Molecular Characterization of PauR and Its Role in Control of Putrescine and Cadaverine Catabolism through the γ-Glutamylation Pathway in Pseudomonas aeruginosa PAO1

Abstract

Pseudomonas aeruginosa PAO1 grows on a variety of polyamines as the sole source of carbon and nitrogen. Catabolism of polyamines is mediated by the γ-glutamylation pathway, which is complicated by the existence of multiple homologous enzymes with redundant specificities toward different polyamines for a more diverse metabolic capacity in this organism. Through a series of markerless gene knockout mutants and complementation tests, specific combinations of pauABCD (polyamine utilization) genes were deciphered for catabolism of different polyamines. Among six pauA genes, expression of pauA1, pauA2, pauA4, and pauA5 was found to be inducible by diamines putrescine (PUT) and cadaverine (CAD) but not by diaminopropane. Activation of these promoters was regulated by the PauR repressor, as evidenced by constitutively active promoters in the pauR mutant. The activities of these promoters were further enhanced by exogenous PUT or CAD in the mutant devoid of all six pauA genes. The recombinant PauR protein with a hexahistidine tag at its N terminus was purified, and specific bindings of PauR to the promoter regions of most pau operons were demonstrated by electromobility shift assays. Potential interactions of PUT and CAD with PauR were also suggested by chemical cross-linkage analysis with glutaraldehyde. In comparison, growth on PUT was more proficient than that on CAD, and this observed growth phenotype was reflected in a strong catabolite repression of pauA promoter activation by CAD but was completely absent as reflected by activation by PUT. In summary, this study clearly establishes the function of PauR in control of pau promoters in response to PUT and CAD for their catabolism through the γ-glutamylation pathway.

INTRODUCTION

Biogenic polyamines are essential for cell growth and participate in a wide spectrum of physiological functions in living organisms (1, 2). Common compounds in this group include the diamines 1,3-diaminopropane (DAP), putrescine (PUT), and cadaverine (CAD), the triamines spermidine (SPD) and norspermidine, and the tetramine spermine (SPM). In microbes, PUT and CAD are two most common biogenic diamines. Intracellular concentrations of these compounds are finely tuned through biosynthesis, degradation, and transport (3). PUT and CAD are synthesized directly from decarboxylation of ornithine (speC) and lysine (ldcA), respectively (4, 5). Alternatively, PUT can also be generated through arginine and agmatine degradation (6, 7). On the other hand, exogenous PUT and CAD can be taken as carbon and nitrogen sources by many bacteria, including Pseudomonas aeruginosa (4, 7, 8). It was first reported for Escherichia coli that PUT is degraded through the γ-glutamylation pathway (9, 10). This pathway consists of four consecutive reactions with consumption of ATP in the first step for γ-glutamylputrescine synthesis (Fig. 1A), and one single set of puuABCD genes encodes these four enzymes in the pathway.

Fig 1.

γ-Glutamylation pathway for polyamine catabolism and schematic presentation of pau loci in P. aeruginosa. (A) The pathway shown is representative for diamines only. The n values of DAP, PUT, and CAD are 3, 4, and 5, respectively, and these three diamines are converted into β-alanine, γ-aminobutyrate, and δ-aminovalerate through this pathway. (B) Gene organizations of loci containing pauA1 to pauA6, pauB1 to pauB4, pauC, and pauD1 and pauD2 for polyamine utilization. Genes in these loci are presented as horizontal arrows and labeled with the corresponding PA gene numbers according to the PAO1 genome annotation. Putative functions of uncharacterized and hypothetical proteins are in parentheses. Promoters under PauR regulation as analyzed in this study are indicated with stars.

The γ-glutamylation pathway is further expanded in Pseudomonas to accommodate a more diverse metabolic capacity. Different from that in E. coli, the γ-glutamylation pathway is more complex because of the existence of multiple homologous enzymes with redundant specificities toward different polyamines for a larger metabolic capacity in P. aeruginosa (11). Through transcriptome analysis, we reported identification of a set of redundant pauABCD genes that are essential for polyamine utilization via the γ-glutamylation pathway in P. aeruginosa (7, 11). These genes include six puuA (pauA1 to pauA6), four puuB (pauB1 to pauB4), one puuC (pauC), and two puuD (pauD1 and pauD2) homologues. From growth phenotype analysis of a series of unmarked pauA mutants, specific combinations of pauA genes were assigned to catabolism of DAP, PUT, CAD, and SPM (11).

Expression of puuABCD genes in E. coli is subjected to regulation by the PuuR repressor (10). Due to the redundancy of pau genes for a more diverse scope of substrates, regulation of polyamine catabolism in P. aeruginosa was expected to be more complicated than that in E. coli. The BauR protein was identified as a transcriptional activator of the pauA3B2 operon for DAP catabolism. At least three other regulatory genes were located in physical proximity to the scattered pau genes. Based on sequence comparison, the PA5301 gene of P. aeruginosa was proposed to encode a putative homologue of E. coli PuuR.

This study was undertaken to further decipher the genetic combinations of all pauABCD genes for the catabolism of polyamines. Expression of four PUT- and CAD-dependent pauA promoters was subjected to detailed analysis to elucidate the inducer signal molecules of these promoters. Genetic and biochemical characterization of PA5301 and its encoded PauR protein established their roles in control of PUT- and CAD-responsive pauA promoters, independent of BauR in control of DAP-responsive genes.

MATERIALS AND METHODS

Bacterial strains, plasmids, media, and growth conditions.

Luria-Bertani (LB) medium was used with the following supplements as required: ampicillin at 100 μg ml⁻¹, tetracycline at 12.5 μg ml⁻¹, gentamicin at 10 μg ml⁻¹, and 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) at 0.03% (wt/vol) for E. coli and carbenicillin at 100 μg ml⁻¹, streptomycin at 500 μg ml⁻¹, tetracycline at 100 μg ml⁻¹, and gentamicin at 100 μg ml⁻¹ for P. aeruginosa. Minimal medium P (MMP) (12) containing carbon sources at 20 mM and nitrogen sources at 5 mM was used for the growth of P. aeruginosa.

Cloning of pauB, pauD, and spuC genes for complementation tests.

In general, each gene was amplified by PCR with a pair of primers that covers the entire sequence of its structural gene and the ribosomal binding site. The PCR products were cloned into appropriate restriction sites on pUCP18 so that expression of the cloned gene was driven by the upstream lac promoter. Positive clones were first identified in E. coli and confirmed by sequencing, followed by transformation into strains of P. aeruginosa.

Construction of knockout mutants.

For the pauR mutant, two flanking regions of the targeted gene were PCR amplified and cloned into pRTP1 using primers pauR1F (5′-CGG GAT CCC GAT CAG AAA TTT GCG GGC GTG G-3′), pauR1R (5′-CGG AAT TCC GGC GCG GTG TCC ATG CGC CTG-3′), pauR2F (5′-CGG AAT TCC GGT CGA TCA GTT CGT CTG CCT-3′), and pauR2R (5′-CCA AGC TTG GGG TGC TCG TCT GCA AAC CAT-3′). The PCR products were cloned into pRTP1. A cassette carrying a gentamicin resistance gene from pGMΩ1 was inserted into the conjunction of the two DNA fragments (13). For gene replacement, E. coli SM10 served as the donor in biparental mating with PAO1-Sm^r (14). The desired knockout mutants were selected on LB plates containing streptomycin and either gentamicin or tetracycline, and the mutation was confirmed by PCR. For the construction of a series of pauB and pauD mutants, the protocol for gene replacement and in vivo excision by the Flp-FRT recombination system (15) was used to generate unmarked mutants of PAO1. Expected deletions in these mutants were confirmed by PCR; the locations of the deleted regions in each gene are described in Table 1.

Table 1.

Strains and plasmids used in this study

Strain or plasmid	Genotype or descriptiona	Source or reference
E. coli strains
DH5α	F− Φ80dlacΔM15 Δ(lacZYA-argF)U169 deoR recA1 endA1 hsdR17 (rK− mK−) supE44 λ− thi-1 gyrA96 relA	Bethesda Research Laboratories
TOP10	F− mcrA Δ(mrr-hsdRMS-mcrBC) Φ80lacZΔM15 ΔlacX74 recA1 araD139 Δ(ara-leu)7697 galU galK rpsL (Smr) endA1 nupG	Invitrogen
SM10	thi-1 thr leu tonA lacY supE recA::RP4-2-Tc::Mu (Kmr)	22
P. aeruginosa strains
PAO1	Wild type	12
PAO1-Smr	Spontaneous Smr mutant strain of PAO1	23
PAO5722	pauR::Gmr	This study
PAO5725	ΔpauA1A2A3A4A5A6	11
PAO5008	spuC::Tcr	16
PAO5730	ΔpauB1B2B3B4; spuC::Tcr	This study
PAO5731	ΔpauB1; from Ser-63 to Arg-362	This study
PAO5732	ΔpauB2; from His-6 to Leu-413	This study
PAO5733	ΔpauB3; from Lys-123 to Arg-318	This study
PAO5734	ΔpauB4; from Asn-139 to Leu-331	This study
PAO5735	ΔpauB1B2B3B4	This study
PAO5707	pauC::Gmr (kauB::Gmr)	7
PAO5737	ΔpauD1; from Leu-38 to Cys-243	This study
PAO5738	ΔpauD2; from Arg-50 to Thr-223	This study
PAO5739	ΔpauD1D2	This study
Plasmids
pRTP1	Ampr Sms conjugation vector	24
pGMΩ1	Ampr Gmr; gentamicin resistance gene cassette with omega loop on both ends	13
pQF50	bla; lacZ transcriptional fusion vector	25
pQF52	bla; lacZ translational fusion vector derived from pQF50	26
pGU101	PpauA2::lacZ (spuA) translational fusion of pQF52	16
pGU102	PpauA1(spuI)::lacZ translational fusion of pQF52	16
pHT2040	PpauA4::lacZ transcriptional fusion of pQF50	This study
pHT3356	PpauA5::lacZ transcriptional fusion of pQF50	This study
pBAD-HisD	Modified protein expression vector derived from pBAD-HisA	19
pBAD-PauR	6× His-tagged PauR protein expression vector	This study
pPAUB1	pauB1 (PA0534) in pUCP18	This study
pPAUB2	pauB2 (PA1565) in pUCP18	This study
pPAUB3	pauB3 (PA2776) in pUCP18	This study
pPAUB4	pauB4 (PA5309) in pUCP18	This study
pSPUC	spuC (PA0299) in pUCP18	This study
pPAUD1	pauD1 (PA0297) in pUCP18	This study
pPAUD2	pauD2 (PA1742) in pUCP18	This study
pUCP18	E. coli/P. aeruginosa shuttle vector	27

^aSm^r, streptomycin resistance; Km^r, kanamycin resistance; Gm^r, gentamicin resistance; Tc^r, tetracycline resistance; Amp^r, ampicillin resistance; Sm^s, streptomycin sensitive.

Construction of lacZ fusions.

Plasmids pGU101 and pGU102 were used to measure pauA2 and pauA1 promoter activities, respectively (7, 16). Regulatory regions of pauA4 (PA2040) and pauA5 (PA3356) were amplified by PCR from the genomic DNA of P. aeruginosa PAO1 using the following primers: PpauA5F (5′-GGT GGA TCC GAG AAT CAA CGG CAG TAC TC-3′), PpauA5R (5′-GGT AAG CTT GAT GCA TTG CAG CAG CAC GCCA-3′), PpauA4F (5′-GGT GGA TCC TTG AAC GAT CTT GCT CTT CGT ATC-3′), and PpauA4R (5′-GGT AAG CTT GAA GGA ACA GGC TCG GCT CAG-3′). PCR products were cloned into pQF50 and confirmed by DNA sequencing.

Measurements of LacZ enzyme activity.

The cells were grown in MMP containing carbon and nitrogen sources as indicated below. Cells in the mid-log phase when the optical density at 600 nm reached 0.7 were harvested by centrifugation and then passed through a French press at 8,500 lb/in². The cell debris were removed by centrifugation at 20,000 × g for 10 min at 4°C, and protein concentrations in the crude extracts were determined by the Bradford method (17) using bovine serum albumin as a standard. The levels of β-galactosidase activity were measured at 37°C using o-nitrophenyl-β-galactopyranoside as the reaction substrate, and the formation of o-nitrophenol was determined by spectrophotometry at 420 nm (18).

Overexpression and purification of the histidine-tagged PauR protein.

Full-length pauR was amplified from the genomic DNA of P. aeruginosa using the following primers: pauRF (5′-GAC GTC GGT GCT CGT CTG CAA ACC-3′) and pauRR (5′-GAG GAA TTC TTG TGC TTC AGG TCA GAA ATT TGC-3′). The PCR products were digested with EcoRI restriction endonuclease and assembled onto the SmaI and EcoRI sites of pBAD-His6, a modified pBAD expression vector (19, 20). For overexpression, Escherichia coli TOP10 harboring the recombinant plasmid was grown in LB medium containing ampicillin at 37°C until the optical density at 600 nm reached 0.5, at which point 0.2% arabinose was added to the culture for induction. Culture growth was continued for another 4 h under the same conditions before harvest by centrifugation. A cell extract was obtained by passing through a French press cell at 8,500 lb/in². The soluble fraction was subjected to protein purification using a HisTrap HP column (GE Healthcare) by following the manufacturer's instructions. Protein purity was analyzed by SDS-PAGE, and protein concentration was determined by the method of Bradford (17).

Glutaraldehyde cross-linking.

Purified His-PauR was subjected to glutaraldehyde treatment in 20 mM HEPES buffer at pH 7.5. Reaction mixtures with 25 μg of purified His-PauR in a total volume of 100 μl were treated with 5 μl of 2.3% freshly prepared solution of glutaraldehyde for 2 min at 25°C followed by 3 min at 37°C. The reaction was terminated by addition of 10 μl of 1 M Tris-HCl (pH 8.0). Diamines, γ-aminobutyrate (GABA), or δ-aminovalerate (AMV) was incorporated into the reaction mixture at 1 mM concentration either prior to glutaraldehyde treatment or after quenching with Tris buffer. Cross-linked proteins were analyzed by SDS-PAGE and visualized by ProtoBlue Safe staining (National Diagnostics).

Electrophoretic mobility shift assays.

For the regulatory regions of the pau operons, DNA fragments covering the entire intergenic regions as defined by the genome annotation website www.pseudomonas.com were PCR amplified from the genome by PCR with primers. For radioactively labeled DNA probes, [γ-³²P]dATP was incorporated by polynucleotide kinase. DNA probes (0.2 ng) were allowed to interact with different concentrations of purified proteins in a 20-μl reaction mixture containing 50 mM Tris-HCl (pH 8), 50 mM KCl, 1 mM EDTA, 5% (vol/vol) glycerol, and 150 μg/ml of acetylated bovine serum albumin. A smaller DNA fragment amplified from the lipA promoter region was used as a negative control for nonspecific binding. The reaction mixtures were incubated for 30 min at room temperature before being applied to a 4% native polyacrylamide gel in 0.5× Tris-borate-EDTA (TBE) (pH 7.5) buffer. After being dried, the gel was autoradiographed by exposure to a phosphorimager plate, scanned using FLA-7000, version 1.1, and analyzed by Multi Gauge, version 3.0 (Fujifilm). For nonradioactive DNA probes, 1 ng of DNA was used in the reaction mixture and visualized from the polyacrylamide gel by SYBR green I staining (Invitrogen) followed by an Omega UltraLum imaging system with 473-nm excitation and 520-nm emission wavelengths.

RESULTS AND DISCUSSION

Gene knockout analyses decipher specific combinations of pauABCD genes for polyamine catabolism.

As shown in Fig. 1A, diamine catabolism through the γ-glutamylation pathway requires four steps of enzymatic reactions, which serve to convert DAP, PUT, and CAD into β-alanine, GABA, and AMV, respectively. As evidenced by genetic studies and transcriptome analysis (11, 21), multiple enzymes of redundant substrate specificities were proposed to participate in this pathway of polyamine catabolism—six PauA, four PauB, one PauC, and two PauD enzymes. We have reported that PauA3 is essential for DAP catabolism, PauA1A2A4 for PUT, PauA1A4A5 for CAD, and PauA2 for triamine SPD and tetramine SPM (11). PauC is the only enzyme of broad substrate specificity catalyzing the third step in the pathway, and the pauC mutant cannot grow on all polyamines mentioned above.

To further differentiate the functions of PauB and PauD enzymes in catabolism of these polyamines, a series of pauB and pauD knockout mutants were constructed and subjected to growth phenotype analysis. The results in Table 2 indicate that a single ΔpauB1 lesion caused partial growth retardation on CAD, ΔpauB2 mutation abolished DAP utilization, and the ΔpauB4 mutant cannot grow on SPD and SPM. Surprisingly, strain PAO5731 devoid of four pauB genes did not exhibit any apparent growth defect on PUT, and its growth on CAD was comparable to that of the ΔpauB1 mutant. These results indicated the presence of an additional enzyme(s) for polyamine deamination.

Table 2.

Growth phenotypes of pauB, pauD, and spuC mutants on polyamines

Strain	Genotype	Growth responsea in MMP containing:
		Glu	DAP	PUT	CAD	SPD
PAO1	Wild type	+	+	+	+	+
PAO5722	pauR	+	+	+	+	+
PAO5731	ΔpauB1	+	+	+	±	+
PAO5732	ΔpauB2	+	−	+	+	+
PAO5733	ΔpauB3	+	+	+	+	+
PAO5734	ΔpauB4	+	+	+	+	−
PAO5735	ΔpauB1B2B3B4	+	−	+	±	−
PAO5008	spuC	+	+	±	±	+
PAO5730	ΔpauB1B2B3B4 spuC	+	−	−	−	−
PAO5370/pSPUC	ΔpauB1B2B3B4 spuC/spuC+	+	+	+	+	−
PAO5730/pPAUB1	ΔpauB1B2B3B4 spuC/pauB1+	+	−	+	+	−
PAO5730/pPAUB2	ΔpauB1B2B3B4 spuC/pauB2+	+	+	−	−	−
PAO5730/pPAUB3	ΔpauB1B2B3B4 spuC/pauB3+	+	−	+	−	−
PAO5730/pPAUB4	ΔpauB1B2B3B4 spuC/pauB4+	+	−	−	−	−
PAO5734/pPAUB4	ΔpauB4/pauB4+	+	+	+	+	+
PAO5737	ΔpauD1	+	+	+	+	+
PAO5738	ΔpauD2	+	−	−	−	+
PAO5739	ΔpauD1D2	+	−	−	−	−
PAO5739/pPAUD1	ΔpauD1D2/pauD1+	+	−	−	−	±
PAO5739/pPAUD2	ΔpauD1D2/pauD2+	+	+	+	+	+

^a+, growth in 24 h; ±, growth in 48 h; −, no growth in 48 h.

The PauB enzymes were proposed to catalyze FAD-dependent oxidative deamination of γ-glutamylpolyamine. Besides PauB enzymes, genetic analysis from our previous studies (7, 16) supported participation of a putative putrescine:pyruvate transaminase SpuC in PUT catabolism. As shown in Table 2, while PAO5008 with a lesion on spuC was partially defective in growth on PUT or CAD as the sole source of carbon or nitrogen in the liquid medium, growth on these two diamines was completely abolished in PAO5730 devoid of spuC and all four pauB genes. The growth phenotypes of these three mutant strains were also analyzed on agar plates. As shown in Fig. 2, the results were consistent with those in the liquid medium (Table 2) except for PAO5730 on PUT, which exhibited growth of distinct colonies derived from the inoculation spot. It is very likely that these outgrowth colonies were from spontaneous suppressors of PAO5730 when exposed to PUT; however, the nature of this specific stress was not clear.

Fig 2.

Growth phenotype analysis of pau mutants on agar plates. Aliquots of cell suspension with comparable numbers of cells were prepared as described in Materials and Methods and spotted on the minimal medium plates with the indicated compounds (10 mM) as the sole sources of carbon and nitrogen. Strains are indicated as follows: WT, wild-type PAO1; pauB, PA5735 devoid of all four pauB genes; spuC, PA5008; pauB spuC, PA5730 devoid of spuC and all four pauB genes. Glu, l-glutamate; Put, putrescine; Cad, cadaverine; Spd, spermidine.

Complementation tests were conducted to further characterize the physiological functions of spuC and pauB genes on polyamine utilization. These genes were cloned individually into pUCP18 and introduced into PAO5730. The recombinant strains of PAO5730 were subjected to growth phenotype analysis on polyamines. As shown in Table 2, growth of PAO5730 on DAP can only be recovered by the pauB2 or spuC clone, growth on PUT was recovered by pauB1, pauB3, or spuC, and growth on CAD was recovered by pauB1 or spuC. Surprisingly, none of the pauB and spuC genes can make PAO5730 to regain growth on SPD and SPM. However, PAO5374 (ΔpauB4) harboring pPAUB4 indeed restored growth on SPD and SPM. From these results, we concluded that the PauB4 enzyme may catalyze the reaction to break the C—N bond of the internal secondary amine of SPD and SPM to split it into entities with either a three-carbon or four-carbon chain length, which can be further degraded by the same route as either DAP or PUT.

For the pauD1 and pauD2 genes, the results in Table 2 demonstrate that pauD2 is essential for growth on diamines DAP, PUT, and CAD. While growth on SPD or SPM was sustained in mutants with a single knockout in pauD1 or pauD2, no growth on these two compounds can be observed in the strain with pauD1 and pauD2 double lesions. In the γ-glutamylation pathway for diamines, PauD1 and PauD2 were proposed to hydrolyze the final γ-glutamylated intermediates to release and ensure recycle of glutamate moiety. The results of complementation tests of the pauD1D2 double mutant with pPAUD1 and pPAUD2 (Table 2) support the proposed functions of pauD1 and pauD2 on polyamine utilization, with pauD2 playing a more important role than pauD1. These results were also consistent with the current model of SPD and SPM degradation in which γ-glutamylation may take place on either of the two terminal primary amines, followed by the cleavage of the secondary amine. While more work is needed to elucidate details from this point on, γ-glutamyl-GABA and γ-glutamyl-β-alanine are two apparent products of this catabolic pathway, and therefore only the ΔpauD1D2 double mutant was able to block completely SPD and SPM catabolism in this pathway.

Through the exhaustive genetic analyses reported here, we were able to decipher specific combinations of genes for catabolism of different diamines via the γ-glutamylation pathway in P. aeruginosa: pauA3-pauB2-pauC-pauD2 for DAP, pauA1A2A4-pauB1B3 (SpuC)-pauC-pauD2 for PUT, and pauA1A4A5-pauB1 (SpuC)-pauC-pauD2 for CAD. In addition, since PUT in excess can be channeled into SPD through the SPD biosynthetic pathway, it may therefore provide another route for PUT utilization after conversion into SPD.

PauR effect on pauA promoters.

The five pauA genes (pauA1 to pauA5) involved in diamine catabolism are expressed from five independent transcriptional units (Fig. 1B). The BauR protein, a transcriptional regulator of the LysR family, is required for DAP- and SPD-dependent induction of the pauA3 promoter (11), and the bauR mutant affects growth on DAP and SPD but not on CAD and PUT. Therefore, we hypothesized that four PUT- and CAD-related pauA genes (pauA1A2A4A5) are subjected to a regulatory mechanism different from BauR on pauA3. One promising candidate gene was PA5301, which encodes a transcriptional regulator exhibiting 44% sequence identity to E. coli PuuR for putrescine utilization (10). As described below, we provided several lines of evidence to support PA5301 in regulation of polyamine utilization, and hence PA5301 is designated pauR from now on.

Strain PAO5722, a pauR knockout mutant, was constructed as described in Materials and Methods and subjected to growth phenotype analysis for polyamine utilization. As shown in Table 2, no apparent growth defect was observed in this mutant on any of the polyamines tested (DAP, PUT, CAD, and SPD). Although this result indicated that PauR is not essential for pau gene expression, we proposed that PauR may serve as a transcriptional repressor for pau operons. To test this hypothesis, the expression patterns of pauA promoters fused to the lacZ reporter gene on recombinant plasmids were determined by measurements of β-galactosidase activities in the wild-type strain PAO1 and the pauR mutant. The cells were grown in glucose minimal medium with either ammonium or polyamines as the nitrogen source. As shown in Fig. 3, the promoters of pauA1, -A2, -A4, and -A5 were specifically induced by PUT and CAD in the wild-type strain PAO1. When introduced into the pauR mutant, the basal levels of these promoters increased significantly, and polyamine-dependent induction of these promoters was diminished in this mutant. In comparison, the pauA3 promoter was specifically activated by DAP in PAO1 (11), and the expression pattern of this promoter remained the same in the pauR mutant (data not shown). These results provided the first line of evidence for PauR as a transcriptional repressor of the pauA1, -A2, -A4, and -A5 promoters but not the pauA3 promoter (11).

Fig 3.

Effect of pauR on expression profiles of the pauA promoters. Specific activities of β-galactosidase expressed from pauA promoter-lacZ fusions were measured from the host strains wild-type PAO1 (solid bars) and pauR mutant PAO5722 (empty bars). The cells were grown in minimal medium P supplemented with glucose (G) as the sole source of carbon and putrescine (P), cadaverine (C), or ammonia (N) as the sole source of nitrogen.

Enhanced PUT- and CAD-dependent induction of the pauA promoters in the mutant devoid of PauA enzymes.

To test if PUT and CAD are inducers of the pauA promoters, the activities of these promoters from recombinant plasmids as described above in response to exogenous PUT and CAD were compared in the wild-type strain PAO1 and strain PAO5725, in which six pauA genes were deleted. Without functional PauA enzymes, PAO5725 was expected to accumulate PUT or CAD supplement inside the cells to levels higher than that in PAO1, and synthesis of the downstream intermediate compounds in the γ-glutamylation pathway would be blocked in this mutant. As shown in Fig. 4, the promoters of pauA1, pauA2, pauA4, and pauA5 were inducible by either PUT (>5-fold) or CAD (>2.5-fold) in PAO1 grown in the glucose-ammonia minimal medium. In PAO5725, while the basal expression levels of these four pauA promoters were comparable to those in PAO1, the fold induction by PUT and CAD for the pauA1, pauA2, and pauA4 promoters was further enhanced significantly in PAO5725. These data support PUT and CAD as the inducer molecules of the corresponding pauA promoters.

Fig 4.

Expression profile of pauA promoters in P. aeruginosa PAO1. Specific activities of β-galactosidase, expressed from pGU102 for the pauA1 promoter, pGU101 for the pauA2 promoter, pHT2040 for the pauA4 promoter, or pHT3356 for the pauA5 promoter, were measured from the host strains wild-type PAO1 (solid bars) and ΔpauA1A2A3A4A56 mutant PAO5725 (empty bars) devoid of six glutamyl-polyamine synthetase genes. The cells were grown in glucose-ammonia (GN) minimal medium in the presence or absence of putrescine (P) and cadaverine (C).

In contrast, we have reported that DAP-dependent induction of the pauA3 promoter was greatly diminished in the mutant without functional PauA enzymes, and that β-alanine derived from DAP through the γ-glutamylation pathway may serve as the signal molecule for the BauR-dependent regulatory circuit (11).

Specific interactions of PauR with the promoter regions of pauABCD genes in vitro.

To further support PauR as a transcriptional repressor of pauA promoters, PauR with a hexahistidine tag at its N terminus was constructed and purified to homogeneity as described in Materials and Methods. The DNA-binding activity of this recombinant PauR protein was subjected to analysis by electrophoretic mobility shift experiments using the same regulatory regions of the four pauA promoter fusions in expression measurements. As shown in Fig. 5, the presence of nucleoprotein complexes with a retarded mobility clearly demonstrated binding of PauR specifically to these pauA promoter regions with high affinity.

Fig 5.

PauR regulation on pau promoters. Interactions of purified His-PauR with pau promoters were demonstrated by electrophoretic mobility shift assays as described in Materials and Methods. Ctrl, negative control DNA.

Potential interactions of PauR with the promoter regions of other pauABCD genes (Fig. 1B) were also tested. As shown in Fig. 5, PauR binds to the regulatory regions of all pau operons except pauA3B2 and pauB1. We have reported that expression of the pauA3B2 operon is regulated by BauR in control of DAP and β-alanine utilization (11). In the case of pauB1, it might be subjected to regulation by a putative transcriptional regulator encoded by the gene immediately downstream of pauB1.

In the current hypothesis, the repression effect of PauR on the affected promoters would be released upon its interactions with the inducer molecule PUT or CAD. Interactions of PauR with the promoter regions were also tested with inclusion of PUT or CAD in the reactions. However, no apparent effect was detectable on the DNA-binding activity of PauR with either 1 mM or 5 mM diamines in the reactions (data not shown). We did not continue the experiments with higher concentrations of polyamines, as it was well known that polyamines can bind to DNA through charge interactions and therefore may potentially interfere with PauR activity nonspecifically.

Cross-linkage analysis revealed potential interactions of PauR with PUT and CAD.

Cross-linking analysis with glutaraldehyde was also conducted to detect any conformational changes of PauR in response to polyamines. As shown in Fig. 6, an additional distinct peptide band corresponding to the dimeric PauR formation was detected after cross-linking. When PUT or CAD was added into the reaction, the peptide bands for monomeric and dimeric PauR became diffused, with a slight increase in molecular weight, on SDS-PAGE. This change is specific to PUT and CAD but not to DAP, a diamine with a shorter methylene chain. In the control group, polyamines added after completion of cross-linking reactions did not generate detectable band diffusion. These data supported PauR interaction specifically with PUT and CAD but not DAP.

Fig 6.

PauR conformational change in response to putrescine and cadaverine. PauR showed distinct dimer formation after glutaraldehyde cross-linkage (CL). The presence of PUT or CAD, but not DAP, in the reactions generated more diffused cross-linking products. In the control group, polyamines were added after the termination of cross-linking reactions. Reactions including polyamines but no glutaraldehyde are also shown as a negative control.

Catabolite repression on CAD-dependent induction of pauA promoters.

Although PUT- and CAD-dependent regulation of pauA expression involved PauR, induction of these pauA promoters by PUT or CAD responded differently to other nutrient sources. As shown in Fig. 7, PUT-dependent induction of pauA promoters in wild-type PAO1 was not affected by the addition of glucose and ammonia to the minimal medium. In comparison, CAD-dependent induction of all these four pauA promoters was significantly suppressed by exogenous glucose and ammonia. Similar patterns of repression by exogenous glutamate were also observed (data not shown). The molecular nature of this repression effect was not clear. It might be mediated through differential effects on PUT-PauR and CAD-PauR complexes or on uptake of exogenous CAD and PUT.

Fig 7.

Glucose-ammonia effect on expression profile of pauA promoters. Specific activities of β-galactosidase expressed from pauA promoter-lacZ fusions were measured from the host strain (wild-type PAO1). The cells were grown in minimal medium P in the presence of putrescine (Put) or cadaverine (Cad). Empty bars represent expression levels when 10 mM putrescine or cadaverine was used as the sole source of nutrient for growth. Solid bars show repression on cadaverine utilization by 10 mM glucose and 5 mM ammonia as background nutrients, while pauA expression with putrescine remained unaffected.

In summary, specific genetic combinations of pauABCD genes for catabolism of different polyamines were deciphered in this study. The PauR repressor was characterized as a transcriptional regulator in response to PUT- and CAD-dependent induction of pauA1, -A2, -A4, and -A5 promoters. The PauR protein was able to bind to the regulatory regions of all pau operons except pauA3B2 and pauB1. In conjunction with our previous report of BauR on the pauA3 promoter for DAP catabolism (11), we established a clear regulatory circuit by PauR and BauR in control of diamines catabolism. Further studies are in progress to elucidate the regulatory mechanism of pauABCD genes in response to SPD and SPM.