Transcriptional Organization and Regulatory Elements of a Pseudomonas sp. Strain ADP Operon Encoding a LysR-Type Regulator and a Putative Solute Transport System

Abstract

The atzS-atzT-atzU-atzV-atzW gene cluster of the Pseudomonas sp. strain ADP atrazine-degradative plasmid pADP-1, which carries genes for an outer membrane protein and the components of a putative ABC-type solute transporter, is located downstream from atzR, which encodes the LysR-type transcriptional regulator of the cyanuric acid-degradative operon atzDEF. Here we describe the transcriptional organization of these genes. Our results show that all six genes are cotranscribed from the PatzR promoter to form the atzRSTUVW operon. A second, stronger promoter, PatzT, is found within atzS and directs transcription of the four distal genes. PatzT is σ^N dependent, activated by NtrC in response to nitrogen limitation with the aid of IHF, and repressed by AtzR. A combination of in vivo mutational analysis and primer extension allowed us to locate the PatzT promoter and map the transcriptional start site. Similarly, we used deletion and point mutation analyses, along with in vivo expression studies and in vitro binding assays, to locate the NtrC, IHF, and AtzR binding sites and address their functionality. Our results suggest a regulatory model in which NtrC activates PatzT transcription via DNA looping, while AtzR acts as an antiactivator that diminishes expression by interfering with the activation process.

INTRODUCTION

Pseudomonas sp. strain ADP (31) is the model organism for bacterial degradation of the s-triazine herbicide atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine). Pseudomonas sp. strain ADP uses atrazine as the sole nitrogen source via a six-step hydrolytic pathway, carried by the 108-kbp catabolic plasmid pADP-1 (35). The six genes involved in atrazine mineralization are localized in two distinct regions of the plasmid: atzA, atzB, and atzC, responsible for atrazine conversion to cyanuric acid (2,4,6-trihydroxy-1,3,5-triazine), are harbored at three distant positions within a large (>40-kbp) unstable region featuring multiple direct repeats and transposable elements. The genes involved in cyanuric acid cleavage and ammonium release are clustered in the stable portion of pADP-1 to form the atzDEF operon (reviewed in reference 51).

Our previous work has shown that atrazine utilization by Pseudomonas sp. strain ADP is regulated by nitrogen availability in a manner resembling general nitrogen control (15), and the cyanuric acid utilization operon atzDEF is one of the targets for such regulation (16). Expression of atzDEF is induced by the substrate of the pathway, cyanuric acid, and repressed in the presence of a preferential nitrogen source. The product of atzR, the gene transcribed divergently from atzDEF, is a LysR-type transcriptional regulator (LTTR) required for both nitrogen- and cyanuric acid-dependent control. Transcription of atzR is initiated from the σ^N-dependent PatzR promoter, activated by the general nitrogen control protein NtrC, and repressed by AtzR (16). NtrC-dependent activation of atzR is unusual in that it does not require interaction with any upstream or downstream sequence elements (41). AtzR binds a single site overlapping the PatzR promoter σ^N RNA polymerase (E-σ^N) recognition element and competes with E-σ^N for DNA binding, resulting in decreased promoter occupancy (41). Cyanuric acid interacts with AtzR to alter its conformation on the divergent PatzR-PatzDEF promoter region and activate atzDEF transcription (40). Despite this effect, cyanuric acid does not alter the ability of AtzR to repress its own synthesis (16). AtzR also stimulates atzDEF transcription in response to nitrogen limitation by an unknown mechanism involving protein-protein interaction with the P_II protein GlnK (17; A. López-Sánchez and F. Govantes, unpublished results). Regulation of the atrazine utilization pathway was recently reviewed (21).

A specific transport system for s-triazines has not been described thus far. Several loci at the pADP-1 plasmid appear to carry genes for transport proteins (35). The product of orf46 is similar to a family of xanthine/uracil permeases, and that of orf69 is homologous to a secondary magnesium/citrate transporter. Finally, the orf98-orf97-orf96-orf95-orf94 cluster (renamed here as atzS-atzT-atzU-atzV-atzW) gene products display significant similarity to a family of outer membrane proteins containing a nucleoside binding channel (atzS) and to the subunits of the ABC family of solute transporters (atzT, atzU, atzV, and atzW). This, and the fact that this group of genes is located immediately downstream from atzR and in the same orientation (Fig. 1A), drew our attention to the possibility that their function may be related to the atrazine-degradative pathway.

Fig 1.

RT-PCR analysis of the atzRSTUVW intergenic regions. (A) Schematic of the atzRSTUVW cluster. The scale indicates the coordinates of the pADP-1 plasmid, as described by Martinez et al. (35). (B) Agarose gel images of RT-PCR products obtained from Pseudomonas sp. ADP cultures grown on ammonium, serine, or cyanuric acid as the sole nitrogen source, using primers flanking each intergenic region. As a control, 16S rRNA primers were used. cDNA (25 ng) for each experimental condition was used as a template. (C) Positive and negative experimental controls. The - sign denotes RT-PCRs performed in the absence of template RNA. The DNA lanes show PCRs performed using cloned atzRSTUVW operon DNA as the template. One representative image of at least three repeat experiments using independent RNA preparations is shown for each primer set.

In the present work, we explored the transcriptional organization and regulation of the atzR-atzS-atzT-atzU-atzV-atzW gene cluster. Our work documents the presence of two promoters, the previously characterized PatzR, providing low-level transcription of the complete atzRSTUVW operon, and the internal PatzT promoter, which fosters high-level transcription of the four distal genes. Here we describe and characterize in detail the physiological regulation, as well as the cis- and trans-acting elements, involved in transcriptional control of the latter.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Bacterial strains used in this work and their relevant genotypes are summarized in Table 1. Minimal medium containing 25 mM sodium succinate as the sole carbon source was used for in vivo gene expression analysis (30). Nitrogen sources were ammonium chloride, l-serine (1 g/liter), or cyanuric acid (3.3 mM). When required, 0.1 mM cyanuric acid was added as a nonmetabolizable inducer. Luria-Bertani (LB) medium was used as rich medium (46). Liquid cultures were grown in culture tubes or flasks with shaking (180 rpm) at 30 or 37°C (for Pseudomonas or Escherichia coli strains, respectively). For solid medium, Bacto agar (Difco, Detroit, MI) was added to a final concentration of 18 g/liter. Antibiotics and other additions were used, when required, at the following concentrations: ampicillin (100 mg/liter), kanamycin (20 mg/liter), carbenicillin (500 mg/liter), rifampin (10 mg/liter), chloramphenicol (15 mg/liter), tetracycline (5 mg/liter), and 5-bromo-4-chloro-3-indoyl-β-d-galactopyranoside (X-Gal; 25 mg/liter). All reagents were purchased from Sigma-Aldrich.

Table 1.

Bacterial strains and plasmids used in this work

Strain or plasmid	Genotype or phenotype	Reference or source
Bacterial strains
E. coli DH5α	ϕ80dlacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17 (rK− mK+) supE44 thi-1 gyrA relA1	23
E. coli KT5746	N5271 [galK ilv his(λ cIts5857 N7N53 ΔBamΔHI)]/pPLhiphimA-5; Apr	37
E. coli NCM631	hsdS gal λDE3::lacI lacUV5::gen1(T7 RNA polymerase) Δlac linked to Tn10	20
Pseudomonas. sp. ADP	Prototroph, pADP-1	31
P. putida KT2440	mt-2 hsdR1 (r− m+) Cmr	13
P. putida KT2440-IHF3	mt-2 hsdR1 (r− m+) ΔihfA::Tcr	33
P. putida KT2442	mt-2 hsdR1 (r− m+), Rifr	13
P. putida MPO201	mt-2 hsdR1 (r− m+) Cmr Rifr ΔntrC::Tcr	16
Plasmids
pIZ227	pACYC184-derived plasmid containing lacIq and the T7 lysozyme gene; Cmr	20
pMPO109	atzR coding sequence and promoter region cloned in pKT230; Kmr	16
pMPO135	pET23b plasmid derivative for overexpression of AtzR-His6; Apr	40
pMPO216	Fragment from pADP-1 containing atzT, atzS, atzR, andatzDEF operon cloned in pBluescript II SK(+); Apr	This work
pMPO234	Broad-host range lacZ transcriptional fusion vector, based on pBBR1MCS-4; Apr	41
pMPO805	atzS-lacZ transcriptional fusion in pMPO234 carrying the sequence between positions −218 to and +319; Apr	This work
pMPO806	PatzR-atzR-atzS-lacZ transcriptional fusion in pMPO234; Apr	This work
pMPO807	Promoterless atzR-atzS-lacZ transcriptional fusion in pMPO234; Apr	This work
pMPO810	atzS-lacZ transcriptional fusion in pMPO234 carrying the sequence between positions −150 and +319; Apr	This work
pMPO811	atzS-lacZ transcriptional fusion in pMPO234 carrying the sequence between positions −123 and +319; Apr	This work
pMPO813	atzS-lacZ transcriptional fusion in pMPO234 carrying the sequence between positions −103 and +319; Apr	This work
pMPO814	atzS-lacZ transcriptional fusion in pMPO234 carrying the sequence between positions −76 and +319; Apr	This work
pMPO815	atzS-lacZ transcriptional fusion in pMPO234 containing the AtzR-L mutant sequence; Apr	This work
pMPO816	atzS-lacZ transcriptional fusion in pMPO234 containing the AtzR-R mutant sequence; Apr	This work
pMPO817	atzS-lacZ transcriptional fusion in pMPO234 containing PatzT with a mutant E-σN binding motif; Apr	This work
pMPO818	atzS-lacZ transcriptional fusion in pMPO234 containing the NtrC1-L mutant sequence; Apr	This work
pMPO819	atzS-lacZ transcriptional fusion in pMPO234 containing the NtrC1-R mutant sequence; Apr	This work
pMPO821	atzS-lacZ transcriptional fusion in pMPO234 containing the AtzR-LR mutant sequence; Apr	This work
pMPO822	atzS-lacZ transcriptional fusion in pMPO234 containing the NtrC1-LR mutant sequence; Apr	This work
pMPO825	atzS-lacZ transcriptional fusion in pMPO234 containing the NtrC2-L mutant sequence; Apr	This work
pMPO826	atzS-lacZ transcriptional fusion in pMPO234 containing the NtrC1-LR+NtrC2-L mutant sequence; Apr	This work
pMPO831	PatzT template plasmid for in vitro transcription, based on pTE103; Apr	This work
pMPO882	PatzR-atzR-atzS-lacZ transcriptional fusion in pMPO234 containing PatzT with a mutant E-σN binding motif; Apr	This work
pMPO883	Promoterless atzR-atzS-lacZ transcriptional fusion in pMPO234 containing PatzT with a mutant E-σN binding motif; Apr	This work
pRK2013	Helper plasmid used in triparental conjugation; Kmr Tra+	12
pTE103	Vector for in vitro transcription assays; Apr	9

Plasmid construction.

Plasmids and oligonucleotides used in this work are summarized in Tables 1 and 2, respectively. All DNA manipulations were performed according to standard procedures (46). Restriction enzymes, DNA polymerases, and T4 DNA ligase were purchased from Roche Applied Science. The Klenow fragment or T4 DNA polymerase was routinely used to fill in recessed 3′ ends and trim protruding 3′ ends of incompatible restriction sites. Plasmid DNA preparation and DNA purification kits were purchased from Sigma-Aldrich, GE Healthcare, or Macherey-Nagel and used according to the manufacturers' specifications. In all cloning procedures involving PCR amplification, the presence of the desired mutations and the absence of unwanted alterations were determined by commercial sequencing (Secugen, Madrid, Spain). Sequence comparison was performed using the BLAST package (2), available at the NCBI Web server (http://www.ncbi.nlm.nih.gov/blast). E. coli DH5α was used as the host in all cloning procedures. Plasmid DNA was transferred to E. coli and Pseudomonas putida strains by transformation (27) or by triparental mating (10).

Table 2.

Oligonucleotides used in this study

Oligonucleotide	Sequence (5′–3′)
2mutAtzR1	GCAACGTTTCGTTGCCGGTGTGATT
2mutAtzR2	AATCACACCGGCAACGAAACGTTGC
2mutNtrC	ACGTGACATCATATGTAATATTCGTGTGGC
2mutNtrC2	GCCACACGAATATTACATATGATGTCACGT
3mutNtrC	ATTTCCCTGCATATTTTTGAGATCC
3mutNtrC2	GGATCTCAAAAATAAGCAGGGAAAT
BendP2	AACCTGGTCGACGAACACATAAAAAAGG
del-ABS	TATCAGGGTTATTGTCTCATGAGCGG
del-RBS	TTGAATGGGCAAATATTATACGCAA
ext1-rev	TGGTCGCATTGCGTGAGAGG
f27	AGAGTTTGATCMTGGCTCAG
fpIHF2	AAGGCGGTCGACTCGGATGCAATCTTT
fpIHF3	ACCCACGTCGACACGTGACATCACCAG
fpIHF4	AGACAGGCGGTGCGGACGGT
fpNtrC1	TTACAAGTCGACTTATGAGCTTGATATC
fpNtrC2	TATCGTTATGAAAGGCACTGCGTT
fpNtrC3	GAATTCGTCGACGGCAAGCTCGCTGATACG
fwd-lacZ	GTTTTCCCAGTCACGAC
mut1-fwd	TTGAATCACACCGGCAACGAAACGGCACCGG
mut10-rev	ACGTGACATCATATGTAAGGTGCGTGTGGC
mut2-rev	CCGGTGCCGTTTCGTTGCCGGTGTGATTCAA
mut5-fwd	CCTTGTTATAAAAGAGCGGGCCAGCCTCAAC
mut6-rev	GTTGAGGCTGGCCCGCTCTTTTATAACAAGG
mut7-fwd	GCCACACGAATATTACTGGTGATGTCACGT
mut8-rev	ACGTGACATCACCAGTAATATTCGTGTGGC
mut9-fwd	GCCACACGCACCTTACATATGATGTCACGT
orf95-94a	ATCATCGTGGCCTACGTCGG
orf95-94b	GTCCGAGGAAGGCCGACAGG
orf96-95a	TGCTGATCCGCCAGGCGATC
orf96-95b	GTACAGGCTGGTGCACGGCT
orf97-96a	CGCCGCCAAACTGTTCGAGG
orf97-96b	GTTCTCGCCCAGCACCGCAT
orf98-0	ACACGGCATAGGCTT
orf98-2	GGGAATTCTGGCATTTCCCTGCAC
orf98-3	TTGAATTCCCGGTGCCGTTTCGGC
orf98-5	CGGAATTCGTGTGATTCAAAAAAC
orf98-6	TTGAATTCTTCGGTCAACAGGTTC
orf98-97a	GACCGCAGTCGAGATCAACG
orf98-97b	GTTCTTCTCCATCGCCAGGC
PEX98-2	GACGGTCGTCTTGGACTTCATGCGGGTTC
Porf98XS	GCGATCGGATCCAACCTTTAAATGCCTTG
R-SalI	GGGATGTCGACCAAGGCGATT
r519	GWATTACCGCGGCKGCTG
ret1	ATTCGTCGCCGGCAAG
ret2	ATGCAATCTTTAAATG
secIHFa	GGCGATCGGATGCAATCTTT
secIHFb	ATGCAACGTGACATCACCAG
secNtrCa	TGACCGAACACATAAAAAAGG
secNtrCb	TCAAATTATGAGCTTGATATC

The 8.3-kb BalI fragment from the atrazine catabolic plasmid pADP-1 carrying atzT, atzS, atzR, and the atzDEF operon was cloned into XbaI-cleaved pBluescript II KS(+) to yield pMPO216. To construct the atzS-lacZ transcriptional fusion plasmid pMPO806, a 1,539-bp HindIII-BamHI fragment from pMPO216 carrying PatzR, atzR, and the 5′ end of atzS was inserted into EcoRI- and BamHI-cleaved pMPO234. Similarly, a 1,459-bp SphI-BamHI fragment and a 532-bp EcoRI-BamHI fragment from pMPO216 spanning the complete but promoterless atzR and the 5′ end of atzS, or the 3′ end of atzR and the 5′ end of atzS, were cloned into EcoRI- and BamHI-digested pMPO234 to yield pMPO807 and pMPO805, respectively. PatzT deletion mutants were constructed by PCR amplification of the Δ1, Δ2, Δ3, and Δ4 PatzT fragments by using orf98-2, orf98-3, orf98-5, or orf98-6 as the forward primers and orf98-0 as the reverse primer. The PCR products obtained were cleaved with EcoRI and BamHI and ligated to EcoRI- and BamHI-digested pMPO234 to yield pMPO810, pMPO811, pMPO813, and pMPO814, respectively. Short derivatives of the wild-type and Δ2 fusions were generated by PCR amplification using pMPO805 and pMPO811 as the templates with primers delABS and Porf98XS. The resulting PCR products were cloned similarly into pMPO234 to construct pMPO829 and pMPO830. The in vitro transcription template plasmid pMPO831 was constructed by cloning the 532-bp PatzT EcoRI-BamHI insert from pMPO805 into EcoRI- and BamHI-digested pTE103.

Site-directed mutagenesis by overlap extension with PCR was performed essentially as described previously (1). pMPO805 was used as a template except when noted, and oligonucleotides fwd-lacZ and ext1-rev were used as external primers in all cases. The AtzR-L and AtzR-R mutant derivatives were generated using the mutagenic oligonucleotide pairs mut2-rev/mut1-fwd and mut4-rev/mut3-fwd, respectively. The final PCR products were cleaved with EcoRI and BamHI and ligated into EcoRI- and BamHI-digested pMPO234 to generate pMPO815 and pMPO816, respectively. Plasmid pMPO816 was used as the template to generate the AtzR-LR mutant derivative, along with the mutagenic oligonucleotide pair 2mutAtzR1/2mutAtzR2. The mutant PCR product was cloned into pMPO234 to yield pMPO821. Similarly, mutagenic primer pairs mut10-rev/mut9-fwd and mut8-rev/mut7-fwd were used to generate the NtrC1-L and NtrC1-R mutant derivatives, which were cloned into pMPO234 to yield pMPO818 and pMPO819, respectively. pMPO818 was used as the template to generate the NtrC1-LR mutant by using the mutagenic primer pair 2mutNtrC/2mutNtrC2. Cloning of the mutant fragment into pMPO234 yielded pMPO822. The NtrC2-L mutant derivative was also obtained by the same procedure, using the mutagenic oligonucleotide pair 3mutNtrC/3mutNtrC2 to produce pMPO825. The NtrC1-LR+NtrC2-L mutant was derived from pMPO822 by using the mutagenic oligonucleotide pair 3mutNtrC/3mutNtrC2, and subsequent cloning into pMPO234 yielded pMPO826. Finally, the σ^N site mutant was constructed with oligonucleotides mut6-rev/mut5-fwd and an ∼500-bp NaeI-BamHI fragment cloned to replace the wild-type sequences in pMPO805, pMPO806, and pMPO807 to generate pMPO817, pMPO882, and pMPO883, respectively.

β-Galactosidase assays.

Steady-state β-galactosidase assays were used to examine the expression of the different atzS-lacZ fusions in P. putida KT2440 and its derivatives. Preinocula of bacterial strains harboring the relevant plasmids were grown to saturation in minimal medium under nitrogen-sufficient conditions (ammonium chloride at 1 g/liter), and cells were then diluted in minimal medium containing the appropriate nitrogen sources (1 g/liter ammonium chloride for nitrogen excess, or 1 g/liter l-serine for nitrogen limitation). Diluted cultures were shaken for 16 to 20 h to mid-exponential phase (optical density at 600 nm, 0.25 to 0.5). Growth was then stopped, and β-galactosidase activity was determined from SDS- and chloroform-permeabilized cells as previously described (36).

Protein purification.

AtzR-His₆ was purified from the overproducing strain NCM631 harboring pMPO135 and pIZ227 by nickel affinity chromatography as previously described (40). IHF was purified from the overproducing strain E. coli K5746 by ammonium sulfate fractionation and affinity chromatography on heparin-Sepharose as described previously (37). Pure P. putida NtrC^D55E,S161F (25) and σ^N (28) were kind gifts of A. B. Hervás and V. Shingler. Core E. coli RNA polymerase was purchased from Epicenter Biotechnologies (Madison, WI).

Gel mobility shift assays.

Gel mobility shift assays with AtzR were performed essentially as described previously (42). Probes containing the wild-type and the mutant versions of the AtzR binding site (AtzR-L, AtzR-R, and AtzR-LR) were obtained by PCR using the corresponding lacZ fusion plasmids pMPO805, pMPO815, pMOPO816, and pMPO821 as templates and oligonucleotides Del-RBS and R-SalI as primers. The PCR products were subsequently digested with EcoRI and BamHI and gel purified. DNA fragments were labeled by filling in 5′-overhanging ends using the Klenow fragment in a reaction mixture containing [α-³²P]dCTP. AtzR-DNA complexes were formed at room temperature in 20-μl reaction mixtures containing 10 ng of the probe, 100 μg/ml salmon sperm DNA, 250 μg/ml bovine serum albumin (BSA), and increasing amounts (0 to 160 nM) of purified AtzR-His₆ (40) in binding buffer (35 mM Tris-acetate [pH 7.9], 70 mM potassium acetate, 20 mM ammonium acetate, 2 mM magnesium acetate, 1 mM calcium chloride, 1 mM dithiothreitol [DTT], 5% glycerol) for 20 min. Reactions were stopped with 4 μl of loading buffer (0.125% [wt/vol] bromophenol blue, 0.125% [wt/vol] xylene cyanol, 10 mM Tris-HCl [pH 8], 1 mM EDTA, 30% glycerol), and samples were separated on a 5% polyacrylamide native gel in Tris-borate-EDTA buffer at 4°C. Gels were dried and exposed to phosphoscreens, which were scanned in a Typhoon 9410 scanner (GE Healthcare) and analyzed with the ImageQuant software (GE Healthcare).

A probe containing the IHF binding site was obtained by PCR amplification using pMPO805 as the template and primers ret-1 and ret-3. The PCR product was subsequently digested with ClaI, gel purified, and radiolabeled, as described above. IHF-DNA complexes were formed at room temperature in 15-μl reaction mixtures containing 10 ng of the probe, 20 μg/ml of poly(dI-dC), 300 μg/ml BSA, and increasing amounts (0 to 500 nM) of purified IHF in binding buffer (10 mM Tris acetate [pH 8], 100 mM potassium acetate, 27 mM ammonium acetate, 8 mM magnesium acetate, 1 mM DTT, 5% glycerol) for 20 min. Reactions were stopped, and samples were separated on an 8% polyacrylamide native gel and processed as described above.

DNase I footprinting assays.

DNase I footprinting assays with AtzR were performed essentially as described previously (42). Probes for DNase I footprinting were obtained by PCR amplification using pMPO805 as the template and primers delABS and fpIHF2 for the top strand and fpNtrC1 and fpIHF4 for the bottom strand. The PCR products were subsequently digested with SalI and gel purified. Binding reactions were performed in binding buffer (35 mM Tris-acetate [pH 7.9], 70 mM potassium acetate, 20 mM ammonium acetate, 2 mM magnesium acetate, 1 mM calcium chloride, 1 mM DTT, 5% glycerol, 100 μg/ml salmon sperm DNA, 250 μg/ml BSA; pH 7.9) containing 10 ng of the radiolabeled probe and 0 to 400 nM AtzR in a final volume of 20 μl. After a 20-min incubation at room temperature, partial digestion of the DNA was initiated by the addition of 1 μl of an empirically determined dilution (typically 10⁻² to 10⁻³) of a DNase I stock solution (10 U/ml; Roche Diagnostics, Basel, Switzerland). Incubation was continued for 30 additional seconds, and reactions were stopped by the addition of 5 μl stop buffer (1.5 M sodium acetate [pH 5.2], 130 mM EDTA, 1 mg/ml salmon sperm DNA, 2.4 mg/ml glycogen). DNA was subsequently ethanol precipitated, resuspended in 5 μl loading buffer (0.125% [wt/vol] bromophenol blue, 0.125% [wt/vol] xylene cyanol, 20 mM EDTA, 95% [vol/vol] formamide) and separated by gel electrophoresis on a 6% polyacrylamide–6 M urea denaturing sequencing gel. Sequencing reactions were performed with the Sequenase 2.0 kit (USB) using the primers secIHFa for the upper strand and secNtrCb for the bottom strand and run in parallel as size markers. Gels were processed and analyzed as described above.

DNase I footprinting assays with NtrC and IHF were performed essentially as described for AtzR. Probes for DNase I footprinting were obtained by PCR amplification using pMPO805 as the template and primers pairs delABS and BendP2 (NtrC, top strand), fpNtrC1 and fpNtrC2 (NtrC, bottom strand), or fpIHF3 and fpIHF4 (IHF, bottom strand). The PCR products were subsequently digested with SalI and gel purified. Binding reactions were performed in binding buffer (10 mM Tris-acetate [pH 8], 100 mM potassium acetate, 27 mM ammonium acetate, 8 mM magnesium acetate, 0.67 mM CaCl₂, 0.33 mg/ml salmon sperm DNA, 250 μg/ml BSA, 1 mM DTT, 5% glycerol) containing 10 ng of the radiolabeled probe and 0 to 2 μM NtrC or 0 to 4 μM IHF in a final volume of 15 μl. Sequencing reactions were performed using primers secNtrCa (NtrC, top strand), secNtrCb (NtrC, bottom strand), or secIHFb (IHF, bottom strand), and run in parallel as size markers.

RNA preparation and RT-PCR.

Pseudomonas sp. strain ADP was grown to mid-exponential phase under nitrogen excess. Cultures were then washed three times, suspended in the same volume of medium containing ammonium, serine, or cyanuric acid as the sole nitrogen source, and incubated for an additional 2 h prior to harvesting. The total RNA preparation procedure was as previously described (16). Reverse transcription (RT) of total RNA (3 μg) was carried out using the high-capacity cDNA Archive kit (Applied Biosystems), with random hexamers as primers. To detect transcript segments corresponding to the atzR-atzS, atzS-atzT, atzT-atzU, atzU-atzV, and atzV-atzW intergenic regions, primer pairs fpNtrC3/fpIHF4, orf98-97a/orf98-97b, orf97-96a/orf97-96b, orf96-95a/orf96-95b, and orf95-94a/orf95-94b were used with 10 to 50 ng of cDNA as template in a 25-cycle PCR program. Amplification of a 455-bp band of the 16S ribosomal gene with primers f27 and r519 (29) was performed to ensure integrity of the cDNA. Negative and positive controls were performed with no template or pMPO216 as a template, respectively. The RT-PCR products obtained were resolved by 1% agarose gel electrophoresis and visualized by ethidium bromide staining.

Primer extension.

Primer extension reactions were performed as previously described (16) with 50 μg of RNA from each condition as the template and ³²P-end-radiolabeled primer PEX98-2 in reaction mixtures containing SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). Sequencing reactions were performed with the Thermo Sequenase cycle sequencing kit (USB, Cleveland, OH), according to the manufacturer's instructions. Samples were run on 6% polyacrylamide–urea sequencing gels and subsequently processed as described above.

In vitro transcription.

Multiround in vitro transcription reactions were performed as previously described (41) in a final volume of 20 μl containing 35 mM Tris-acetate (pH 7.9), 70 mM potassium acetate, 20 mM ammonium acetate, 5 mM magnesium acetate, 1 mM DTT, 10% glycerol, 250 mg/liter BSA, 20 nM E-σ^N, and 0.5 μg of supercoiled plasmid pMPO831 containing PatzT. E. coli core RNA polymerase (100 nM; Epicentre), P. putida σ^N factor (200 nM), and 4 mM ATP were added. When required, IHF was added at a final concentration of 75 nM. All mixtures were incubated for 10 min at 30°C. Open complex formation was then stimulated by the addition of different concentrations of NtrC D^55E,S161F, and the reaction mixtures were incubated for an additional 10 min at 30°C. Subsequently, a mixture of ATP, GTP, CTP (final concentration, 0.4 mM each), UTP (0.07 mM), and [α-³²P]UTP (0.033 mM; Perkin Elmer) was added to initiate multiround in vitro transcription. After 5 min of incubation at 30°C, reinitiation was prevented by the addition of heparin (0.1 mg ml, final concentration). The samples were incubated for an additional 5 min at 30°C, and the reactions were terminated by the addition of 5 μl of stop buffer (150 mM EDTA, 1.05 M NaCl, 14 M urea, 3% glycerol, 0.075% xylene cyanol, and 0.075% bromophenol blue). The samples were run in 6% polyacrylamide–urea gels in Tris-borate-EDTA buffer at room temperature, and gels were processed for radiolabel detection as described above.

RESULTS

atzR, atzS, atzT, atzU, atzV, and atzW are cotranscribed in Pseudomonas sp. strain ADP.

The intergenic regions between atzR, atzS, atzT, atzU, atzV, and atzW are short, spanning 50 bp (atzR-atzS), 38 bp (atzS-atzT), or 6 bp (atzT-atzU), or they are absent (the atzU-atzV and atzV-atzW gene pairs overlap by 14 and 11 bp, respectively). To test whether these six genes form an operon, RT-PCR was performed using primers designed to amplify the five gene junctions and total RNA from Pseudomonas sp. strain ADP grown under nitrogen excess (ammonium as a nitrogen source), nitrogen limitation (serine as a nitrogen source), or with the inducer of AtzR, cyanuric acid, as the sole nitrogen source as the template. A drawback to this approach is the high rate of pADP-1 loss in the absence of selection (8, 15). To minimize this effect, a single culture of Pseudomonas sp. strain ADP was grown in medium containing ammonium as the nitrogen source to mid-exponential phase. Cells were washed thoroughly and split into three cultures containing the indicated nitrogen sources, and incubation was resumed for two additional hours before harvesting for RNA preparation. Growth during the 2-h incubation was typically less than one doubling, implying that minimal plasmid loss occurred during this period. Plasmid loss in the initial culture was 30 to 70%, as assessed by dilution plating and scoring for clear halos on atrazine-containing minimal medium.

The results of the RT-PCR showed that PCR products spanning all five gene junctions were efficiently amplified (Fig. 1B), suggesting that readthrough transcription occurs along all five gene junctions and, therefore, the six genes form an operon. The intensity of the band corresponding to the PCR product spanning the atzR-atzS intergenic region was similar under all three conditions, suggesting that the RT-PCR assay is not sensitive enough to detect the relatively weak nitrogen regulation of the PatzR promoter in the presence of AtzR. In contrast, the products corresponding to the other four gene junctions displayed changes in intensity consistent with induction under nitrogen limitation, as expression was low in ammonium and similarly increased in serine and cyanuric acid, both of which are poor nitrogen sources for Pseudomonas sp. strain ADP (16, 40). This change in the regulatory pattern suggests that a promoter directing nitrogen-regulated transcription occurs downstream from atzR and upstream from atzT.

A nitrogen-regulated promoter within the atzS coding sequence.

To test directly the presence of a transcript derived from the atzR-atzS intergenic region, primer extension analysis was performed using RNA from Pseudomonas sp. ADP grown on ammonium, serine, or cyanuric acid as the sole nitrogen source (Fig. 2) and an oligonucleotide annealed to the atzS sequence as primer. Surprisingly, no extension products compatible with transcription initiation at the atzR-atzS intergenic region were obtained. Instead, the highest molecular weight products consistently obtained in these experiments corresponded to transcripts starting at two consecutive adenine residues located 117 and 118 bp downstream from the atzS start codon. The relative intensities of these products under the different growth conditions reproduced the nitrogen response observed in the RT-PCR assays described above, and primer extension analysis performed with P. putida KT2442 bearing this region of pADP1 cloned in plasmid pMPO805 (see below) yielded equivalent results (data not shown). These results strongly suggest that transcription initiation within the atzS coding sequence contributes to the expression of the downstream genes. Interestingly, a region displaying high similarity to the consensus for σ^N-dependent promoters (3) (TGGCCCGCTCTTTGC versus TGGCAC-N₅-TTGC [conserved positions are underlined]) is present immediately upstream (positions −28 to −14) from the transcriptional start mapped above (the most upstream transcript 5′ end is used as the +1 coordinate here). The coincidence of nitrogen regulation and a σ^N recognition motif supports the notion that this sequence represents a bona fide promoter, designated PatzT, that is likely coregulated with the upstream promoter PatzR (see below).

Fig 2.

Primer extension analysis of the atzS transcripts. Total RNA was extracted from Pseudomonas sp. strain ADP grown on ammonium (NH), serine (SE), or cyanuric acid (CN) as the sole nitrogen source. A sequencing ladder, denoted in the G, A, T, and C lanes, is shown alongside the primer extension reaction results. The sequence around the transcriptional start site is indicated, and the initiating nucleotides are shown in bold within this sequence.

Transcriptional organization of the atzRSTUVW operon.

The results above imply an unanticipated complexity in the transcriptional organization of the atzRSTUVW operon. To analyze the relative contributions of the identified promoters, a set of plasmids bearing transcriptional atzS-lacZ fusions at the BamHI site at position +319 relative to the PatzT transcriptional start site and containing different lengths of upstream sequences were constructed in the broad-host-range fusion vector pMPO234 (Fig. 3A). The insert in plasmid pMPO806 harbors the PatzR promoter region, the complete atzR, the 50-bp atzR-atzS intergenic region, and 146 codons of the atzS coding region (including the putative PatzT promoter). Plasmid pMPO807 is identical to pMPO806 but lacks the PatzR promoter region. A third plasmid, pMPO805, contains only the distal 48 bp of the atzR sequence upstream from the intergenic region and the atzS-lacZ fusion. All three constructs were transferred by mating to Pseudomonas putida KT2442, a genetically tractable strain that we have routinely used as a surrogate host with this system (16, 40), and expression in response to nitrogen availability was tested by means of β-galactosidase assays (Fig. 3B).

Fig 3.

Expression of atzS-lacZ fusions with deletions in atzR. (A) Schematic of the atzRS region fused to lacZ. Black bars below the genes denote the extension of the fragments present in each of the indicated constructs. Inactivation of the PatzT promoter is denoted by an X. (B) Expression of atzS-lacZ transcriptional fusions in P. putida KT2442. Bars represent the averages and standard deviations from at least three independent measurements.

Expression from all three atzS-lacZ fusions was low in medium containing ammonium as the sole nitrogen source. The β-galactosidase levels from pMPO806 were increased 17-fold in cells grown in nitrogen-limited medium, and induction under nitrogen limitation was increased further in the constructs lacking the PatzR promoter (pMPO807 and pMPO805), to 106-fold and 134-fold, respectively. These results indicated that the 538-bp insert in pMPO805 displayed a strong promoter activity that was induced under nitrogen limitation. The increase in β-galactosidase activity in pMPO807 and pMPO805, which do not produce AtzR, relative to AtzR-producing pMPO806, may be an indication that AtzR negatively regulates this promoter. To test this possibility, expression levels from pMPO807 and pMPO805 were also tested in the presence of the AtzR-producing plasmid pMPO109. Under nitrogen limitation, a 4- to 5-fold decrease in expression was observed with both constructs, to levels equivalent to those with pMPO806, thus confirming that AtzR represses transcription from both constructs. Expression was also tested in the presence of 0.1 mM cyanuric acid, the inducer of the atzDEF operon and an effector of AtzR. However, no significant effect was observed, indicating that cyanuric acid does not influence the regulation of the atzRSTUVW operon (data not shown).

To test whether the predicted PatzT promoter is accountable for the high levels of gene expression observed in our atzS-lacZ fusions, we performed site-directed mutagenesis to replace the conserved GC dinucleotide at the −12 box of PatzT with TA. This set of mutations was expected to prevent transcription initiation from PatzT. Plasmids pMPO882, pMPO883, and pMPO817 are pMPO234-based atzS-lacZ fusion plasmids equivalent to pMPO806, pMPO807, and pMPO805, respectively, but carry the mentioned substitutions at PatzT (Fig. 3A). When expression from these fusion plasmids was tested in P. putida KT2442 (Fig. 3B), basal levels in ammonium-containing medium were similarly low. PatzT promoter inactivation considerably decreased the induced activity levels (6-fold) in the construct containing the PatzR promoter (pMPO882 versus pMPO806) and nearly abolished expression of the shorter constructs (82-fold decrease in pMPO883 versus pMPO807; 119-fold in pMPO817 versus pMPO805). These results confirmed that PatzT is the functional promoter present in pMPO805.

Taken together, our results indicated that PatzT is likely responsible for most of the transcription of the genes downstream from atzS under nitrogen limitation. Much of the residual expression observed in the absence of PatzT can be attributed to PatzR, which is responsible for transcription of atzR and atzS but only a minor contributor to the expression of the downstream genes. In addition, our results indicate that cis-acting sequences required for high-level transcription (except for an additional 1.6-fold change that may require upstream sequences), nitrogen control, and AtzR-dependent repression of PatzT are contained within the 538-bp insert in pMPO805.

Transcription from PatzT is subjected to general nitrogen control.

Transcription from the PatzR promoter is activated by the general nitrogen control activator NtrC (16, 41). To test whether PatzT transcription was also dependent on NtrC, expression from the PatzT promoter in pMPO805 was tested in P. putida strain MPO201, a KT2442 derivative harboring a deletion of ntrC (16). In addition, since the DNA bending protein IHF acts as a coactivator in many promoters regulated by NtrC and other σ^N-dependent activators, PatzT-lacZ expression was also assessed in a Δihf mutant of P. putida KT2440 (KT2442 is a rifampin-resistant derivative of KT2440). The results are displayed in Table 3.

Table 3.

PatzT-lacZ expression in ΔntrC and Δihf mutant backgrounds

Fusion plasmid	Strain	Background	β-Galactosidase activitya with nitrogen source
			Ammonium	Serine
pMPO805	KT2442	Wild type (Rifr)	346 ± 62	40,800 ± 6,710
pMPO805	MPO201	ΔntrC	236 ± 36	445 ± 106
pMPO805	KT2440	Wild type (Rifs)	262 ± 8	45,300 ± 4,830
pMPO805	KT2440Δihf	Δihf	128 ± 35	6,220 ± 930

^aExpression of the PatzT-lacZ transcriptional fusion in pMPO805 in the wild type and null ntrC and ihf mutants of P. putida. Data (reported in Miller units) represent averages and standard deviations of at least three independent measurements.

The expression pattern of pMPO805 in the wild-type strains KT2440 and KT2442 was as described above: expression was low during nitrogen-sufficient growth and greatly induced (over 100-fold) under nitrogen limitation. Induction was abolished when assayed in the ΔntrC strain, indicating that PatzT is directly or indirectly subjected to NtrC-dependent nitrogen control. Expression in the Δihf background was also substantially lowered (7-fold) relative to the wild-type strains, indicating that PatzT activation is also dependent on this auxiliary protein.

Functional characterization of the NtrC UAS at the PatzT promoter region.

Sequence analysis of the PatzT promoter region revealed a possible upstream activation sequence (UAS) for NtrC, consisting of two putative NtrC binding sites centered at positions −160 and −130 relative to the PatzT transcriptional start (Fig. 4A). The promoter-distal element, designated NtrC-1, shows high conservation on both half-sites relative to the P. putida NtrC binding site consensus (24), (TCACCAGTAAGGTGC versus GCACCAW-N₄-GTGC [conserved positions are underlined]), while the promoter-proximal element, designated NtrC-2, displays a highly conserved left half-site and a degenerate right half-site (GCACCATTTTGAGAT versus GCACCAW-N₄-GTGC [conserved positions are underlined]) (Fig. 4A). To address the relevance of these and other possible cis-acting elements for transcriptional regulation of PatzT, a series of PatzT-lacZ transcriptional fusion plasmids bearing deletions at the PatzT promoter region was constructed. We showed above that the insert in the control construct, pMPO805, spanning positions −217 to +321 relative to the PatzT transcriptional start, harbors all the sequences required for correct transcriptional regulation. Plasmids pMPO810, pMPO811, pMPO813, and pMPO814 also harbor PatzT-lacZ fusions at position +321, but the upstream endpoints of the inserts are located at positions −150, −123, −103, and −76, respectively (Fig. 4A). These fusions are designated Δ1, Δ2, Δ3, and Δ4, respectively. Expression from all fusion constructs was monitored by means of β-galactosidase assays with P. putida KT2442 (Fig. 4B).

Fig 4.

Expression of PatzT-lacZ fusions bearing deletions and point mutations at putative cis-acting elements. (A) Schematic of the PatzT promoter region and lacZ fusion constructs. Putative NtrC and AtzR binding sites are denoted by open boxes and shaded boxes, respectively (the degenerate right half-site of NtrC-2 is surrounded by a broken line). (Top) Sequence of the relevant region containing the putative NtrC and AtzR binding sites. The identities of the mutations constructed are indicated above the sequence. Upstream ends of deletion mutants are shown by bent arrows. The atzS start codon is underlined. (Bottom) Schematic of the fusion constructs of the PatzT promoter region (drawn to scale). The putative PatzT σ^N-dependent promoter is denoted by a closed arrow. Mutationally inactivated elements are crossed. (B to D) β-Galactosidase activities from PatzT-lacZ fusions bearing deletions at the PatzT promoter (B), point mutations at the NtrC binding sites (C), or point mutations at the AtzR binding site (D). Bars represent the averages and standard deviations from at least three independent measurements.

Expression from the wild-type fusion in pMPO805 was essentially as described above: activity levels were low in nitrogen-sufficient medium and greatly increased (129-fold) under nitrogen-limited conditions. A large decrease in induction (down to 19-fold decrease) was observed with fusion Δ1, indicating that sequences present in the wild type but not in Δ1 are involved in PatzT activation in response to nitrogen limitation. Deletion of additional sequences decreased induction further, to a residual ∼10-fold, which was maintained in fusions Δ2, Δ3, and Δ4. Induction under nitrogen limitation was completely abolished when tested in a ΔntrC strain (data not shown). Taken together, these results indicate that a major determinant for activation of PatzT expression by NtrC is located between positions −218 and −150, containing the proposed NtrC1 site. In addition, the region between −50 and −123, containing the proposed NtrC2 site, also appears to be involved in activation. To determine whether sequences downstream from the PatzT promoter are involved in NtrC-dependent activation, fusions wt-short and Δ2-short (equivalent to the wild-type and Δ2 fusions, but with lacZ fused to position +4) were tested. These constructs displayed 100-fold and 7-fold induction under nitrogen limitation (data not shown), similar to the 129-fold and 10-fold induction observed with the wild type and Δ2 fusions. Thus, sequences downstream from PatzT appear to have a marginal contribution to its regulation, and the observed residual NtrC-dependent activation is likely to occur in a UAS-independent fashion (i.e., with the activator not bound or nonspecifically bound to DNA) (53), as documented previously for other Pseudomonas NtrC-dependent promoters (41).

To test directly the possible implication of the NtrC1 and NtrC2 elements in PatzT activation in vivo, sets of point mutations were generated by site-directed mutagenesis. These sets of mutations are intended to impair NtrC binding by diminishing the similarity of each half-site to the consensus. Mutants NtrC1-L and NtrC1-R bear three substitutions at the left and right half-sites of the NtrC1 element, respectively, while mutant NtrC1-LR bears mutations at both half-sites of NtrC1. Mutant NtrC2-L bears three substitutions at the left half-site of NtrC2, and mutant NtrC1-LR+2-L combines mutations at both half-sites of NtrC1 and the left half-site of NtrC2. Mutations at the right half-site of NtrC-2 were not deemed necessary, as it bears no resemblance to the consensus. Fragments equivalent to that in pMPO805 bearing the indicated mutations were transferred to pMPO234 to generate the PatzT-lacZ fusion plasmids pMPO818, pMPO819, pMPO822, pMPO825, and pMPO826. These plasmids were transferred to P. putida KT2442 by mating, and expression was monitored in β-galactosidase assays as described above. Results are shown in Fig. 4C.

Disruption of either one or both half-sites of the NtrC1 element provoked a 6- to 7-fold decrease in expression relative to the wild-type fusion under nitrogen limitation, similar to that obtained with the deletion of the complete NtrC1 element (fusion Δ1). Interestingly, the NtrC2-L mutation at the left half-site of the NtrC2 element was sufficient to elicit a 13-fold decrease in expression under these conditions, to levels equivalent to those observed with the Δ2, Δ3, and Δ4 deletions. Furthermore, the combination of mutations at both NtrC1 half-sites with the NtrC2-L mutation (NtrC1-LR+2-L) failed to diminish induced expression levels significantly below those obtained with NtrC2-L alone. All mutations tested provoked a slight (≤2-fold) but reproducible decrease in activity under nitrogen excess and failed to significantly affect expression in a ΔntrC background (data not shown). Taken together, our results indicate that both NtrC1 and NtrC2 are required for full PatzT activation, and therefore they represent a bona fide NtrC UAS.

NtrC interaction with the PatzT promoter region.

To characterize the interaction of NtrC with the PatzT promoter region further, we used pure NtrC^D55E,S161F to perform DNA binding assays. NtrC^D55E,S161F is a constitutively active mutant that does not require phosphorylation for in vivo or in vitro activity (25). Initially, gel mobility shift assays were attempted, but NtrC-DNA complexes were not consistently obtained, likely due to low stability under electrophoresis conditions (data not shown). Next, the NtrC-DNA interaction was tested by means of DNase I footprinting assays (Fig. 5). NtrC^D55E,S161F protected the PatzT promoter region bottom strand at positions −174 to −171, −166 to −160, and −156 to −150, overlapping with the NtrC1 site. Hypersensitive positions were observed in this region at −178, −177, −176, −158, −149, −148, and −147. Several protected positions were also observed at the bottom strand around the NtrC2 site, at −138 to −134 and −124 to −122. Two strongly hypersensitive positions were observed in this region, at positions −127 and −128. Protection was also observed on the top strand at positions −168 to −164, −158 to −154, −148 to −145, −139 to −134, and −130 to −124. Hypersensitive positions at this strand occurred at positions −174, −173, −171, −161, −141, −140, −133, and −131. Taken together, the in vitro footprinting results confirmed the location of the two NtrC binding sites at the PatzT promoter region. In addition, the occurrence of hypersensitive positions suggested that NtrC binding causes deformation in the DNA strand at both of its binding sites.

Fig 5.

DNase I footprinting analysis of NtrC at the PatzT promoter region. Gel images were obtained from assays performed with radiolabeled the bottom-strand (left) or top-strand (right) PatzT promoter region as a probe. The predicted NtrC-1 and NtrC-2 sites are shown as open boxes. Protected regions are denoted by closed bars, and hypersensitive positions are denoted by closed circles. Coordinates are relative to the PatzT transcriptional start site. NtrC^D55E,S161F concentrations used were 0 (lanes 1), 0.25 μM (lanes 2), 0.5 μM (lanes 3), 1 μM (lanes 4), and 2 μM (lanes 5).

Functional characterization of the AtzR binding site at the PatzT promoter region.

The PatzT promoter region contains the sequence GGTGCCGTTTCGGCACC, centered at position −112 (Fig. 4A). The perfect heptameric inverted repeat (positions underlined) is identical to that found at the AtzR repressor binding site (RBS), responsible for the primary interaction of AtzR with the atzR-atzDEF promoter region (40). To explore the role of this and other cis-acting sequences involved in AtzR-dependent repression, expression from the deleted derivatives of the PatzT promoter was also determined in the presence of the AtzR-producing plasmid pMPO109 (Fig. 4B). Under nitrogen excess, expression was essentially unchanged in all constructs compared to that obtained in the absence of AtzR. Under nitrogen limitation, removal of the sequences between −217 and −150 in the Δ1 fusion fully abolished AtzR-dependent repression, and similar results were obtained with the shorter Δ2, Δ3, and Δ4 fusions, suggesting that an essential determinant for repression lies between positions −217 and −150. The striking coincidence of this region with the location of the NtrC1 site (see above) and the fact that the predicted AtzR binding element is located far downstream from this region suggest that NtrC binding (rather than AtzR binding) to this region may be a requisite for AtzR-dependent repression (for further clarification, see below).

The fact that repression is abolished when sequences upstream from −150 are removed precludes the analysis of the role of the putative RBS motif centered at −112 in the deletion analysis above. To test directly whether this element is involved in AtzR-dependent repression, site-directed mutagenesis was performed separately on each half-site of the motif. Mutant promoter AtzR-L contains three substitutions in the left half-site that alter the sequence to GGCAACG (mutated positions underlined). Mutant promoter AtzR-R contains three substitutions in the right half-site that alter the sequence to CGTTGCC (mutated positions underlined). Mutant promoter AtzR-RL harbors the combination of both sets of mutations. Plasmids pMPO815, pMPO816, and pMPO821 contain fusions identical to pMPO805 but harbor the promoter variants AtzR-L, AtzR-R, and AtzR-LR, respectively (Fig. 4A). The effects of these mutations were determined by means of β-galactosidase assays in P. putida KT2442 and P. putida KT2442 bearing the AtzR-producing plasmid pMPO109 (Fig. 4D).

Expression levels of the PatzT-lacZ fusions bearing mutations at the putative AtzR RBS in the absence of AtzR were similar to that of the control wild-type fusion, indicating that the mutations did not significantly affect basal transcription or NtrC-dependent activation. In contrast, AtzR-dependent repression was nearly abolished in all three mutant promoters, as the repression ratio was reduced from 4-fold to less than 1.5-fold in all cases. This result indicated that the putative RBS is indeed an essential element for AtzR-dependent repression.

AtzR interaction with the PatzT promoter region.

To characterize the interaction of AtzR with the PatzT promoter region, gel mobility shift assays were performed using the wild-type promoter region as a probe (Fig. 6A and E). Upon addition of increasing concentrations of AtzR, a specific retarded band was observed, strongly suggesting that AtzR binds the PatzT promoter region strongly at a single site. The apparent dissociation constant (K_d) for AtzR binding was ∼9 nM. To assess the importance of the RBS element identified for the AtzR-DNA interaction, assays were also performed with probes harboring the AtzR-L, AtzR-R, and AtzR-LR mutations (Fig. 6B, C, D, and E). A similar retarded band was observed with the wild-type and AtzR-L and AtzR-R mutant probes, but the affinity of AtzR for the mutant probes was diminished >10-fold relative to the wild-type fragment. Very weak binding was detected only at the highest protein concentrations when the AtzR-LR probe was used. These results indicated that the heptameric palindrome is an essential determinant for AtzR binding to the PatzT promoter region. Integrity of both half-sites of the putative recognition is essential for optimal high-affinity DNA binding by AtzR and in vivo repression. However, the presence of a single functional half-site may promote a low level of AtzR-DNA interaction.

Fig 6.

Gel mobility shift analysis results for AtzR at the PatzT promoter region. (A to D) Gel images obtained from assays performed using the wild-type (A), AtzR-L (B), AtzR-R (C), or AtzR-LR (D) PatzT promoter fragments as a probe are shown. The arrowheads indicate the retarded complexes. AtzR-His₆ concentrations were 0 (lanes 1), 5 nM (lanes 2), 10 nM (lanes 3), 20 nM (lanes 4), 40 nM (lanes 5), 80 nM (lanes 6), and 160 nM (lanes 7). (E) The percentage of retarded probe plotted against the AtzR-His₆ concentration. Curves represent the best-fitting rectangular hyperbola curves for each data set. Values and error bars represent the averages and standard deviations of at least three independent experiments.

DNase I footprinting assays were also performed in order to characterize further the interaction of AtzR with its binding site at the PatzT promoter region (Fig. 7). AtzR protected positions around the identified RBS (−128 to −109 at the bottom strand and −120 to −97 at the top strand). In addition, hypersensitive positions were noted at −139, −136, −129, −100, −99, −90, and −85 (bottom strand) and −130, −96, −95, −87, −86, and −85 (top strand). Continuous protection of the RBS and the presence of strongly hypersensitive positions downstream from the RBS were reminiscent of the AtzR interaction with the PatzDEF promoter region and suggested that AtzR bends DNA upon binding (40, 42). However, unlike the PatzDEF promoter region, additional protected positions extending toward the PatzT promoter were not observed, suggesting that the interaction of AtzR with DNA outside the strong RBS element is likely very weak or absent. Accordingly, our in vivo and in vitro analyses of the AtzR binding site did not support the notion that AtzR interacts with the PatzT promoter region upstream from position −150, and therefore the lack of repression observed with the Δ1 mutant (see above and Fig. 4B) is likely due to a requirement for the NtrC interaction in this region for repression to occur (see Discussion).

Fig 7.

DNase I footprinting analysis of AtzR at the PatzT promoter region. Gel images obtained from assays performed with radiolabeled bottom-strand (left) or top-strand (right) PatzT promoter region as a probe are shown. The predicted AtzR RBS is shown as a pair of open boxes. Protected regions are denoted by closed bars, and hypersensitive positions are denoted by closed circles. Coordinates are relative to the transcriptional start site. AtzR-His₆ concentrations used were 0 (lanes 1), 50 nM (lanes 2), 100 nM (lanes 3), 200 nM (lanes 4), and 400 nM (lanes 5).

IHF assists NtrC-dependent PatzT activation.

The region between −100 and −50 contains several A-T-rich stretches showing partial matches to the IHF recognition sequence consensus, WATCAA-N₄-TTR (14), although poor conservation of the consensus in IHF binding sites precludes unequivocal identification (19). To assess the hypothesis that IHF interacts with the PatzT promoter, gel mobility shift assays were performed using purified IHF from Escherichia coli (Fig. 8A). A clear retarded complex was observed at IHF concentrations ranging between 200 and 500 nM. However, the complex only represented a small fraction of the labeled probe, and the assay was not improved by changing buffers or the IHF preparation. DNase I footprinting of IHF was also performed on a bottom-strand-radiolabeled PatzT promoter fragment (Fig. 8B). A window of partial protection was observed at positions −82 to −55. The protected region is A-T rich (66%) and contains the best match (6 out of 9 positions) to the IHF consensus (GGTCAACAGGTTC [conserved positions underlined]) within the PatzT promoter region, suggesting that this region indeed represents a true IHF binding site. Footprinting assays with top-strand-radiolabeled PatzT failed to reveal a reproducible protection pattern (data not shown). Taken together, our binding assays suggest that IHF interacts with the PatzT promoter region. However, IHF affinity for its binding site is low, and the complex appears to be unstable under the set of conditions tested.

Fig 8.

DNA binding assays of IHF at the PatzT promoter region. (A) Gel mobility shift analysis. Shown is the gel image obtained from an assay performed using the wild-type PatzT promoter fragment as a probe and 0 (lane 1), 100 nM (lane 2), 200 nM (lane 3), 300 nM (lane 4), 400 nM (lane 5), and 500 nM (lane 6) pure IHF. The arrowhead indicates the retarded complex. (B) DNase I footprinting analysis. Gel images obtained from assays performed on the bottom-strand radiolabeled PatzT promoter region. The protected region is denoted by a closed bar. Coordinates are relative to the transcriptional start site. AtzR-His₆ concentrations used were 0 (lane 1), 0.5 μM (lane 2), 1 μM (lane 3), 2 μM (lane 4), and 4 μM (lane 5).

To reinforce the notion that IHF interacts with, and contributes to positive control of, the PatzT promoter, we directly tested the ability of IHF to stimulate PatzT transcription in vitro. To this end, we set up a multiround in vitro transcription assay using supercoiled plasmid pMPO831, which bears the wild-type PatzT promoter region, as a template, pure E. coli core RNA polymerase and IHF and P. putida σ^N and NtrC^D55E,S161F. The results are shown in Fig. 9. The PatzT transcript was not detected in the absence of NtrC, regardless of the addition of IHF. In contrast, increasing transcript levels were detected when the NtrC concentration was increased gradually between 100 and 500 nM, consistent with the notion that NtrC is the direct activator of the PatzT promoter. Addition of 75 nM IHF to NtrC-containing reaction mixtures greatly enhanced transcription at all NtrC concentrations. These results strongly suggest that IHF is a coactivator that stimulates NtrC-dependent activation of PatzT, although it is not strictly required for activation.

Fig 9.

In vitro activation of the PatzT promoter. (A) Gel image obtained from a representative multiround in vitro transcription assay using the wild-type PatzT promoter region as a template and 0 nM (lanes 1), 100 nM (lanes 2), 200 nM (lanes 3), 300 nM (lanes 4), 400 nM (lanes 5), and 500 nM (lanes 6) NtrC^D55E,S161F. Reaction mixtures contained 75 nM IHF (+; left) or no IHF (−; right). (B) Levels of PatzT transcript obtained in the in vitro transcription assays plotted against the NtrC^D55E,S161F concentration. The signal obtained with the highest concentration of NtrC^D55E,S161F (500 nM) in the presence of IHF was set as 100%. Transcript levels are expressed relative to this value. Bars represent the averages and standard deviations of three independent measurements.

DISCUSSION

LTTRs and σ^N-dependent promoters are widespread elements involved in the control of bacterial transcription initiation. While σ^N-dependent promoters are obligately activated by a group of transcription factors designated enhancer binding proteins (EBPs), belonging to the family of AAA+ ATPases (49), very few of them are also subjected to negative regulation exerted by members of other regulatory families. AtzR is one of these rare proteins that has been shown to repress a σ^N-dependent promoter (41). Here, we characterized the regulatory cis- and trans-acting elements involved in activation and repression of PatzT, a σ^N-dependent internal promoter responsible for transcription of the four distal genes of the atzRSTUVW operon.

Two promoters are accountable for most of the transcription of the atzRSTUVW operon. The upstream PatzR promoter was previously characterized as σ^N dependent, NtrC activated, and AtzR repressed (16, 41). Transcript levels seen in RT-PCR and primer extension assays, as well as β-galactosidase actvity obtained from PatzR-lacZ fusions, suggested that PatzR is a weak promoter, as frequently observed for genes that encode transcription factors. Consistently, AtzR, the product of the first gene in the operon, is known to be present in limiting concentrations within the cell (16, 40). Translation of the second gene of the operon, atzS, is initiated from a suboptimal GUG start codon, suggesting that high levels of AtzS are also avoided. We found that a second promoter, PatzT, is present within the atzS coding region. Our results are consistent with this promoter being responsible for the bulk of transcription of the four distal genes, atzT, atzU, atzV, and atzW. The regulatory patterns of the PatzR and PatzT promoters are similar: they are both σ^N dependent, activated by NtrC under nitrogen limitation, and repressed by AtzR in a cyanuric acid-independent fashion. Thus, the regulatory responses are consistent throughout the complete operon, although the overall levels of transcription are higher for the four distal genes. This may reflect the requirement of higher concentrations of the atzT, atzU, atzV, and atzW gene products. These four proteins are the subunits of a ABC-type solute transporter, and the presence of very short or absent intergenic regions between them may reflect tight translational coupling in order to maintain the correct stoichiometry in the transporter complex (20).

Analysis of the PatzT promoter region revealed the characteristic architecture of a σ^N-dependent promoter (45), including a σ^N binding motif with high similarity to the −12/−24 consensus, a pair of binding sites for the enhancer binding protein NtrC, and a binding site for the auxiliary protein IHF (Fig. 10). Activation of PatzT in vivo under nitrogen limitation is strictly dependent on NtrC and partly dependent on IHF. In vitro transcription assays confirmed that dependence on both factors is direct and likely due to the interaction of both proteins with the promoter region. We identified two NtrC binding sites, centered at positions −160 and −130. Deletion and point mutation analyses showed that both sites are required for activation, as NtrC1 did not suffice to promote activation in the absence of a functional NtrC2 and inactivation of NtrC1 yielded only 2-fold activation above the levels observed in the absence of both sites. Interestingly, the NtrC1 site is a better match to the NtrC binding consensus in P. putida (24), while NtrC2 has a conserved left half-site but a highly degenerate right half-site (Fig. 10). We propose that the NtrC1 element facilitates the NtrC interaction with NtrC2 via protein-protein contacts, based on the following observations: (i) both sites are on the same side of the helix, offset by approximately three helix turns, (ii) both sites bind NtrC at the same concentrations, despite the apparent differences from the consensus sequence, and (iii) multiple hypersensitive sites occurred in the DNase I footprinting assays in the region between the NtrC1 and NtrC2 sites, indicative of DNA deformation that may be required for interaction between NtrC dimers bound to each site. The presence of UAS elements composed of pairs of activator binding sites located on the same face of the helix is a conserved feature of σ^N-dependent promoters (7), including the NtrC-activated glnK, ureD, and codB promoter regions of P. putida (25), and has been linked to cooperative binding and oligomerization of the activator (43). Interestingly, significant (up to 10-fold) NtrC-dependent activation is observed when both NtrC binding sites are removed. As downstream sequences appear to make a marginal contribution to the regulation of the system, we conclude that the observed residual induction is likely due to UAS-independent activation. This phenomenon has been documented for several σ^N-dependent promoters (5, 39, 44, 55), including PatzR, which completely lacks NtrC binding sites and is activated exclusively in a UAS-independent fashion (41).

Fig 10.

Schematic of all cis-acting elements at the PatzT promoter region. The sequence of the PatzT promoter region is shown along with the location of the NtrC-1, NtrC-2, AtzR, and IHF binding sites and the PatzT σ^N-dependent promoter. Conserved binding motifs for NtrC, AtzR, IHF, and σ^N-RNA polymerase are denoted by open boxes. A broken line surrounds the box corresponding to the nonconserved right half-site of the NtrC-2 site. The consensus for the corresponding binding sites are shown below the sequence. Protected regions from the DNase I footprinting assays for NtrC (closed bars), AtzR (open bars), or IHF (shaded bar) are displayed above (top strand) or below (bottom strand) the sequence. Hypersensitive positions in DNase I footprints are also displayed for NtrC (closed circles) and AtzR (open circles). The mapped transcription start sites are denoted by bent arrows. Coordinates are relative to the first (upstream) transcription start site.

We also showed that IHF interacts with the PatzT promoter region at a single site centered at position −70 (Fig. 10) and stimulates NtrC-dependent activation, both in vivo and in vitro. As an auxiliary protein in the activation of σ^N-dependent promoters, IHF usually binds at A-T-rich stretches within the intervening region between the activator binding sites and the promoter. In this position, IHF-induced bending acts as a hinge to bring the activator into close contact with RNA polymerase. This activation model has been documented for multiple examples of σ^N-dependent promoters (4, 6, 26).

Our results showed that AtzR is a repressor of the PatzT promoter. Repression was mild (∼4-fold) and did not require the cognate inducer of AtzR, cyanuric acid. Similarly, AtzR represses its own synthesis in an inducer-independent fashion (41), a trait that is conserved in other LTTRs (47, 50). AtzR binds the PatzT promoter region at a single site containing a strong binding determinant, the RBS, which is essential for repression (Fig. 10). The sequence of the 14-bp interrupted palindrome in this element is identical to that found at the atzR-atzDEF divergent promoter region (40). AtzR, like most LTTRs, is a tetramer that interacts with an extended (>50-bp) site at the atzR-atzDEF promoter region. This form of binding requires a second, weaker recognition element, designated the activator binding site (ABS), which is essential for PatzDEF activation. In contrast, DNase I protection of the PatzT promoter region by AtzR does not extend far beyond the limits of the RBS. Sequence comparison revealed that the sequences of the three individual binding sites within the ABS, designated ABS-1, ABS-2, and ABS-3 (42), are poorly conserved in the PatzT promoter region. Due to this fact, interaction in this region may be too weak to be detected by means of DNase I footprinting. AtzR has been shown to generate a bend on its binding site at the atzR-atzDEF promoter (40). Two lines of evidence support the notion that AtzR also bends DNA at the PatzT promoter region. First, two sets of DNase I-hypersensitive positions occur between positions −100 and −80 of PatzT in one-helix turn intervals on both DNA strands. This hypersensitivity pattern is very reminiscent of that observed at the atzR-atzDEF promoter region. Second, an A-tract is present at an identical location in both binding sites (positions −94 to −89 relative to the PatzT transcriptional start). We recently demonstrated that this element is intrinsically bent and contributes to AtzR binding, DNA bending, and PatzDEF activation (O. Porrúa and F. Govantes, unpublished results). Experiments aimed to address the presence and relevance of an AtzR-induced DNA bend at the PatzT promoter region are under way.

Only a few examples of negatively regulated σ^N-dependent promoters have been reported thus far (11, 32, 34, 52), including the AtzR-repressed PatzR promoter (41). Although PatzR and PatzT are both σ^N-dependent promoters repressed by AtzR, our results suggest that the underlying mechanisms of repression are strikingly different. In PatzR, AtzR prevents transcription by competing with E-σ^N for DNA binding. However, the location of the AtzR binding site at the PatzT promoter region, >80 bp upstream from the E-σ^N binding motif, makes this possibility unlikely. As discussed above, NtrC displays significant UAS-independent activation at the PatzT promoter. Strikingly, deletions that eliminated one or both of the NtrC binding sites while leaving the AtzR recognition sequences intact completely prevented AtzR-dependent repression of the UAS-independent transcription. DNase I footprinting failed to identify any interactions of AtzR, at least in the region overlapping the NtrC-1 site, and in vivo and in vitro evidence strongly suggests that AtzR binds primarily to the RBS motif centered at position −112, far downstream from this region. The simplest explanation for this phenomenon is that AtzR is an antiactivator that interferes with the action of NtrC when activating transcription from its specific binding sites and, conversely, UAS-independent activation observed in the promoter deletion derivatives is not sensitive to AtzR-dependent repression. Antiactivation as a mechanism for repression of σ^N-dependent promoters has been previously documented (11, 32, 34).

The functions of the outer membrane protein and ABC transport system encoded by atzS, atzT, atzU, atzV, and atzW operon are as of yet unknown. Transport systems for alternative nitrogen sources are in fact one of the most abundant functional classes among genes subjected to general nitrogen control, both in E. coli (55) and P. putida (24). Although s-triazine transport by degrading strains has not been studied in detail, indirect evidence of an s-triazine transporter has been documented for several organisms (18, 48, 54), including Pseudomonas sp. strain ADP (15, 22, 38). Proximity to the cyanuric acid utilization atzDEF operation and coregulation by the general nitrogen control system and AtzR suggest that they may be involved in the transport of cyanuric acid. On the other hand, our finding that atrazine uptake by Pseudomonas sp. ADP is strongly regulated by nitrogen availability, even though synthesis of the enzymes required for atrazine conversion to cyanuric acid is constitutive, strongly suggests the existence of an atrazine transport system that is subject to nitrogen regulation (15). The possibility that the atzRSTUVW operon encodes a dedicated atrazine or cyanuric acid transport system is certainly attractive, and experimental work aimed to address these questions is under way.