Introduction

Manganese (Mn) oxides are highly reactive minerals formed via Mn(II) oxidation. Mn oxides are associated with the morphology, migration and transformation processes of many substances in the environment and play a crucial role in the biogeochemical cycle of many elements1. In soils, Mn oxide minerals are derived from both the chemical and biological oxidation of soluble Mn(II) to stable Mn(IV) oxides. Chemogenesis resulting from spontaneous thermodynamic processes slowly mineralizes Mn oxides through chemical oxidation, surface catalysis and collochemical agglomeration. By contrast, the reaction rates associated with the biological processes (mediated by various enzymes) that oxidize Mn(II) to form different types of Mn oxide minerals are markedly higher than chemical processes1,2. In general, biological Mn oxide mineralization occurs in surficial soils because biogenic Mn(II) oxidation of soluble Mn(II) to Mn(IV) oxides is an aerobic process that requires O23,4,5,6. Under anaerobic conditions, Mn(II) is favorably converted to certain other Mn compounds, such as rhodochrosite (MnCO3)7 and MnS8,9.

Various microorganisms, particularly bacteria and fungi, are considered the primary driving forces in the biological formation of Mn oxides. Previous studies of several Mn(II)-oxidizing bacteria from aquatic habitats, such as Leptothrix discophora SS-110, Pseudomonas putida MnB111, P. putida GB-112 and Bacillus sp. SG-113, have characterized Mn(II) oxidation as an enzymatically catalyzed biochemical process. The genes associated with Mn(II) oxidation in these strains have been identified as mofA, mofB and mofC in SS-110, the ccm operon containing five or more genes in MnB111 and mnxA, B, C, D, E, F and G, which constitute a cluster in an operon in SG-114. However, only a few studies have investigated the Mn oxidation by bacteria in either surface soils15 or soils surrounding Fe–Mn nodules16. Furthermore, the genes associated with Mn oxidation in terrestrial soil-borne bacteria have not yet been cloned or characterized. The elucidation of the mechanisms underlying biogenic Mn oxidation and the biomineralization of manganese oxides in soils continues to be a research goal.

Interestingly, certain Mn(II)-oxidizing genes in marine bacteria, including moxA in Pedomicrobium sp. ACM306717 and mnxG in Bacillus SG-114, share significant sequence similarity with bacterial multicopper oxidase (MCO). MCO is a widespread enzyme capable of catalyzing the oxidation of many inorganic and organic substrates. Several studies have reported that the bacterial oxidation of Mn(II) to form Mn(IV) oxides is strongly influenced by different MCOs18,19. Bacterial MCOs are typically composed of three distinct cupredoxin domains. Only the C-terminal domain III containing a type I copper (Cu) site transfers electrons from the substrate to the trinuclear Cu center composed of a type II Cu atom and a pair of type III Cu atoms20. In Bacillus sp. strain PL-12 and Bacillus SG-1, the oxidation of Mn from soluble Mn(II) to Mn(IV) oxides has been described as a two-step reaction catalyzed by an MCO-containing complex, in which a single electron is transferred from Mn(II) to the type 1 Cu site and relayed to O2 at the type 3 Cu site3,21. Moreover, various bacterial MCOs have a wide range of substrate specificities, including aromatic amine and annular polyol (benzenediol) and the substrates may even include low-quasi-valence transition metals, such as Fe(II), Cu(I) and Mn(II)22. For example, an MCO from Pseudomonas aeruginosa is capable of oxidizing Fe(II) to Fe(III) after the acquisition of Fe(II) by the cell23. Escherichia coli CueO is also capable of oxidizing Cu(I) to Cu(II) using T1-type copper at the substrate-binding site and the trinuclear copper center at the oxygen-binding site24,25. A region rich in methionine near the T1 copper site that senses copper is required for the binding activity of the fifth copper24,25,26. These MCO-mediated oxidation reactions all appear to be related to the cell surface; thus, investigating the role of MCOs in Mn oxidation on the cell surface will improve the understanding of the biogenesis of Mn oxide minerals in the soil environment.

The objective of the current study was to define the role of an MCO protein (MCO266) from a soil-borne Mn(II)-oxidizing bacterium in bacterial cell-surface-orientated Mn oxidation and the biomineralization of Mn oxides. For this purpose, an MCO heterologous expression system for Mn(II) oxidation was established using the surface-display system of E. coli. Using this approach, we confirmed the biological formation of Mn oxides (III and IV) catalyzed by surface-expressed MCO266 proteins. Following continuous Mn enrichment in laboratory shake-flask trials, recombinant strains with MCO266 were able to catalyze the formation of Mn oxide deposits on the cell surface, resulting in the successive formation of bulky, regular, microspherical aggregates. Therefore, MCO266 is involved in the biological nucleation of Mn oxides under laboratory conditions.

Results

Characterization of Mn(II)-oxidizing E. coli MB266

We previously identified a soil-borne E. coli strain (named MB266) in a brown soil that surrounded Fe−Mn nodules27. Mn oxide formation was detectable in this strain after culturing in Lept medium for 48 h. Within 120 h, the concentration of Mn oxides [quantified using a standard Leucoberbelin Blue (LBB) spectrophotometry assay] reached 65 μM. Scanning electron microscopy (SEM) revealed that the surfaces of the MB266 cells exhibited rough and irregular deposit layers (Fig. 1A). Transmission electron microscopy (TEM) also showed a thick non-transparent deposit layer on the surface of MB266 cells in conjunction with dark-brown acicular deposits (Fig. 1B). This strain was subjected to continuous Mn(II) enrichment in a laboratory shake-flask trial for 14 d to investigate its micromorphological and physiological responses. Unexpectedly, regular microspherical aggregates with diameters of approximately 10 μm to 20 μm were observed via SEM and the bacteria attached to and embedded in the aggregates were easily distinguished (Fig. 1C,D). Energy-dispersive X-ray (EDX) spectroscopic scanning assays revealed that the surfaces of the aggregates were mainly covered with Mn, in addition to the elements carbon (C) and oxygen (O) (Fig. 1E).

Figure 1
figure 1

SEM and TEM micrographs of MB266 and EDX spectrogram.

A: SEM micrograph of the dissociated MB266 and Mn oxide sediments; B: TEM micrograph of MB266 and Mn oxide deposits; C: SEM micrograph of the Mn oxide aggregates produced by MB266; D: SEM micrograph of the Mn oxide aggregates produced by MB266; E: EDX spectrogram of a randomly selected EDX assay position (indicated by the arrow in Fig. 1D).

Full size image

Multicopper oxidase is involved in Mn oxidation

To identify the genes responsible for the Mn(II)-oxidizing activity of MB266, a mutant library of MB266 was constructed via the conjugal transfer of mini-Tn5 (Fig. S1). From the approximately 1,500 transconjugants, we obtained three candidate mutants with significantly reduced Mn(II)-oxidizing activities, as detected through LBB assays. Nucleotide sequencing and BLASTN analysis of the inverse polymerase chain reaction (PCR)-amplified fragment flanking the Tn5 insertion site from one mutant revealed that the insertion occurred in the multicopper oxidase (MCO266) gene locus. This gene (mco266) (GenBank accession number: JF682492) was amplified from the MB266 genome. The sequence analysis of mco266 revealed that the putative MCO266 protein contained four typical Cu-binding sites located in regions surrounding the aa residues H101−H103, H141−H143, H443−H448 and H499−H505 in its intermolecular structure. The molecular organization of the putative MCO266 revealed that the N-terminal moiety contains one type 2 and two type 3 copper-binding sites coordinated by eight histidine residues. By contrast, the C-terminal moiety contains a type 1 copper-binding site consisting of two histidine residues and one cysteine residue, as well as a methionine-rich region. Although MCO266 showed an aa sequence similarity of less than 20% compared with several other well-known Mn(II)-oxidizing proteins (e.g., MnxG and MofA), two conserved copper-oxidizing superfamily domains and two conserved Cu(II)-binding sites corresponding to domains in the MnxG structure were identified in this protein. Thus, we selected mco266 as a target gene for further study.

A mutant strain of E. coli MB266 with a site-specific gene disruption of mco266 was developed using the Red recombination strategy28. The Mn(II)-oxidizing activity of the mco266-disrupted mutant strain MB267 was observed to be approximately 37% lower than that of MB266 (Fig. 2). The plasmid-mediated intracellular mco266 expression in MB267 completely compensated for the reduced activity and the resultant E. coli MB268 strain exhibited even higher activity than MB266 (Fig. 2). However, after introducing the mco266 gene into the activity-negative recipient strain E. coli JM109, the resultant MB252 strain failed to exhibit any Mn(II)-oxidizing activity (Fig. 2). Surprisingly, SEM images revealed that the mutant strain MB267 formed irregular and loose cell aggregates after continuous Mn2+-enriched culturing for 14 d (Fig. 3A). This result differs significantly from the regular and compact microspherical aggregates composed of rare bacterial cells formed by MB268 (Fig. 3B) and wild-type MB266 (Fig. 1D), although similar rough and non-transparent deposit layers were observed in association with MB267 cells (Fig. 3B) and MB268 cells (Fig. 3D). These results indicate that the MCO266 protein from the strain MB266 is involved in Mn(II) oxidation.

Figure 2
figure 2

Mn(II)-oxidizing activity measurement of the recombinant E. coli strains.

All of the strains (Table S3) were grown in liquid Lept medium in the presence of 1 mM Mn(II) for 5 d. The Mn oxide concentrations produced by these strains were then measured using the LBB method. For comparison purposes, a cell-free liquid Lept medium, heat-killed E. coli MB266, JM109, MB253 and MB261, wild-type MB266 and the recipient strain JM109 were subjected to the same measurement protocol in parallel.

Full size image
Figure 3
figure 3

SEM and TEM micrographs of the mco266-disrupted mutant MB267 and recombinant MB268 compensated with MCO266.

A and C: Aggregates formed by MB267 and MB268, respectively. B and D: Mn oxide deposit layers produced by MB267 and MB268 cells, respectively.

Full size image

Surface display of MCO266 and Mn(II)-oxidizing activity of intact cells and cell fractions

Next, we verified the surface occurrence of MCO266-mediated Mn(II) oxidation. We performed cell-surface-display experiments to immobilize the MCO266 protein on the surfaces of both E. coli JM109 and MB266 cells using the N-terminal domain of an ice nucleation protein (InaQ)29 from Pseudomonas syringae as the anchor. Immunomicroscopic examination (Fig. S2A), Western blot analysis of the cell fractions (Fig. S2B) and fluorescence-activated cell sorting (Fig. S2C) of the resultant E. coli MB253 and E. coli MB261 cells confirmed the surface localization of the fusion protein InaQ-N/MCO266 in both strains.

The whole-cell Mn(II)-oxidizing activities of all recombinant E. coli strains, the wild-type strain E. coli MB266 and the recipient strain E. coli JM109 were assayed under comparably normalized conditions. As shown in Fig. 2, MB266 retained an activity level comparable to that observed in its initial state27. By contrast, a cell-free substrate control, JM109 and heat-killed MB266, JM109, MB253 and MB261 exhibited nearly zero activity. However, the MB253 cells expressing surface-displayed MCO266 proteins exhibited very high activity, strongly contrasted with the activity-negative parent strain JM109 and the MB252 cells expressing intracellular MCO266. Significantly increased activity was also exhibited by recombinant MB261 cells expressing surface-displayed MCO266 and these cells exhibited higher activity than the parent strain MB266. These results confirm the biogenesis of Mn oxidation and further demonstrate that MCO266 is involved in Mn(II) oxidation on the cell surface.

Samples of the purified MCO266 protein and the cell outer membrane component (COMC) fractions of MB266 and JM109 (separately or combined and with or without proteinase K proteolysis) were obtained to investigate their Mn(II)-oxidizing activities. In contrast to the limited background activity of the JM109 COMC fraction alone, the MB266 COMC fraction exhibited substantial activity (Fig. S3), further verifying the surface occurrence of Mn(II) oxidation in this strain. Notably, the purified MCO266 protein alone failed to exhibit the expected activity, but the level of activity was significantly higher in a mixture of purified MCO266 and the MB266 COMC fraction and/or the activity-negative JM109 COMC fraction (Fig. S3). However, the treatment of these fractions with proteinase K significantly reduced the joint Mn(II)-oxidizing activity of MCO266 with either the JM109 COMC or MB266 COMC fractions to background levels (Fig. S3, columns titled “MCO266+[JM109COMC+K]” and “MCO266+[MB266COMC+K]”). The proteinase K treatment also had a significant activity-reducing effect on the MB266 COMC fraction. These results suggest that MCO266-mediated Mn oxidation is associated with certain COMCs that appear to contain proteins. To determine the functional domains responsible for Mn(II) oxidation in the MCO266 molecules, the surfaces of E. coli JM109 cells were forced to express either the truncated N-terminal (MCO266-N, from aa 1 to aa 333) or the C-terminal moiety (MCO266-C, from aa 334 to aa 516) of MCO266 using the same anchor protein. Neither the resultant E. coli MB254 (displaying InaQ-N/MCO266-N) nor MB255 cells (displaying InaQ-N/MCO266-C) exhibited Mn(II)-oxidizing activity (Fig. 2), which indicates that the full-length MCO266 is required for Mn(II) oxidation.

Microspherical aggregates of Mn oxides formed by recombinant E. coli MB261 and MB253 cells

Recombinant E. coli MB253 and MB261 cells with surface-displayed MCO266 proteins were subjected to continuous 14-d cultivation with 1 mM Mn(II). Not unexpectedly, SEM images showed that microspherical aggregates were formed by both MB253 and MB261 cultures (Fig. 4A,D). The regular shape and diameter of these MB261 aggregates were very similar to those of the MB266 aggregates (Fig. 1C,D). Conversely, the MB253 aggregates were irregular in shape and smaller (<5 μm in diameter) compared with those of MB261. The MB253 and MB261 cells were wrapped with rough and non-transparent deposit layers (Fig. 4B,D). The TEM images showed brown acicular deposits surrounding the MB253 and MB261 cells, consistent with the observations from the MB266 cells (Fig. 1B). EDX scanning of randomly selected sites on the microspherical aggregate surfaces revealed that Mn was the dominant element, but other elements, including C and O, were also present (Fig. 4B, position 1 for the MB253 aggregate; 4D, position 2 for the MB261 aggregate).

Figure 4
figure 4

SEM and TEM micrographs and EDX spectrograms of the recombinant E. coli strains MB253 and MB261.

A and D: SEM micrographs of the aggregates produced by MB253 and MB261, respectively; B and E: TEM micrographs of the Mn oxide deposits produced by MB253 and MB261, respectively; C and F: EDX spectrograms of randomly selected EDX assay positions, which are indicated by the arrows in A and D respectively.

Full size image

Kinetic analysis of Mn oxidation and aggregation formation by MB253 and MB261

Due to their apparent high Mn(II)-oxidizing activity and their ability to catalyze aggregate formation, MB253 and MB261 cells were selected as the target strains for a kinetic examination using a 14-d culturing period with an initial Mn(II) concentration of 1 mM. The whole-cell Mn(II)-oxidizing activities of MB253 and MB261 steadily increased over the first 5 d to the maximum observed values and maintained these activities throughout the experiment (Fig. 5Aa). Thus, the pattern of these cells was synchronous with that of the wild-type strain MB266. The pH levels of all the culture suspensions were maintained at 7.3 to 7.8 (Fig. 5Ab) throughout the initial 5-d period and reached a stable level of <8.0 over the course of the 14-d culturing process. Therefore, these results demonstrate that these strains exhibited biological Mn-oxidation activity. SEM images (Fig. 5B) showed that the aggregate formation processes in MB266, MB253 and MB261 were similar, consistent with their Mn(II)-oxidizing activities during the culturing period. In addition to the production and accumulation of Mn oxides, the bacterial cells gradually gathered during days 2 to 5, formed loose aggregates during days 7 to 11 and ultimately formed more compact microspherical aggregates on day 14. TEM images (Fig. S4) also captured the gradual formation of deposit layers and acicular deposits from zero deposition (day 1) to deposit spots (day 2), deposit layers (days 5 to 7) and layers and acicular deposits (days 11 to 14) during the culturing of all three strains.

Figure 5
figure 5

Kinetic analysis of (A) Mn(II) oxidation and (B) aggregate formation by the wild-type strain MB266 and recombinant strains MB253 and MB261 expressing surface-located MCO266.

A: The Mn(II)-oxidizing activities (a) and pH (b) were measured over a period of 14 d during which the cells were grown in liquid Lept medium in the presence of 1 mM Mn(II). B: The cultures were examined via SEM on days 1, 2, 5, 7, 11 and 14.

Full size image

XPS and XRD characterization of Mn oxides formed by MB261 and MB253 cells

X-ray photoelectron spectroscopy (XPS) analysis was performed to calculate the relative quantities of Mn(III), Mn(IV) oxides and Mn(II) near the surface relative to the total quantities of Mn oxides produced by the E. coli MB261, MB253 and MB266 cultures. Figure 6 shows the XPS narrow scans of Mn (2p3/2) for the Mn oxides formed by MB266, MB261 and MB253. The Mn (2p3/2) spectra demonstrated the existence of Mn(IV) and Mn(III) oxides, as well as Mn(II), in the biotic Mn oxides. The binding energies, relative intensities of all of the peaks in each multiplet set and the full width at half maximum were determined and are presented in Table S1. The proportions of Mn(IV)-oxides, Mn(III)-oxides and Mn(II) in the total quantity of Mn oxides formed by MB266 were found to be 33.68%, 34.02% and 32.29%, respectively; those of MB261 were 31.74%, 34.73% and 33.51%, respectively; and those of MB253 were 24.57%, 52.8% and 22.63%, respectively.

Figure 6
figure 6

XPS patterns of the Mn (2p3/2) spectrogram of the Mn oxides produced by the E. coli MB266, MB253 and MB261 cultures.

The upper circles represent the observed data. The thick, solid curve indicates the best fit of the data. The dashed–dotted curves represent the Mn(IV) multiplet peaks, the dotted lines represent Mn(III) and the thin solid lines represent Mn(II).

Full size image

X-ray diffraction (XRD) assays were also conducted to verify the types of Mn oxides in the microspherical aggregates formed by MB261, MB253 and MB266 cells. As shown in Fig. 7, the power XRD profile of the natural pure bixbyite exhibits sharp peaks at positions corresponding to those of Mn2O3 (JCPDS 02-0896). As expected, the XRD patterns of the Mn oxides from the microspherical aggregates from all three strains exhibit four distinct peaks at 0.38, 0.27, 0.17 and 0.14 nm, which correspond to the (220), (222), (440) and (622) crystal planes, respectively. These results indicate the presence of Mn2O3 in all three strains’ microspherical aggregate samples. However, all the strains’ Mn oxide peaks were relatively low in intensity and exhibited relatively broad areas compared with those of bixbyite; thus, the biogenic Mn oxides may be poorly crystallized bixbyite.

Figure 7
figure 7

XRD patterns of Mn oxides produced by the recombinant E. coli MB261 and MB253 strains and the wild-type MB266 strain.

The experiments were performed using dried powdered Mn oxide samples. The dashed lines indicate the peaks that overlap with standard bixbyite (Mn2O3).

Full size image

Effect of Mn(II) concentrations and different chemicals on Mn(II)-oxidizing activity and aggregate formation

To facilitate the assays, the strain MB253 derived from the activity-negative recipient strain JM109 was used to investigate the effects of Mn(II) concentrations and different chemicals on Mn(II)-oxidizing activity and aggregate formation. As shown in Fig. S5A, as the initial Mn(II) concentrations increased from 0.1 mM to 2.0 mM, both MB253 and MB266 exhibited steadily increasing Mn(II)-oxidizing activities. By contrast, at 0.05 Mn(II), no activity was detected in either strain. SEM micrographs showed that both MB266 (Fig. S5B, a) and MB253 (Fig. S5B, e) formed regular and compact microspherical aggregates following incubation with 1 mM Mn(II). By contrast, incubations with 0.5 mM Mn(II) (Fig. S5B, b for MB266; Fig. S5B, f for MB253) and 0.2 mM Mn(II) (Fig. S5B, c for MB266; Fig. S5B, g for MB253) resulted in irregular and incompact aggregates and an incubation with 0.1 mM Mn(II) produced no observable aggregates (Fig. S5B, d for MB266; Fig. S5B, h for MB253). These results confirm that Mn(II) oxidation does not idiomorphically occur in these strains but requires the concentration of Mn(II) to exceed certain thresholds. Furthermore, because the process is synchronous with Mn(II) oxidation, the formation of aggregates requires the production of a certain quantity of Mn oxides by these strains.

The effects of various chemical compounds on the Mn(II)-oxidizing activities of MB253 and MB266 were investigated. Table S2 shows that a well-known inhibitor of MCOs30, sodium azide, significantly inhibited MB253 activity (approximately 7% and 2% residual activities following treatment with 1 mM and 10 mM sodium azide, respectively), suggesting that MCO266 is the primary contributor to the Mn(II)-oxidizing activity of this strain. However, in MB266, the activity was higher (30% and 18%, following treatment with 1 mM and 10 mM sodium azide, respectively). This result suggests that this strain possesses one or more Mn(II)-oxidizing enzyme(s) in addition to MCO266. Moreover, higher concentrations of SDS (10 mM), cytochrome C (100 μM) and ATP (2000 μM) exhibited certain inhibitory effects on the Mn(II)-oxidizing activities of both strains, whereas three coenzymes (PQQ, NADH and NAD+) increased the Mn(II)-oxidizing activities of both strains. Thus, Mn oxidation is potentially associated with the energy metabolic pathways in the host cells.

Discussion

This study demonstrates that MCO266 is an Mn(II)-oxidizing protein that actuates Mn oxide deposition onto the bacterial cell surface and successively forms oxide aggregates. Both wild-type E. coli MB266 and the recombinant E. coli strains with surface-displayed MCO266 are capable of producing microspherical aggregates of Mn oxides. Thus, soil-borne Mn(II)-oxidizing bacteria have the potential to serve as rudimentary or preliminary nucleating centers for the formation of ferromanganous nodules in natural soils. Various metallic/nonmetallic ions are localized in these centers and are adsorbed to the bacterial aggregates due to the strong oxidation and adsorption activities of the Mn oxides. Thus, the aggregates can ultimately initiate the formation of ferromanganese nodules and/or Mn oxide minerals in soil following polyfunctional interactions of various microorganisms and their chemical reactions with various elements31,32,33.

Although many studies have investigated the molecular mechanism of Mn oxidation in marine bacteria, soil-borne bacteria have rarely been studied. To identify the Mn(II)-oxidizing gene in E. coli MB266 that results in high Mn(II) oxidation activity, we constructed a transposon Tn5-insertion mutant library for MB266. The mco266 gene located in the genomic sequences flanking a Tn5 insertion site manifested a positive response. Therefore, the Mn(II) oxidation activity of the strain MB266 is closely linked to the mco266 gene. Sequence analysis revealed that mco266 is a monocistronic gene, whereas the mco266 genes of marine bacteria are located in polycistronic operons10,12,14. Because different gene organizations may lead to significant discrepancies between aqueous-phase and soil-borne bacteria in terms of the mechanisms of Mn(II) oxidation, an intensive study of MCO266 via heterologous expression and in vitro characterization of Mn(II)-oxidizing activity was performed. Significant catalytic activity was detected when MCO266 was mixed with the COMCs or immobilized on the surfaces of the target cells. Moreover, neither purified nor intracellularly expressed MCO266 was capable of Mn(II) oxidation. The results of mco266 disruption and compensation in trans showed an obvious decrease in Mn(II)-oxidizing activity; however, the activity was not completely eradicated. Although the results demonstrated in principle that MCO266-actuated Mn oxidation is a surface-engendered event, other oxidative routes may also exist as a result of the adaptation of the cells to varied and inconstant soil environments.

In the natural environment, both abiotic and biological oxidation of Mn(II) to Mn(III/IV) may occur simultaneously; however, the abiotic oxidation of Mn(II) is known to occur preferentially at pH values greater than 8.034,35. In the current study, the Mn(II) oxidation associated with several target strains was kinetically examined and substantial Mn(II)-oxidizing activities were observed during the experiments at pH values less than 8.0 (Fig. 5Ab). Moreover, the cell-free substrate and all heat-killed strains failed to exhibit Mn(II)-oxidizing activity (Fig. 2). Therefore, Mn oxidation apparently occurs through a biotic transformation process. Furthermore, bacterial Mn oxidation has been found to be an aerobic biotransformation process4. For example, the Mn oxidation rate in L. discophora SS1 is directly proportional to the O2 concentration6 and Mn(II) failed to be converted into Mn(III/IV) in L. discophora SS16 and P. putida GB-136 under anaerobic conditions. To investigate whether MCO266-mediated Mn oxidation is also dependent on oxygen, we performed comparative measurements of the Mn(II)-oxidizing activity of both the wild-type stain MB266 and the engineered strains with surface-displayed MCO under aerobic and anaerobic conditions in the presence of 1 mM Mn(II). Interestingly, although E. coli strains are well-known facultative anaerobes, all of the tested strains exhibited very poor cell growth under the anaerobic conditions and no Mn oxide could be detected, verifying that Mn oxidation by MCO266 requires O2 and that the presence of 1 mM Mn(II) may act as a restraining factor for cell growth in the absence of O2. Conceivably, the oxidation of Mn(II) to Mn oxide by MCO266 occurs through electron transfer to O2, which acts as the final electron receptor, as suggested by the results from in vitro experiments. These experiments demonstrated that the mixture of purified MCO266 and COMC complex exhibited Mn(II)-oxidizing activity, but the purified MCO266 alone did not (Fig. S3).

The SEM and TEM images obtained in this study revealed that the Mn oxide layers were located in close proximity to MB266 and the recombinant strains with surface-displayed MCO266 (Figs. 1,3, 4, 5). This result further supports the conclusion that the Mn(II) oxidation process occurs on the cell surface. Moreover, the surface location of MCO266 conferred activity to intrinsically inactive E. coli JM109. The increased activity observed in MB266 strongly suggests the existence of interactions between MCO266 and cell-surface components. The requirement for COMCs from either JM109 or MB266 for Mn(II) oxidation by purified MCO266 further confirms this interaction (Fig. S3). In general, MCO266-mediated oxidation is initiated by the receipt of an electron from substrates via type I Cu bound to MCO266-C and the subsequent electron transfer to other substrates is conducted via three additional copper atoms bound to MCO266-N37. Thus, the cell-surface Mn(II) oxidation process mediated by MCO266 in the MB266 strain is most likely associated with the electron transfer process.

Mn oxidation from Mn(II) to Mn(III) or Mn(IV) by electron transfer from a series of redoxases has been previously described in other bacteria. Webb et al.38 revealed that Mn(II) oxidation by Bacillus SG-1 is mediated by two consecutive single-electron transfer reactions that require MnxG. In this process, Mn(III) is an instantaneous intermediate product and the oxidation process from Mn(II) to Mn(III) is the rate-limiting step for the entire reaction. Soldatova et al.3 proposed that the MnxG-gene-product active site may be organized to couple successive electron transfers for the formation of polynuclear Mn(IV) complexes in this strain. In all of these cases, MCO is the key enzyme that directly catalyzes Mn(II) oxidation into Mn(III), whereas other proteins encoded in the same operon appear to be organized for the full oxidation of Mn(II) to Mn(IV). However, unlike other MCOs (e.g., MnxG and MofA) encoded in an operon comprising several genes, the MCO266 in MB266 is monocistronic. MCO266 also catalyzes only two single-electron transfer reactions and not the usual four electron transfers to O239.

Therefore, we speculate that MCO266 catalyzes one-electron oxidation by converting Mn(II) into Mn(III) and then cooperates with other cellular membrane-bound electron transporters to perform successive electron transfers that oxidize Mn(III) to Mn(IV). The Mn(III) and Mn(IV) oxides produced by the recombinant E. coli strains with surface-displayed MCO266 (Fig. 6 and Table S1) may arise from an uncoordinated electron transfer from MCO266 to subsequent membrane electron transporters. Despite the instability of Mn(III) in its ionic state, it can form stable Mn(III) oxides in the presence of oxygen; this process is facilitated when Mn(III) is adsorbed onto a solid medium or combined with organic substrates to form chelating complexes40.

The formation of the aggregates appeared to require joint interactions between Mn oxides, MCO266 and other molecules attached to the bacterial cell surface. However, the MCO266 protein plays a key role in Mn oxidation and the successive formation of deposits and aggregates, as confirmed by the introduction of Mn(II)-oxidizing activity and aggregate formation into the activity-negative JM109 strain expressing MCO266 on the cell surface (Figs. 2,4). The data shown in Figs. 4 and S5 confirm that Mn(II) oxides contribute to aggregate formation. These results are consistent with those reported by Miller et al.41, who confirmed that naturally occurring Mn oxides bind to certain proximal proteins, microelements and microorganisms to form aggregates. Moreover, the self-aggregation of MCO266 proteins and the induction of aggregation by MCO266 should contribute to aggregate formation. Several previous investigations have confirmed that CotA laccase not only self-aggregates42,43 but also induces the conjugation between proteins and polysaccharides44. Thus, we propose that surface MCO266 not only directly oxidizes Mn(II) to form Mn oxides but also self-aggregates and induces multiple aggregations of proteins, polysaccharide macromolecules and intact cells during aggregate biogenesis. During the continuous culturing process, the MCO266 proteins released from lytic cells can intensify the aggregations. Therefore, the strains expressing and/or displaying MCO266 on the cell surface initiate the formation of large, compact and multicomponent aggregates. However, disruption of the mco266 gene in the genome of strain MB267, which resulted in fewer Mn oxides and the disappearance of MCO266 on the cell surfaces, correlated with the formation of smaller and incompact cell aggregates by MB267 after a continuous 14-d culturing process (Fig. 3).

The aggregates formed by the recombinant E. coli strains with surface-displayed MCO266 and the wild-type MB266 may initiate the formation of bulkier agglomerates through the participation of various types of ions and organic materials in the natural environment, ultimately forming Mn micronodules under the influence of pressure, temperature and other factors. Based on the results obtained in this study, we propose the following outline for MCO-related Mn(II) oxidation and aggregate formation (Fig. S6): A two-component regulatory system (i.e., the Mnx system45) in the cells may trigger or accelerate oxidase expression and transmembrane transport when the environmental Mn(II) concentration reaches a certain threshold value. Prior to Mn(II) oxidation, Mn(II) may be combined with organic (i.e., oxalate, malonate, etc.) or inorganic (i.e., pyrophosphate) substrates to form Mn chelates46,47. The binding of Mn chelates to cell-surface-associated MCO266 activates the entire Mn(II) oxidation reaction. MCO266 catalyzes the conversion of Mn(II) into Mn(III) and Mn(III) is then transported to other redoxases (i.e., the CCM) to be oxidized into Mn(IV) in association with the reduction reaction of O2 to H2O. As the reaction proceeds, more Mn(III) and Mn(IV) oxides assemble on the cell surface, forming an Mn-oxide deposit layer. The layers on the cell surface may become gradually superimposed, resulting in various microspherical aggregates. These bacterial spheroids with high-potential Mn oxides may further serve as preliminary nucleation centers to oxidize and adsorb other metal ions or organic substrates, thereby successively forming bulkier aggregates. These aggregates can ultimately form ferromanganese or polymetallic micronodules under the influence of natural factors, such as pressure and temperature.

Materials and Methods

Plasmids, strains, culture conditions and Mn(II)-oxidizing activity assay

Table S3 shows the strains, plasmids and oligonucleotide primers used in this study. The wild-type E. coli MB266 was isolated from a surface soil with Fe−Mn nodules in Shangdong Province, China. The bacteria were cultured on Lept medium containing 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES, pH 7.2) and 1 mM Mn(II) at 30 °C48. The medium was supplemented with 100 μg ml−1 ampicillin (Amp) and 0.1 mM isopropyl-β-D-thiogalactoside (IPTG) once the culture OD600 of the E. coli constructs reached 0.6. For the production of Mn oxides, the overnight-cultured bacteria were inoculated into Lept medium at a ratio of 2% (V/V) and MnCl2 was added to the medium to achieve a final concentration of 1 mM. The cultures were cultivated continuously at 30 °C with shaking at 200 rpm for 5 d or 14 d, unless stated otherwise. The LBB method was used for Mn oxide quantification49,50 with the addition of 1.0 mM MnCl2 (final concentration) in Lept medium. The cells were cultured for 120 h at 37 °C unless otherwise stated. For the preparation of the heat-killed E. coli cell samples, the cells were incubated to an OD600 of 1.2 and the cultures were then shifted to a water bath at 65 °C for further incubation for 24 h. Twenty microliters of each treated cell suspension were inoculated into fresh LB broth to examine the lethal effect. Cell-free Lept medium was used in parallel as a control.

Screening and identification of Mn(II)-oxidizing activity-deficient mutants

E. coli S17 λ-pir carrying pUTmini-Tn5 Km2 was used as the donor strain for the conjugative transfer of the mini-Tn5 transposon to E. coli MB266. Overnight cultures of S17 λ-pir and MB266 were resuspended in 10 mM MgSO4. Each culture containing 200 μl of S17 λ-pir (approximately 4.5 × 108 CFUs ml−1) and 400 μl of MB266 (approximately 9 × 108 CFUs ml−1) was plated and grown on an LB plate at 37 °C for 24 h. After conjugation, the cells were harvested and spread onto an LB plate containing 50 μg ml−1 of kanamycin (Kan) and 100 μg ml−1 of rifampicin to screen for resistant colonies. Each resistant colony was inoculated onto an LB plate containing 100 μg ml−1 Amp and 50 μg ml−1 Kan and an LB plate containing only 50 μg ml−1 Kan. The Amp-sensitive and Kan-resistant colonies were selected as the Tn5-inserted transconjugants (Kanr, Amps). The Mn(II)-oxidizing activities of the transconjugants were then measured following the method described above. For inverse PCR, the total DNA of the transconjugants with little or no Mn(II)-oxidizing activity was extracted using a DNeasy DNA isolation kit (TaKaRa Bio, Inc., Dalian, China), digested with EcoR I and self-ligated using standard procedures51. The resultant self-ligated DNA samples were used as templates for inverse PCR in combination with the CKYP1 and CKYP2 primers, which are complementary to the internal sequence of mini-Tn5 for the first-round amplification. The amplified products were analyzed through agarose electrophoresis. The corresponding bands were incised, purified and used as templates for the second-round PCR amplification with the CKJP1 and CKJP2 primers. The amplified products were sequenced using an ABI 373A automatic sequencer and were identified through a BLASTN search of the GenBank nucleotide sequence databases on the National Center for Biotechnology Information (NCBI) server (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Cloning and disruption of the MB266 mco266 gene

The mco266 gene was PCR-amplified from the genome of MB266 using the mcoF1 and mcoR1 primers. The amplified products were subjected to sequencing. Disruption of the mco266 gene in the MB266 genome was performed using the one-step chromosome gene inactivation method described by Baba et al.28.

Plasmid construction

The Bgl II/EcoR I-digested PCR-amplified product from the mco266 gene was ligated into the Bgl II/EcoR I sites of the E. coli expression vector pTrcHis B (Invitrogen, Carlsbad, CA92008, USA) to yield the pMB252 recombinant plasmid for the intracellular expression and in vitro purification of MCO266. A similarly digested fragment was inserted into the Bgl II/EcoR I sites of a previously constructed plasmid (pMB102)29 to yield the pMB253 recombinant plasmid expressing the InaQ-N/MCO266 fusion protein. Both the N- and C-terminal moieties of mco266 (mco266-N and mco266-C) were amplified from the mco266 gene using the mcoF1/mconR and mcocF/mcoR1 primers, respectively. The moieties were digested with BglII/EcoRI and inserted into the BglII/EcoRI sites of pMB102 to yield the pMB254 and pMB255 recombinant plasmids expressing the fusion proteins InaQ-N/MCO266-N and InaQ-N/MCO266-C, respectively.

Expression, purification, protein cell-surface display and cell fraction preparation

Protein expression in the recombinant E. coli strains was induced with 0.1 mM IPTG and protein purification was performed using a His-tag purification system (Invitrogen) following the manufacturer’s instructions. To induce protein cell-surface display, the recombinant strains were induced with 0.1 mM IPTG. The surface localization of the corresponding proteins was detected by Western blot assays, immunofluorescence microscopy and flow cytometry using polyclonal InaQ-N antiserum52 following previously described procedures29. Preparation of the COMC fraction was performed following previously described procedures53.

TEM and SEM/EDX assays

Sample preparation, TEM and SEM/EDX assays were performed following previously described procedures19,54.

XPS analysis

The Mn(II)-oxidizing bacterial cultures were centrifuged and stored in a 20 mM HEPES buffer solution (pH 7.2). The suspensions were smashed six times at 15,000 psi using a French pressure cell (Thermo). Mn oxides were then collected by centrifugation at 5000 × g for 10 min and freeze-dried. XPS analysis of the obtained Mn oxides was conducted on a VG Multilab2000 X-ray photoelectron spectrometer with an Al Kα X-ray source (1486 eV) and a base pressure of 3 × 10−9 Torr in the analytical chamber24. The charge effect was corrected by adjusting the binding energy of adventitious C (1 s) to 284.62 eV. The Shirley-type background was subtracted before deconvolution and data fitting using the parameters used by Nesbitt et al.55 for the multiplet peaks of Mn (2p3/2) for spectra fitting. A 20 : 80 ratio of the Lorentzian–Gaussian mix-sum function was used for all data fittings56.

XRD analysis

The culture products were freeze-dried to obtain biogenic manganese oxide for powder X-ray diffraction (XRD) analysis (Bruker D8 Advance X-ray diffractometer). The samples were placed on low-background quartz plates and scanned over the range 2θ = 5°– 85°. The measurement conditions included the following: a LynxEye array detector, a Ni filter plate, Cu Kα (λ = 0.15418 nm), a tube voltage of 40 kV, a tube current of 40 mA, step scan, a step size of 0.02° and a scan speed of 10°/min.

Additional Information

How to cite this article: Zhang, Z. et al. Surface Mn(II) oxidation actuated by a multicopper oxidase in a soil bacterium leads to the formation of manganese oxide minerals. Sci. Rep. 5, 10895; doi: 10.1038/srep10895 (2015).