Diazotrophic Anaeromyxobacter Isolates from Soils

Anaeromyxobacter is globally distributed in soil environments, especially predominant in paddy soils. Current studies based on environmental DNA/RNA analyses frequently detect gene fragments encoding nitrogenase of Anaeromyxobacter from various soil environments. Although the importance of Anaeromyxobacter as a diazotroph in nature has been suggested by culture-independent studies, there has been no solid evidence and validation from genomic and culture-based analyses that Anaeromyxobacter fixes nitrogen. This study demonstrates that Anaeromyxobacter harboring nitrogenase genes exhibits diazotrophic ability; moreover, N₂-dependent growth was demonstrated in vitro and in the soil environment. Our findings indicate that nitrogen fixation is important for Anaeromyxobacter to survive under nitrogen-deficient environments and provide a novel insight into the environmental function of Anaeromyxobacter, which is a common bacterium in soils.

ABSTRACT

Biological nitrogen fixation is an essential reaction in a major pathway for supplying nitrogen to terrestrial environments. Previous culture-independent analyses based on soil DNA/RNA/protein sequencing could globally detect the nitrogenase genes/proteins of Anaeromyxobacter (in the class Deltaproteobacteria), commonly distributed in soil environments and predominant in paddy soils; this suggests the importance of Anaeromyxobacter in nitrogen fixation in soil environments. However, direct experimental evidence is lacking; there has been no research on the genetic background and ability of Anaeromyxobacter to fix nitrogen. Therefore, we verified the diazotrophy of Anaeromyxobacter based on both genomic and culture-dependent analyses using Anaeromyxobacter sp. strains PSR-1 and Red267 isolated from soils. Based on the comparison of nif gene clusters, strains PSR-1 and Red267 as well as strains Fw109-5, K, and diazotrophic Geobacter and Pelobacter in the class Deltaproteobacteria contain the minimum set of genes for nitrogenase (nifBHDKEN). These results imply that Anaeromyxobacter species have the ability to fix nitrogen. In fact, Anaeromyxobacter PSR-1 and Red267 exhibited N₂-dependent growth and acetylene reduction activity (ARA) in vitro. Transcriptional activity of the nif gene was also detected when both strains were cultured with N₂ gas as a sole nitrogen source, indicating that Anaeromyxobacter can fix and assimilate N₂ gas by nitrogenase. In addition, PSR-1- or Red267-inoculated soil showed ARA activity and the growth of the inoculated strains on the basis of RNA-based analysis, demonstrating that Anaeromyxobacter can fix nitrogen in the paddy soil environment. Our study provides novel insights into the pivotal environmental function, i.e., nitrogen fixation, of Anaeromyxobacter, which is a common soil bacterium.

IMPORTANCE Anaeromyxobacter is globally distributed in soil environments, especially predominant in paddy soils. Current studies based on environmental DNA/RNA analyses frequently detect gene fragments encoding nitrogenase of Anaeromyxobacter from various soil environments. Although the importance of Anaeromyxobacter as a diazotroph in nature has been suggested by culture-independent studies, there has been no solid evidence and validation from genomic and culture-based analyses that Anaeromyxobacter fixes nitrogen. This study demonstrates that Anaeromyxobacter harboring nitrogenase genes exhibits diazotrophic ability; moreover, N₂-dependent growth was demonstrated in vitro and in the soil environment. Our findings indicate that nitrogen fixation is important for Anaeromyxobacter to survive under nitrogen-deficient environments and provide a novel insight into the environmental function of Anaeromyxobacter, which is a common bacterium in soils.

INTRODUCTION

Biological nitrogen fixation is an essential step in the nitrogen cycle. Apart from anthropogenic processes such as fertilization (1), biological nitrogen fixation is the major pathway to supply nitrogen to terrestrial environments. Ammonium converted from ambient nitrogen gas via nitrogen fixation is pivotal for the nitrogen nutritional needs of plants; thus, terrestrial ecosystems and sustainable agriculture are dependent on nitrogen fixation. Since nitrogen-fixing microorganisms, i.e., diazotrophs, were first discovered in the root nodules of leguminous plants in the late 19th century (2), nitrogen fixation activity has been well demonstrated in various plant-associated and free-living prokaryotes, particularly in bacteria belonging to the Alpha-, Beta-, and Gammaproteobacteria, Cyanobacteria, Firmicutes, and Actinobacteria (3–6). Moreover, recent genomic analyses show that a broader taxonomic range of bacteria harbors nitrogen fixation genes, suggesting that the phylogenetic diversity of diazotrophs is higher than previously believed (7).

We recently evaluated microbial communities involved in reductive nitrogen transformations, including denitrification (NO₃⁻→NO₂⁻→NO→N₂O→N₂), dissimilatory nitrate reduction to ammonium (DNRA; NO₃⁻→NO₂⁻→NH₄⁺), and nitrogen fixation (N₂→NH₄⁺) in paddy soils using shotgun sequencing of soil DNA and RNA, metagenomics, and metatranscriptomics (8). Nitrogen fixation genes (nif) and nif gene transcripts of Geobacter and Anaeromyxobacter (within the class Deltaproteobacteria) were detected much more frequently than those of other bacterial genera, suggesting that Geobacter and Anaeromyxobacter are major diazotrophs in paddy soils (8). nif gene fragments and Nif protein of Anaeromyxobacter have also been detected by culture-independent methods from other soil environments such as upland agricultural, crop rhizospheric, grassland, and polar soils (9–15), suggesting that diazotrophic Anaeromyxobacter occurs commonly in soil environments and has a pivotal role in nitrogen fixation. Although the reports have been limited, N₂ gas-dependent growth and nitrogen fixation activity have been demonstrated by culture-dependent methods in two species of Geobacter, G. metallireducens and G. sulfurreducens (16, 17). However, to date, there has been no research on the genetic and enzymatic backgrounds of Anaeromyxobacter nitrogenase; and the nitrogen fixation activity of Anaeromyxobacter has not been reported.

Anaeromyxobacter, a member of the Myxococcales order in the class Deltaproteobacteria, is a facultative anaerobe with the abilities to reduce various metals such as iron and uranium, dechlorinate aromatic compounds, and perform NO₃⁻ reductions to NO₂⁻ and ammonium via DNRA (18–20). Recent culture-dependent and -independent studies have found that Anaeromyxobacter is a key player driving the reduction of N₂O to N₂ in terrestrial environments (8, 21, 22). Despite possessing no nitrite reductase for catalyzing NO₂⁻ reduction to NO, which is an essential reaction for denitrification, Anaeromyxobacter can completely denitrify by itself through the chemical transformation of NO₂⁻ to NO via Fe²⁺ oxidation (23). Thus, Anaeromyxobacter carries out multiple activities to transform nitrogen, especially the reductive processes of the nitrogen cycle. Notably, DNA fragments of Anaeromyxobacter except for nif genes have also been commonly found in various soil environments such as agriculture, forest, tundra, desert, grassland, permafrost, polar, and sedimentary soils (21, 22, 24–30). Especially in paddy soils, river sediments, and some upland agricultural soils, Anaeromyxobacter has been reported as one of the most predominant members of the bacterial community by some culture-independent studies (8, 31–34). These observational studies highlight that Anaeromyxobacter is not rare but ubiquitous in soil environments and strongly suggest that Anaeromyxobacter would be a common and important driver of nitrogen fixation as well as other nitrogen reductive reactions in soil environments. However, as described above, the nitrogen-fixing activity of Anaeromyxobacter has not yet been verified at all despite its ubiquitousness in soils. Therefore, the demonstration of diazotrophic activity in Anaeromyxobacter is valuable for the comprehensive understanding of microbial drivers of the nitrogen cycle in soil environments. In this study, we verified the diazotrophy of Anaeromyxobacter based on both genomic and culture-dependent analyses using Anaeromyxobacter strains isolated from soils.

RESULTS

Isolation and identification of Anaeromyxobacter derived from paddy soil.

Prior to this study, genomic information from five cultured strains of Anaeromyxobacter, namely, PSR-1, 2CP-1^T, 2CP-C, Fw109-5, and K, isolated from sediment, contaminated soil, and forest soil has been available to the public (Table 1). Genomic information about Anaeromyxobacter derived from paddy soils has been lacking, although this organism is one of the predominant diazotrophs in paddy soils as suggested by our previous metatranscriptomic analysis (8). To obtain such information, we first isolated an Anaeromyxobacter strain from paddy soil. After 2 weeks of incubating a paddy soil slurry, the reduction of Fe³⁺ to Fe²⁺ changed the soil’s color from reddish brown to gray; Fe²⁺ concentration shifted from 0.75 ± 0.03 to 498 ± 39 mg/100 g soil slurry before and after incubation, respectively. The gray soil was spread on R2Af medium and anaerobically incubated, resulting in the isolation of one anaerobic strain, designated Red267. The strain formed a red colony on R2Af, similar to that of Anaeromyxobacter dehalogenans 2CP-1^T (19). Comparison of a partial sequence (1,454 bp) of the 16S rRNA gene of strain Red267 to database sequences showed the closest similarity (96.56%) to A. dehalogenans 2CP-1^T. Based on these results, the entire genome of Red267 was sequenced and analyzed. The full 16S rRNA gene sequence of Red267 is also most similar (96.72%) to that of A. dehalogenans 2CP-1^T. Phylogenetic analysis based on the full-length 16S rRNA gene demonstrates that Red267 belongs to the Anaeromyxobacter cluster (Fig. 1A). The Red267 genome has an extremely high G+C content (74.5%), similar to those of other Anaeromyxobacter strains (73.5% to 74.9%) (Table 1). These results indicate that Red267 belongs to the genus Anaeromyxobacter.

TABLE 1.

Genomically sequenced strains of Anaeromyxobacter

Strain	Isolation origin	Stage of genomic sequencing (accession no.)	Size (Mb)	GC content (%)	No. of genes	nif genes	Reference(s)
Anaeromyxobacter sp. PSR-1	Arsenic-contaminated soil	Draft (BAZG00000000)	4.86	74.4	4,670	nifBHDKENXA	35, 36
Anaeromyxobacter sp. Red267	Paddy soil	Draft (BJTG00000000)	4.83	74.5	4,301	nifBHDKENXAW	This study
A. dehalogenans 2CP-1T	Stream sediment	Complete (CP001359)	5.03	74.7	4,560	None	19
A. dehalogenans 2CP-C	Rain forest soil	Complete (CP000251)	5.01	74.9	4,522	None	19, 66
Anaeromyxobacter sp. Fw109-5	Nitrate- and uranium-contaminated sediment	Complete (CP000769)	5.28	73.5	4,662	nifBHDKENXA	37
Anaeromyxobacter sp. K	Terrestrial	Complete (CP001131)	5.06	74.8	4,528	nifBHDKENXA	—a

^a—, deposited in INSD database, but no valid publication.

FIG 1.

Phylogenetic relationship of Anaeromyxobacter strains PSR-1 and Red267. Neighbor-joining phylogenies were inferred from the aligned nucleotide and amino acid sequences of 16S rRNA genes (A) and NifHDK (B), respectively, retrieved from complete genomic information (except for strains PSR-1 and Red267) listed in Table S2 in the supplemental material. ●, strains harboring a nif cluster comprising at least the minimum set of genes for nitrogen fixation (nifHDKENB). Only bootstrap values higher than 70% are shown at each branching node. Strains PSR-1 and Red267 examined in this study are highlighted by black boxes.

Genetic and enzymatic backgrounds of nitrogenase in Anaeromyxobacter.

Based on our analysis of the published and validated genomic data of six cultured Anaeromyxobacter strains, we found that strains PSR-1 and Red267, isolated in our previous and present studies (35, 36), respectively, as well as strains Fw109-5 and K harbor nifBHDKENXA (37) (Table 1). No nif gene was found in the complete genomes of strains 2CP-1^T and 2CP-C (Table 1). Based on an extensive survey of genomic information of the orders in the class Deltaproteobacteria, we found that among the order Myxococcales, only some representatives of Anaeromyxobacter harbor nif genes (Fig. 1A). In contrast, most strains in the order Desulfuromonadales, including the genera Geobacter, Pelobacter, and Desulfuromonas, have at least nifHDKENB (Fig. 1A). The NifHDK (comprising the essential and catalytic compartments of nitrogenase [7]) sequences in Anaeromyxobacter showed the closest phylogenetic relationship to those in Geobacter, within the order Desulfuromonadales (Fig. 1B).

The structure of the nif gene cluster of strains PSR-1 and Red267 also resembles that of other Anaeromyxobacter sequences; six consecutive genes (nifHDKENX-fdxN) are conserved in all Anaeromyxobacter nif gene clusters (Fig. 2A). The nif gene clusters of strains PSR-1 and Red267 are composed of genes involved in regulation (draT and draG in Red267, fdxN, and nifA [38–40]), construction of nitrogenase (nifHDK), and biosynthesis of FeMo cofactor (nifBENX). In addition, nifBHDKENX-fdxN shares similar features with corresponding sequences in diazotrophic Geobacter and Pelobacter (16, 17, 41). Moreover, the nif gene cluster of Anaeromyxobacter is more compact than that of other Desulfuromonadales; that is, all genes are located closer to each other in Anaeromyxobacter than in other genera (Fig. 2A). All Anaeromyxobacter strains consistently contain nifA and the gene encoding the GCN-related N-acetyltransferase, while these genes are absent in closely related species.

FIG 2.

Genetic and enzymatic backgrounds of nitrogenase in Anaeromyxobacter. (A) Comparison of the nif gene cluster in Anaeromyxobacter and its relatives. Regions a to g (not shown) contain 13, 1, 4, 5, 3, 7, and 26 genes, respectively. Homologous genes are shown by the same colors. (B) Alignments around crucial residues of NifHDK in Anaeromyxobacter and other diazotrophs. Amino acid residues shaded by red indicate the MgATP binding motif (P-loop) and the [4Fe-4S] cluster ligating cysteines in NifH, FeMo cofactor ligands in NifD, and P-cluster ligands in NifK (42). Numbering of each crucial residue is derived from Azotobacter vinelandii, a typical diazotroph. *, amino acids conserved in sequences of all strains.

The catalytic site of nitrogenase is made from the NifHDK complex and comprises the gamma subunit of the iron protein and beta and alpha subunits of the molybdenum-iron protein (42). Alignment of NifH sequences showed that the P-loop/MgATP binding motif and cysteinyl ligands (Cys97 and Cys132) of the [4Fe-4S] cluster are conserved in the NifH of Anaeromyxobacter sp. strains PSR-1 and Red267 as well as those of other diazotrophs (Fig. 2B). Furthermore, crucial residues of the FeMo cofactor binding site in NifD (Cys275 and His442) and the P cluster in NifK (Cys70, Cys95, and Cys153) are also conserved in NifDK of all strains of Anaeromyxobacter (Fig. 2B).

Fixation and assimilation of N₂.

Although strains PSR-1 and Red267 are cultivable in minimum medium containing NH₄⁺ with acetate and fumarate as an electron donor and acceptor, respectively (MMaf; see Materials and Methods), both strains also grew under culture conditions with N₂ as a sole nitrogen source (Fig. 3). The abundance levels of strains PSR-1 and Red267 harboring the nif gene cluster increased from (5.8 ± 0.5) × 10⁸ and (7.5 ± 0.6) × 10⁸ cells/vial to (1.3 ± 0.1) × 10¹⁰ and (1.0 ± 0.1) × 10¹⁰ cells/vial (mean ± standard deviation [SD]; n = 3), respectively, after 5 days of culturing. Exponential growth of strains PSR-1 and Red267 lasted for approximately 2 days before the cell abundance peaked. Neither strain grew with Ar in replacement of N₂ (data not shown). In contrast, A. dehalogenans 2CP-1^T harboring no nif gene did not grow in culture with N₂ as the sole nitrogen source; the abundance of A. dehalogenans 2CP-1^T remained constant at 10⁸ cells/vial even after 5 days of culture (Fig. 3).

FIG 3.

Time-dependent growth of Anaeromyxobacter strains cultured with N₂ as a sole nitrogen source. Anaeromyxobacter strains with nif (PSR-1 [●] and Red267 [▲]) and without nif (2CP-1^T [■]) were grown in minimal medium without NH₄Cl containing acetate and fumarate as an electron donor and acceptor, respectively, in serum bottles with an N₂ gas atmosphere. Means ± standard deviations (n = 3) are shown.

Nitrogen fixation activity and transcription of the nif gene.

Nitrogen fixation activity was estimated based on C₂H₂ reduction activity (ARA) as described in Materials and Methods. Cultured strains PSR-1 and Red267 showed ARA ([6.2 ± 1.0] × 10⁻¹⁰ and [1.5 ± 0.1] × 10⁻¹⁰nmol C₂H₄/h/cell, respectively) in the absence of NH₄⁺, indicating that both nif-containing Anaeromyxobacter strains fix nitrogen (Fig. 4A). Neither strain produced ethylene in the absence of acetylene (data not shown). No ARA was detected in strain 2CP-1^T, a strain harboring no nif gene. There have been reports that NH₄⁺ inhibits the nitrogenase activity of well-known diazotrophs such as Azospirillum spp., Azotobacter spp., and Rhizobium spp. (43, 44) as well as diazotrophic Geobacter (16). The ARA levels (i.e., nitrogenase activity) of strains PSR-1 and Red267 also decreased as the NH₄⁺ concentration in the medium increased, and ARA was completely inhibited by more than 0.7 and 0.5 mM NH₄⁺, respectively (Fig. 4A). Transcriptional activities of nifD in strains PSR-1 and Red267 were significantly higher in the absence of NH₄⁺ than with 1 mM NH₄⁺ (Fig. 4B). Strains PSR-1 and Red267 could utilize Fe(III)-nitrilotriacetic acid (NTA) and ferrihydrite (γ-FeOOH) but not goethite (α-FeOOH) and hematite (α-Fe₂O₃) as an electron acceptor (see Table S1 in the supplemental material). The ARAs of both strains were (1.7 ± 0.1) × 10⁻⁷ and (1.2 ± 0.3) × 10⁻⁷nmol C₂H₄/h/cell when cultured with 2 mM acetate and 10 mM Fe(III)-NTA as electron donor and acceptor, respectively. Strains PSR-1 and Red267 also reduced acetylene ([7.8 ± 0.3] × 10⁻⁹ and [1.6 ± 0.3] × 10⁻⁸nmol C₂H₄/h/cell) when 10 mM ferrihydrite (γ-FeOOH) was added as an electron acceptor in replacement of 10 mM Fe(III)-NTA.

FIG 4.

Inhibition effect of NH₄⁺ on the nitrogen fixation by Anaeromyxobacter. (A) Nitrogen fixation activity of strains PSR-1 (●) and Red267 (▲) cultured with 0 to 1 mM NH₄⁺ was estimated based on ARA. The actual maximum ARAs of strains PSR-1 and Red267 were (6.2 ± 1.0) × 10⁻¹⁰ and (1.5 ± 0.1) × 10⁻¹⁰nmol C₂H₄/h/cell at 0 mM NH₄, respectively. (B) Transcriptional activity of nifD in Anaeromyxobacter strains cultured with or without NH₄⁺, normalized by rpoB transcription level. *, P < 0.05 by Mann-Whitney U test. Means ± standard deviations (n = 3) are shown.

In situ nitrogen fixation activity of Anaeromyxobacter.

After demonstrating the nitrogen-fixing abilities of strains PSR-1 and Red267 in vitro, we investigated whether Anaeromyxobacter fixes nitrogen in paddy soil environments by using a soil microcosm. We inoculated Anaeromyxobacter in sterilized paddy soil and incubated the soil. The number of copies of the 16S rRNA gene and ARA in the soil slurries inoculated with strain PSR-1 or strain Red267 were below the detection limit (<10³ copies/g soil and <0.1 nmol/vial, respectively) before the incubation, i.e., immediately after the inoculation of each strain (Table 2). After 3 days of incubation at 30°C, the number of copies of the 16S rRNA genes of strain PSR-1 and strain Red267 in the soil slurries increased to (2.4 ± 2.1) × 10⁵ and (1.4 ± 0.6) × 10⁵ copies/g soil, respectively, indicating that both strains grew in the soil slurries (Table 2). The reduction of acetylene was detected in slurries inoculated with strains PSR-1 and Red267 (0.24 ± 0.33 and 0.58 ± 0.38 nmol C₂H₄/g-soil/h, respectively), but not in the uninoculated slurries (Table 2).

TABLE 2.

Nitrogen fixation activity of Anaeromyxobacter spp. in the paddy soil microcosm

Inoculum	No. of 16S rRNAs (copies/g soil)	C2H4 (nmol/g soil/h)
Before incubationa	NDb	0
Distilled water (control)	ND	0
Anaeromyxobacter sp. PSR-1	(2.38 ± 2.14) × 105	0.24 ± 0.33
Anaeromyxobacter sp. Red267	(1.39 ± 0.56) × 105	0.58 ± 0.38

^aJust after the inoculation of each Anaeromyxobacter strain.

^bND, not detected (under the detection limit [<10³ copies/g soil]).

DISCUSSION

In the present study, the diazotrophy of Anaeromyxobacter, which had been suggested by our previous metatranscriptomics study in paddy soils, was proven by genomic and culture-dependent analyses.

The analysis of the complete genomes of four published Anaeromyxobacter strains (2CP-1^T, 2CP-C, K, and Fw109-5) revealed that some Anaeromyxobacter strains (K and Fw109-5) possess nif genes. In addition to strain PSR-1, derived from arsenic-contaminated soil, we successfully isolated a new Anaeromyxobacter strain (Red267) from paddy soil, and the subsequent genomic analysis revealed that both strains harbor a nif gene cluster (Table 1). Remarkably, the 16S rRNA gene of strain PSR-1 is 99.68% and 99.74% similar to those of strains 2CP-1^T and 2CP-C, respectively, suggesting that strain PSR-1 must belong to the same species as 2CP-1^T and 2CP-C (i.e., A. dehalogenans). However, strain PSR-1 harbors a nif gene cluster, which is lacking in strains 2CP-1^T and 2CP-C. Therefore, we cannot conclude whether diazotrophs comprise a major group within the genus Anaeromyxobacter because of the limited genomic information and low number of isolates currently available (19, 35, 45). Moreover, even if the 16S rRNA gene of Anaeromyxobacter has been detected in environmental samples during polymorphism analysis of PCR amplicons of the 16S rRNA gene, it is too early to conclude about the role that Anaeromyxobacter plays as one of the diazotrophs in environments. Such a conclusion would require us first to expand the genomic information by isolating more Anaeromyxobacter strains from soil environments.

The structure of the nif gene cluster of Anaeromyxobacter is highly conserved in the genus, and it is comprised of nifHDKENBXA, fdxN, and draT (Fig. 2A). Although such a structure is extremely compact compared to those of the nif gene clusters in facultative anaerobic diazotrophs such as Azotobacter vinelandii, Pseudomonas stutzeri, Cyanothece sp., etc. (46), Anaeromyxobacter contains the minimum gene set for nitrogenase, nifHDKENBX, which is also conserved in diazotrophic Geobacter and Pelobacter (Fig. 2A) (16, 17, 41). In addition, the essential residues of the active site of nitrogenase have also been conserved (Fig. 2B). Therefore, it is plausible that translational products from the nif gene cluster in Anaeromyxobacter constitute active nitrogenase.

A functional nif gene cluster in Anaeromyxobacter strains PSR-1 and Red267 is suggested by N₂-dependent growth and ARA in vitro. These traits are not observed with strain 2CP-1^T, a strain with no nif genes (Fig. 3). Also, the nitrogen-fixing activity of Anaeromyxobacter was suppressed by NH₄⁺ (Fig. 4A), a response observed in known diazotrophs, including Geobacter (16, 43). Transcriptional activity of the nif gene was also detected when strains PSR-1 and Red267 were cultured with N₂ gas as a sole nitrogen source (Fig. 4B). These results clearly demonstrated that Anaeromyxobacter can fix and assimilate N₂ gas by its nitrogenase. Previous culture-dependent studies validated the physiological features of Anaeromyxobacter, including metal reduction, dechlorination, N₂O reduction, and denitrification (19–21, 23), all of which are also known to play pivotal roles in soil ecosystems. This study revealed that the nitrogen fixation is also one of the physiological features of Anaeromyxobacter, which is a common soil bacterium, enhancing the importance of Anaeromyxobacter in soil ecosystems.

Ferric iron compounds are the representative electron acceptors for Anaeromyxobacter and a typical component in soil environments (47, 48). Strains PSR-1 and Red267 were not able to use all ferric iron compounds tested in this study as an electron acceptor; they used Fe(III)-NTA and ferrihydrite (γ-FeOOH) but not goethite (α-FeOOH) and hematite (α-Fe₂O₃) (see Table S1 in the supplemental material), suggesting that the nitrogen-fixing activity of Anaeromyxobacter in soils would be greatly affected by the forms of ferric iron compounds in each soil. In addition to ferric iron compounds, soils contain other substances that are candidates for electron acceptors, e.g., nitrate, Mn⁴⁺, humic acid, etc. Therefore, it is necessary to further investigate the details regarding Anaeromyxobacter’s profile for availability of other electron acceptors than irons and carefully determine whether Anaeromyxobacter exhibits nitrogen fixation in each soil environment. To accomplish these tasks, we need to isolate more strains of Anaeromyxobacter and clarify their physiological characteristics, as mentioned above.

Anaeromyxobacter is known to predominate in paddy soils worldwide (8, 31, 32), and its nif gene transcripts and nitrogenase have been detected in previous metatranscriptomics and metaproteomics surveys of paddy soil (8, 9). The question of whether Anaeromyxobacter actually fixes nitrogen to survive in paddy soils was answered by a soil microcosm experiment conducted using only native paddy soil without any additives. The microcosm was inoculated with Anaeromyxobacter PSR-1 and Red267, and both exhibited nitrogen fixation activity (ARA). The growth of each of the inoculated strains was indicated by RNA-based analysis (Table 2), demonstrating that Anaeromyxobacter can fix and assimilate nitrogen in the paddy soil environment. In our previous field survey, the NH₄⁺ concentration temporarily increased after fertilization; however, no or less NH₄⁺ was detected both before fertilization and 4 weeks after fertilization (0 to 0.2 mmol/kg soil [49]). Acetate is present in some paddy fields at a concentration of 7.8 ± 1.9 mM throughout the year (50), a concentration at which PSR-1 and Red267 exhibited ARA in vitro (see Fig. S1). Moreover, they were able to use Fe(III)-NTA and ferrihydrite as an electron acceptor and showed ARA when cultured with acetate as the electron donor (Table S1). Thus, it is quite conceivable that Anaeromyxobacter fixes nitrogen in paddy fields where the NH₄⁺ concentration is less but acetate and ferric irons adequately exist (49).

We have demonstrated that some strains of Anaeromyxobacter have a nif gene cluster in their genomes and show diazotrophic traits in the soil environment. Considering the global distribution of Anaeromyxobacter as described above, Anaeromyxobacter is likely to be involved in nitrogen fixation in various soil environments. However, it remains unclear how much they actually contribute to nitrogen fixation in nature compared to that of other diazotrophs. A PCR-based method may underestimate the amount of Anaeromyxobacter nif gene in soil environments because of its extremely high G+C content, as also discussed in a previous study (8). In fact, the functional gene of Anaeromyxobacter failed to amplify during PCR despite the improved PCR primers (51). Future studies aiming to estimate the contribution of Anaeromyxobacter to nitrogen fixation will find it worthwhile to summarize and reanalyze the diversity of diazotrophic microbiota using soil metagenomic and metatranscriptomic data from around the world. Moreover, stable isotope probing will be a useful addition (52, 53) to expanding culture-independent investigations.

MATERIALS AND METHODS

Soil sample.

Soil was collected from paddy fields at the Niigata Agricultural Research Institute (Nagaoka, Niigata, Japan: 37°26′N, 138°52′E). This soil was also used in previous RNA-based studies (8, 48). Detailed characteristics of the soil, total carbon (12.4 g/kg), total nitrogen (0.85 g/kg), etc., were described previously (54). The soil was air dried at 25°C for >1 week and then finely crushed by a pestle and mortar. After debris and plant residues were removed by a 2-mm sieve, the prepared soil was used for the isolation of Anaeromyxobacter. For in situ assessment of nitrogen fixation activity of Anaeromyxobacter as described below, the soil was used without air drying. The Fe²⁺ concentration of soil samples was measured as described previously (49).

Anaeromyxobacter strains.

In addition to Anaeromyxobacter sp. strain PSR-1 previously isolated from arsenic-contaminated soil in Japan (35), Anaeromyxobacter sp. strain Red267 was newly isolated from paddy soil by the following method. The air-dried paddy soil was suspended in distilled water (soil/water, 1/1.5 [wt/vol]), and 5 ml of the resulting soil suspension was transferred to a 15-ml serum bottle. The serum bottle was autoclaved and supplemented with 0.1 g raw paddy soil and 5 μl of vitamin solution (liter⁻¹: 0.020 mg d-biotin, 0.045 mg folic acid, 0.100 mg pyridoxine-HCl, 0.050 mg thiamine-HCl, 0.050 mg nicotinic acid, 0.050 mg dl-calcium pantothenate, 0.001 mg vitamin B₁₂, 0.050 mg p-aminobenzoic acid, 0.050 mg dl-alpha lipoic acid). Bottles were sealed with a butyl rubber stopper with an aluminum crimp. The gaseous phase of the serum bottle was exchanged with N₂/CO₂ (80:20 [vol/vol]), and then the bottle was kept at 30°C without shaking. After 2 weeks of incubation, soil slurry was spread on R2A broth “DAIGO” (Nihon Pharmaceutical, Tokyo, Japan) medium supplemented with 1.5% agar and 5 mM disodium fumarate (R2Af) (19) and incubated at 30°C for 10 days under anaerobic conditions using a sealed jar equipped with an AnaeroPack (Mitsubishi Gas Chemical, Tokyo, Japan). After single-colony isolation, the partial sequence of the 16S rRNA gene of strain Red267 was determined and identified as described previously (55). Strains PCR-1 and Red267 were deposited to the Marine Culture Collection of China, as strains MCCC 1K04378 and MCCC 1K04211, respectively.

Both strains, PSR-1 and Red267, were subcultured using minimal medium (MM; liter⁻¹: 0.535 g NH₄Cl, 0.136 g KH₂PO₄, 0.204 g MgCl₂·6H₂O, 0.147 g CaCl₂·2H₂O, 2.52 g NaHCO₃, 1 ml trace mineral element solution [liter⁻¹: 64 mg nitrilotriacetic acid, 6.750 mg FeCl₂·6H₂O, 0.500 mg MnCl₂·6H₂O, 0.120 mg CoCl₂·2H₂O, 0.500 mg CaCl₂·2H₂O, 0.500 mg ZnCl₂, 0.125 mg CuCl₂·2H₂O, 0.050 mg H₃BO₃, 0.120 mg NaMoO₄·2H₂O, 5.000 mg NaCl, 0.600 mg NiCl₂·6H₂O, 0.130 mg Na₂S₂O₃·5H₂O], and 1 ml vitamin solution [35]) supplemented with 2 mM acetate and 40 mM fumarate (MMaf) and anaerobically incubated in serum bottles under an N₂/CO₂ (80:20 [vol/vol]) atmosphere at 30°C without shaking. A type strain of Anaeromyxobacter, A. dehalogenans 2CP-1^T (DSM21875^T), was obtained from the Leibniz-Institut Deutsche Sammlung von Mikroorganismen und Zellkulturen and used as a negative control in this study because it lacks nif gene in its genome (Table 1). Ability of Anaeromyxobacter to use ferric iron as an electron acceptor was tested in MM with 2 mM acetate as an electron donor based on visual change of the medium (red turbid to clear with/without black precipitation) derived from the reduction of ferric iron to ferrous iron, as described previously (56, 57).

Sequence analyses.

For the draft genomic sequencing, genomic DNA was extracted from strain Red267 cells grown on R2Af using the DNeasy blood and tissue kit (Qiagen, Germany), according to the manufacturer’s instructions for Gram-negative bacteria. Resulting DNA was sequenced using the Illumina HiSeq 2500 system at Beijing Novogene Bioinformatics Technology Co. Ltd. (Beijing, China), and sequence reads were assembled using SOAPdenovo version 2.04 (58). The draft genome of strain Red267 was analyzed using the DDBJ Fast Annotation and Submission Tool ver. 1.1.0 with default settings (59), with annotations added manually.

The nucleotide and amino acid sequences of the 16S rRNA gene and nif gene of Anaeromyxobacter spp. and related strains were obtained from complete and fine genomic information, which are available in the KEGG database (https://www.genome.jp/kegg/) (see Table S2 in the supplemental material). The draft genomic information of Anaeromyxobacter sp. PSR-1 (36) was retrieved from the DDBJ/ENA/GenBank database. Then, multiple alignments of nucleotide and amino acid sequences of the 16S rRNA gene and NifHDK, respectively, were generated using the ClustalW algorithm (60). Phylogenies were inferred based on the neighbor-joining and maximum likelihood methods with 1,000 replications of bootstrap tests using MEGA ver. 7.0.26 (61).

In vitro assessment of nitrogen fixation ability.

After preculturing in MMaf medium for 3 days as described above, cells of Anaeromyxobacter spp. were rinsed with and transferred into 20 ml of nitrogen-free MMaf (i.e., MMaf without NH₄Cl) in serum bottles in which the gaseous phase had been exchanged with N₂/CO₂ gas (80:20 [vol/vol]). As a control experiment, the same system was set up using pure Ar instead of N₂/CO₂ gas. During incubation at 30°C without shaking, cell growth was monitored by direct counting of the cells using a Multisizer 3 (Beckman Coulter, CA, USA). Cells cultured with or without NH₄⁺ under the same conditions were immediately frozen in liquid nitrogen and stored at −80°C until RNA extraction. RNA was extracted and purified from the cultured cells using ISOGEN with Spin Column (Nippon Gene, Tokyo, Japan) and RNA Clean and Concentrator (Zymo Research, CA, USA), respectively, according to the manufacturers’ instructions and used for determining the transcriptional activity of nif gene by quantitative PCR (qPCR) analysis as described below.

To complement the monitoring of cell growth, nitrogen fixation activity was also measured by the traditional ARA method based on C₂H₂ reduction into C₂H₄ by nitrogenase (62). ARA experimental conditions were the same as described above, except the atmosphere in bottles was Ar/C₂H₂ gas (90:10 [vol/vol]). The negative control was set up using pure Ar gas in replacement of Ar/C₂H₂ gas. C₂H₄ production in the gaseous phase was measured by gas chromatography equipped with a fused-silica column (RT-U PLOT; Restek) as described previously (63). The inhibitory effect of NH₄⁺ on nitrogen fixation activity was checked in cultures maintained at 30°C for 3 days in nitrogen-free MM added with 0 to 1 mM NH₄Cl. The effects of different concentrations of acetate on nitrogen fixation activity were evaluated based on ARA as described above on cultures incubated at 30°C for 3 days in MM supplemented with 40 mM fumarate as an electron acceptor.

In situ assessment of nitrogen fixation activity.

To construct a soil microcosm, soil suspension (soil/water, 1/2.3 [wt/vol]) was placed in a serum bottle, sealed with a butyl rubber stopper and an aluminum crimp, and autoclaved. Then, the gaseous phase was exchanged with N₂/CO₂ (80:20 [vol/vol]). Anaeromyxobacter spp. were precultured in nitrogen-free MM supplemented with 2 mM acetate and 40 mM fumarate under N₂/CO₂ (80:20 [vol/vol]) at 30°C for 4 days. Cultured cells were collected from 20 ml of culture broth by centrifugation at 6,000 rpm, rinsed, and then resuspended in 500 μl sterilized O₂-free water. This suspension was spiked into the soil microcosm prepared as described above. A negative control consisted of the same setup spiked with only 500 μl of sterilized O₂-free water. After incubating at 30°C for 3 days, 2 g of soil slurry was collected, immediately frozen in liquid nitrogen, and stored at −80°C until RNA extraction. The remaining 8 g of soil slurry was subjected to ARA assays as described above. RNA was extracted from soil slurries before and after incubation as described previously (64) and used for qPCR analysis for the 16S rRNA gene of the spiked Anaeromyxobacter as described below.

qPCR analysis.

cDNA was synthesized from prepared RNA using DNA-free RNA using the ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, Tokyo, Japan) according to the manufacturer’s instructions. qPCR was performed to specifically amplify nifD, rpoB, and 16S rRNA genes of Anaeromyxobacter using the respective specific primer sets (Table 3) and the StepOnePlus system (Life Technologies, CA, USA) with TB Green Premix Ex Taq GC (TaKaRa Bio, Shiga, Japan). The thermal cycling for all genes was performed as follows: preheat at 95°C for 30 s followed by 40 cycles of 95°C for 5 s, 52°C for 30 s, and 72°C for 30 s.

TABLE 3.

PCR primer sets used in this study

Target gene	Direction	Sequence (5′→3′)	Reference
rpoB	Forward	GTGGTCGATGTCGTCGAYGT	This study
	Reverse	CAACCTGTTCTTCAACGCCG	This study
nifD	Forward	CGWCATCTTCTTCGCGAC	This study
	Reverse	TGCAGWTSACSTCCGACTTC	This study
16S rRNA	Forward (515F)	GTGCCAGCMGCCGCGGTAA	55
	Reverse (806R)	GGACTACHVGGGTWTCTAAT	55

Statistical analysis.

Mann-Whitney U tests were performed using R software ver. 3.0.1 (65) to analyze data on the transcriptional activity of nifD.

Data availability.

The partial sequence of the 16S rRNA gene and the draft genome of Anaeromyxobacter sp. strain Red267 have been deposited in the DDBJ/GenBank/EBI databases under accession numbers LC488723 and BJTG01000001 to BJTG01000017, respectively.