Identification of a Two-Component Regulatory Pathway Essential for Mn(II) Oxidation in Pseudomonas putida GB-1

Abstract

Bacterial manganese(II) oxidation has a profound impact on the biogeochemical cycling of Mn and the availability of the trace metals adsorbed to the surfaces of solid Mn(III, IV) oxides. The Mn(II) oxidase enzyme was tentatively identified in Pseudomonas putida GB-1 via transposon mutagenesis: the mutant strain GB-1-007, which fails to oxidize Mn(II), harbors a transposon insertion in the gene cumA. cumA encodes a putative multicopper oxidase (MCO), a class of enzymes implicated in Mn(II) oxidation in other bacterial species. However, we show here that an in-frame deletion of cumA did not affect Mn(II) oxidation. Through complementation analysis of the oxidation defect in GB-1-007 with a cosmid library and subsequent sequencing of candidate genes we show the causative mutation to be a frameshift within the mnxS1 gene that encodes a putative sensor histidine kinase. The frameshift mutation results in a truncated protein lacking the kinase domain. Multicopy expression of mnxS1 restored Mn(II) oxidation to GB-1-007 and in-frame deletion of mnxS1 resulted in a loss of oxidation in the wild-type strain. These results clearly demonstrated that the oxidation defect of GB-1-007 is due to disruption of mnxS1, not cumA::Tn5, and that CumA is not the Mn(II) oxidase. mnxS1 is located upstream of a second sensor histidine kinase gene, mnxS2, and a response regulator gene, mnxR. In-frame deletions of each of these genes also led to the loss of Mn(II) oxidation. Therefore, we conclude that the MnxS1/MnxS2/MnxR two-component regulatory pathway is essential for Mn(II) oxidation in P. putida GB-1.

In living cells, manganese (Mn) is an essential trace element, required for enzymes such as superoxide dismutase and in photosystem II (7). In the environment, Mn cycles between a soluble reduced form [Mn(II)] and an insoluble oxidized form [Mn(III, IV)] that can adsorb other trace metals from the environment and serve as potent oxidizing agents. Thus, redox cycling of Mn has a profound effect on the bioavailability and geochemical cycling of many essential or toxic elements (40). Microorganisms, particularly bacteria, are capable of catalyzing the oxidation of Mn(II), thereby increasing the rate of formation of Mn(III, IV) by several orders of magnitude (39). Since Mn(III, IV) oxides are able to bind trace metals, the bacteria that catalyze their formation are good candidates for bioremediation of heavy metal contaminated sites (26, 39).

Although bacterial Mn(II) oxidation is widespread, little is known about the physiological function of oxidation (40). The oxidation of Mn(II) to Mn(III) or Mn(IV) is thermodynamically favorable; thus, bacteria may derive energy from this reaction, although this has never been unequivocally proven (40). In addition, Mn(II) oxidation could protect cells from reactive oxygen species (4) or UV irradiation (11). Since oxidation occurs on the cell surface, the bacteria become coated with the solid Mn(IV) oxides, which may also provide protection from toxic heavy metals, predation, or phage infection (40). As a strong oxidant, Mn(IV) oxides could allow the bacteria to degrade refractory organic matter to low-molecular-weight compounds that could then be used to support bacterial growth (38). Conversely, Mn(II) oxidation may be a side reaction or the result of nonspecific interactions with cellular products (15). Identifying signals or conditions that regulate oxidation could provide some insight into the role of Mn(II) oxidation in the cell. Aside from a requirement for oxygen (28) and iron (27, 30), as well as the observation that oxidation occurs in stationary phase (23), very little is known about this regulation.

The enzymes responsible for Mn(II) oxidation have been tentatively identified from some species of bacteria and in several cases the enzyme is a putative multicopper oxidase (MCO). MCOs are a family of enzymes that use four Cu ion cofactors to catalyze oxidation of diverse substrates such as metals and organic compounds (33). This family of enzymes is found in plants and fungi (laccase) and humans (ceruloplasmin), as well as in bacteria (35). Some fungi have been shown to use a laccase enzyme to oxidize Mn(II) (20). In both Leptothrix discophora SS-1 and Pedomicrobium sp. strain ACM 3067, the Mn(II)-oxidizing MCO was identified genetically (mofA [10] and moxA [31], respectively). A third MCO—MnxG—was identified both biochemically and genetically as the Mn(II) oxidase in Bacillus sp. strain SG-1 and related strains (14, 43). Recent work with the Mn(II)-oxidizing alphaproteobacterium Aurantimonas manganoxydans SI85-9A1 and Erythrobacter sp. strain SD21 has identified a second class of enzyme involved in Mn(II) oxidation: the heme-binding peroxidase named MopA (3). This class of enzyme had previously been shown to be used by fungi to oxidize Mn(II) (29), in some cases in concert with an MCO (34).

Pseudomonas putida GB-1 is a Mn(II)-oxidizing bacterium (9) whose genetic tractability and ease of growth under standard laboratory conditions make it an ideal model system for studying the physiology and mechanism of Mn(II) oxidation. Consequently, several random transposon mutagenesis screens have been undertaken with this organism to identify genes required for Mn(II) oxidation. These screens have identified several categories of genes as important for oxidation or the export of the oxidase to the cell surface: the ccm operon of c-type cytochrome synthesis genes (8, 13), genes encoding components of the trichloroacetic acid (TCA) cycle and the tryptophan biosynthesis pathway (8) and genes encoding a general secretory pathway (12). The Mn(II) oxidation-defective mutant GB-1-007 has a transposon insertion in the gene cumA that encodes a putative MCO (6). Therefore, P. putida GB-1 has been thought to use a similar mechanism as L. discophora SS-1, Pedomicrobium sp. strain ACM 3067, and Bacillus sp. to oxidize Mn(II).

Because the available data suggested that CumA was an MCO essential for Mn(II) oxidation, we wanted to study its function in greater detail. We were hampered in this, however, by the fact that the transposon insertion in cumA resulted in a growth defect due to its polar effect on expression of the downstream cumB gene (6). In order to assess the role of CumA in Mn(II) oxidation without the complications arising from polarity, we generated an in-frame deletion of cumA and tested the ability of the resulting ΔcumA strain to form Mn(IV) oxides. Our results showed that cumA is dispensable for Mn(II) oxidation and have instead revealed a complex two-component regulatory pathway essential for Mn(II) oxidation in P. putida GB-1.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The strains used in the present study are described in Table 1. Escherichia coli strains were grown on Luria-Bertani (LB) medium containing antibiotics as needed. P. putida GB-1 strains were grown on LB medium with antibiotics, if needed, or Leptothrix medium (Lept) supplemented with 100 μM MnCl₂ (5). E. coli strains were grown at 37°C, while P. putida GB-1 strains were grown either at room temperature or at 30°C. Antibiotics were added to the following concentrations (μg ml⁻¹): ampicillin (Ap), 100; kanamycin (Km), 30; and gentamicin (Gm), 50.

TABLE 1.

Strains and plasmids used in the study

Strain or plasmid	Genotype, characteristics, or constructiona	Source or reference
Strains
E. coli TAM1	mcrA Δ(mrr-hsdRMS-mcrBC) φ80lacZΔM15 ΔlacX74 recA1 araD139 (ara-leu)7697 galU galK rpsL endA1 nupG	Active motif
P. putida
GB-1	Wild type	9
GB-1-002	Spontaneous mutant of GB-1; Smr	13
GB-1-007	Non-Mn2+ oxidizing Tn5 mutant of GB-1-002; Smr Kmr	13
KG60	GB-1 ΔcumA	This study
KG67	GB-1-002 ΔcumA	This study
KG68	GB-1-007 ΔcumA	This study
KG127	GB-1 ΔPputGBI_2521	This study
KG129	GB-1 ΔPputGBI_2519	This study
KG133	GB-1 ΔPputGBI_2520	This study
Plasmids
Cosmid 2A	Cosmid from the screen of a cosmid library for complementation of the GB-1-007 oxidation defect	This study and reference 13
pEX18Gm	Gene replacement vector; Gmr; oriT sacB	19
pJET1.2/blunt	Commercial cloning vector	Fermentas
pRK2013	Helper plasmid for conjugation	16
pUCP22	Broad-host-range vector, Gmr; ColE1	44
pKG169	The ∼500 bp upstream of cumA fused to the ∼500 bp downstream and cloned into pEX18Gm, for generating an in-frame deletion	This study
pKG179	EcoRI fragment from cosmid 2A cloned into EcoRI-cut pUCP22; carries PputGB1_2525 through PputGB1_2534	This study
pKG181	EcoRI/HindIII fragment from cosmid 2A cloned into EcoRI/HindIII-cut pUCP22; carries PputGB1_2511 through PputGB1_2517	This study
pKG206	pUCP22 carrying mnxS2 under the control of PlacZ	This study
pKG207	pUCP22 carrying mnxR under the control of PlacZ	This study
pKG208	pUCP22 carrying mnxS1 under the control of PlacZ	This study
pKG210	The ∼500 bp upstream of PputGB1_2520 fused to the ∼500 bp downstream and cloned into pEX18Gm, for generating an in-frame deletion	This study
pKG211	The ∼500 bp upstream of PputGB1_2521 fused to the ∼500 bp downstream and cloned into pEX18Gm, for generating an in-frame deletion	This study
pKG212	The ∼500 bp upstream of PputGB1_2519 fused to the ∼500 bp downstream and cloned into pEX18Gm, for generating an in-frame deletion	This study
pKG216	XhoI fragment from cosmid 2A cloned into XhoI-cut pUCP22; carries PputGB1_2517 through PputGB1_2524	This study

^aGm^r, gentamicin resistance; Sm^r, streptomycin resistance; Km^r, kanamycin resistance.

Determination of Mn(II) oxidation.

To assess Mn(II) oxidation by P. putida GB-1 strains, the bacteria were grown overnight in LB liquid containing antibiotics, if necessary, to maintain the presence of plasmids. Portions (5 μl) of the overnight cultures were then spotted onto Lept plates supplemented with 100 μM MnCl₂. No antibiotics were added to the Lept plates because antibiotics have been shown to interfere with oxidation (8, 32). After incubation at 30°C for 2 to 3 days, the plates were photographed to record the extent of oxidation.

Conjugation.

The cosmid library and the deletion constructs were mobilized to P. putida GB-1 cells via triparental mating using the helper plasmid pRK2013 (Table 1). Mid-exponential-phase cultures of donor, helper, and recipient strains were mixed at a ratio of 2:1:2 and pelleted, and the resuspended pellets were plated on LB overnight at room temperature. Next, the cells were scraped from the LB plate and resuspended in 1 ml of LB medium. Conjugants were selected by plating on Pseudomonas isolation agar (Fluka Analytical/Sigma Aldrich, St. Louis, MO) containing Gm.

Generation of in-frame deletions.

The deletion constructs were generated by first amplifying ∼500 bp upstream and downstream of the target gene (for primers, see Table 2) using a high-fidelity DNA polymerase (Phusion Hot Start high-fidelity DNA polymerase (New England Biolabs, Ipswich, MA). Performing the PCR with “junction” primers that fuse the sequence at the 5′ end of the gene in frame to the sequence at the 3′ end resulted in two PCR products that share ∼40-bp homology at one end. These two PCR products were then joined by PCR splicing by overlap extension (21), resulting in an ∼1-kb PCR product in which the upstream and downstream regions of the gene were fused. This deletion construct was then cloned into the plasmid pEX18Gm (Table 1), which cannot be stably maintained in P. putida GB-1. After conjugation of the plasmid into P. putida GB-1, conjugants were screened for gentamicin resistance (Gm^r) and sucrose sensitivity. Such colonies represent those that had undergone a single recombination event, resulting in integration of the plasmid into the chromosome. Colonies were then streaked onto LB containing 5% sucrose and selected for sucrose resistance. Sucrose-resistant colonies were screened for Gm sensitivity; such colonies represent those that had undergone a second recombination event, removing the plasmid backbone from the chromosome. To determine whether the gene had been successfully deleted, genomic DNA was isolated from candidate strains (UltraClean tissue and cell DNA isolation kit; Mo Bio Laboratories, Carlsbad, CA), and PCR was performed using the upstream and downstream flanking primers. The presence of a truncated PCR product indicated that the gene had been deleted.

TABLE 2.

Primers used in this study

Primer	Sequence (5′-3′)
cumA_upstream-F	CATCCTGCGTCGCGCCGAC
cumA_downstream-R	GGGGTGAACAAGGACAAGCCGGC
cumA_junction-F	CGTCGACAAATGCTCAAGGGCGCCGCGATCGCGGTGGTC
cumA_junction-R	GACCACCGCGATCGCGGCGCCCTTGAGCATTTGTCGACG
2519_1-F	GGCGGCGCGCATTCACAATG
2519_2-R	TGGCCTGAATGAACTGCGCGG
2519_3-F	TGGAGTCCACCATGGACCTCCGCTTGGCTAATACGCTCTGAATTACATTAAG
2519_4-R	CTTAATGTAATTCAGAGCGTATTAGCCAAGCGGAGGTCCATGGTGGACTCCA
2519_5-F	CCGGCCCTTATCGGCAAGC
2519_6-F	GTGCCCAGCAGTGGAGTC
2519_7-R	CCCCACATTTGGGTAATTTCCACC
2520_1-F	GGGCATGAACGAAGCGCAGC
2520_2-R	GTCTCGCCTTCGACCAGCACTG
2520_7-F	GAACCACGCGATGCAGATGCTGGGAGGGGACCATGGAGCACAG
2520_8-R	CTGTGCTCCATGGTCCCCTCCCAGCATCTGCATCGCGTGGTTC
2520_9-F	GCAATTCTGAACGGACGCCCTG
2520_10-R	TAACCTCGTAGCCCTTGAGGC
2521_1-F	GGTGGACCGCATGTCGCG
2521_2-R	CATCATGTTGCTGTCGCTGGTGGTG
2521_7-F	GTCGAGGATGATGAAATCCTCGCCCGTCAGTGGTAATCATGGCCATG
2521_8-R	CATGGCCATGATTACCACTGACGGGCGAGGATTTCATCATCCTCGAC
2521_9-F	CAATATCGCAGCGGGAGGGG
2521_10-R	CTTTCGCCGCACAAGGCC

Plasmid constructs for complementation.

The gene of interest was PCR amplified (primers, Table 2) using a high-fidelity DNA polymerase (Phusion Hot Start high-fidelity DNA polymerase), and the resulting PCR product was cloned into pJET1.2/blunt (CloneJet PCR cloning kit; Fermentas, Glen Burnie, MD). The genes were subsequently subcloned into the broad-host-range plasmid pUCP22 (Table 1). The mnxS2 gene was amplified with 2520_9-F and 2520_10-R and ligated into pJET1.2/blunt. It was then subcloned by digestion with XbaI and XhoI and ligation of the resulting fragment into pUCP22 digested with XbaI and SalI. The mnxR gene was amplified with 2521_9-F and 2521_10-R and then subcloned in a similar manner. The mnxS1 gene was amplified with 2519_9-F and 2519_10-R and ligated into pJET1.2/blunt. It was then subcloned by digestion with BglII and ligation of the resulting fragment into pUCP22 digested with BamHI. The orientation of the mnxS1 insert was determined by restriction digestion. The genes inserted into pUCP22 were expressed constitutively from the plasmid-borne promoter P_lac.

Transformation of P. putida GB-1.

P. putida GB-1, or derivatives thereof, was grown overnight in LB medium at room temperature. The cells were made competent by washing 1 ml of overnight culture with 1 ml of ice-cold 100 mM MgCl₂, followed by incubation in 1 ml of TG salts (75 mM CaCl₂, 6 mM MgCl₂, 15% glycerol) on ice for 10 min. The cells were then pelleted and resuspended in 200 μl of TG salts. To transform the cells, 2 to 4 μl of plasmid DNA was added to 100 μl of cells, and the cell-DNA mixture was incubated on ice 10 to 30 min. Next, the cell-DNA mix was incubated at 37°C for 2 min before being placed back on ice for 2 more minutes. Next, 800 μl of LB was added to each mix before incubation at 30°C for 1 to 1.5 h. Finally, the cell mixture was pelleted and plated on LB plus Gm, followed by incubation at 30°C.

Sequencing GB-1-007.

Genomic DNA was isolated from GB-1-007 by using an UltraClean tissue and cell DNA isolation kit (Mo Bio Laboratories), and the region containing PputGB1_2519 through PputGB1_2521 was PCR amplified by using Phusion Hot Start high-fidelity DNA polymerase and the primers 2519_1-F and 2521_2-R (Table 2). The PCR product was cloned into pJET1.2/blunt, and plasmid DNA was isolated from two separate clones (Qiaprep spin miniprep kit; Qiagen, Valencia, CA). The entire insert region was sequenced from each plasmid using the Molecular Biology Core Sequencing Facility of the Oregon Health & Science University. The presence of the frameshift mutation in PputGB1_2519 was confirmed by sequencing the gene from two additional clones derived from an independent PCR.

RESULTS

cumA is dispensable for Mn(II) oxidation.

GB-1-007 has a growth defect that was attributed to polarity of the transposon insertion on the downstream gene cumB (6), which encodes a putative cytosine/adenosine deaminase. In order to study the role of CumA in Mn(II) oxidation without this growth defect, we generated a strain of P. putida GB-1 carrying an in-frame cumA deletion. The resulting strain should lack cumA but have its cumB expression unaltered. In agreement with this, the ΔcumA mutant does not have the growth defect associated with GB-1-007 (data not shown). To assess the ability of ΔcumA to oxidize Mn(II), we plated the ΔcumA strain on a solid medium containing MnCl₂. On this medium, P. putida GB-1 produces orange/brown colonies, as a result of the deposition of Mn(IV) oxides on the cell surface, while GB-1-007 fails to oxidize Mn(II) and forms white colonies (Fig. 1A). The ΔcumA deletion strain formed brown colonies that were indistinguishable from wild type (Fig. 1A). Thus, the two cumA alleles cumA::Tn5 and ΔcumA do not produce the same phenotype, calling into question the relationship between the Tn5 insertion in cumA and the oxidation defect of GB-1-007.

FIG. 1.

(A) Deletion of cumA does not affect Mn(II) oxidation. The indicated strains were spotted onto Lept plates containing 100 μM MnCl₂, followed by incubation at 30°C for 48 h. Each strain was spotted in duplicate. (B) Complementation of GB-1-007. “Empty vector” refers to pUCP22, the plasmid backbone used in all of the complementation constructs.

A possible explanation for the difference between cumA::Tn5 and ΔcumA genotypes is that the experiments were performed in different strain backgrounds; GB-1-007 was generated in a spontaneous streptomycin-resistant derivative of GB-1 called GB-1-002 (6), whereas our cumA deletion was generated in GB-1. It is technically possible that CumA is required for Mn(II) oxidation in GB-1-002 but not GB-1. To test this, we deleted cumA from GB-1-002. The resulting GB-1-002 ΔcumA strain oxidized Mn(II) to the same extent as GB-1-002 (Fig. 1A). Thus, differences in strain background do not explain the difference between GB-1-007 and the GB-1 ΔcumA strain.

Another explanation for the discrepancy between GB-1-007 and the ΔcumA mutant is that the transposon insertion in GB-1-007 results in a truncated CumA protein whose activity interferes with Mn(II) oxidation; in other words, cumA::Tn5 is a dominant-negative allele. The transposon insertion in cumA::Tn5 is located in the 3′ end of the gene, between the third and fourth copper-binding motifs of this putative MCO (6). Such an insertion could leave the bulk of the protein intact, with an altered and possibly truncated carboxy terminus. If the failure to oxidize Mn(II) is due to a dominant-negative effect of the mutant CumA protein, deleting cumA::Tn5 from GB-1-007 should alleviate this effect. However, Mn(II) oxidation by GB-1-007 ΔcumA was indistinguishable from GB-1-007 (Fig. 1A). Thus, the truncated CumA is not required to inhibit Mn(II) oxidation.

Complementation of the Mn(II) oxidation defect of GB-1-007.

Given our results above, the most likely explanation for the difference in phenotype between the GB-1-007 and ΔcumA strains is the existence of an unmarked second site mutation in GB-1-007 that is responsible for its oxidation defect, while the transposon insertion in cumA has little, if any, effect on Mn(II) oxidation. In order to map the location of the mutation responsible for the oxidation defect of GB-1-007, we screened a library of cosmid clones generated from GB-1-002 (13) for cosmids that restore Mn(II) oxidation. The cosmid library was moved via conjugation into the GB-1-007 ΔcumA strain, and the resulting transconjugants were screened for their ability to oxidize Mn(II) on solid medium. GB-1-007 ΔcumA was used because of its less severe growth defect. From this screen, we isolated a cosmid with an ∼27-kb insert that restored oxidation (cosmid 2A, Fig. 1B). We sequenced the endpoints of the insert present on the cosmid and showed that it carried a region of the GB-1 genome encompassing the genes PputGB1_2511 through 2534 (Table 3). cumA (PputGB1_1031) is not present on the complementing cosmid.

TABLE 3.

Genes present on cosmid 2A^a

Locus taga	Annotationb
PputGB1_2511	Efflux transporter RND family
PputGB1_2512	Rhodanese domain protein
PputGB1_2513	Cysteine dioxygenase type I
PputGB1_2514	Aliphatic sulfonate family ABC transporter
PputGB1_2515	Nitrilotriacetate monooxygenase component A
PputGB1_2516	Type II secretion system protein
PputGB1_2517*	xcmT1, GB-1-009
PputGB1_2518	Lytic transglycosylase
PputGB1_2519	Sensor histidine kinase
PputGB1_2520*	Sensor histidine kinase psk2, GB-1-005
PputGB1_2521	σ54-Dependent response regulator
PputGB1_2522	Hypothetical protein
PputGB1_2523	Gluconate transporter
PputGB1_2524	Gluconate kinase
PputGB1_2525	LacI family transcription regulator
PputGB1_2526	Chemotaxis sensor
PputGB1_2527	Hybrid sensor histidine kinase
PputGB1_2528	LuxR family response regulator
PputGB1_2529	DUF1272
PputGB1_2530	Precorrin-4 C11-methyltransferase
PputGB1_2531	Cobalamin biosynthesis protein CobE
PputGB1_2532	Cobalt transporter subunit CbtA
PputGB1_2533	Cobalt transporter subunit CbtB
PputGB1_2534	Alkanesulfonate monooxygenase

^aThe genes indicated by an asterisk (*) were previously identified through transposon mutagenesis as affecting Mn(II) oxidation and their annotation follows that in reference 12.

^bGene annotations are from the Integrated Microbial Genomes website (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi) (25).

Two other oxidation-defective strains have transposon insertions mapped to the region on the cosmid (Table 3) (12). GB-1-009 is defective in the secretion of the Mn(II) oxidase activity and has a transposon inserted in the general secretion pathway gene xcmT1 (PputGB1_2517). GB-1-005 fails to oxidize Mn(II) and has a transposon inserted in the sensor histidine kinase gene psk2 (PputGB1_2520). To identify the gene mutated in GB-1-007, we subcloned portions of the cosmid insert into the multicopy plasmid pUCP22 and screened for the ability of the resulting plasmids to restore Mn(II) oxidation. Only the plasmid carrying the region from PputGB1_2517 through PputGB1_2524 restored Mn(II) oxidation, while those carrying PputGB1_2511 to PputGB1_2518 and PputGB1_2525 to PputGB1_2534 did not (Fig. 1B). Thus, the mutation in GB-1-007 likely falls between PputGB1_2519 and PputGB1_2524.

These results, coupled with the similarity in phenotype between GB-1-005 and GB-1-007, led us to focus on the central portion of the cosmid insert. In particular, we focused on the genes PputGB1_2519 through PputGB1_2521 (Fig. 2A). These genes encode two sensor histidine kinases (PputGB1_2519 and PputGB1_2520) and a response regulator (PputGB1_2521), suggesting they may work together in a two-component regulatory pathway required for Mn(II) oxidation. Therefore, we sequenced the region from ∼500 bp upstream of PputGB1_2519 to ∼500 bp downstream of PputGB1_2521 from GB-1-007 and discovered an insertion of an extra cytidine residue within a run of eight C residues within PputGB1_2519 (Fig. 2B) relative to the published P. putida GB-1 sequence. The C insertion was the only mutation present in the region. As a control, we also sequenced this region from the strain GB-1 and confirmed that the wild-type genome encodes eight C residues at this location. In support of the identification of the C insertion as the cause of its oxidation defect, multicopy expression of PputGB1_2519 restored Mn(II) oxidation to GB-1-007 (Fig. 2C).

FIG. 2.

Mapping the causative mutation in GB-1-007. (A) Diagram depicting the three putative TCR genes located on the complementing plasmid. Sensor histidine kinase genes are shown in red block arrows, while the response regulator gene is in blue. The numbers within the genes are their locus tags, while in italics above are the proposed new gene names. Also shown is the spacing between the genes. (B) Alignment of nucleotides 2830193 to 2830201 from the P. putida GB-1 sequence with the corresponding region from GB-1-007. Shown beneath the DNA is the amino acid sequence of the region. Shown in boldface is the inserted C nucleotide. (C) Complementation of GB-1-007 with p2519 and the phenotype of Δ2519. The assay was performed as described for Fig. 1A.

The ninth C residue present in GB-1-007 results in a frameshift mutation at amino acid 188 and a premature stop codon after 30 additional amino acids. PputGB1_2519 encodes a 486-amino-acid sensor histidine kinase with two predicted transmembrane helices (41, 42; http://www.enzim.hu/hmmtop/html/submit.html) that together delineate a 171-amino-acid periplasmic sensor domain and a cytoplasmic histidine kinase/ATPase domain. The GB-1-007 mutation results in a truncation of the protein within the periplasmic domain, and thus the truncated protein lacks the kinase domain. Therefore, we predict this is a null mutation. In agreement with this prediction, a strain with an in-frame deletion of PputGB1_2519 had the same oxidation phenotype as GB-1-007 (Fig. 2C).

The Mnx two-component regulatory pathway is required for Mn(II) oxidation.

The identification of two mutations within the PputGB1_2519-2521 region that result in a loss of Mn(II) oxidation suggests that these genes play a central role in regulating Mn(II) oxidation. The group that first isolated GB-1-005 and mapped its transposon insertion to PputGB1_2520 named these genes psk1, psk2, and ptr (for protein sensor kinases 1 and 2 and the phosphorylation transcription regulator [12]). Because these genes are required for Mn(II) oxidation, we propose renaming them mnxS for Mn(II) oxidation sensor and mnxR for Mn(II) oxidation regulator. Accordingly, PputGB1_2519, or psk1, is now mnxS1. PputGB1_2520, or psk2, is mnxS2, and PputGB1_2521, formerly ptr, is now mnxR.

The strain GB-1-005 has a transposon insertion in mnxS2 that is likely polar on mnxR because the two genes overlap by 31 nucleotides (Fig. 2A). As a result, the loss of oxidation exhibited by this strain may be due to disruption of mnxS2, mnxR, or both. To examine the role of these two regulators, we generated in-frame deletions of each gene independently. In each case, the in-frame deletion resulted in a loss of Mn(II) oxidation (Fig. 3). Oxidation could be restored in the deletion strains by complementation with the deleted gene on pUCP22 (Fig. 3). Therefore, mnxS1, mnxS2, and mnxR are each required for Mn(II) oxidation.

FIG. 3.

Individual roles of the Mnx two-component regulatory genes in Mn(II) oxidation. The assay was performed as described for Fig. 1A except that the photographs were taken after 3 days of incubation at 30°C.

Sensor kinase/response regulator cognate pairs are frequently found with their genes present in adjacent loci on the chromosome (1). Thus, the cognate response regulator for MnxS1, MnxS2, or both may be MnxR (Fig. 2A). If MnxR is downstream of either sensor in the regulatory pathway, we predicted that multicopy expression of mnxR would restore oxidation to strains defective in either sensor histidine kinase. Therefore, we examined the effect of the mnxR plasmid on Mn(II) oxidation by the ΔmnxS1 and ΔmnxS2 strains. The mnxS2 deletion was clearly complemented by MnxR (Fig. 3), supporting the conclusion that these two proteins work together. However, in the case of the mnxS1 deletion, the mnxR plasmid had little effect on Mn(II) oxidation.

DISCUSSION

In this study we have shown that the Mn(II) oxidation defect exhibited by strain GB-1-007 is not due to the transposon insertion present in this strain but instead is caused by a point mutation in the sensor histidine kinase gene mnxS1. Our original aim was to use GB-1-007 to study one of the key enzymes required for P. putida GB-1 to oxidize Mn(II), the putative MCO encoded by cumA. Instead, our investigation of this mutant strain has uncovered a two-component regulatory pathway that is essential for Mn(II) oxidation.

CumA is not required for Mn(II) oxidation. The deletion of cumA did not affect oxidation, regardless of strain background (Fig. 1A). Previous work showed that the oxidation defect of GB-1-007 could not be complemented by plasmids carrying cumA (6). Furthermore, we showed that the mnxS1 gene of GB-1-007 carried a frameshift mutation not present in GB-1 and that deletion of mnxS1 from GB-1 resulted in a defective oxidation phenotype similar to that of GB-1-007. Previous work from our group failed to detect a correlation between the presence of cumA and the ability to oxidize Mn(II) in various Pseudomonas strains (17). In light of our results here, the presence of cumA in nonoxidizing Pseudomonas strains further supports the conclusion that CumA is not required for Mn(II) oxidation.

The Mn(II) oxidase enzymes in Bacillus sp. strain SG-1, L. discophora SS-1, and Pedomicrobium sp. strain ACM 3067 have been identified as MCO enzymes (10, 14, 31); therefore, does P. putida GB-1 also use an MCO as its Mn(II) oxidase? Since previous work showed that the addition of copper substantially increased Mn(II) oxidation rates by P. putida GB-1 (6), we tested the effect of Cu²⁺ on the cumA deletion mutant and Cu²⁺ was still able to stimulate Mn(II) oxidation (data not shown). Thus, an MCO other than CumA may the Mn(II) oxidase in this species. P. putida GB-1 has a gene with a low level of homology to the Bacillus MCO Mn(II) oxidase gene mnxG (PputGB1_2447, 22% identical, E = 4 × 10⁻²²). However, when this gene is deleted from GB-1, the resulting strain retains the ability to oxidize Mn(II), albeit at a lower rate than the wild type (J. McCarthy, unpublished data). A third MCO gene (PputGB1_2665) has been identified via transposon mutagenesis as involved in Mn(II) oxidation. Again, oxidation was only slowed, but not lost, in this mutant strain (45). In A. manganoxydans SI85-9A1 and Erythrobacter sp. strain SD-21, the Mn(II) oxidase enzyme is a heme-binding peroxidase (3), suggesting that the Mn(II) oxidase in P. putida GB-1 is also a peroxidase. The P. putida GB-1 genome does encode a homolog to the Mn(II)-oxidizing peroxidase gene mopA; however, in-frame deletion of this gene does not impair Mn(II) oxidation (data not shown). Thus, P. putida GB-1 may have multiple Mn(II) oxidase enzymes, with alternate enzymes dominating under different growth conditions.

Our results showed that the direct cause of the oxidation defect exhibited by GB-1-007 is a frameshift mutation in the sensor histidine kinase gene mnxS1. Furthermore, a neighboring sensor histidine kinase (mnxS2) and a response regulator gene (mnxR) are also essential for Mn(II) oxidation. Thus, our results have revealed a two-component regulatory (TCR) pathway that is required for Mn(II) oxidation. TCR pathways are composed of a sensor histidine kinase (SK) and a response regulator (RR). In response to an environmental signal, the SK undergoes autophosphorylation within its kinase domain. The phosphoryl group is then transferred to a target RR, which is often a transcription factor. Phosphorylation ultimately alters transcription of downstream genes, thus producing a physiological response to the environmental signal (36).

By sequence homology, MnxR is predicted to be a transcriptional regulator that interacts with σ⁵⁴-containing RNA polymerase (2), similar to the σ⁵⁴-dependent response regulator NtrC (37). Therefore, activation of the MnxS/R signal transduction pathway likely results in the modulation of expression of downstream genes transcribed from σ⁵⁴-specific promoters. Using the P. putida GB-1 genome sequence (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi) and the PromScan website (http://molbiol-tools.ca/promscan/) (37), we analyzed genes previously identified as important for Mn(II) oxidation, such as general secretion pathway genes, the cytochrome synthesis operon ccm, and genes encoding components of the TCA cycle and the tryptophan biosynthesis pathway (8, 12, 13) for the presence of σ⁵⁴-dependent promoters. Of these genes only the general secretion gene xcmT1 and mnxS2 are predicted to have σ⁵⁴ promoters (data not shown). This suggests that many of the previously identified genes are not directly regulated by the Mnx TCR pathway or that mutations in these genes indirectly affect the ability of the cell to oxidize Mn(II) by affecting the metabolic state of the cell.

The MnxS/R TCR pathway appears to require input from both of the two SKs and the RR MnxR to trigger Mn(II) oxidation, whereas multicopy expression of mnxR restores robust oxidation to the ΔmnxS2 strain and has little effect on the ΔmnxS1 strain (Fig. 3). The most parsimonious explanation for these observations is that MnxS2 phosphorylates MnxR, while MnxS1 signals through a second, as-yet-unidentified RR. Another possibility is that the two sensors each signal through MnxR in a branched TCR pathway (24), similar to the sporulation regulatory pathway in Bacillus subtilis (22), and the quorum-sensing network in Vibrio harveyi (18). In these pathways, the individual sensor kinase proteins are thought to respond to different environmental or intracellular signals to regulate the levels of the phosphorylated RR (24). Regardless of whether the MnxS/R TCR pathway converges on the single RR MnxR or there is a second RR in the pathway, it appears that expression of the genes required for Mn(II) oxidation requires signaling through both MnxS1 and MnxS2, since deletion of either gene results in complete loss of oxidation under the conditions screened.

TCR pathways regulate cell physiology in response to environmental signals (36). It would be logical to assume that the Mnx TCR system senses Mn(II) concentration; however, the production of the Mn(II) oxidase enzyme does not require growth in the presence of Mn(II) (23). Mn(II) oxidation does require oxygen (28), begins as the cells enter stationary phase (23), and is blocked by growth in rich media (23) or the presence of antibiotics (8, 32). Therefore, the signals sensed by MnxS1 and MnxS2 may be related to nutritional state, growth phase, oxygen concentration, or population density, among other things. Unfortunately, although MnxS1 and MnxS2 are 58% identical to each other along the full length of the proteins, they have little homology to characterized SK proteins, and both lack conserved sequences such as PAS or HAMP domains (25) that could offer a clue to the signal they detect. Through microarray analysis of the genes regulated by the Mnx TCR pathway, the Mn(II) oxidation field now has a powerful tool for studying the signals that trigger oxidation, the nature of the genes coregulated with oxidation and possibly even the Mn(II) oxidase enzyme(s) itself, ultimately revealing the role of Mn(II) oxidation in the physiology of P. putida GB-1.