeric fusion constructs was cultured in LB medium (4 ml) to mid-exponential phase and induced with 2 mM IPTG with or without 100 mM urea for 1 h. -Galactosidase levels measured in Miller units (18) as an index of expression from psulA-lacZ. The Repression factor was calculated by dividing the Miller units for strain only (JL1436) by Miller units for each construct in JL1436. All constructs were assayed in duplicate for at least three independent experiments.

Binding of the UreR-LexA fusions to target DNA assayed by gel shift. Gel shift experiments were performed to show that the fusion proteins as well as the LexA control bound to p_sulA, providing an explanation for the inhibition of transcription initiation from p_sulA-lacZ. A 64-bp double-stranded oligonucleotide comprising the LexA binding site upstream of sulA (+26 to 38 relative to the transcriptional start of sulA) was synthesized and labeled with DIG-11-ddUTP by terminal transferase.

View larger version (62K): [in this window] [in a new window]   FIG. 6.   Gel mobility shift assay for chimeric protein and the LexA DNA-binding site upstream of psulA. A 64-bp double-stranded oligonucleotide was synthesized based on the nucleotide sequence of psulA. Whole-cell extracts were obtained by inducing a 100-ml culture, in mid-exponential phase, with 2 mM IPTG for 3 h. Culture (50 ml) was centrifuged, and bacteria were resuspended in 1 ml of TEN buffer and lysed in a French press (18,000 lb/in2). All lanes contained the labeled 64-bp dsDNA fragment. The lane containing no extract contained labeled DNA only. JL1436, LexA, and UreR1-182-LexA1-87 whole-cell extracts were added at volumes of 10, 5, and 1 µl.

PCR site-directed mutagenesis of leucines in the putative UreR dimerization domain. AraC uses three critical leucines (Leu150, Leu151, and Leu161) for dimerization via an antiparallel coiled-coil in the absence and presence of arabinose (15). Due to the conservation of these three leucines in both overall location and spatial orientation within UreR, we hypothesized that these leucine residues are important for dimerization of UreR. To directly test biologically whether these Leu residues are required for dimerization of UreR monomers, PCR site-directed mutagenesis (9) was used to create leucine to alanine mutants of UreR in Leu147 (L147A) alone, Leu148 (L148A) alone, Leu158 (L158A) alone, Leu147 and Leu148 (L147A-L148A), and Leu147, Leu148, and Leu158 (L147A-L148A-L158A). -Galactosidase expression from the p_ureD-lacZ reporter plasmid was measured for each of the UreR leucine mutants in the presence and absence of urea (Fig. 7). In the presence of urea, the L147A, L148A, and L158A single mutants had levels of expression that were 71, 89, and 137%, respectively, of the wild-type UreR level. Thus, dimerization, required for activity, was not dramatically altered. The L147A-L148A double mutant, however, had an expression level that was only 15% of the wild-type UreR level in the presence of urea, a significant drop in activity (P < 0.001). The expression level observed for the L147A-L148A-L158A triple mutant, 6% of the wild-type UreR level (P < 0.001) in the presence of urea, tended to be even lower than that of the double mutant (P = 0.052).

View larger version (17K): [in this window] [in a new window]   FIG. 7.   -Galactosidase assays using the pureD-lacZ reporter construct to measure activation by the UreR leucine mutants. E. coli transformed with the Leu mutant constructs was cultured to mid-log phase in LB medium (4 ml) and induced with 0.02% arabinose with or without 100 mM urea for 1 h. -Galactosidase activity represented expression from pureD-lacZ and is expressed in Miller units. All constructs were assayed in duplicate for at least three independent experiments.

Dimer formation in UreR derivatives. To determine whether the low activity of the triple Leu mutant of UreR-Myc-His was due a loss of dimerization, we compared the elution profile of this protein to that of UreR-Myc-His on a Sephadex G-75 gel filtration column. The two proteins, purified from induced cell lysates on a Ni-nitrilotriacetic acid column, were applied to the column. Both UreR-Myc-His and the L147A-L148A-L158A-Myc-His derivative of UreR eluted at fractions corresponding to the dimerized protein (Fig. 8). Thus, loss of activity in the site-directed mutant was not due to the inability to dimerize.

View larger version (23K): [in this window] [in a new window]   FIG. 8.   Gel filtration chromatography of UreR-Myc-His and its triple Leu mutant. Purified UreR-Myc-His and UreR-L147A-L148A-L158A-Myc-His (3 X Leu mutant) (approximately 10 µg of protein) were applied to a Sephadex G-75 column. Protein elution was monitored at 280 nm. Fractions (0.5 ml) were collected, and aliquots of the 3 X Leu mutant (0.15 ml) were assayed by immunoblotting using anti-Myc antibodies (positioned above peak fractions). The elution volumes of bovine serum albumin (BSA) and carbonic anhydrase (CA) are indicated by arrows. The inset shows Western blot of lysates used for purification, run on a denaturing gel, and developed using anti-Myc antibodies. The apparent molecular size is noted.

Constitutive induction of urease genes by His mutants of UreR. We did not observe urea inducibility using the chimeric fusion proteins. We had reasoned earlier, however, that histidine residues were likely involved in either urea binding or transmission of a structural alteration induced by urea activation to the DNA-binding domain. Indeed, the active site of urease coordinates urea by interaction with four His, one Cys, one Asp, and one Lys residue (18). Thus, urea could be coordinated by similarly configured His residues within UreR. Using site-directed mutagenesis, eight His residues were changed to Ala. Six mutations (H5A, H73A, H107A, H186A, H129A, and H152A) did not significantly alter urea inducibility of the mutated UreR (data not shown). Two of eight His-to-Ala mutants tested, however, H102A and H175A, constitutively induced urease genes (i.e., in the absence of urea), as assayed using the ureD-lacZ translational fusion to levels that were not significantly different from the wild-type UreR level in the presence of urea (Table 3). The His102 residue resides in the N-terminal domain, the region predicted to bind urea. Interestingly, the His175 residue resides in the linker region that joins the dimerization domain and DNA-binding domain.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

UreR chimeric proteins were constructed to identify and localize functional domains of the AraC-like transcriptional activator of the P. mirabilis urease gene cluster. Our studies led us to conclude that the N-terminal half of UreR contains the dimerization domain and the C-terminal half of UreR serves as the DNA-binding domain. Leucine residues in the putative dimerization domain of UreR, conserved with respect to AraC and other UreR homologues, were required to fully activate transcription. Site-directed mutagenesis studies were consistent with their involvement in dimerization. While it is proposed that urea binding by UreR is required for activation, construction of the chimeric proteins did not allow us to elucidate this role. Two His mutants of UreR, however, displayed constitutive induction of urease genes.

Our experimental results support the hypothesis that the dimerization domain of UreR localizes to the N-terminal half of UreR. Repression of p_sulA-lacZ by UreR_1-182-LexA_1-87 is consistent with the presence of a dimerization domain supplied by UreR. LexA_1-87 alone was unable to repress transcription from p_sulA-lacZ (Fig. 5). Only when the N terminus of UreR was fused to LexA_1-87 was repression evident, indicating the requirement for dimerization. Other studies have demonstrated that the LexA_1-87 as well as the  repressor requires a dimerization domain for full function (8, 10, 22). That dimerization occurs in the fusion protein is further supported by the observation that UreR_1-182-LexA_1-87 is capable of retarding the mobility of a double-stranded oligonucleotide containing the sulA promoter in a gel shift assay (Fig. 6). LexA_1-87-Myc-His, a stably expressed protein, does not retard the mobility of the target DNA. Thus, the addition of a dimerization domain, provided by the UreR N-terminal amino acid sequences, to LexA_1-87 is necessary and sufficient to restore the DNA-binding capability of LexA_1-87 for its target, p_sulA.

A number of observations also led us to conclude that the DNA-binding domain resides in the C-terminal half of UreR. We demonstrated that a functional chimeric protein, C/EBP_302-350-UreR_164-293, binds to a double-stranded oligonucleotide containing the mapped UreR-binding site (25) within p_ureD (Fig. 4). The UreR-binding site, mapped by Thomas and Collins, is 57 to 34 upstream of the transcriptional start of ureD in the P. mirabilis ureR-ureD intergenic region (25). Interestingly, C/EBP_302-350-UreR_164-293 activated transcription from p_ureD to the same degree in both the absence and presence of urea (Fig. 2). This finding is consistent with the hypothesis that the putative urea-binding site resides in the nonhomologous N-terminal domain. The C/EBP_302-350-UreR_164-293 fusion has slightly less than optimal activation in comparison to UreR. This may have resulted from constraints imparted on the UreR_164-293 by the heterologous C/EBP dimerization domain. The ability of the C/EBP_302-350-UreR_164-293 to bind and activate transcription from p_ureD is consistent with DNA-binding specificity residing in the putative UreR DNA-binding domain. This assertion is further supported by the observation that in the absence of a dimerizing mechanism, the UreR DNA-binding domain alone was unable to activate transcription from p_ureD-lacZ even though UreR_164-293 could recognize its target sequence (Fig. 2) when dimerized. Furthermore, these results suggest that a dimerization domain, provided by C/EBP in this case, is necessary and sufficient to allow the binding of the C-terminal portion of UreR to the ureD promoter. Taken together, these results suggest that, as in other AraC family members, dimerization of UreR is required for activation of urease promoters and that the DNA-binding domain, rather than the dimerization domain, resides in the C-terminal portion of UreR.

UreR activates transcription from both p_ureD and p_ureR in a urea-inducible manner (6), leading to the hypothesis that UreR binds urea. Likely, this mechanism involves a structural change induced by urea binding. We also hypothesized that because the AraC family of transcriptional regulators contains little or no homology in the N-terminal portion of the proteins, binding specificity for an inducer molecule would likely reside in the N-terminal portion of the protein (7). Although UreR_1-182-LexA_1-87 represses transcription from p_sulA-lacZ, the hypothesized urea requirement for UreR function was not observed in the -galactosidase assays (Fig. 5). The tertiary structure of the UreR_1-182-LexA_1-87 fusion appears to allow for UreR dimerization to occur but clearly eliminates the urea-inducible mechanism. If UreR dimerizes via two different conformations depending on whether urea is present or not, then possibly only one of the dimerization conformations is attainable when fused to LexA_1-87, allowing for binding of the UreR_1-182-LexA_1-87 fusion to p_sulA-lacZ. This may account for the unresponsiveness of UreR_1-182-LexA_1-87 to urea.

Nevertheless, some clues as to the mechanism of urea inducibility by UreR were uncovered by site-directed mutagenesis. Interestingly, among eight His-to-Ala mutants of UreR tested for urea inducibility, two resulted in constitutive transcriptional activation from the ureD promoter in the absence of urea. His102 resides in the putative urea-binding (N-terminal) domain. His175 resides in the amino acid sequence that comprises a linker region between the N-terminal and C-terminal domains. It could be speculated that this latter residue transmits the structural alteration that follows urea binding to the C-terminal DNA-binding domain, resulting in binding to specific DNA sequences within the ureR-ureD intergenic region.

AraC dimerizes via an antiparallel coiled-coil containing three leucines that fits into a knobs-into-holes conformation at both ends of the coil (23, 15). UreR retains these three conserved leucines in the same spatial orientation and relative location. PCR site-directed mutagenesis of Leu147, Leu148, and Leu158 in UreR demonstrates the requirement of these residues for transcriptional activation of p_ureD-lacZ (Fig. 7). The levels of transcriptional activation for each of the single leucine mutants of UreR are not significantly different from the wild-type UreR level in the presence of urea. The double and triple leucine mutants, however, show a dramatic decrease in activation in comparison to both wild-type UreR and the single leucine mutants in both the presence and absence of urea (as little as 6% of native UreR activity). These results indicate that alteration of each one of the leucines alone is not sufficient to disrupt the protein structure to any degree in the presence of urea. However, because both Leu147 and Leu148 from one monomer may flank Leu158 from the other monomer and act as a dimerization anchor at each end of the coiled-coil helix, the mutation of both Leu147 and Leu148 would be expected to perturb the native dimerization state. A mutation in one of the leucines still may allow for the interaction between the nonmutated leucine and Leu158 from the other monomer. However, mutation of both Leu147 and Leu148 may eliminate the pocket into which Leu158, from the other monomer, is anchored. Surprisingly, while the native dimerization state may have been altered, the triple Leu mutant remained a dimer. The purified Myc-His derivative of the triple mutant eluted at the identical fraction as purified UreR-Myc-His on a gel filtration column (Fig. 8), indicating that under physiologic conditions the two monomers remained associated. More drastic amino acid substitutions of the homologous residues (Leu to Lys in combination with Leu to Ser) in a construct expressing the N-terminal half of AraC did disrupt the dimeric state to yield the monomers (15). Likely, this would be the case for UreR as well. In this study, the Leu residues were conservatively changed to only Ala. While this study is the first to show the biological relevance of these critical leucine residues in UreR, clearly additional studies are required to substantiate their precise role.

The UreR-Myc-His fusion was constructed to detect and follow a protein that mimics wild-type UreR in vivo and in vitro. Although UreR-Myc-His contains the c-Myc and His₆ epitopes at the C terminus of UreR, -galactosidase reporter activity demonstrated that UreR-Myc-His activated transcription from p_ureD-lacZ as well as wild-type UreR (Fig. 2). Overexpression of UreR-Myc-His and detection by anti-Myc and anti-His antibodies on a Western blot show that the fusion protein was produced (Fig. 3). Furthermore, gel mobility shift assays demonstrated that UreR-Myc-His was able to bind to p_ureD, which resulted in a strong shift of labeled target DNA comparable to the shift seen with wild-type UreR (data not shown).

Using protein fusion technology, we have provided evidence that the AraC-like transcriptional activator UreR consists of dimerization and DNA-binding domains. We have also elucidated a possible dimerization mechanism based on the knobs-into-holes conformation facilitated by three leucine residues that are conserved within both UreR and AraC. Our chimeric proteins were unresponsive to urea activation, and thus a urea-binding domain could not be assigned although we identified two key His residues that were involved in urea inducibility.