Magnetosome gene cluster containing bacteria in oxygen-stratified freshwater ecosystems of northern landscapes

Ji, Runjia; Wan, Juan; He, Kuang; Goswami, Pranami; Shen, Jianxun; Pan, Yongxin; Lin, Wei

doi:10.1038/s43247-025-02495-8

Download PDF

Article
Open access
Published: 08 August 2025

Magnetosome gene cluster containing bacteria in oxygen-stratified freshwater ecosystems of northern landscapes

Communications Earth & Environment volume 6, Article number: 644 (2025) Cite this article

1368 Accesses
10 Altmetric
Metrics details

Subjects

Abstract

Magnetotactic bacteria (MTB) are capable of magnetic navigation (magnetotaxis) through the intracellular biomineralization of membrane-bound magnetic nanoparticles, known as magnetosomes, which are regulated by magnetosome gene clusters (MGCs). Despite their ubiquity near the oxic-anoxic transition zone in various aquatic environments, the diversity, abundance, and biogeochemical roles of MTB remain poorly understood due to limited large-scale studies. Here, we conducted a metagenomic survey of MGC-containing bacteria across 38 oxygen-stratified freshwater environments of northern landscapes. We reconstructed 63 MGC-containing genomes affiliated with eight bacterial phyla, including Myxococcota. Putative MTB exhibited high relative abundance (up to 15.4% of metagenomic reads), with Deltaproteobacteria identified as the most ubiquitous and predominant group. Vertical distribution profile revealed that magnetotactic Deltaproteobacteria primarily inhabit low-oxygen niches, suggesting their magnetofossils might serve as a potential proxy for oxygen depletion. Our findings underscore the important ecological roles of putative MTB as key components of oxygen-stratified freshwater ecosystems of northern landscapes.

Magnetotactic bacteria and magnetofossils: ecology, evolution and environmental implications

Article Open access 01 June 2022

Renaissance for magnetotactic bacteria in astrobiology

Article Open access 17 August 2023

Recovery and genome reconstruction of novel magnetotactic Elusimicrobiota from bog soil

Article 27 October 2022

Introduction

Freshwater ecosystems are a crucial component of Earth’s aquatic systems, playing an important role in sustaining biodiversity and regulating global biogeochemical cycles¹. Microbes are fundamental to these processes, driving nutrient cycling, organic matter decomposition, and overall ecosystem functionality^2,3. Among these microbes, magnetotactic bacteria (MTB) are unique for their ability to orient and navigate using the geomagnetic field, typically inhabiting the oxic-anoxic transition zone (OATZ) of stratified aquatic environments with worldwide distribution and exhibiting high morphological and phylogenetic diversity^{4,5,6,7,8,9,10,11,12}. However, their abundance, diversity, and biogeochemical roles remain largely unexplored due to a lack of large-scale systematic analyses.

Magnetotaxis, or magnetic navigation in MTB, is driven by magnetosomes—unique bacterial organelles composed of membrane-bound, nano-sized magnetite (Fe₃O₄) and/or greigite (Fe₃S₄) crystals. The morphology of Fe₃O₄ magnetosomes varies among different phylogenetic groups, including cubo-octahedral, prismatic, and bullet-shaped. The chain-like arrangement of magnetosomes optimizes the magnetic dipole moment of MTB cells, enhancing their alignment with the geomagnetic field. Due to their unique characteristics and high biocompatibility, magnetosomes hold great potential for applications in biotechnology and nanotechnology^13,14,15,16.

After the death of MTB, magnetosomes can be preserved in sediments as magnetofossils (fossils of magnetosomes), which can serve as indicators of paleoenvironmental and paleomagnetic conditions. For instance, elongated magnetofossils (prismatic and bullet-shaped magnetosomes) have often been associated with anoxic environments, while isotropic forms (cubo-octahedral magnetofossils) were considered to link to suboxic conditions¹⁷. An increase in the proportion of elongated magnetofossils has been suggested to reflect less oxygenated environments^18,19,20. Despite previous studies connecting MTB to hypoxic and anoxic layers within the OATZ^21,22,23,24, the relationship between environmental factors and magnetosome morphology remains unclear. Understanding the diversity and ecology of MTB in modern stratified habitats is key to elucidating the paleoenvironmental implications of magnetofossils.

The biosynthesis of magnetosomes is regulated by magnetosome gene clusters (MGCs), a collection of genes involved in the biomineralization and arrangement of magnetosomes^25,26. Magnetosome biogenesis is thought to have originated in the Archean Eon, with magnetosomes being the only confirmed magnetoreceptors to date^7,9,27, making MTB an ideal model for studying the origin and evolution of magnetoreception and organelle biosynthesis. Recent omics-based approaches have expanded MTB diversity to at least 16 bacterial phyla^9,28,29. The continuous discovery of new MTB lineages suggests that our understanding of their diversity and environmental functions is still very limited. Traditional methods for studying MTB diversity and ecology often rely on their magnetotactic behavior, with pre-processing by magnetic enrichment of MTB cells from environmental samples. MTB within the Magnetococcia (also known as Etaproteobacteria) of the phylum Pseudomonadota (previously known as Proteobacteria) have long been considered the predominant MTB members in both freshwater and marine environments^30,31. In some cases, MTB from the Nitrospirota (previously known as Nitrospirae) phylum have also been identified as dominant in specific environments^32,33. However, the reliance on magnetic enrichment may lead to a biased understanding of MTB diversity and distribution, as it depends on the magnetotaxis capability and motility of MTB cells³⁴.

In this study, we report the high abundance of MGC-containing bacteria (putative MTB) in oxygen-stratified freshwater ecosystems of northern landscapes, revealed through genome-resolved metagenomic analysis of 267 metagenomes³⁵ without magnetic enrichment pre-processing. Our results show that putative MTB account for up to 15.4% of metagenomic reads at optimal niche depths within oxygen-stratified water columns, indicating an underappreciated ecological role for these bacteria. Deltaproteobacteria was found as the most ubiquitous and predominant MGC-containing group across these samples. We further examined the vertical distribution and metabolic potential of dominant MGC-containing members, uncovering mechanisms that shape the vertical niche differentiation of MTB communities. Additionally, we provide the first genomic evidence of putative MTB within Myxococcota, a bacterial phylum renowned for its extraordinary social behaviors. Overall, our findings highlight the importance of MTB in oxygen-stratified freshwater ecosystems of northern landscapes and offer new clues for reconstructing paleoenvironments using magnetofossils.

Results

Metagenomic dataset description

To investigate the putative MTB population in oxygen-stratified freshwater environments, we utilized a publicly available metagenomic dataset (Project accession: PRJEB38681) that was reported and deposited to the European Nucleotide Archive by Buck et al.³⁵. This dataset includes 267 metagenomes from 38 oxygen-stratified freshwater environments of northern landscapes, including 25 lakes, 12 ponds, and one reservoir (Fig. 1a and Supplementary Data 1). These samples were collected across five countries and regions, spanning tropical to polar zones (Fig. 1b). Of the 267 metagenomes, 225 were collected from freshwater lakes, 34 from ponds, and 8 from a reservoir (Fig. 1c and Supplementary Data 1). Except for nine metagenomes from sediments in Lake Alinen Mustajärvi, all other metagenomes were collected from water columns (Fig. 1c and Supplementary Data 1). In total, the dataset comprised over 9.5 billion paired-end Illumina sequencing reads, corresponding to 2.85 terabases (150 bp/reads × 9.5 billion reads × 2).

**Fig. 1: Distribution and overview of sampling sites.**

Phylogenomic diversity of MGC-containing genomes

Following sequence assembly, binning, and MGC screening (see “Methods” for details), a total of 63 MGC-containing genomes were obtained. Among these, 27 genomes were identified as high quality (completeness ≥90% and contamination <5%), and 36 were classified as medium quality (completeness ≥50% and contamination <10%), according to the Minimum Information about a Metagenome-Assembled Genome (MIMAG) standard³⁶ (Supplementary Data 2). Using an average nucleotide identity (ANI) cutoff of 99%³⁷, the 63 MGC-containing genomes were dereplicated into 41 non-redundant (nr) genomes, comprising 16 high-quality and 25 medium-quality genomes (Supplementary Data 2). Based on the GTDB release R207, the 41 nr MGC-containing genomes were classified into eight bacterial phyla: Desulfobacterota (16 genomes), Pseudomonadota (12 genomes), Omnitrophota (5 genomes), Bdellovibrionota (4 genomes), Desulfobacterota_I (1 genome), Desulfobacterota_G (1 genome), Myxococcota (1 genome), and Hydrogenedentota (1 genome) (Fig. 2, Supplementary Fig. 1, and Supplementary Data 2).

**Fig. 2: Phylogeny and quality of reconstructed MGC-containing genomes.**

One MGC-containing genome belonging to the phylum Myxococcota (SFLP_MTB_KT4_bin17) was reported for the first time. In addition, contrary to the previous notion that Magnetococcia of the phylum Pseudomonadota are often the dominant MTB members, we detected only one genome (SFLP_MTB_Umea3p4_bin52) belonging to this class (Fig. 2 and Supplementary Data 2). Of the 41 nr MGC-containing genomes, 23 belong to the Deltaproteobacteria, which has recently been reclassified into four phyla: Desulfobacterota, Bdellovibrionota, Myxococcota, and SAR324³⁸. For consistency and comparability with prior studies, we will continue to use the term “Deltaproteobacteria” in subsequent sections.

All 41 nr MGC-containing genomes reconstructed in this study contained Fe₃O₄-type MGCs, and their magnetosome gene composition and arrangement exhibited a distinct lineage-specific feature. In general, the magnetosome gene content and organization were more conserved among phylogenetically close members compared to those in distantly related lineages (Fig. 3). Several key magnetosome genes (mam-E, -P, -L, -M, -B, -A, -Q) were identified in the incomplete MGC of the Myxococcota genome (SFLP_MTB_KT4_bin17). Although Myxococcota are phylogenetically close to Bdellovibrionota (Fig. 2, Supplementary Fig. 1), the MGC fragment of the Myxococcota genome exhibited a unique gene synteny that differed from those within Bdellovibrionota and other phyla (Fig. 3).

**Fig. 3: Magnetosome gene clusters (MGCs) in 41 non-redundant MGC-containing genomes.**

Distribution and relative abundance of MGC-containing bacteria

To assess the distribution and relative abundance of putative MTB communities in the oxygen-stratified freshwater ecosystems studied here, we conducted a genome-wide quantitative read recruitment analysis using the 41 nr MGC-containing genomes across all metagenomes (see “Methods” for details). Putative MTB were identified in 142 metagenomes obtained from 22 lakes and ponds in Canada, Finland, Switzerland, and Sweden (Supplementary Data 3). In the remaining 16 lakes and ponds where no MGC-containing genome was detected, the samples primarily consisted of surface water characterized by relatively high levels of dissolved oxygen. Therefore, the absence of MGC-containing genomes in these environments is likely due to inadequate sampling of low-oxygen areas, and thus, the presence of MTB in those environments cannot be ruled out.

We then profiled the relative abundance of MGC-containing bacteria across the lakes and ponds. Samples collected on the same day from the same lake or pond were grouped together, referred to as “lake-time groups”. In this manner, 267 samples were categorized into 80 lake-time groups (Supplementary Data 4). The total number of samples (N_t) and the number of putative MTB-containing samples (N_mtb) were counted for each lake-time group. After discarding the lake-time groups with N_t < 3 or N_mtb < 2 (denoted as inadequate sampling groups), 26 lake-time groups containing 165 samples passed the filtering (Fig. 4 and Supplementary Data 4). For each metagenome, the percentage of reads retrieved during read recruitment was used as a proxy for the relative abundance of MGC-containing bacteria. Considering the vertical distribution characteristics of MTB (Supplementary Figs. 2 and S3), within each lake-time group, the sample with the highest relative abundance was considered the optimal niche for putative MTB communities, and this percentage value was used to represent the relative abundance of MGC-containing bacteria in that group in Fig. 4. Our results showed that the relative abundance of MGC-containing bacteria ranged from 0.5% to 15.4%, with a median of 3.5% across the 26 lake-time groups (Fig. 4 and Supplementary Data 4).

**Fig. 4: Distribution and relative abundance of MGC-containing bacteria.**

Predominance of Deltaproteobacteria in putative MTB communities

The composition and distribution of putative MTB communities across different lakes were then profiled, where MGC-containing Deltaproteobacteria were found most ubiquitous (Fig. 5). Among the 165 samples from the filtered 26 lake-time groups, MGC-containing genomes were detected in 103 samples. Specifically, MGC-containing bacteria belonging to the Deltaproteobacteria, Alphaproteobacteria, and Magnetococcia were found in 76, 38, and 2 samples, respectively (Fig. 5a, Supplementary Fig. 2 and Supplementary Data 4). The distribution and relative abundance pattern demonstrated the remarkable dominance of MGC-containing Deltaproteobacteria (Fig. 5a). To determine whether the difference in relative abundances between MGC-containing Deltaproteobacteria and Alphaproteobacteria (with Magnetococcia excluded due to limited data) was significant, the nonparametric Mann-Whitney U test was employed (see “Methods”). The results revealed that MGC-containing Deltaproteobacteria were significantly more abundant than MGC-containing Alphaproteobacteria within the 26 lake-time groups (p value 0.0036) (Fig. 5b).

**Fig. 5: Distribution and abundance profile of MGC-containing *Alphaproteobacteria* and *Deltaproteobacteria* across 26 lake-time groups.**

The distribution and abundance of MGC-containing Deltaproteobacteria across 26 lake-time groups were further investigated (Fig. 5c), which revealed a predominance of Desulfobacterota, noting that here the phyla Desulfobacterota_G and Desulfobacterota_I are considered part of the phylum Desulfobacterota. MGC-containing Desulfobacterota were detected in 76 metagenome samples from 25 lake-time groups, followed by Bdellovibrionota that were identified in 16 samples from 7 lake-time groups, and Myxococcota in 6 metagenome samples from 3 lake-time groups (Fig. 5c). In contrast, the absence of MGC-containing SAR324 members in this study suggests that this group might not be a common MTB population in oxygen-stratified freshwater environments of northern landscapes.

Vertical niche differentiation of MGC-containing bacteria

MTB are generally microaerobic or anaerobic bacteria that inhabit areas near OATZs, where rapid declines in dissolved oxygen and substantial changes in redox potential occur. In this study, we profiled the vertical distribution of MGC-containing bacteria and analyzed their distribution patterns together with a group of environmental factors. To ensure a robust analysis, ten lake-time groups with sufficient physicochemical records (including dissolved oxygen, pH, electrical conductivity, temperature, ferric, ferrous, sulfate, nitrate, and ammonium) were selected for in-depth analysis (Fig. 6 and Supplementary Fig. 3). Dissolved oxygen (DO) levels were used to delineate the ecological niches of the collected samples, which were classified into three zones: samples with DO ≥ 2 mg/L were classified as the oxic zone, 0 mg/L < DO < 2 mg/L as the hypoxic zone (i.e., OATZ), and DO = 0 mg/L as the anoxic zone. A distinct vertical distribution pattern of MGC-containing bacteria (gray bars in Fig. 6a and Supplementary Fig. 3) was observed across all lake-time groups, with abundance peaking near or below the OATZs (shaded areas in Fig. 6a and Supplementary Fig. 3). The relative abundance of MGC-containing bacteria in each zone was profiled (Fig. 6b), revealing that their relative abundance was significantly higher in the hypoxic and anoxic zones than in the oxic zone. Although no significant difference was observed between hypoxic and anoxic zones, some anoxic metagenomes showed relatively high abundances of MGC-containing bacteria.

**Fig. 6: Vertical distribution of MGC-containing bacteria within oxygen-stratified freshwater columns of northern landscapes.**

The taxonomic compositions of MGC-containing bacteria were further investigated, revealing distinct vertical shifts in the dominant MGC-containing taxa within the ten lake-time groups (Fig. 6a and Supplementary Fig. 3). For instance, MGC-containing Alphaproteobacteria dominated the upper layer of Lake Alinen Mustajärvi (Fig. 6a), while the dominant MGC-containing taxa shifted to Deltaproteobacteria and Omnitrophota (previously known as candidate division OP3) in deeper layers, indicating a vertical niche-specific distribution of different MGC-containing taxa.

To further define the optimal ecological niches for major MGC-containing taxa, the distributions and relative abundances of 41 nr MGC-containing genomes were profiled in 131 MTB-containing samples with DO records (Supplementary Fig. 4). The dominant MGC-containing Deltaproteobacteria and Alphaproteobacteria exhibited clear ecological niche differentiation. Most MGC-containing Deltaproteobacteria flourish in low-oxygen to anoxic environments, at a DO level of ~0.14 mg/L where they are most frequently detected in the dataset (Fig. 6c and Supplementary Fig. 4). MGC-containing Alphaproteobacteria mostly thrive in hypoxic samples, with a DO level around 0.39 mg/L (Fig. 6c and Supplementary Fig. 4). Moreover, the relative abundance of MGC-containing Deltaproteobacteria is significantly higher in the anoxic zone than in the hypoxic zone, while no significant difference is observed for MGC-containing Alphaproteobacteria between these two zones (Fig. 6d). These findings suggest that MGC-containing Deltaproteobacteria primarily thrive near anoxic zone, while MGC-containing Alphaproteobacteria mostly inhabit hypoxic zone in oxygen-stratified freshwater columns of northern landscapes.

Metabolic potential of dominant MGC-containing bacteria

Key metabolic pathways of the 41 nr MGC-containing genomes were reconstructed from their genomes (Supplementary Fig. 5 and Supplementary Data 5), and the metabolic potentials of the dominant MGC-containing members (Deltaproteobacteria and Alphaproteobacteria) were predicted (Fig. 7) and discussed below. The core carbohydrate metabolism pathways were present in all 41 nr MGC-containing genomes, and some genomes contained complete carbon fixation pathways (Supplementary Fig. 5). Specifically, the Calvin–Benson–Bassham (CBB) pathway, which includes the key CO₂-fixing enzyme, ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO), was found in MGC-containing Alphaproteobacteria and Gammaproteobacteria (Fig. 7 and Supplementary Fig. 5). MGC-containing Deltaproteobacteria, on the other hand, could be furnished with other autotrophic metabolic strategies, such as the reductive tricarboxylic acid (rTCA) cycle, the Wood-Ljungdahl (WL) pathway, and the dicarboxylate-hydroxybutyrate cycle (Fig. 7). The Wood-Ljungdahl pathway has previously been identified in Nitrospirota and Elusimicrobiota MTB genomes^11,33,39, which may represent a common carbon fixation pathway in deep-branching lineages of anaerobic autotrophic MTB. The dicarboxylate-hydroxybutyrate cycle, typically restricted to anaerobic bacteria and archaea⁴⁰, suggests an anaerobic lifestyle of deltaproteobacterial MTB.

**Fig. 7: Metabolic predictions of MGC-containing *Alphaproteobaceria* and *Deltaproteobacteria*.**

Both MGC-containing Alphaproteobacteria and Deltaproteobacteria exhibit nitrogen fixation potential (Fig. 7), evidenced by the presence of the nitrogenase genes (nifDKH) in most genomes. Specifically, 14 out of 23 deltaproteobacterial MGC-containing genomes and 7 out of 9 alphaproteobacterial MGC-containing genomes contain nifDKH genes.

Key enzymes associated with the dissimilatory nitrate reduction to ammonium (DNRA) pathway, including nitrate reductases NarGHI and NapAB, were identified in both MGC-containing Alphaproteobacteria and Deltaproteobacteria, suggesting that both groups can perform anaerobic nitrate respiration via DNRA (Fig. 7). However, the nitrite reductase enzymes differed between the two groups: NirBD was found exclusively in MGC-containing Alphaproteobacteria genomes, while NrfAH was exclusive to MGC-containing Deltaproteobacteria. Although both NirBD and NrfAH are dissimilatory nitrite reductases, NirBD was previously found to remain functional during aerobic growth in certain Actinomycetota⁴¹ and Gammaproteobacteria^42,43 species. These findings imply that MGC-containing Alphaproteobacteria may perform DNRA under both aerobic and anaerobic conditions.

In addition to DNRA, genes associated with denitrification, including nirK/nirS, norBC, and nosZ (encoding nitrite, nitric oxide, and nitrous oxide reductases, respectively), were identified in both MGC-containing Alphaproteobacteria and Deltaproteobacteria genomes. This suggests that both groups are capable of complete denitrification under anaerobic conditions (Fig. 7), using nitrate as the terminal electron acceptor.

The nitrite oxidoreductase (NxrAB), which catalyzes the final step of nitrification by oxidizing nitrite to nitrate⁴⁴, was found in 4 out of 9 alphaproteobacterial MGC-containing genomes. Nitrite-oxidizing bacteria (NOB), such as the genus Nitrobacter (affiliated with Alphaproteobacteria), are well-known for this function⁴⁵. However, the nitrite oxidizers identified here belong to two families, Magnetospirillaceae and Azospirillaceae (within the orders Rhodospirillales and Azospirillales, respectively), distinct from the genus Nitrobacter (which belongs to the family Xanthobacteraceae within the order Rhizobiales). The fact that nitrite oxidation typically occurred in oxic environments suggests a dependency of MGC-containing Alphaproteobacteria on oxygen (Fig. 7).

Regarding sulfur metabolism, both MGC-containing Alphaproteobacteria and Deltaproteobacteria genomes contain genes for key enzymes related to assimilatory sulfate reduction, such as Sat, CysND, CysC, CysH, and CysJI. Additionally, they also contain genes for dissimilatory sulfate reduction/oxidation pathways, specifically aprAB and dsrAB genes. The dissimilatory sulfite reductase DsrAB catalyzes sulfite reduction in sulfate/sulfite-reducing microorganisms, but can also operate in reverse during sulfur oxidation in some sulfur-oxidizing bacteria⁴⁶. Phylogenetic analyses have found that oxidative-type DsrAB enzymes, while homologous to reductive-type enzymes, form distinct evolutionary lineages⁴⁷. Most Deltaproteobacteria harbor reductive DsrAB, with the exception of the SAR324 clade, which phylogenetically branches with Alphaproteobacteria in the distinct oxidative DsrAB family⁴⁷. We did not identify MGC-containing SAR324 members in this study, suggesting that MGC-containing Deltaproteobacteria likely perform dissimilatory sulfate reduction, while Alphaproteobacteria may function as sulfur oxidizers. A recent study reported intracellular calcium carbonate (iACC) biomineralization in diverse MTB populations (_iACCMTB), with Gammaproteobacteria _iACCMTB exhibiting the dissimilatory sulfate reduction pathway but likely performing reverse sulfide oxidation⁴⁸. Differences were observed in thiosulfate metabolism: sox genes encoding the sulfur-oxidizing enzyme (SOX) system involved in thiosulfate oxidation were found in MGC-containing Alphaproteobacteria genomes but were absent in Deltaproteobacteria genomes (Fig. 7). Research on chemolithoautotrophic Magnetovibrio strains MV-1 and MV-2 revealed that they could grow microaerobically with S₂O₃²⁻ as an electron source and O₂ as the electron acceptor⁴⁹. Thus, the sox genes-bearing MGC-containing Alphaproteobacteria might be able to harness energy from thiosulfate oxidation under aerobic conditions. Conversely, MGC-containing Deltaproteobacteria identified here were prone to perform thiosulfate reduction to sulfide via thiosulfate reductase (EC 1.8.5.5), which was found in 16 out of 23 genomes. Thiosulfate reductase is commonly found in anaerobic bacteria that use thiosulfate as electron acceptors in the absence of oxygen⁵⁰, as seen in Desulfovibrio species from the phylum Desulfobacterota⁵¹.

In summary, both MGC-containing Alphaproteobacteria and Deltaproteobacteria investigated here were capable of anaerobic respiration, using nitrate or sulfate as electron acceptors. However, distinct metabolic traits reveal that MGC-containing Alphaproteobacteria are also capable of aerobic respiration in the presence of oxygen (Fig. 8).

**Fig. 8: Conceptual model of carbon, nitrogen, and sulfur cycling driven by dominant MTB (*Deltaproteobacteria* and *Alphaproteobacteria*) members near the oxygen-stratified water column.**

Discussion

Due to their special abilities of magnetotaxis and organelle biosynthesis, MTB have long been of great interest to both biologists and geologists. A deeper understanding of their abundance, distribution, diversity, and ecology can illuminate the ecophysiological roles of MTB communities.

Predominance of Deltaproteobacteria in putative MTB communities from oxygen-stratified freshwater environments of northern landscapes

Conventionally, MTB abundance in aquatic environments has been considered low, ranging from 10³ to 10⁶ cells/mL^31,52,53, with a few exceptions^23,32. For decades, magnetotactic cocci within the class Magnetococcia (also known as Etaproteobacteria^7,30,54,55) of the phylum Pseudomonadota have been regarded as the predominant MTB member^{30,31,52,56,57,58,59,60,61,62,63}. As of January 2024, a total of 146 MTB or MGC-containing genomes affiliated with five bacterial phyla have been reported from over 50 various freshwater environments, of which ~42% (61 genomes) belong to the class Magnetococcia (Supplementary Data 6). Besides Magnetococcia, MTB from the phylum Nitrospirota have occasionally been identified as predominant in certain lake sediments and acidic peatland soils^9,32.

Our study shows that MGC-containing bacteria can constitute a considerable portion (0.5% to 15.4%) of the metagenomic reads at optimal niche depths within oxygen-stratified water columns of northern landscapes. Note that the relative abundance values applied here were estimated based on reads mapped to MGC-containing MAGs, not direct cell concentrations, and non-specific mapping from closely related non-MTB microbes may occur. In most oxygen-stratified freshwater environments investigated here, Deltaproteobacteria, rather than the conventionally recognized Magnetococcia and Nitrospirota, emerged as the most widespread and abundant MGC-containing group. Samples from Lake Loclat, collected across seasons from autumn 2008 to autumn 2009, were further analyzed to investigate the putative MTB community distribution and abundance over a full annual cycle (Supplementary Fig. 6). These results reveal consistent Deltaproteobacteria dominance across all seasons, indicating that their prevalence in Lake Loclat is not influenced by seasonal variation. Additionally, of the 26 lake-time groups analyzed, nine were dominated by Alphaproteobacteria and Omnitrophota, as shown in Figs. 5, S2 and S3, highlighting these three taxa as the most dominant members. As most studied lakes are located in Northern Europe, the composition and distribution patterns we observed here may not apply to all oxygen-stratified freshwater ecosystems globally. For example, unlike the Deltaproteobacteria-dominated pattern observed here, Bidaud et al.⁶⁴ investigated the MTB community in Lake Pavin, France, a ferruginous meromictic lake, and reported a Magnetococcia-dominated pattern. This indicates that variations exist among lakes across different geographical locations.

Beyond location and seasonal variations, methodological biases may also explain discrepancies between our study and previous research. Traditional MTB studies often use magnetic enrichment methods, such as capillary racetrack⁶⁵ or MTB trap⁶⁶, which apply magnetic fields stronger than the geomagnetic field to concentrate MTB cells from sediment or water based on their magnetotactic motility^34,67. These methods can introduce biases, potentially underestimating total MTB abundance³⁴ and the taxonomic diversity within these communities, where fast-swimming cells like magnetotactic cocci (e.g., Magnetococcia) may be preferentially enriched over slower-swimming populations (e.g., Deltaproteobacteria). Additionally, a recent study identified a functionality retained yet silent MGC in a non-MTB Alphaproteobacteria strain⁶⁸, suggesting hidden diversity missed by magnetotaxis-dependent approaches. In contrast, our study analyzed metagenomes directly from water columns, avoiding magnetic enrichment. This enabled us to detect both active magnetosome-forming bacteria and dormant (silent) MGCs. While it remains unclear whether silent magnetosome genes in non-MTB can be activated or utilized by their hosts, our findings provide valuable gene resources for producing biogenic magnetic nanoparticles^69,70,71 and new clues for investigating the evolutionary history of MGCs within the Bacteria domain.

Magnetotaxis potential in predatory bacteria

Recent discoveries of novel MTB-containing phyla highlight the ongoing expansion of our knowledge regarding MTB taxonomy^7,9,11,28. Our study contributes to this growing understanding by reconstructing 63 MGC-containing genomes across eight bacterial phyla. We provide the first genomic evidence of MGC within the phylum Myxococcota, although the MGC identified in this phylum is incomplete. Future studies are necessary to confirm whether members of Myxococcota possess magnetotactic capabilities. In addition to Myxococcota, we also obtained seven MGC-containing genomes from the Bdellovibrionota phylum. Both Myxococcota and Bdellovibrionota phyla are known for their predatory lifestyle^72,73. Further annotation of key predatory-related genes within the MGC-containing Bdellovibrionota and Myxococcota genomes revealed potential predatory capabilities (see “Methods” and Supplementary Data 7). Specifically, two Bdellovibrionota MGC-containing MAGs harbor homologs to five predation-related proteins from Bdellovibrio bacteriovorus HD100⁷⁴, while the Myxococcota MGC-containing MAG harbors myxovirescin biosynthesis genes, suggesting potential antibiotic-mediated prey killing^75,76,77. Together with a previous study⁹ that reconstructed six Bdellovibrionota MTB genomes, our findings suggest that MGCs, and thus magnetotaxis, might be widespread among predatory bacteria.

Predatory Myxococcota and Bdellovibrionota are considered to play important roles in the microbial food web⁷⁸, and bacterial predation may act as an unrecognized selective force in microbial ecosystems⁷³. Thus, MTB with a potential predatory lifestyle could take advantage of the geomagnetic field to efficiently prey on bacteria residing in or near OATZs and further influence microbial communities and trophic networks. However, links between magnetotaxis and predation activities need further investigation, which will deepen our understanding of their ecological roles of magnetotaxis in shaping microbial community structures and nutrient cycling.

Vertical distribution of putative MTB in oxygen-stratified freshwater columns and implications for paleoenvironmental reconstructions

The vertical distribution analysis across redox gradients reveals a niche differentiation pattern of MGC-containing bacteria in water columns. Using genome-based metabolic predictions, we identified clear associations between metabolic strategies and ecological niches of MGC-containing bacteria for the two dominant groups: Deltaproteobacteria and Alphaproteobacteria (Fig. 8). The relative abundances of these MGC-containing groups are significantly correlated with several physicochemical parameters (Supplementary Fig. 7), which potentially play crucial roles in shaping their vertical distribution. Specifically, MGC-containing Alphaproteobacteria, capable of thiosulfate and nitrite oxidation, mostly inhabit hypoxic water columns. In contrast, MGC-containing Deltaproteobacteria, which perform anaerobic respiration, stay close to anoxic layers (Fig. 8). The abundance of MGC-containing Deltaproteobacteria positively correlates with ammonium, indicating a preference for reduced environments (Supplementary Fig. 7).

No Fe₃S₄-type MGC-containing genomes were identified near or below the OATZ, contrary to the classical model, which suggests that Fe₃S₄-producing populations typically occur in anoxic conditions^23,79,80. This result indicates that Fe₃S₄-type MTB are not the dominant populations across the analyzed freshwater lakes and ponds, and their abundance may be too low to be detected using a genome-resolved metagenomics approach. Thus, our current understanding of the distribution of Fe₃S₄-type MTB remains very limited and warrants further investigation.

Our findings have further implications for reconstructing paleoenvironmental conditions using magnetofossils, particularly the bullet-shaped forms that biomineralized by Deltaproteobacteria MTB. Previous studies often compare magnetofossils abundance with other paleoredox proxies to infer their paleoenvironmental significance. For instance, Chang et al.²⁰ found that elevated levels of elongated magnetofossils correlate with high concentrations of authigenic uranium (aU), Cd/Al, Mo/Al, U/Al, and low Δδ¹³C, suggesting that these magnetofossils are reliable indicators of low-oxygen conditions. Similarly, Wagner et al.⁸¹ demonstrated a strong correlation between the abundance of elongated magnetofossils (e.g., bullet-shaped forms) and benthic foraminiferal assemblages, such as Fursenkoina aquiensis and Pseudouvigerina wilcoxensis, from the same stratigraphic horizons in the Wilson Lake core—species known for their tolerance to low-oxygen conditions. Bullet-shaped magnetofossils are also evident in late and middle Cretaceous sediments^82,83. Wagner et al.⁸³ found that the abundance of small bullet-shaped magnetofossils increased after Oceanic-Anoxic Event 2 (OAE2) relative to the pre-OAE2 interval, potentially due to enhanced seasonal stratification coupled with increased nutrient and organic matter inputs. Here, our study establishes another possible link between bullet-shaped magnetosomes and low-oxygen environments by revealing the vertical niche of Fe₃O₄-type MGC-containing Deltaproteobacteria, which most likely synthesize bullet-shaped magnetosomes, as evidenced in previous studies^84,85. It should be noted that magnetosome morphology was inferred from gene clusters rather than direct observation in this study. Further research is needed to confirm these associations.

In summary, our study reveals high relative abundance of MGC-containing bacteria in oxygen-stratified freshwater environments of the northern landscapes³⁵, with Deltaproteobacteria emerging as the most ubiquitous and predominant MGC-containing group. These findings highlight a gap in our understanding of the abundance, diversity, distribution, and ecological roles of MTB in natural environments, even after several decades of research. Additionally, the prevalence of Fe₃O₄-type MGC-containing Deltaproteobacteria in oxygen-depleted zones suggests that bullet-shaped magnetofossils could serve as a valuable proxy for oxygen-depleted conditions in paleoenvironmental studies.

Methods

Shotgun metagenomes

A total of 267 shotgun metagenomes were downloaded from the European Nucleotide Archive (ENA) under project accession number PRJEB38681 using AsperaSRAgetter v2.1⁸⁶, a Python package developed here to facilitate the download of sequencing data (fastq.gz format) from ENA via Aspera-CLI (Command Line Interface for IBM Aspera products) (https://github.com/IBM/aspera-cli). These metagenomes were sampled and reported by Buck et al.³⁵ (see Supplementary Data 1). Briefly, all metagenomes were sequenced on the Illumina NovaSeq platform using a paired-end 2 × 150 bp sequencing strategy at the Science for Life Laboratory (Uppsala University, Uppsala, Sweden)³⁵.

Reads quality control, metagenomic assembly, binning, and bin refinement

Raw reads of all metagenomes were trimmed and filtered using the Read_qc module of metaWRAP v1.1.5⁸⁷ with the “--skip-bmtagger” flag. For each metagenome, clean reads were then assembled using metaSPAdes v3.13.0⁸⁸ with the minimum contig length set to 2000 bp (“-l 2000”). Assemblies were then binned with the metaWRAP Binning module using three binners: metaBAT2 v2.12.1⁸⁹, MaxBin2 v2.2.6⁹⁰, and CONCOCT v1.0.0⁹¹. For each metagenome, the three original bin sets were then compared, and a refined bin set was generated using the metaWRAP Bin_refinement module with the minimum completion of bins set to 50% and the maximum contamination set to 10% (“-c 50 -x 10”).

Annotation and identification of MGC-containing genomes

Refined bins were annotated and screened for putative magnetosome gene clusters (MGCs) using MagCluster v0.2.2⁹². The process involved: (i) annotating bins with Prokka v1.13.3⁹³ with the e-value set to 1e⁻⁵; (ii) screening for putative MGCs with the mgc_screen module of MagCluster using default parameters; (iii) retrieving putative magnetosome protein sequences and comparing them against the NCBI non-redundant protein sequence database using NCBI Blast+ suite v2.12.0 (https://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/), using the Position-Specific Iterated (PSI-BLAST) algorithm with the number of iterations set to 1 (“-num_iterations 1”); (iv) manually checking PSI-BLAST results of all putative magnetosome proteins; and (v) visualizing MGCs using clinker v0.0.23⁹⁴.

Quality assessment of MGC-containing genomes

The quality of MGC-containing genomes, including completeness and contamination, was estimated with CheckM v1.0.12⁹⁵ using the “lineage_wf” workflow. Genomic statistics, including genome size, GC content, contigs N50/L50, and maximum contig length, were analyzed using QUAST v5.0.2⁹⁶.

Dereplication of redundant MGC-containing genomes

A non-redundant MGC-containing genomic dataset was generated using dRep v3.4.0⁹⁷. The quality filter for genomes was set to a minimum completeness of 50% and a maximum contamination of 10%, corresponding to medium-quality MAGs as defined by the MIMAG standard³⁶. The ANImf algorithm was used for ANI computation with thresholds for primary clusters set at 0.9 and for secondary clusters at 0.99. After dereplication, representative MGC-containing genomes formed a non-redundant dataset for further analysis.

Taxonomic classification and phylogenomic analysis of MGC-containing genomes

Taxonomic classification of 41 nr MGC-containing genomes was performed using the Genome Taxonomy Database Toolkit (GTDB-Tk) v2.1.0⁹⁸ with database release 207_v2 using the “classify_wf” workflow. The phylogeny of these genomes was analyzed together with 33 previously published MTB genomes and 60 non-MTB genomes. For each genome, 120 bacterial single-copy marker protein⁹⁹ sequences were identified, aligned, and concatenated using GTDB-Tk. The concatenated sequences were then used to infer the phylogeny. The phylogenomic tree was constructed using FastTree v2.1.10¹⁰⁰ under the WAG model (“-wag”). Bootstrap evaluation was performed using the Shimodaira-Hasegawa (SH)-like local bootstrap test with 1,000 replicates. The tree was rooted using three Patescibacteria genomes (NCBI accession numbers MFYJ01000001, MNWP00000000, CAIKAD010000001) and visualized with GC content, genome size, completeness, and contamination data using Interactive Tree Of Life (iTOL) v6¹⁰¹ and anvi’o v7.1¹⁰².

Profiling of MGC-containing genomes across all metagenomes

The 41 nr MGC-containing genomes were used to profile the distribution and abundance of putative MTB across all metagenomes by conducting a read recruitment procedure using anvi’o v7.1¹⁰² as described by Delmont et al.³⁷. Briefly, all contigs and scaffolds from the nr MGC-containing genomes were first concatenated into a single FASTA file. Next, Bowtie2 v2.3.5¹⁰³ was used to build the index files. Short reads from each metagenome were then mapped against all contigs and scaffolds with the “--no-unal” parameter, and the alignment results were stored as SAM files. Then, samtools¹⁰⁴ was used to convert the SAM files to BAM files. anvi’o v7.1 was used to profile each BAM file, estimating coverage and detection statistics for each scaffold. All mapping profiles were combined into a merged profile database for each metagenome. The “anvi-profile” program was used to calculate the coverage value per nucleotide position of each BAM file and stored as an anvi’o profile database. The “anvi-merge” program then combined all profile databases into a single merged profile database. Information on contigs or scaffolds was imported into their corresponding MTB MAGs using the “anvi-import-collection” command. Finally, “anvi-summarize” was used to report read recruitment statistics, including the coverage and detection data for each MGC-containing genome across all metagenomes.

Characterization of the presence and abundance of putative MTB across all metagenomes

Detection and coverage statistics of nr MGC-containing genomes across all metagenomes were calculated after read recruitment by anvi’o v7.1¹⁰² (Supplementary Data 3). A genome was considered detected in a metagenome when more than 25% of its sequence length was covered by sequencing reads^105,106. Coverage values below this limit were set to zero. To further eliminate non-specific mapping, nucleotide positions with coverage values falling within the 1st and 4th quartiles were excluded, and the interquartile mean coverage value was then used as a proxy for the relative abundance of putative MTB.

Metabolic characterization prediction

Gene function annotation was performed using KofamScan v1.3.0¹⁰⁷ by mapping protein sequences against the HMMs within the KOfam database, assigning KO identifiers to each gene. By default, an asterisk “*” is added if the mapping score is higher than the predefined threshold by KofamScan (see https://github.com/takaram/kofam_scan). Significantly mapped genes were filtered using a custom bash script (kofam.get.sig.sh, see https://github.com/RunJiaJi/MTBProfiling/). The completeness of key metabolic pathways was assessed and illustrated using KEGGDecoder v1.3¹⁰⁸. Pathway maps were then generated with the online KEGG Mapper Reconstruct tool (https://www.genome.jp/kegg/mapper/reconstruct.html) and manually verified.

Predatory-related function gene annotation

To investigate the predatory potential of MGC-containing MAGs, we analyzed seven Bdellovibrionota and one Myxococcota MAGs by comparing their protein sequences to those of model predatory strains Bdellovibrio bacteriovorus HD100 (GCA_000196175.1) and Myxococcus xanthus DK1622 (GCA_000012685.1). Using DIAMOND (v2.1.8)¹⁰⁹, protein sequences from the Bdellovibrionota MAGs were aligned against 11 lifecycle-associated proteins from Bdellovibrio bacteriovorus HD100⁷⁴ (Supplementary Data 7). For the Myxococcota MAG, 50 predation-related proteins from Myxococcus xanthus DK1622, encompassing type IV pilin proteins, Tad-like apparatus (Kil proteins), and myxovirescin biosynthetic enzymes, were annotated (Supplementary Data 7). Alignments were filtered with stringent criteria (E-value < 1e-5, identity > 30%, alignment length > 50) to identify homologs indicative of predatory traits.

Data processing and visualization

Data processing and statistical analyses were conducted in the interactive development environment of JupyterLab v3.3.2 (https://jupyter.org/). Python v3.9.5 (https://www.python.org/) was used, along with several libraries and packages. Data organization, integration, and cleaning were performed with pandas v1.3.4 (https://pandas.pydata.org/) and numpy v1.22.4 (https://numpy.org/). Spearman Correlation analysis between MTB abundance and environmental factors was carried out using scipy v1.8.1 (https://scipy.org/) with the scipy.stats.spearmanr function. The map of sample sites was created with QGIS v3.30 (https://qgis.org/en/site/). Data visualization for figures was conducted using matplotlib v3.5.2 (https://matplotlib.org/), altair v4.2.0 (https://altair-viz.github.io/), and seaborn v0.11.2 (https://seaborn.pydata.org/). All figures were finally organized using the open-source vector graphics editor Inkscape v1.2.2 (https://inkscape.org/).

Declaration of Generative AI and AI-assisted technologies in the writing process

The authors used ChatGPT to improve the English language of the introduction, results, and discussion sections of the manuscript. The authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Data availability

The raw metagenomic sequencing data are available in the European Nucleotide Archive (ENA) under project accession PRJEB38681. All MGC-containing genomes reconstructed in this study have been deposited in the NCBI BioProject under PRJNA400260 (accession numbers pending) and the eLMSG (an eLibrary of Microbial Systematics and Genomics, https://www.biosino.org/elmsg/index) under accession numbers LMSG_G000036956.1-LMSG_G000037018.1. Supplementary Data 1–7 are available with the online version of the article and have also been deposited on the Figshare repository at https://doi.org/10.6084/m9.figshare.29254040.v3.

Code availability

Codes and scripts used in the metagenomic analysis and data processing are freely available under the MIT License at https://github.com/RunJiaJi/MTBProfiling.

References

Carpenter, S. R., Stanley, E. H. & Vander Zanden, M. J. State of the world’s freshwater ecosystems: physical, chemical, and biological changes. Annu. Rev. Environ. Resour. 36, 75–99 (2011).
Article Google Scholar
Linz, A. M. et al. Freshwater carbon and nutrient cycles revealed through reconstructed population genomes. PeerJ 6, e6075 (2018).
Article Google Scholar
Shabarova, T. et al. Recovery of freshwater microbial communities after extreme rain events is mediated by cyclic succession. Nat. Microbiol. 6, 479–488 (2021).
Article CAS Google Scholar
Bazylinski, D. A. & Frankel, R. B. Magnetosome formation in prokaryotes. Nat. Rev. Microbiol. 2, 217–230 (2004).
Article CAS Google Scholar
Spring, S. & Bazylinski, D. A. Magnetotactic bacteria. Prokaryotes 2, 842–862 (2006).
Article Google Scholar
Jogler, C. et al. Cultivation-independent characterization of ‘Candidatus Magnetobacterium bavaricum’ via ultrastructural, geochemical, ecological and metagenomic methods. Environ. Microbiol 12, 2466–2478 (2010).
Article CAS Google Scholar
Lin, W. et al. Genomic expansion of magnetotactic bacteria reveals an early common origin of magnetotaxis with lineage-specific evolution. ISME J. 12, 1508–1519 (2018).
Article CAS Google Scholar
Koziaeva et al. Genome-Based Metabolic Reconstruction of a Novel Uncultivated Freshwater Magnetotactic Coccus “ Ca. Magnetaquicoccus Inordinatus” UR-1, and Proposal of a Candidate Family “ Ca. Magnetaquicoccaceae. Front. Microbiol. 10, 2290 (2019).
Article Google Scholar
Lin, W. et al. Expanding magnetic organelle biogenesis in the domain Bacteria. Microbiome 8, 152 (2020).
Article CAS Google Scholar
Wang, Y. et al. Complete Genome Sequence of Magnetospirillum sp. Strain XM-1, Isolated from the Xi’an City Moat, China. Genome Announc. 4, https://doi.org/10.1128/genomeA.01171-16 (2016).
Uzun, M. et al. Recovery and genome reconstruction of novel magnetotactic Elusimicrobiota from bog soil. ISME J. 17, 204–214 (2023).
Article CAS Google Scholar
Uebe, R. & Schüler, D. Magnetosome biogenesis in magnetotactic bacteria. Nat. Rev. Microbiol. 14, 621–637 (2016).
Article CAS Google Scholar
Gareev, K. G. et al. Magnetotactic bacteria and magnetosomes: Basic properties and applications. Magnetochemistry 7, 86 (2021).
Article CAS Google Scholar
Vargas, G. et al. Applications of magnetotactic bacteria, magnetosomes and magnetosome crystals in biotechnology and nanotechnology: mini-review. Molecules 23, 2438 (2018).
Article Google Scholar
Arakaki, A., Tanaka, M. & Matsunaga, T. Molecular mechanism of magnetic crystal formation in magnetotactic bacteria. in Biological Magnetic Materials and Applications 23–51 (Springer Singapore, 2018).
Basit, A., Wang, J., Guo, F., Niu, W. & Jiang, W. Improved methods for mass production of magnetosomes and applications: a review. Microb. Cell Factories 19, 197 (2020).
Article Google Scholar
Hesse, P. P. Evidence for bacterial palaeoecological origin of mineral magnetic cycles in oxic and sub-oxic Tasman Sea sediments. Mar. Geol. 117, 1–17 (1994).
Yamazaki, T. & Kawahata, H. Organic carbon flux controls the morphology of magnetofossils in marine sediments. Geology 26, 1064–1066 (1998).
Article CAS Google Scholar
He, K. & Pan, Y. Magnetofossil abundance and diversity as paleoenvironmental proxies: a case study from Southwest Iberian Margin Sediments. Geophys. Res. Lett. 47, e2020GL087165 (2020).
Article Google Scholar
Chang, L. et al. Indian Ocean glacial deoxygenation and respired carbon accumulation during mid-late Quaternary ice ages. Nat. Commun. 14, 4841 (2023).
Article CAS Google Scholar
Bazylinski, D. A. et al. Controlled biomineralization of magnetite (Fe(inf3)O(inf4)) and greigite (Fe(inf3)S(inf4)) in a magnetotactic bacterium. Appl. Environ. Microbiol. 61, 3232–3239 (1995).
Article CAS Google Scholar
Simmons, S. L. & Edwards, K. J. in Magnetoreception and Magnetosomes in Bacteria 77–102 (Springer, 2006).
Simmons, S. L., Sievert, S. M., Frankel, R. B., Bazylinski, D. A. & Edwards, K. J. Spatiotemporal distribution of marine magnetotactic bacteria in a seasonally stratified coastal salt pond. Appl. Environ. Microbiol. 70, 6230–6239 (2004).
Article CAS Google Scholar
Bazylinski, D. A. Synthesis of the bacterial magnetosome: the making of a magnetic personality. Int. Microbiol. 2, 71–80 (1999).
CAS Google Scholar
Grünberg, K., Wawer, C., Tebo, B. M. & Schuler, D. A large gene cluster encoding several magnetosome proteins is conserved in different species of magnetotactic bacteria. Appl. Environ. Microbiol. 67, 4573–4582 (2001).
Article Google Scholar
Lin, W., Pan, Y. & Bazylinski, D. A. Diversity and ecology of and biomineralization by magnetotactic bacteria. Environ. Microbiol. Rep. 9, 345–356 (2017).
Article CAS Google Scholar
Lin, W. et al. Origin of microbial biomineralization and magnetotaxis during the Archean. Proc. Natl. Acad. Sci. USA 114, 2171–2176 (2017).
Article CAS Google Scholar
Uzun, M., Alekseeva, L., Krutkina, M., Koziaeva, V. & Grouzdev, D. Unravelling the diversity of magnetotactic bacteria through analysis of open genomic databases. Sci. Data 7, 252 (2020).
Article Google Scholar
Goswami, P. et al. Magnetotactic bacteria and magnetofossils: ecology, evolution and environmental implications. npj Biofilms Microbiomes 8, https://doi.org/10.1038/s41522-022-00304-0 (2022).
Liu, P. et al. Diverse phylogeny and morphology of magnetite biomineralized by magnetotactic cocci. Environ. Microbiol. 23, 1115–1129 (2021).
Article CAS Google Scholar
Lin, W., Bazylinski, D. A., Xiao, T., Wu, L. F. & Pan, Y. Life with compass: diversity and biogeography of magnetotactic bacteria. Environ. Microbiol. 16, 2646–2658 (2014).
Article CAS Google Scholar
Spring, S. et al. Dominating role of an unusual magnetotactic bacterium in the microaerobic zone of a freshwater sediment. Appl. Environ. Microbiol. 59, 2397–2403 (1993).
Article CAS Google Scholar
Lin, W. et al. Genomic insights into the uncultured genus ‘Candidatus Magnetobacterium’ in the phylum Nitrospirae. ISME J. 8, 2463–2477 (2014).
Article CAS Google Scholar
Lin, W., Tian, L., Li, J. & Pan, Y. Does capillary racetrack-based enrichment reflect the diversity of uncultivated magnetotactic cocci in environmental samples?. FEMS Microbiol. Lett. 279, 202–206 (2008).
Article CAS Google Scholar
Buck, M. et al. Comprehensive dataset of shotgun metagenomes from oxygen stratified freshwater lakes and ponds. Sci. Data 8, 131 (2021).
Article CAS Google Scholar
Bowers, R. M. et al. Minimum information about a single amplified genome (MISAG) and a metagenome-assembled genome (MIMAG) of bacteria and archaea. Nat. Biotechnol. 35, 725–731 (2017).
Article CAS Google Scholar
Delmont, T. O. et al. Nitrogen-fixing populations of Planctomycetes and Proteobacteria are abundant in surface ocean metagenomes. Nat. Microbiol. 3, 804–813 (2018).
Article CAS Google Scholar
Waite, D. W. et al. Proposal to reclassify the proteobacterial classes Deltaproteobacteria and Oligoflexia, and the phylum Thermodesulfobacteria into four phyla reflecting major functional capabilities. Int. J. Syst. Evol. Microbiol. 70, 5972–6016 (2020).
Article CAS Google Scholar
Zhang, W., Wang, Y., Liu, L., Pan, Y. & Lin, W. Identification and genomic characterization of two previously unknown magnetotactic nitrospirae. Front. Microbiol. 12, https://doi.org/10.3389/fmicb.2021.690052 (2021).
Garritano, A. N., Song, W. & Thomas, T. Carbon fixation pathways across the bacterial and archaeal tree of life. PNAS Nexus 1, https://doi.org/10.1093/pnasnexus/pgac226 (2022).
Akhtar, S., Khan, A., Sohaskey, C. D., Jagannath, C. & Sarkar, D. Nitrite reductase NirBD is induced and plays an important role during in vitro dormancy of Mycobacterium tuberculosis. J. Bacteriol. 195, 4592–4599 (2013).
Article CAS Google Scholar
Huang, X. et al. Nitrate assimilation, dissimilatory nitrate reduction to ammonium, and denitrification coexist in Pseudomonas putida Y-9 under aerobic conditions. Bioresour. Technol. 312, 123597 (2020).
Article CAS Google Scholar
Huang, X., Luoluo, Xie, D. & Li, Z. Dissimilatory nitrate reduction to ammonium in four Pseudomonas spp. under aerobic conditions. Heliyon 9, e14983 (2023).
Article CAS Google Scholar
van Kessel, M. A. et al. Complete nitrification by a single microorganism. Nature 528, 555–559 (2015).
Article Google Scholar
Daims, H., Lücker, S. & Wagner, M. A new perspective on microbes formerly known as nitrite-oxidizing bacteria. Trends Microbiol. 24, 699–712 (2016).
Article CAS Google Scholar
Loy, A. et al. Reverse dissimilatory sulfite reductase as phylogenetic marker for a subgroup of sulfur-oxidizing prokaryotes. Environ. Microbiol. 11, 289–299 (2009).
Article CAS Google Scholar
Müller, A. L., Kjeldsen, K. U., Rattei, T., Pester, M. & Loy, A. Phylogenetic and environmental diversity of DsrAB-type dissimilatory (bi) sulfite reductases. ISME J. 9, 1152–1165 (2015).
Article Google Scholar
Mangin, C. C. et al. Magnetotactic bacteria affiliated with diverse Pseudomonadota families biomineralize intracellular Ca-carbonate. ISME J. 19, https://doi.org/10.1093/ismejo/wrae260 (2025).
Bazylinski, D. A. et al. Chemolithoautotrophy in the marine, magnetotactic bacterial strains MV-1 and MV-2. Arch. Microbiol. 182, 373–387 (2004).
Article CAS Google Scholar
Barrett, E. L. & Clark, M. A. Tetrathionate reduction and production of hydrogen sulfide from thiosulfate. Microbiol Rev. 51, 192–205 (1987).
Article CAS Google Scholar
Aketagawa, J., Kobayashi, K. & Ishimoto, M. Purification and properties of thiosulfate reductase from Desulfovibrio vulgaris, Miyazaki F. J. Biochem. 97, 1025–1032 (1985).
Article CAS Google Scholar
Busigny, V. et al. Mass collection of magnetotactic bacteria from the permanently stratified ferruginous Lake Pavin, France. Environ. Microbiol. https://doi.org/10.1111/1462-2920.15458 (2021).
Faivre, D. & Schüler, D. Magnetotactic bacteria and magnetosomes. Chem. Rev. 108, 4875–4898 (2008).
Article CAS Google Scholar
Ji, B. et al. The chimeric nature of the genomes of marine magnetotactic coccoid-ovoid bacteria defines a novel group of Proteobacteria. Environ. Microbiol. 19, 1103–1119 (2017).
Article CAS Google Scholar
Zhang, W. et al. Two metagenome-assembled genome sequences of magnetotactic bacteria in the order Magnetococcales. Microbiol. Resour. Announc. 9, https://doi.org/10.1128/mra.00363-20 (2020).
Moench, T. T. & Konetzka, W. A novel method for the isolation and study of a magnetotactic bacterium. Arch. Microbiol. 119, 203–212 (1978).
Article CAS Google Scholar
Spring, S., Amann, R., Ludwig, W., Schleifer, K.-H. & Petersen, N. Phylogenetic diversity and identification of nonculturable magnetotactic bacteria. Syst. Appl. Microbiol. 15, 116–122 (1992).
Article Google Scholar
Spring, S. et al. Phylogenetic affiliation and ultrastructure of uncultured magnetic bacteria with unusually large magnetosomes. Arch. Microbiol. 169, 136–147 (1998).
Article CAS Google Scholar
Flies, C. B., Peplies, J. & Schüler, D. Combined approach for characterization of uncultivated magnetotactic bacteria from various aquatic environments. Appl. Environ. Microbiol. 71, 2723 (2005).
Article CAS Google Scholar
Lin, W. & Pan, Y. Uncultivated magnetotactic cocci from yuandadu park in beijing, china. Appl. Environ. Microbiol. 75, 4046–4052 (2009).
Article CAS Google Scholar
Lins, U. & Farina, M. Magnetosome chain arrangement and stability in magnetotactic cocci. Antonie van Leeuwenhoek 85, 335–341 (2004).
Article CAS Google Scholar
Abreu, F. et al. Culture-independent characterization of novel psychrophilic magnetotactic cocci from Antarctic marine sediments. Environ. Microbiol. 18, 4426–4441 (2016).
Article CAS Google Scholar
Schulz-Vogt, H. N. et al. Effect of large magnetotactic bacteria with polyphosphate inclusions on the phosphate profile of the suboxic zone in the Black Sea. ISME J. 13, 1198–1208 (2019).
Article CAS Google Scholar
Bidaud, C. C. et al. Biogeochemical niche of magnetotactic cocci capable of sequestering large polyphosphate inclusions in the anoxic layer of the Lake Pavin Water Column. Front. Microbiol. 12, 789134 (2021).
Article Google Scholar
Wolfe, R., Thauer, R. & Pfennig, N. A ‘capillary racetrack’method for isolation of magnetotactic bacteria. FEMS Microbiol. Ecol. 3, 31–35 (1987).
Article Google Scholar
Jogler, C. et al. Toward cloning of the magnetotactic metagenome: identification of magnetosome island gene clusters in uncultivated magnetotactic bacteria from different aquatic sediments. Appl. Environ. Microbiol. 75, 3972 (2009).
Article CAS Google Scholar
Pan, Y. et al. Reduced efficiency of magnetotaxis in magnetotactic coccoid bacteria in higher than geomagnetic fields. Biophys. J. 97, 986–991 (2009).
Article CAS Google Scholar
Dziuba, M. V. et al. Silent gene clusters encode magnetic organelle biosynthesis in a non-magnetotactic phototrophic bacterium. ISME J. https://doi.org/10.1038/s41396-022-01348-y (2022).
Dziuba, M. V., Müller, F.-D., Pósfai, M. & Schüler, D. Exploring the host range for genetic transfer of magnetic organelle biosynthesis. Nat. Nanotechnol. 19, 115–123 (2024).
Article CAS Google Scholar
Kolinko, I. et al. Biosynthesis of magnetic nanostructures in a foreign organism by transfer of bacterial magnetosome gene clusters. Nat. Nanotechnol. 9, 193–197 (2014).
Article CAS Google Scholar
Awal, R. P., Lefevre, C. T. & Schüler, D. Functional expression of foreign magnetosome genes in the alphaproteobacterium Magnetospirillum gryphiswaldense. Mbio 14, e03282-22 (2023).
Mohr, K. I. Diversity of myxobacteria—we only see the tip of the iceberg. Microorganisms 6, 84 (2018).
Article Google Scholar
Cohen, Y. et al. Community and single cell analyses reveal complex predatory interactions between bacteria in high diversity systems. Nat. Commun. 12, 5481 (2021).
Article CAS Google Scholar
Caulton, S. G. & Lovering, A. L. Moving toward a better understanding of the model bacterial predator Bdellovibrio bacteriovorus. Microbiology 169, https://doi.org/10.1099/mic.0.001380 (2023).
Wang, Y., Li, X., Zhang, W., Zhou, X. & Li, Y. Z. The groEL2 gene, but not groEL1, is required for biosynthesis of the secondary metabolite myxovirescin in Myxococcus xanthus DK1622. Microbiology160, 488–495 (2014).
Article CAS Google Scholar
Perez, J., Contreras-Moreno, F. J., Munoz-Dorado, J. & Moraleda-Munoz, A. Development versus predation: transcriptomic changes during the lifecycle of Myxococcus xanthus. Front. Microbiol. 13, 1004476 (2022).
Article Google Scholar
Simunovic, V. et al. Myxovirescin A biosynthesis is directed by hybrid polyketide synthases/nonribosomal peptide synthetase, 3-hydroxy-3-methylglutaryl-CoA synthases, and trans-acting acyltransferases. ChemBioChem 7, 1206–1220 (2006).
Article CAS Google Scholar
Keane, R. & Berleman, J. The predatory life cycle of Myxococcus xanthus. Microbiology 162, 1–11 (2016).
Article CAS Google Scholar
Chen, A. et al. Magnetic properties of uncultivated magnetotactic bacteria and their contribution to a stratified estuary iron cycle. Nat. Commun. 5, 4797 (2014).
Article CAS Google Scholar
Moskowitz, B. M., Bazylinski, D. A., Egli, R., Frankel, R. B. & Edwards, K. J. Magnetic properties of marine magnetotactic bacteria in a seasonally stratified coastal pond (Salt Pond, MA, USA). Geophys. J. Int. 174, 75–92 (2008).
Article CAS Google Scholar
Wagner, C. L., Stassen, P., Thomas, E., Lippert, P. C. & Lascu, I. Magnetofossils and benthic foraminifera record changes in food supply and deoxygenation of the coastal marine seafloor during the Paleocene-Eocene Thermal Maximum. Paleoceanogr. Paleoclimatol. 37, e2022PA004502 (2022).
Article Google Scholar
Montgomery, P., Hailwood, E. A., Gale, A. S. & Burnett, J. A. The magnetostratigraphy of Coniacian-Late Campanian chalk sequences in southern England. Earth Planet. Sci. Lett. 156, 209–224 (1998).
Article CAS Google Scholar
Wagner, C. L. et al. Discovery of giant and conventional magnetofossils bookending Cretaceous Oceanic Anoxic Event 2. Commun. Earth Environ. 5, https://doi.org/10.1038/s43247-024-01540-2 (2024).
Lefèvre, C. T. et al. Comparative genomic analysis of magnetotactic bacteria from the Deltaproteobacteria provides new insights into magnetite and greigite magnetosome genes required for magnetotaxis. Environ. Microbiol. 15, 2712–2735 (2013).
Article Google Scholar
Amor, M., Mathon, F. P., Monteil, C. L., Busigny, V. & Lefevre, C. T. Iron-biomineralizing organelle in magnetotactic bacteria: function, synthesis and preservation in ancient rock samples. Environ. Microbiol 22, 3611–3632 (2020).
Article CAS Google Scholar
Ji, R. AsperaSRAgetter: a Python package to download sequencing data (fastq.gz format) from European Nucleotide Archive (ENA) by using Aspera-CLI (2023).
Uritskiy, G. V., DiRuggiero, J. & Taylor, J. MetaWRAP-a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome 6, 158 (2018).
Article Google Scholar
Nurk, S., Meleshko, D., Korobeynikov, A. & Pevzner, P. A. metaSPAdes: a new versatile metagenomic assembler. Genome Res. 27, 824–834 (2017).
Article CAS Google Scholar
Kang, D. D., Froula, J., Egan, R. & Wang, Z. MetaBAT, an efficient tool for accurately reconstructing single genomes from complex microbial communities. PeerJ 3, e1165 (2015).
Article Google Scholar
Wu, Y.-W., Simmons, B. A. & Singer, S. W. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics 32, 605–607 (2016).
Article CAS Google Scholar
Alneberg, J. et al. Binning metagenomic contigs by coverage and composition. Nat. methods 11, 1144–1146 (2014).
Article CAS Google Scholar
Ji, R., Zhang, W., Pan, Y. & Lin, W. MagCluster: a tool for identification, annotation, and visualization of magnetosome gene clusters. Microbiol. Resour. Announc. 11, e01031–01021 (2022).
Article CAS Google Scholar
Seemann, T. Prokka: rapid prokaryotic genome annotation. Bioinformatics 30, 2068–2069 (2014).
Article CAS Google Scholar
Gilchrist, C. L. M. & Chooi, Y.-H. clinker & clustermap.js: automatic generation of gene cluster comparison figures. Bioinformatics 37, 2473–2475 (2021).
Article CAS Google Scholar
Parks, D. H., Imelfort, M., Skennerton, C. T., Hugenholtz, P. & Tyson, G. W. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 25, 1043–1055 (2015).
Article CAS Google Scholar
Gurevich, A., Saveliev, V., Vyahhi, N. & Tesler, G. QUAST: quality assessment tool for genome assemblies. Bioinformatics 29, 1072–1075 (2013).
Article CAS Google Scholar
Olm, M. R., Brown, C. T., Brooks, B. & Banfield, J. F. dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. ISME J. 11, 2864–2868 (2017).
Article CAS Google Scholar
Chaumeil, P.-A., Mussig, A. J., Hugenholtz, P. & Parks, D. H. GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database. Bioinformatics 36, 1925–1927 (2019).
Article Google Scholar
Parks, D. H. et al. Recovery of nearly 8,000 metagenome-assembled genomes substantially expands the tree of life. Nat. Microbiol. 2, 1533–1542 (2017).
Article CAS Google Scholar
Price, M. N., Dehal, P. S. & Arkin, A. P. FastTree 2–approximately maximum-likelihood trees for large alignments. PLoS ONE 5, e9490 (2010).
Article Google Scholar
Letunic, I. & Bork, P. Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res. https://doi.org/10.1093/nar/gkae268 (2024).
Eren, A. M. et al. Community-led, integrated, reproducible multi-omics with anvi’o. Nat. Microbiol. 6, 3–6 (2021).
Article CAS Google Scholar
Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).
Article CAS Google Scholar
Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Article Google Scholar
Delmont, T. O. Discovery of nondiazotrophic Trichodesmium species abundant and widespread in the open ocean. Proc. Natl. Acad. Sci. USA 118, e2112355118 (2021).
Article CAS Google Scholar
Delmont, T. O. et al. Heterotrophic bacterial diazotrophs are more abundant than their cyanobacterial counterparts in metagenomes covering most of the sunlit ocean. ISME J. 16, 927–936 (2022).
Article CAS Google Scholar
Aramaki, T. et al. KofamKOALA: KEGG Ortholog assignment based on profile HMM and adaptive score threshold. Bioinformatics 36, 2251–2252 (2020).
Article CAS Google Scholar
Graham, E. D., Heidelberg, J. F. & Tully, B. J. Potential for primary productivity in a globally-distributed bacterial phototroph. ISME J. 12, 1861–1866 (2018).
Article CAS Google Scholar
Buchfink, B., Reuter, K. & Drost, H. G. Sensitive protein alignments at tree-of-life scale using DIAMOND. Nat. Methods 18, 366–368 (2021).
Article CAS Google Scholar

Download references

Acknowledgements

We thank Moritz Buck for permission to use the metagenomic data reported in their study. We also thank the Supercomputing Laboratory of the Institute of Geology and Geophysics, Chinese Academy of Sciences, for their support in data transfer and storage. We are thankful to Wensi Zhang for assistance with downloading the metagenomic datasets and to Jia Liu for her helpful advice on drafting the figures. Additionally, we appreciate Yinzhao Wang for his advice on the metabolism analysis. This work was supported by the National Natural Science Foundation of China (NSFC) grants 42388101 and T2225011, and by the CAS Project for Young Scientists in Basic Research (YSBR-097).

Author information

Authors and Affiliations

Key Laboratory of Planetary Science and Frontier Technology, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, China
Runjia Ji, Juan Wan, Pranami Goswami, Jianxun Shen, Yongxin Pan & Wei Lin
France-China Joint Laboratory for Evolution and Development of Magnetotactic Multicellular Organisms, Chinese Academy of Sciences, Beijing, China
Runjia Ji, Juan Wan, Pranami Goswami, Jianxun Shen, Yongxin Pan & Wei Lin
College of Earth and Planetary Sciences, University of Chinese Academy of Sciences, Beijing, China
Runjia Ji, Juan Wan, Pranami Goswami, Yongxin Pan & Wei Lin
Frontiers Science Center for Deep Ocean Multispheres and Earth System, Key Laboratory of Submarine Geosciences and Prospecting Techniques, MOE and College of Marine Geosciences, State Key Laboratory of Physical Oceanography, Ocean University of China, Qingdao, China
Kuang He
Laoshan Laboratory, Laboratory for Marine Mineral Resources, Qingdao Marine Science and Technology Center, Qingdao, China
Kuang He
Research School of Earth Sciences, Australian National University, Canberra, ACT, Australia
Pranami Goswami

Authors

Runjia Ji
View author publications
Search author on:PubMed Google Scholar
Juan Wan
View author publications
Search author on:PubMed Google Scholar
Kuang He
View author publications
Search author on:PubMed Google Scholar
Pranami Goswami
View author publications
Search author on:PubMed Google Scholar
Jianxun Shen
View author publications
Search author on:PubMed Google Scholar
Yongxin Pan
View author publications
Search author on:PubMed Google Scholar
Wei Lin
View author publications
Search author on:PubMed Google Scholar

Contributions

W.L. and R.J.J. designed the study. R.J.J. conducted the metagenomic analysis and statistical analysis. R.J.J. and W.L. drafted the manuscript with input from J.W., K.H., P.G., J.X.S., and Y.X.P. All authors contributed to critical revisions and approved the final manuscript.

Corresponding author

Correspondence to Wei Lin.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Communications Earth & Environment thanks Marina Dziuba and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Somaparna Ghosh and Heike Langenberg. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Transparent Peer Review file

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Ji, R., Wan, J., He, K. et al. Magnetosome gene cluster containing bacteria in oxygen-stratified freshwater ecosystems of northern landscapes. Commun Earth Environ 6, 644 (2025). https://doi.org/10.1038/s43247-025-02495-8

Download citation

Received: 27 September 2024
Accepted: 17 June 2025
Published: 08 August 2025
DOI: https://doi.org/10.1038/s43247-025-02495-8