Creatine utilization as a sole nitrogen source in Pseudomonas putida KT2440 is transcriptionally regulated by CahR

Abstract

Glutamine amidotransferase-1 domain-containing AraC-family transcriptional regulators (GATRs) are present in the genomes of many bacteria, including all Pseudomonas species. The involvement of several characterized GATRs in amine-containing compound metabolism has been determined, but the full scope of GATR ligands and regulatory networks are still unknown. Here, we characterize Pseudomonas putida ’s detection of the animal-derived amine compound creatine, a compound particularly enriched in muscle and ciliated cells by a creatine-specific GATR, PP_3665, here named CahR (Creatine amidohydrolase Regulator). cahR is necessary for transcription of the gene encoding creatinase (PP_3667/creA) in the presence of creatine and is critical for P. putida’s ability to utilize creatine as a sole source of nitrogen. The CahR/creatine regulon is small, and an electrophoretic mobility shift assay demonstrates strong and specific CahR binding only at the creA promoter, supporting the conclusion that much of the regulon is dependent on downstream metabolites. Phylogenetic analysis of creA orthologues associated with cahR orthologues highlights a strain distribution and organization supporting probable horizontal gene transfer, particularly evident within the genus Acinetobacter . This study identifies and characterizes the GATR that transcriptionally controls P. putida ’s metabolism of creatine, broadening the scope of known GATR ligands and suggesting GATR diversification during evolution of metabolism for aliphatic nitrogen compounds.

Introduction

Like many primarily soil-dwelling microbes, the Gram-negative bacterium Pseudomonas putida has evolved a vast and diverse array of transport and metabolic machinery to fuel its organoheterotrophic lifestyle [1]. Within the rhizosphere, fast and efficient adjustment to the surrounding environment ensures successful acquisition of essential nutrients and activation of stress responses that are crucial to P. putida ’s survival [2]. An effective response to the extracellular environment is partly afforded by the large number of transcription regulatory proteins encoded by the P. putida genome. About 2 % of the predicted genes in the 6.15 Mbp P. putida KT2440 genome contain a conserved AraC-type DNA-binding helix–turn–helix motif, characteristic of many catabolism-related transcription regulators [3]. These AraC-family transcription regulators control a large number of metabolic processes within P. putida , although for many the cognate inducing ligands and target regulons have not been identified.

Creatine is an amine compound found primarily in animal tissues where it serves to buffer charging of high-energy carriers during rapid ADP to ATP conversion. It is particularly abundant in tissues that require large pools of ATP and ADP for periods of intense energy expenditure, such as fast-twitch skeletal muscle and ciliated cells [4]. Metabolism of dietary creatine in animals occurs via partial metabolism and modification by gut-residing bacteria creating 1-methylhydantoin, a metabolite that is not easily metabolized by the body and can lead to tissue inflammation. To prevent creation and accumulation of 1-methylhydantoin, the major avenue of creatine homeostasis in animals is the excretion of creatine and creatinine in the urine, where it can be used by soil microbes. Pseudomonas species have been observed to metabolize creatine and creatinine, as well as the microbial breakdown product, 1-methylhydantoin [4–12]. The products of creatine and creatinine metabolism by Pseudomonas vary depending on the specific degradation pathway used, which is species- and sometimes strain-dependent. For creatinine, the process often begins with breakdown of creatinine to creatine by a creatininase (creatinine amidohydrolase), though there are alternativee pathways of metabolism. Creatine is then further metabolized to sarcosine and urea by creatinase (creatine amidinohydrolase), and sarcosine is then converted to glycine and formaldehyde by the tetrameric sarcosine oxidase (SoxBDAG) (Fig. 1a), or to methylamine by sarcosine reductase and/or glycine reductase [4, 5, 13, 14].

Fig. 1.

The creatine metabolic pathway and the role of PP_3665/cahR in creatine metabolism. (a) The creatine metabolic pathway in P. putida intersects with the choline metabolic pathway at sarcosine. The enzymes primarily discussed in this study are noted in blue. (b) Genomic organization around the creatinase gene (PP_3667/creA) in P. putida KT2440. Gene colours match those used in Fig. 6. (c) Net growth of P. putida wild-type (WT), ∆PP_3665/cahR (∆) and the complemented strain (C), compared to nitrogen-free minimal media controls. The compounds present as the sole nitrogen source are listed below each group of bars. Error bars denote standard deviation from between three and six experiments.

Production of creatinase is inducible in P. putida when the bacterium is supplied with creatine as the sole carbon or nitrogen source, but the induction mechanism is currently unknown [8]. P. putida PP_3665 encodes an AraC-type transcription regulator convergently transcribed with an operon containing the creatinase gene, creA (Fig. 1b). Among characterized Pseudomonas genomes, only a few, including particular strains of P. putida and P. resinovorans, maintain probable orthologues of both the creatinase gene and the associated AraC-family transcription regulator in a similar syntenic arrangement [3]. Both of these species of Pseudomonas are present primarily in soils and are active participants in the rhizobiome, i.e. bacteria within the rhizosphere. It is within the rhizosphere that P. putida most probably encounters creatine, as creatine and its anhydrous form, creatinine, are present in soils, deposited via animal urine or faeces as well as through degradation of animal tissue [15–18].

Here, we describe identification of a regulator, PP_3665, required for transcriptional induction of creatinase in P. putida in response to creatine and utilization of creatine as a sole nitrogen source, leading us to name PP_3665 as the Creatine amidohydrolase Regulator, CahR.

Methods

Bacterial strains and growth conditions

P. putida strain KT2440 (ATCC 47054) and mutants made from this parent strain were grown at 30 °C with shaking (170 r.p.m.) in MOPS media [19] with modification we have previously reported [20], supplemented with 25 mM pyruvate, 5 mM arginine and 20 µg gentamicin ml⁻¹ when needed for plasmid maintenance. Escherichia coli strains NEB5α (New England Biolabs) and S17λpir [21], used for cloning and conjugation with P. putida , respectively, and E. coli strain T7Express (New England Bioloabs), used for recombinant protein expression, were maintained in lysogeny broth (LB), Lennox formulation, supplemented with 7 µg gentamicin ml⁻¹ or 150 µg carbenicillin ml⁻¹ as appropriate. E. coli strains were grown at 37 °C with shaking (170 r.p.m.).

Construction of the P. putida ∆PP_3665 deletion strain and complementation construct

A 980 bp fragment upstream of the coding region of cahR was amplified with primers PP3665KO_F1_HindIII (5′-AAGCTTCGGGATCGTTCCAGATGCGT-3′) and PP3665KO_R1 (5′-GATCCAGGTGCTGCCCGATGCCA-3′). A 1 kb fragment downstream of the coding region of cahR was amplified with primers PP3665KO_F2 (5′- GTTGGGTCAGGTATTGGCATTG-3′) and PP3665KO_R2_EcoRI (5′- GAATTCCCGAGGCGGAAAACCCCTGT-3′). The two fragments flanking the cahR coding region were spliced together via splice-overlap extension (SOE) PCR using primers PP3665KO_F1_HindIII and PP3665KO_R2_EcoRI, digested, and ligated into similarly cut plasmid pMQ30 [22] containing a gentamicin resistance cassette for initial selection and the sacB gene for counter selection. pMQ30 containing the flanking regions of cahR (pMQ30:cahR-KO) was maintained in E. coli strain DH5α, and transformed into E. coli S17λpir for conjugation with P. putida . Single crossover integrants were selected by gentamicin resistance (50 µg ml⁻¹) and double crossover to the deletion or revertant was carried out on LB agar with no salt and amended with 10 % sucrose, as previously described [23, 24], to yield an unmarked deletion of PP_3665/cahR in P. putida strain KT2440 (strain LAH111). Deletion or reversion was determined using primers PP3665 KO screen F (5′-ATTTCACCACCATCGGCCTT-3′) and PP3665 KO screen R (5′-AGCGGTAGCCTTTGAGCAAT-3′), which yielded a 3.3 kb product in the wild-type (WT) and a 2 kb product in the deletion strain. cahR complementation in the ∆cahR strain was achieved by plasmid expressed cahR. Briefly, the coding region of cahR and divergently transcribed gene pssA (to ensure inclusion of the full coding region and promoter region of cahR) were amplified using primers PP_3665-EXP-F_EcoRI (5′-ATAGAATTCGACATCAATGCGCGGTGC-3′) and PP_3665-EXP-R-HindIII (5′-AAAAAGCTTCTGCACTGGCTTCTTCTCAC-3′). The ~2.5 kb fragment was digested with EcoRI and HindIII and ligated into the similarly cut pMQ80 plasmid. pMQ80 is a yeast–E. coli–Pseudomonas shuttle vector containing the Pseudomonas stabilization fragment and carries the gentamicin resistance gene [25]. The resulting pMQ80:pssA/cahR complementing plasmid was then electroporated into P. putida KT2440 ∆cahR as previously described [23].

Nitrogen source growth assays

P. putida KT2440 WT and ∆cahR strains carrying the pMQ80 empty vector, and P. putida KT2440 ∆cahR complemented with pMQ80:pssA/cahR were grown overnight at 30 °C with shaking in 1× MOPS buffer amended with 25 mM pyruvate, 5 mM arginine and 20 µg gentamicin ml⁻¹. Cells from overnight culture were collected by centrifugation, washed with 1× MOPS medium lacking nitrogen (MOPS no nitrogen), and adjusted in 1× MOPS no nitrogen to an optical density at 600 nm (OD₆₀₀) of 0.5. These normalized cell suspensions were added to pre-warmed 1× MOPS no nitrogen media amended with 20 mM pyruvate, 20 µg gentamicin ml⁻¹ and 2 mM of one of the following nitrogen sources: choline, creatine, creatinine, sarcosine or arginine. A control for residual growth in media lacking a nitrogen source was also included. Cells were added to each condition to a final OD₆₀₀ of 0.05 in a 500 µl final volume in wells of a 48-well plate. Cultures were incubated at 30 °C with horizontal shaking at 170 r.p.m. for 18 h. The optical densities of cultures were measured at times 0 h and 18 h using a Synergy H1 plate reader (BioTek). Growth was reported as the net growth of each strain amended with a nitrogen source after subtraction of the optical density in the 0 mM nitrogen condition, as there is always a small amount of residual growth from stored intracellular nitrogen.

cahR-dependent transcriptional induction assays and PP_3667/creA promoter mapping

Transcriptional reporter fusions of the full creA promoter, −25 to −227 bp from the predicted creA translational start site, and a promoter truncation including −25 to −195 bp from the predicted creA translational start site, to the coding region of GFP were constructed following a protocol for creation of recombinant plasmids similar to that described by Bryksin and Matsumura [26]. The full promoter region of creA was amplified using primers PP_3667PromF (5′- CGGGTACCGAGCTCGTTCAGGCCGGCCGC-3′) and PP_3667PromR (5′-TAAGATTAGCGGATCAGACTTTGTGGC-3′), and the −169 truncation fragment was amplified using primers PP_3667PromF (5′- CGGGTACCGAGCTCGTTCAGGCCGGCCGC-3′) and PP_3667Prom-169R (5′-CGGGTACCGAGCTCGTCATGGAGCTGGACC-3′). Primers for amplification of the full and −195 truncated creA promoters contain 5′-regions complementary to the sequence upstream of GFP’s coding region on plasmid pMQ80. The promoter fragments flanked with pMQ80 complementary ends were mixed with plasmid pMQ80, and the mixtures were denatured, annealed and amplified via PCR to allow for insertion of the promoter region upstream of pMQ80’s GFP. Insertion of the creA promoter fragments resulted in deletion of pMQ80’s NheI and EcoRI restriction sites, so the amplified plasmid pool was digested with EcoRI and NheI to cull pMQ80 that did not incorporate the creA promoters. Following separation and purification of the circular plasmid via gel electrophoresis, the creA promoter-containing plasmids were transformed into E. coli DH5α. The −76 creA promoter truncation was constructed via three sequential PCRs, using the reverse primer pMQ80_GFP_R (5′-TCAGGCTGAAAATCTTCTC-3′) and the following forward primers for a 5′-extension of GFP including the truncated promoter of creA: PP3667-56_1 (5′-CTTCTCAGGCGGCCGGCCTGAACCGAGCTCGGTACCCG-3′), PP_3667–56_2 (5′-GTCGGTTCTGTTGCAATGCTTCTCAGGCGGCCGGCC-3′) and PP_3667–56_3_BamHI (5′-AATGGATCCGTGTCCTCGTCGGTTCTG-3′). This −76 creA promoter-GFP fusion fragment was digested with restriction enzymes BamHI and HindIII and ligated into similarity cut pMQ80. The −113 creA promoter-GFP fusion fragment was created by further extending the −76 creA promoter-GFP fragment using primers pMQ80_GFP_R and PP3667-80_NheI (5′-AATGCTAGCATTCCAGGTCGGATAGATACAAAAGTGTCCTCGTCGGT-3′). The −113 creA promoter-GFP fusion fragment was digested with NheI and HindIII restriction enzymes and ligated into similarly cut pMQ80. All plasmids were propagated in E. coli NEB5α, purified and electroporated into P. putida KT2440 for induction assays.

Each plasmid encoding the creA promoter-gfp fusion, full or truncated, was electroporated into P. putida KT2440 WT or ∆cahR for transcriptional induction assays. To test metabolite-specific induction using the full creA promoter gfp reporter construct, P. putida KT2440 WT and ∆cahR carrying the pMQ80:creA_-195gfp plasmid were grown overnight in 1× MOPS medium amended with 25 mM pyruvate, 5 mM arginine and 20 µg gentamicin ml⁻¹. Overnight cultures were adjusted to a uniform OD₆₀₀ and added to 1× MOPS medium amended with 20 mM pyruvate, 20 µg gentamicin ml⁻¹ and ±2 mM of each nitrogen-containing compound to a final OD₆₀₀ of 0.5 in a 48-well plate. The compounds creatine, creatinine, sarcosine and glycine betaine were tested for their cahR-dependent induction via the creA promoter. Plates were incubated at 30 °C with periodic shaking for 18 h with readings of the OD₆₀₀ and GFP fluorescence (excitation: 485 nm/emission: 528 nm) taken every hour.

To determine the region of the creA promoter essential for creatine-responsive cahR-dependent induction, induction assays using creA-gfp transcriptional reporters engineered with truncated regions of the creA promoter upstream of gfp were conducted. The assays were conducted as described above with the following adjustments: P. putida KT2440 WT strains carrying the pMQ80:creA_-195-gfp, pMQ80:creA_-169-gfp, pMQ80:creA_-80-gfp or pMQ80:creA_-56-gfp plasmids were induced in 1× MOPS media amended with 20 mM pyruvate, 20 µg gentamicin ml⁻¹, and with or without 2 mM creatine. The full (−195) and −169 creA promoters were tested for cahR-dependence by measuring the transcriptional induction of gfp carried on the respective reporter plasmid in the P. putida KT2440 ∆cahR background. Fold induction was reported using the equation: Fold Induction = (Fluorescence Units_{2mM N-source}/Fluorescence Units_{0mM N}).

MBP-CahR fusion protein expression and affinity purification

A maltose binding protein-CahR fusion (MBP-CahR) was engineered in the pMALc2x vector as previously described [13, 27, 28]. Briefly, the coding region of cahR was amplified with primers PP_3665Exp2ndMet_EcoRI (5′-AAAGAATTCGTCCACCCCGCTTCTGCAAAC-3′) and PP_3665ExpR_HindIII (5′-AAACAAGCTTAATGCCAATACCTGACCCAA-3′). The ~1 kb fragment was then digested with restriction enzymes EcoRI and HindIII (New England Biolabs), purified using Fisher’s GeneJET Gel Extraction and DNA Clean Up Kit (ThermoFisher Scientific), and ligated into the similarly cut pMALc2x vector downstream of the MBP coding region. Cloning and propagation of the pMALc2x:MBP-CahR vector was carried out in E. coli DH5α cells grown in LB broth supplemented with 150 µg carbenicillin ml⁻¹. The pMALc2x:MBP-CahR vector was then transformed into E. coli T7 cells, a strain engineered to support protein expression.

Expression and purification of MBP-CahR was conducted as previously described for MBP-tagged GATR family regulators [13, 27, 28]. Briefly, E. coli T7 cells carrying the pMALc2x:MBP-CahR plasmid were grown in a 50 ml volume of LB broth supplemented with 150 µg carbenicillin ml⁻¹, ith shaking at 170 r.p.m. at 37 °C for 3 h. Protein expression was induced by addition of IPTG to 1 mM followed by an additional 3 h of growth. Cells were collected by centrifugation, frozen at −20 °C overnight, thawed and brought up in 4 ml lysis buffer per 1 g of cells [lysis buffer: 20 mM Tris-HCl pH 7.4, 1 mM EDTA, 200 mM NaCl, 1× Halt protease inhibitor (Thermo), 3 mg lysozyme ml⁻¹]. After a 20 min incubation to allow for lysis, an additional five volumes of lysis buffer was added to further dilute the lysate. Cell lysates were clarified by centrifugation (20 min, 21 000 g at 4 °C) and applied to a column with a 2 ml bed volume of amylose resin. After application of lysate, the column was washed with 10× the resin bed volume with wash buffer (20 mM Tris-HCl pH 7.4, 200 mM NaCl, 1 mM EDTA) and the protein was eluted using one resin bed volume of elution buffer (20 mM Tris-HCl pH 7.4, 350 mM NaCl, 1 mM EDTA, 10 mM maltose). Expression and purification of the MBP-CahR fusion protein was visualized via Coomassie staining of SDS-PAGE gels, and protein concentration of the eluate was measured by UV absorbance on a Nanodrop spectrophotometer with an extinction coefficient of 15. The peak of the MBP-CahR typically eluted in fractions 7 and 8 and the fusion protein was >90 % of the protein in these fractions (data not shown). Aliquots of protein were stored at −80 °C in 20 % glycerol for future use in electrophoretic mobility shift assays (EMSAs).

Electrophoretic mobility shift assays

EMSAs using MBP-CahR protein and biotinylated promoter probes were conducted as previously described for related GATR family regulators [27]. The promoter regions of several genes predicted to be involved in creatine metabolism were amplified with 5′-biotinylation using the primers noted: creA/PP_3665 (CreatinaseProm_F_biotin 5′−5′/Biosg/GGTCTTGGGCATTTGCATGG-3′ and CreatinaseProm_R 5′-GTTTGTCCGAGACTTTGTGGC-3′); glyA1/PP_0297 (PP_0297_F_biotin 5′−5′/Biosg/CAGTACGGAACGGGTCGTAT-3′ and PP_0297_R 5′-GGGTAGCTGCTAGGCTCAAA-3′); and tdcG-I/PP_0322 (PP_3022_F_biotin 5′−5′/Biosg/AAACCATCGATTCAGCACTTG-3′ and PP_3022_R 5′-CCTTTGTGGCGATGTTATGA-3′). An additional probe consisting of the promoter region of the atoA/PP_3122 gene, which is unrelated to creatine metabolism, was tested as a negative control of CahR-promoter binding and was amplified with 5′-biotinylation using primers PP_3122_F_biotin (5′−5′/Biosg/CTGGGCGAAGCTCTGGTACT-3′) and PP_3122_R (5′-TGTTTAACCGACGAGGCTGT-3′). All probes were gel purified using a GeneJET Gel Extraction and DNA Clean Up Kit (ThermoFisher). EMSA binding reactions were conducted with the changes previously described with 1 fmol µl⁻¹ of probe and the replacement of poly(dI-dC) with salmon sperm DNA to a final concentration of 500 µg ml⁻¹ [27]. Binding reactions were incubated for 40 min at 37 °C, then run on a precast 5 % TBE acrylamide gel at 4 °C for 45 min at 100 V (Bio-Rad). After transfer to a Biodyne B Modified Nylon Membrane (ThermoFisher), biotinylated probe was visualized as per the manufacturer’s instructions with changes as previously described using the Pierce Lightshift Chemiluminescent EMSA kit (ThermoFisher Scientific) [27].

Bacterial growth conditions and RNA preparation for RNA-Seq

P. putida KT2440 WT and ∆cahR overnight cultures were grown in 1× MOPS amended with 25 mM pyruvate and 5 mM arginine at 30 °C with shaking at 170 r.p.m. Overnight cultures were washed in 1× MOPS and adjusted to an OD₆₀₀ of 1.0 in 1× MOPS with 20 mM pyruvate, and 600 µl of adjusted culture was added to 600 µl pre-warmed 1× MOPS with 20 mM pyruvate ±2 mM creatine in a 24-well plate. Each P. putida strain was incubated in the ±creatine condition in technical duplicate and biological triplicate. Cultures were incubated at 30 °C with shaking with sample collection at 1 h via centrifugation and resuspension in 600 µl of ~60 °C RNAzol RT (Sigma-Aldrich). After lysing cells by pipetting and vortexing in RNAzol, samples were stored until processing at −80 °C. RNA was extracted and purified from these frozen samples using the RNeasy Mini Extraction Kit (Qiagen) as per the manufacturer’s instructions with the following adjustments: after the initial RNA extraction, the RNA samples underwent a DNaseI treatment and an additional sequential RNeasy purification. Semi-quantitative RT-PCR was conducted using this RNA, Superscript IV (Invitrogen), and primers for creA (PP3667_RTF 5′-cggcaatcttcacctcgtat-3′ and PP3667_RTR 5′- ctggatggcgacgaaatagt-3′) and the control transcript ppiD (PP2304_RTF 5′- agtggtacagccgtttctgg-3′ and PP2304_RTR 5′-tcagttcgttctggctgatg-3′).

RNA-Seq library preparation

Purified total RNA samples were depleted of rRNA using the MICROBExpress Bacterial mRNA Enrichment Kit protocol (ThermoFisher), concentrated via precipitation and resuspension, and mRNA concentrations were measured via a BioAnalyzer. Precipitated and depleted mRNA samples were used for construction of Illumina-compatible single-end libraries using the NEXTflex Rapid Directional mRNA-Seq Bundle - Barcodes 1–24 (BIOO Scientific). Barcoded libraries were submitted to the Vermont Genetics Network sequencing facility at the University of Vermont for generation of read counts via the Illumina HiSeq sequencing system. An average of 11.3 million reads per sample were generated on a HiSeq 1500/2500 single-end 85 bp run.

RNA-Seq data processing and analysis

Quality assessment of raw sequencing data was performed using FastQC (v0.11.6). Adapters and low-quality sequences were removed using Trim Galore! (v0.6.4), removing Illumina adapters, and reads <Q20 and a minimum length of 35 bp. Transcript quantification was performed using Rockhopper2 using the default parameters with verbose output. Reads from each sample were mapped to the reference genome of P. putida strain KT2440 pre-packed with the program.

Differential abundance was calculated using DESeq2 by Group, a feature encompassing genotype and treatment (i.e. WT creatine treatment vs WT without nitrogen). Raw counts were adjusted for library size and genes with fewer than 10 counts in at least two samples were removed from further analysis. Normalization was performed using the default settings of DESeq2 with independent filtering and alpha (false discovery rate, FDR) set to <0.05. Genes displaying greater than a 2-fold log₂ change in transcript levels between conditions were considered differentially expressed and those with a P-value of <0.05 were considered significant. Gene expression data are available in the NCBI GEO database under accession GSE163362.

Phylogenetic tree building

The amino acid FASTA sequences of orthologues of CreA (creatinase) and associated CahR orthologues (GATR-subfamily AraC members) were compiled via the STRING protein database and blast protein searches conducted using the National Centers for Biotechnology Information database [29–34]. Sequences were entered into a phylogenetic tree-building pipeline available on phylogeny.lirmm.fr [35]. This pipeline uses FASTA protein sequences to create a neighbour-joining phylogenetic tree using muscle for sequence alignment and PhyML software for tree building. The sequences of a creatinase and GATR from Pyschrobacter sp. 4Dc were used as the out-groups for their respective trees due to their distance in similarity from the majority of sequences analysed. The strains and accession numbers for the amino acid FASTA sequences used for CreA orthologue tree building and associated CahR accessions [in the form ‘strain (CreA accession, CahR accession]’ are: Psychrobacter sp. 4Dc (WP_193834191.1, WP_101206352.1), Thauera sp. 2A1 (WP_194270409.1, WP_153168678.1), Paraburkholderia hospita BS437 (WP_086918772.1, WP_051495264.1), Paraburkholderia caribensis MBA4 (WP_035989123.1, WP_051453741.1), Caballeronia pedi LMG 29323 (WP_061177555.1, WP_061177577.1), Burkholderia multivorans BCC0099 (WP_006415961.1, WP_048994731.1), Burkholderia pseudomultivorans MSMB368WGS (WP_060239820.1, WP_060239822.1), Burkholderia diffusa BCC0109 (WP_151048639.1, WP_151048638.1), Acinetobacter pittii WCHAP100015 (WP_130136211.1, WP_004832906.1), Acinetobacter baumannii 1106579 (EXE17292.1, EXE17291.1), Vitreoscilla massiliensis (WP_187119852.1, WP_058356168.1), Pseudomonas resinovorans MO-1 (WP_077520507.1, WP_077520504.1), Pseudomonas oryzae KCTC 32247 (WP_090349497.1, WP_090349496.1), Pseudomonas knackmussii B13 (WP_043248253.1, WP_043248255.1), Pseudomonas mendocina EF27 EF27-21 (WP_147810201.1, WP_147810202.1), Acinetobacter baumannii NCTC 7422 contig_0064 (WP_005244916.1, WP_005244913.1), Acinetobacter baumannii 173876 (WP_005244916.1, WP_005244913.1), Acinetobacter baumannii NIPH 601 (ENW47798.1, ENW47797.1,), Acinetobacter baumannii 625974 (EXC04443.1, EXC04444.1), Acinetobacter baumannii 21072 (KCY15186.1, KCY15187.1), Acinetobacter baumannii 1288284 (KCY60723.1, KCY60724.1), Acinetobacter nosocomialis NIPH 386 (ENV39029.1, ENV39028.1), Acinetobacter nosocomialis 6411 plasmid (AJB50085.1, AJB50136.1), Acinetobacter baumannii 1106579 (EXE17292.1, EXE17291.1), Acinetobacter baumannii 83444 (EXE77507.1, EXE77506.1), Acinetobacter baumannii WC-692 (EKA71974.1, EKA71975.1), Acinetobacter baumannii 1429530 (EXB17260.1, EXB17259.1), Acinetobacter baumannii 4749 (EXC10249.1, EXC10250.1), Acinetobacter baumannii AYE plasmid p3ABAYE (WP_001094923.1, WP_000941154.1), Acinetobacter pittii AP43 plasmid pAP43-2 (WP_001094923.1, WP_000941154.1), Pseudomonas putida S16 (AEJ12833.1, AEJ12829.1), Pseudomonas monteilii SB3078 (AHC82230.1, AHC82227.1), Pseudomonas monteilii SB3101 (AHC87608.1, AHC87605.1), Pseudomonas putida DLL-E4 (AHZ77043.1, AHZ77039.1).

Results

Identification of PP_3665 (cahR) as essential for P. putida KT2440 utilization of creatine as a sole nitrogen source

The location of the uncharacterized GATR PP_3665 near the creA gene led to the prediction that it would function to regulate creatine metabolism in P. putida . To test this prediction, we evaluated the WT and the ∆PP_3665 deletion strain’s abilities to grow on various nitrogen sources related to creatine metabolism. After 18 h of incubation, P. putida WT, P. putida ∆PP_3665 and the complemented strain all grew equally efficiently on choline, arginine and sarcosine as the sole nitrogen sources (Fig. 1c). Growth of P. putida ∆PP_3665 on creatine was significantly lower when compared to the WT and complemented strains (P<0.0001), showing no net growth compared to the no-nitrogen control media (Fig. 1c). When supplied with creatinine, the anhydrous form of creatine, as the sole source of nitrogen, all strains had lower growth compared to their growth in choline or to the WT in creatine (P<0.01), but growth of the P. putida ∆PP_3665 strain was not different than the WT in creatinine (Fig. 1c), suggesting general poor growth is probably due to inefficient creatinine utilization and potentially an alternativee route in this strain of P. putida . Based on its essential role in creatine-dependent growth and as a transcription regulator, we named PP_3665 as cahR (creatine aminohydrolase regulator) and use that nomenclature for the remainder of this report.

CahR induces creA transcription in the presence of creatine

Compounds related to the creatine metabolic pathway were tested for their ability to induce gfp in a cahR-dependent manner from a creA promoter-gfp fusion. Significant transcriptional induction was observed in the presence of creatine, ~13 fold over the no-inducer condition (P<0.0001) (Fig. 2a). Induction of the creA promoter in the presence of creatine was also significantly higher in the WT compared to the ∆cahR deletion strain, in which fluorescence was similar to the no-inducer condition, indicating that creatine-dependent transcription induction from the creA promoter is cahR-dependent and that CahR functions as a transcriptional activator.

Fig. 2.

Induction of a GFP transcriptional reporter to the creA promoter region. (a) Creatine strongly induces the P _creA-gfp reporter in wild-type cells (WT) but not in the ∆cahR strain (∆). Statistical significance was tested using ANOVA with Sidak’s post-test comparing the WT to deletion strain within each condition. ****P<0.0001; all other statistical comparisons are noted in (b). (b) Data replotted from (a) but omitting the creatine condition to emphasize small but replicable changes driven by glycine betaine and creatinine. Statistical analysis used ANOVA with Sidak’s post-test comparing all pairs of data. The glycine betaine condition represses expression significantly independent of cahR (a, P<0.001 in comparison to no inducing compound). Creatinine very slightly but replicably induces the reporter (b, P<0.05 in comparison to no inducing compound), which is dependent on cahR (c, P<0.05 in comparison to WT creatinine). (c) Time course of induction from the creA promoter in the wild-type (WT, filled symbols) and ∆cahR (∆, open symbols) in the presence of no inducer (none), sarcosine (sarc), glycine betaine (GB), creatine (cre) and creatinine (crt). The inset DNA gel image is a representative semi-quantitative RT-PCR from the 1 h time point, demonstrating mRNA induction occurs in advance of robust GFP detection. (d) Fold induction of the WT P _creA-gfp reporter to a range of creatine concentrations compared to no creatine, displayed as a box and whisker plot. Statistical analysis was conducted by ANOVA with Dunnett’s post-test comparing each concentration to the zero creatine concentration (this negative control was graphed at 0.5 µM for display) with *P<0.05 and **P<0.01. Error bars in (a)–(c) represent standard deviation, while in (d) they extend to the minimum and maximum values.

A closer look at induction of the creA promoter in the weakly inducing conditions shows that glycine betaine represses expression independent of cahR, while creatinine mildly induces the reporter in a cahR-dependent manner (Fig. 2b). The creatinine data suggest that creatinine might interact with CahR poorly but in a manner that stimulates transcriptional induction or that our commercial creatinine has trace amounts of creatine contamination. Glycine betaine suppression of the creA promoter is similar to GbdR-dependent suppression of alternate GATR-controlled loci in P. aeruginosa [13, 36].

The transcriptional response of the creA promoter to creatine is specific and is only seen in the WT in the presence of creatine, with normalized reporter activity peaking about 4 h after addition of creatine (Fig. 2c). It is important to note that the activity of the reporter lags behind native creA transcript accumulation, which we assessed by semi-quantitative RT-PCR at 1 h post-induction (Fig. 2c, inset). The creA reporter is also sensitive to creatine, showing statistically significant induction above ~30 µM (Fig. 2d).

CahR binds the upstream regulatory region of creA

MBP-CahR binds to the promoter region of creA, shifting the creA promoter probe in a concentration-dependent manner, but not substantially shifting the non-specific P. putida atoA promoter probe or the creatine metabolism-related genes glyA-1 and tdcG-I (Fig. 3). GATR family regulators are often poorly soluble and while we have purified and examined some without epitope tagging [36, 37], fusion of maltose-binding protein (MBP) to the amino terminus enhances solubility and does not alter DNA binding site specificity for other GATR family members [13, 27, 36–38]. Additionally, from those same studies, including those GATRs purified without epitope tags, ligand binding does not alter GATR association with DNA, a property we also confirmed with CahR (data not shown).

Fig. 3.

EMSAs with purified MBP-CahR. Purified MBP-CahR (concentrations noted at the bottom of each lane) was incubated with the biotinylated promoter probes labelled at the top of each blot. Strong and specific shifts were only noted with the creA promoter.

Identification of the CahR binding site in the creA promoter

cahR-dependent transcriptional induction of the creA promoter is highest when the full promoter region (between −25 and −227 bp from the translational start site) is present. When the promoter is truncated to include only −195 bp upstream of the translational start, transcriptional induction drops substantially compared to the full-length construct and is only ~40 % higher than the uninduced condition (Fig. 4a). Truncations of the creA promoter to −113 bp or beyond eliminate all creatine-dependent transcriptional induction of the creA reporter. The creatine-dependent induction of the −227 and −195 bp reporters is also cahR-dependent, with a significant difference in fold induction observed between the reporters in the P. putida WT and ∆cahR strains (−195 bp, P<0.0001; –169 bp, P<0.001) (Fig. 4b).

Fig. 4.

Mapping the probable CahR binding site in the creA promoter. Promoter truncations based on the P _creA-gfp reporter were used to assess the minimal fragments that retain creatine/cahR-dependent induction. (a) Fold induction of four promoter truncations in WT cells. (b) Expression from the two largest promoter truncations in the WT and the cahR deletion (∆). Numbering for all constructs is from the translation start site of creA.

CahR is required for transcription of creatine and sarcosine metabolism-related genes in P. putida in response to creatine

We used RNA-Seq to determine the genes involved in creatine utilization by P. putida . WT cells were exposed to 2 mM creatine or 0 mM creatine in nitrogen-free minimal media for 1 h. There were 22 transcripts differentially induced more than 4-fold in the 2 mM creatine condition compared to the 0 mM creatine control condition, the majority of which are predicted to be involved in creatine and sarcosine metabolism (Fig. 5, Table 1). The gene with the highest induction over the control condition, 1260-fold, is PP_3667/creA, which encodes the known P. putida creatinase. The other member of the creA-containing operon, PP_3666 encoding a putative metabolite MFS transporter, was also among the creatine-responsive genes, induced 362-fold (Fig. 5a, Table 1).

Fig. 5.

Mean-difference plot of transcript abundance comparing WT in creatine over WT with no added nitrogen source. WT P. putida KT2440 was transferred to media with or without creatine as a sole nitrogen source for 1 h, after which RNA was harvested and transcriptomics by RNA-Seq were conducted. Transcripts that met the fold change (log₂ >2) and P-value (<0.05) are shown in blue and labelled where space allowed. All transcripts meeting these two qualifications are listed in Table 1.

Table 1.

Differentially expressed transcripts in WT cells in the absence and presence of creatine

Gene	log2 fold change	log2 mean	Product function/product relationship
PP_3667 creA	10.3	12.5	Creatinase
PP_3666	8.5	9.3	Major facilitator superfamily transporter
PP_0322 glyA-1	8.3	10.7	Serine hydroxymethyltransferase
PP_4665	8.3	8.1	PA2762 orthologue
PP_0297 tdcG-1	8.3	8.9	l-Serine dehydratase
PP_0323 soxB	7.3	9.8	Sarcosine oxidase subunit beta
PP_0324 soxD	6.9	8.4	Sarcosine oxidase subunit delta
PP_0325 soxA	6.6	10.9	Sarcosine oxidase subunit alpha
PP_0326 soxG	6.4	8.2	Sarcosine oxidase subunit gamma
PP_0327 purU-1	5.1	8.8	Formyltetrahydrofolate deformylase
PP_4638	3.7	6.3	Methylenetetrahydrofolate reductase domain-containing protein
PP_0895	3.3	3.3	Hypothetical protein
PP_3526	2.9	6.2	Predicted SouR orthologue
PP_1617 frmC	2.6	6.5	S-formylglutathione hydrolase
PP_1616 frmA	2.6	7.5	d-Isomer specific 2-hydroxyacid dehydrogenase
PP_0896	2.6	3.7	C/N hydrolase; nitrilase/cyanide hydratase
PP_0299	2.5	6.9	Predicted GbdR orthologue
PP_3543	2.5	6.0	(Fe-S)-binding protein
PP_2183 fdhG	2.4	3.7	Formate dehydrogenase subunit gamma
PP_0236 ssuA	2.2	3.5	NAD(P)H-dependent FMN reductase
PP_0315 gbcA	2.1	4.5	Rieske (2Fe-2S) domain-containing protein; GB metabolism
PP_0894	2.1	4.4	NTF2 family protein

It was not surprising that the creatinase gene was the most highly expressed gene in the presence of creatine, as lysis of creatine into urea and sarcosine is generally the first step in bacterial creatine metabolism. The predicted pathway of P. putida creatine metabolism is outlined in Fig. 1(a) and is supported by our differential expression data. Following the hydrolysis of urea from creatine, the resulting sarcosine molecule is oxidatively demethylated into glycine and formaldehyde by the tetrameric sarcosine oxidase encoded by soxBDAG. The sox operon soxBDAG genes are differentially induced between 64- and 128-fold in 2 mM creatine over the control condition. The glycine that results from sarcosine oxidation can then be converted to serine via the glycine hydroxymethyltransferase encoded by glyA-1, which is expressed approximately 256-fold in 2 mM creatine over the control condition. Finally, serine can be converted to pyruvate via serine dehydratase encoded by the tdcG-I gene (orthologue of P. aeruginosa sdaB) that is expressed 256-fold higher in 2 mM creatine as compared to the control. The pyruvate generated from creatine metabolism is then available for conversion to acetyl-CoA and thus into central metabolism. Taken together, the induction of creatine and sarcosine-metabolic genes provides support for the previously predicted pathway of creatine metabolism in P. putida KT2440 (Fig. 1a).

The role of CahR in creatine-responsive gene induction was also elucidated using RNA-seq and differential expression analysis. The 2 mM creatine versus 0 mM creatine comparison was repeated as above, but with the ∆cahR mutant. When the ∆cahR mutant was exposed to 2 mM creatine, the creatine and sarcosine-metabolic genes were no longer differentially expressed over the control condition, leaving only a single gene that met the cut-off criteria used for the WT – the nitrogen fixation-related gene fixG. The lack of creatine metabolic gene induction in the absence of cahR indicates that CahR is required for the transcription of these genes. The lack of induction of the genes encoding downstream metabolic steps, including sarcosine oxidation in the absence of evidence for their direct control by CahR (see Fig. 3), is not surprising as production of sarcosine is dependent upon a CahR-regulated step. We also confirmed that creatine does not induce a sox operon transcriptional reporter in the absence of creA (data not shown). Complete RNA-seq data areavailable at (NCBI GEO Accession currently in submission).

creA orthologues associated with a cahR orthologue are scattered throughout the alpha- and betaproteobacteria

Orthologues of the creA creatinase cluster into two clades, one with the P. putida and related pseudomonads and the other with betaproteobacteria and Acinetobacter (Fig. 6). A number of Acinetobacter species maintain a genomic region that is orthologous to P. knackmussii, P. oryzae and Burkholderia creatinase-coding regions but is present flanked by transposable element boundaries. For some strains this creatinase-containing transposon is on the chromosome, whereas in three others it is plasmid-borne. The presence of creatine-metabolic genes on a transposon maintained in pathogenic bacteria suggests that creatine metabolism may be beneficial during pathogenesis. The presence of creatinase and related metabolic genes in both environmental and pathogenic species suggests a role in both niches, but the strain-specific carriage of these genes and alternative gene organization in otherwise closely related species suggests strongly that these genes are readily acquired via horizontal gene transfer.

Fig. 6.

Phylogenetic tree of CreA amino acid sequence with associated genomic context of the creA gene. Predicted amino acid sequences from creA genes were examined for nearby cahR orthologues and resultant CreA protein sequences were phylogenetically analysed using a muscle alignment of amino acid sequences and PhyML tree reconstruction. Numbers shown next to branches are approximate likelihood ratio scores. Predicted functions for genes in the creA genomic regions are denoted by colour as shown in the key. Abbreviations: serine hydroxymethyl transferase, shmt; transposable element components including transposases and transposable element flanking sequences, transposon; genes present on plasmids in this strain, P in purple text. Scale bar represents average number of amino acid substitutions per site.

Discussion

Creatine is a nitrogen-rich compound (N/C ratio 1 : 1) present in many of the ecological niches occupied by Pseudomonas species, and existence of creatine-responsive metabolic genes in primarily soil-dwelling bacteria that lack a creatininase suggest that creatine is available within the rhizosphere for utilization. Free creatine may come from a variety of sources, including creatinine breakdown by creatininase-possessing microbes, the in situ degradation of creatine-containing animal tissues or excretion of smaller amounts of creatine in animal urine. Regardless of the source, the ability to metabolize creatine would enable bacteria to access a rich nitrogen source and in some cases an alternative source of carbon [10, 39]. The present paper describes identification of a creatine-responsive transcription regulator, CahR, in P. putida that is critical for utilization of creatine as a sole nitrogen source.

cahR is essential for creatine-dependent induction of P. putida creatine utilization genes, and thus growth on creatine as a sole nitrogen source

The AraC/XylS family of transcriptional regulators are a diverse family generally characterized by a helix–turn–helix (HTH) DNA binding C-terminal domain and an N-terminal domain dedicated to dimerization and/or ligand binding [40]. The creatine-responsive transcriptional regulator of P. putida , CahR, belongs to a subset of AraC/XylS-family regulators that contain a glutamine amidotransferase-1 (GATase) domain. Members of the GATase 1-containing AraC transcriptional regulator (GATR) family have been identified in multiple Gram-negative and Gram-positive bacteria, including Pseudomonas species, although functions have only been described for a limited subset of these regulators. Several of the characterized GATRs in P. aeruginosa participate in metabolic regulation of amine-containing compounds such as arginine, glycine betaine, sarcosine and carnitine [13, 27, 41–44]. The P. putida GATR described in this paper, CahR, controls the metabolism of the amine-containing compound creatine via transcriptional induction of the creatinase CreA.

The participation and necessity of cahR in creatine metabolism is demonstrated in Fig. 1(c), where deletion of cahR results in the inability of P. putida to grow on creatine as the sole nitrogen source. CahR binds with specificity to the promoter region of creA (Fig. 3), from which we conclude direct transcriptional induction of creA, which encodes a creatinase with the ability to efficiently cleave creatine into sarcosine and urea [5–10, 45–47]. Transcription of creA occurs quickly in WT P. putida , detectable by the GFP reporter within 1 h after exposure to creatine, and with rapid increase over the first 5 h post-exposure (Fig. 2c). At 1 h post-creatine exposure, an ≈1260 fold-change in transcript levels of creA is observed in P. putida WT in the presence of creatine versus a pyruvate control (Fig. 5, Table 1). This suggests that lysis of creatine by CreA is the preferential pathway of P. putida creatine metabolism, and the rapidity of creatine metabolic induction compared to rates for other GATRs suggests that creatine utilization is probably a beneficial metabolic strategy for P. putida and/or that creatine is a resource under strong competition.

The metabolism of creatine by P. putida creatinase to the intermediate sarcosine has been observed by multiple groups and is supported by the strong transcriptional induction of the predicted sarcosine metabolic genes, orthologous to P. aeruginosa ’s soxBDAG, 1 h post-creatine exposure in P. putida KT2440 (Fig. 5, Table 1) [5–9, 45–47]. Genes involved in the subsequent steps of sarcosine metabolism, including glyA1 encoding the serine hydroxymethyltransferase and tdcG-I encoding the l-serine dehydratase, are amongst the next most highly transcribed genes in the presence of creatine (Fig. 5, Table 1), providing a more complete picture of creatine metabolism in P. putida KT2440, as outlined in Fig. 1(a). Although multiple genes are induced in the presence of creatine, CahR specifically binds to the promoter region of creA alone and not to the promoter regions of the other metabolic genes most highly induced in the presence of creatine, suggesting a small creatine-specific regulon controlled by CahR (Fig. 3). Based on promoter mapping, CahR’s specific binding site lies within the region −227 to −195 bp from the predicted translational start of creA and probably close to or partially overlapping the −195 position. Unfortunately, CahR’s apparent specificity for a single promoter prevented further prediction of a specific CahR binding site, as there is no additional promoter(s) bound by CahR to use in identifying conserved half-site sequences. We did attempt alignments and motif detection between strains and species and also did not identify a potential conserved CahR binding site.

While P. putida KT2440 is able to utilize creatine and the downstream metabolite sarcosine as sole nitrogen sources, it grows poorly on creatinine (Fig. 1c). This is interesting, as creatinine is generally considered a precursor to creatine in the context of bacterial metabolism (Fig. 1a). However, creatine may or may not be an intermediate in creatinine metabolism in P. putida KT2440, as there are alternatives in some P. putida strains, including creatinine metabolism via N-methylhydantoin and N-carbamoylsarcosine intermediates [48]. Thus, the poor growth of P. putida KT2440 on creatinine as a sole nitrogen source, independent of CahR-dependent creatinase induction, may be due to inefficient creatinine utilization via an intermediate that is not creatine [48, 49]. The hypothesis that creatinine, when available, is converted by P. putida KT2440 into an intermediate that is not creatine, such as N-methylhydantion, is also supported by the negligible induction of creA in the presence of creatinine (Fig. 2a, b). Creatinine metabolism by a non-creatine intermediate is also supported by the observation that P. putida KT2440 does not appear to encode any predicted creatinine amidohydrolases, while several other strains of P. putida, including strain S16 (CP002870.1), DLL-E4 (CP007620.1), HB3267 (CP003738.1) and RS56 (AF170566.3), encode creatininases transcribed divergently from cahR and creA orthologues (Fig. 6). The evidence for creatininase function was demonstrated using the cloned and purified enzyme from P. putida strain RS56 [50]. In addition to the P. putida strains that encode both creatininases and creatinases, there are several P. monteilii strains, a close relative of P. putida , which share syntentic creatininase/creatinase genomic regions, including P. monteilii SB3101 (CP006979.1) and SB3078 (CP006978.1). P. monteilii has been implicated in several opportunistic infections, while P. putida HB3267, a strain isolated from hospitalized patients and also containing this gene arrangement, shows cytolytic activity against human cells [51, 52].

It is also interesting to note that glycine betaine inhibits basal transcription of creA independent of CahR. This may be indicative of an inhibitory feedback mechanism perpetuated by downstream products of creatine metabolism or the direct or tangential involvement of other compound-specific regulators in creatine metabolism. In P. aeruginosa , the glycine betaine/dimethylglycine sensing regulator GbdR is able to repress activation from promoters co-regulated with other GATR family members when glycine betaine is present, best described at the carnitine operon promoter that is induced in a carnitine-dependent manner by the GATR member CdhR [36]. Increased transcription of GbdR and glycine betaine metabolic genes is also observed in WT P. putida KT2440 in the presence of creatine, which supports potential interplay between P. putida GATRs (Fig. 5).

Conservation of creA and cahR synteny illustrate the potential utility of this genetic module in creatine-rich environments

Pathways for creatinine and creatine metabolism are conserved among diverse bacteria. Examining sequence similarity and genomic organization between bacteria, the presence of a CahR-regulated creatine-inducible creatinase appears to be conserved among several species, based on the presence of creA orthologues co-occuring with predicted cahR orthologues in similar genomic organization as P. putida (genomic organization in P. putida KT2440 shown in Fig. 1b). The phylogenetic tree in Fig. 6 is for CreA, while a tree of the associated CahR orthologues has a very similar topology but, as expected for a regulator compared to an enzyme, shows substantially longer branch lengths (data not shown).

The presence of a creA orthologue with an associated cahR orthologue in a limited number of strains within a given species supports a model of horizontal gene transfer of creatinase-responsive metabolic genes between bacteria. The case of cahR and creA orthologues among Acinetobacter species provides the most compelling support for horizontal transfer of cahR and creA. The clade with the shortest branch lengths consists exclusively of Acinetobacter species, including A. baumannii , A. nosocomalis and A. pittii . These Acinetobacter species are members of the Acinetobacter calcoaceticus–Acinetobacter baumannii complex and are known for their ability to cause persistent, multidrug-resistant, nosocomial infections in humans [53, 54]. Of the Acinetobacter isolates possessing cahR/creA orthologues, several were isolated from urine, which is the primary vehicle of creatine excretion in animals [8]. Additionally, all of the Acinetobacter clinical isolates analysed possess cahR/creA orthologues flanked by transposon or integrase flanking sequences, suggesting that the creatine-responsive creatine-metabolic enzymes are part of a transposable element horizontally transferred amongst pathogenic Acinetobacter species.

The majority of the cahR/creA-containing transposons present in the Acinetobacter species are identical, containing cahR, creA and PP_3666 (encoding the MFS transporter) flanked by transposase/integrase flanking regions, while several species include additional genes related to creatine metabolism such as urease genes, within the putative transposable element (Fig. 6). In three cases, the cahR/creA-containing transposon is present on a plasmid that is maintained by Acinetobacter . The strains possessing these plasmids, A. baumannii AYE (SAMEA3138279), A. pittii AP43 (SAMN12612836) and A. nosocomalis 6411 (SAMN03263968), cluster together (Fig. 6). The AP43 plasmid also carries virulence factor blaNDM-1, which confers resistance to carbepenems and cephalosporins, while AYE is associated with multidrug-resistant community-acquired infections [55]. The maintenance of the cahR/creA orthologues, both on plasmids and within transposable elements, suggests that these genes were probably acquired from other sources, such as cahR/creA-possessing gut microbiome members Vitreoscilla massiliensis or Thauera 2A1, or from one of the many environmental or opportunistic pathogen species that share Acinetobacter ’s niches [56, 57]. While the existence of cahR/creA-containing transposons and plasmids suggests that acquisition and maintenance of these genes is advantageous to Acinetobacter ’s lifestyle, the potential benefits of this region for bacterial survival and virulence have yet to be evaluated.

Conclusions

Here we have described the identification of a creatine-responsive transcription regulator, CahR, that is necessary for creatine utilization by regulating creatinase gene induction. There are a number of issues raised by the data presented here that remain to be explored. Based on our count this is now the fourth GATR for which an inducing ligand is known, yet we still do not understand how ligands are bound and how specificity is determined. These GATRs all control organic nitrogen compound utilization and are probably not-so-ancient paralogues that diversified for specialization to structurally related but different small molecules. Thus, the GATRs might provide a good model to understand the evolution of substrate specificity for transcription regulators.