In Vivo Evidence of Single 13C and 15N Isotope-Labeled Methanotrophic Nitrogen-Fixing Bacterial Cells in Rice Roots.

Methane-oxidizing bacteria (methanotrophs) play an ecological role in methane and nitrogen fluxes because they are capable of nitrogen fixation and methane oxidation, as indicated by genomic and cultivation-dependent studies. However, the chemical relationships between methanotrophy and diazotrophy and aerobic and anaerobic reactions, respectively, in methanotrophs remain unclear. No study has demonstrated the cooccurrence of both bioactivities in a single methanotroph bacterium in its natural environment. Here, we demonstrate that both bioactivities in type II methanotrophs occur at the single-cell level in the root tissues of paddy rice (Oryza sativa L. cv. Nipponbare). We first verified that difluoromethane, an inhibitor of methane monooxygenase, affected methane oxidation in rice roots. The results indicated that methane assimilation in the roots mostly occurred due to oxygen-dependent processes. Moreover, the results indicated that methane oxidation-dependent and methane oxidation-independent nitrogen fixation concurrently occurred in bulk root tissues. Subsequently, we performed fluorescence in situ hybridization and NanoSIMS analyses, which revealed that single cells of type II methanotrophs (involving six amplicon sequence variants) in paddy rice roots simultaneously and logarithmically fixed stable isotope gases 15N2 and 13CH4 during incubation periods of 0, 23, and 42 h, providing in vivo functional evidence of nitrogen fixation in methanotrophic cells. Furthermore, 15N enrichment in type II methanotrophs at 42 h varied among cells with an increase in 13C accumulation, suggesting that either the release of fixed nitrogen into root systems or methanotroph metabolic specialization is dependent on different microenvironmental niches in the root. IMPORTANCE Atmospheric methane concentrations have been continually increasing, causing methane to become a considerable environmental concern. Methanotrophy may be the key to regulating methane fluxes. Although research suggests that type II methanotrophs are involved in methane oxidation aerobically and nitrogen fixation anaerobically, direct evidence of simultaneous aerobic and anaerobic bioreactions of methanotrophs in situ is still lacking. In this study, a single-cell isotope analysis was performed to demonstrate these in vivo parallel functions of type II methanotrophs in the root tissues of paddy rice (Oryza sativa L. cv. Nipponbare). The results of this study indicated that methanotrophs might provide fixed nitrogen to root systems or depend on cells present in the spatially localized niche of the root tissue. Furthermore, our results suggested that single type II methanotrophic cells performed simultaneous methane oxidation and nitrogen fixation in vivo. Under natural conditions, however, nitrogen accumulation varied at the single-cell level.

OBSERVATION

Methane is a powerful greenhouse gas, and atmospheric methane concentrations are increasing rapidly (1). However, methane-oxidizing bacteria (methanotrophs) can reduce methane fluxes and thus mitigate climate change (1). Methanotrophs are divided mainly into two major phylogenetic groups: type I (Gammaproteobacteria) and type II (Alphaproteobacteria) (2, 3). Many methanotrophs are involved in nitrogen fixation (3–5) and thus may participate in environmental nitrogen cycling (3). Indeed, the nitrogenase structural genes nifHDK were completely encoded on 88.8% of the genomes of type I methanotrophs and on 98.3% of the genomes of type II methanotrophs in the current publicly available methanotrophic genomes (80 and 37 genomes, respectively; see Table S1 in the supplemental material), suggesting that most of methanotrophs, especially type II, could drive nitrogen fixation. Furthermore, the nitrogenase gene (nifH) sequences of both types of methanotrophs have been detected from terrestrial, freshwater, and marine environments across the world (Fig. S1), suggesting that diazotrophic methanotrophs may be ubiquitously distributed worldwide.

Geographic and habitat distributions of methanotrophs based on the nifH gene in the families Methylococcaceae (type I methanotrophs) and Methylocystaceae (type II methanotrophs). The nifH genes, encoding nitrogenase reductase, were retrieved from the NCBI database and carefully examined as nifH genes of Methylococcaceae and Methylocystaceae. The map (A) and habitat table (B) illustrate the features noted in the data. Methylocystaceae (type II methanotrophs) carrying the nifH gene were widely distributed in many countries (green and yellow in A) and in terrestrial environments, including freshwater habitats (gray in B). Download FIG S1, PDF file, 0.6 MB.

Presence/absence of methane monooxygenase and nitrogen fixation structural genes (nifHDK) in available genomes of methanotrophs. Download Table S1, PDF file, 0.2 MB.

Rice paddy fields are a hot spot for methane metabolism and a habitat of type II methanotrophs (6–8). Methane monooxygenase (MMO) and nitrogenase of type II methanotrophs were simultaneously expressed in rice root-associated bacteria in a low-N paddy field (6). Type II methanotrophs exhibit an endophytic lifestyle in the vascular cylinders and epidermal cell layers of root tissues (6, 9). Nitrogenase for nitrogen fixation requires anoxic conditions, whereas bacterial methane oxidation in type II methanotrophs requires molecular oxygen. Pure culture experiments demonstrated that type II methanotrophs isolated from rice roots fix nitrogen in a methane-dependent manner (methane oxidation-dependent nitrogen fixation) (10). However, oxygen is a key element in both of these reactions. Oxygen regimens in rice roots in paddy fields vary from the aerobic vascular cylinder to the anoxic epidermis (11). This leads to the creation of heterogenous microenvironments with different oxygen levels that affect methane oxidation-dependent nitrogen fixation by individual microbial cells in these microenvironmental niches. Although studies have identified the specific metabolic activities of methanotrophs, they used only test tube-based homogenous methanotroph cultures (4, 10). Therefore, whether methanotrophs can mediate methane oxidation and nitrogen fixation simultaneously at the single-cell level in their natural environments remains unclear (12).

Formaldehyde (HCHO) is a central intermediate in methanotroph metabolism, and its carbon sources are derived from methane through a dissimilation pathway (CH4→CH3OH→HCHO) mediated by (i) methane monooxygenase and methanol dehydrogenase and (ii) subsequent HCHO assimilation, such as that in the serine pathway (2). This pathway indicates that 13CH4 can enrich methanotroph compounds in 13C.

It would be difficult to conduct an in situ experiment in a paddy field due to the continuous production of methane from organic matter by methanogens living in anaerobic soil and the rice rhizosphere, and this would dilute the 13C concentration of the methane by simple isotope addition. Although slightly different from the natural condition, in this study, we performed in vivo stable isotope labeling experiments by using paddy rice roots and single-cell imaging (fluorescence in situ hybridization [FISH] and NanoSIMS). Our results provide in vivo functional evidence of nitrogen fixation by type II methanotrophs residing in the root tissues of paddy rice at the single-cell level (details of root preparation, stable isotope feeding, mass spectrometry, bacterial cell extraction, amplicon sequencing, and FISH-NanoSIMS analyses are provided in Text S1).

Supplementary materials and methods. Download Text S1, PDF file, 0.3 MB.

The root systems of rice plants grown in a paddy field were incubated with 13CH4/15N2 for 24 h with and without difluoromethane (DFM), a methane monooxygenase inhibitor (13). A low concentration of DFM is known to effectively and selectively inhibit methanotrophy by competing as a substrate for MMO (13, 14). 13C concentrations in roots exposed to 13CH4/15N2 with DFM were identical to the natural abundance of 13C (control, 1.07 atom%) in rice roots, and 13C concentrations in the samples without DFM were significantly higher than the natural abundance level. The enrichment of 13C in the sample without DFM could indicate oxygen-dependent 13CH4 oxidation and assimilation by methanotrophs residing in the roots (Fig. 1A). 15N concentrations in roots exposed to 13CH4/15N2 with DFM were significantly higher than the natural abundance of 15N (control, 0.366 atom%). Moreover, the absence of DFM significantly increased 15N concentrations in the roots exposed to 13CH4/15N2 (Fig. 1B). These results suggest that both methane-dependent and methane-independent nitrogen fixation occur in paddy rice roots. On the basis of differences in 15N concentrations in the rice roots with and without DFM, methane-dependent nitrogen fixation was determined to be 0.49 μmol N2 fixed (g root weight)−1 day−1, accounting for 65% of total nitrogen fixation (Table S2).

FIG 1

(A to C) 13CH4 and 15N2 concentrations in rice roots with and without methane oxidization inhibitor and the NanoSIMS experiment. (A and B) 13C (A) and 15N (B) concentrations of rice roots fed with 13CH4 (5% [vol/vol], 99 atom% 13C), 15N2 (39% [vol/vol], 40.8 atom% 15N), and 5% (vol/vol) O2 in Ar balance for 24 h with the addition of difluoromethane (DFM; 0.5% [vol/vol]), a methane monooxygenase inhibitor; “control” indicates root samples before isotope feeding. Average values with the same letter are not significantly different according to Tukey’s honestly significant difference test (P < 0.05). (C) 13C and 15N concentrations in the root samples determined by performing NanoSIMS analysis, with the root systems of field-grown rice plants incubated with a gas phase containing 13CH4 (6% [vol/vol], 99 atom% 13C), 15N2 (35% [vol/vol], 99.4 atom% 15N), and O2 (12% [vol/vol]) in Ar balance for 0, 23, and 42 h. Bolded horizontal bars in A and B indicate the averages of four replicates.

Nitrogen fixation in rice roots estimated by 15N enrichment and N content. Download Table S2, PDF file, 0.2 MB.

For the NanoSIMS analysis, field-grown rice roots were incubated again with a gas containing 13CH4/15N2 (99.4 atom% 15N) for 0, 23, and 42 h. The concentrations of both 13C and 15N increased with incubation time, suggesting that methanotrophic nitrogen fixation occurred in the root samples (Fig. 1C). We then subjected bacterial cells extracted from the rice root tissues (15) to FISH-NanoSIMS analyses. The amplicon sequences of the 16S rRNA gene of the bacterial cells indicated an abundance of type II methanotrophs (Methylocystaceae; average, 7.2%), including six amplicon sequence variants (ASVs; Table S3A and B) that were phylogenetically split equally into two genera (Methylocystis and Methylosinus) (Fig. S2); ASV0004 (belonging to Methylosinus) was most dominant among the samples (average, 6.61%; Table S3A and Fig. S2). In contrast, type I methanotrophs were assigned with only a single ASV (belonging to Methylococcus; average abundance, 0.1%) (Table S3A and Fig. S3). All methanotrophic ASVs were widely positioned in the phylogenetic tree, and each ASV was close to a respective genome that presented particulate MMO (pMMO), soluble MMO (sMMO), and the nifHDK gene cluster (Fig. S2 and S3), suggesting that all ASVs could potentially participate in both methane oxidation and nitrogen fixation. We also confirmed that all ASV sequences of type II methanotrophs were identical to the sequence of the FISH probe Ma450 from a previous study (16), and no other bacterial ASVs were not matched to the probe Ma450. The FISH analyses performed using the Ma450 probe for type II methanotrophs (16) and the EUB338mix probe for total eubacteria (17) indicated that the proportion of type II methanotrophs in total bacterial cells ranged from 6.1% to 7.2% (Fig. S4); this result is in agreement with the amplicon sequences of the 16S rRNA gene.

Phylogenetic analysis of the 16S rRNA gene from six amplicon sequence variants (ASVs) and available genomes of type II methanotrophs. Bootstrap values greater than 50% derived from 1,000 replicates are also shown as black dots. Colored symbols indicate which sequences are matched to a Ma450 probe (green) and which genomes contained particulate methane monooxygenase (pMMO; blue), soluble methane monooxygenase (sMMO; light blue), and the three nitrogenases nifH (yellow green), nifD (orange), and nifK (red). Download FIG S2, PDF file, 0.2 MB.

Phylogenetic analysis of the 16S rRNA gene from one amplicon sequence variants (ASV) and available genomes of type I methanotrophs. Bootstrap values greater than 50% derived from 1,000 replicates are also shown as black dots. Colored symbols indicate which genomes contained particulate methane monooxygenase (pMMO; blue), soluble methane monooxygenase (sMMO; light blue), and the three nitrogenases nifH (yellow green), nifD (orange), and nifK (red). Download FIG S3, PDF file, 0.2 MB.

Proportion (%) of cells identified as type II methanotrophs (green) and other eubacteria (red) based on FISH images at all time points (0, 23, and 42 h). Download FIG S4, PDF file, 0.1 MB.

Relative abundance of type I and II methanotrophs (A) and major abundance sequence variants (ASVs) (B) extracted from rice roots (at 0 h). Table S3, XLSX file, 0.03 MB

The subsequent NanoSIMS analysis revealed the overlapping images of δ13C and δ15N signals in the cells that were merged using Ma450 probe signals in the 23-h and 42-h specimens (Fig. 2A). We determined 13C and 15N atom% of more than 100 cells in the NanoSIMS images, and a significant difference was observed in 13C and 15N concentrations between type II methanotrophs and other eubacteria. After stable isotope feedings, all type II methanotrophs exclusively enriched 13C and 15N concentrations, while other eubacteria did not (Fig. 2B and C). As depicted in the scatterplot in Fig. 2D, a significant positive correlation was noted between 13C and 15N enrichment for type II methanotrophs at all time points. For every mole of 13C that was assimilated, an average 0.76 mol of 15N were fixed at the single-cell level (Fig. 2D), which was three times higher than the previously reported value of 0.25 mol (18, 19).

FIG 2

Methane assimilation and nitrogen fixation of type II methanotrophs and other eubacteria in rice roots at the single-cell level. (A) Example parallel images of FISH, carbon isotope ratio (log10 [13C/12C]), nitrogen isotope ratio (log10 [12C15N/12C14N]), 12C14N counts, and scanning electron micrographs (SEM) for symbiotic microbes in rice roots at 0 h, 23 h, and 42 h of incubation. Green fluorescence in FISH images indicates type II methanotrophic bacteria, and red fluorescence indicates other eubacteria (hybridized with Ma450 and EUB338 mix probes labeled with Alexa 488 and Cy3, respectively). Type II methanotrophic cells hybridized with both probes (yellow signals indicated with white dashed lines). Arrows indicate regions with high ratios of carbon and nitrogen isotopes without FISH signals (suggesting dead cells as the cause because FISH targets rRNA in cells). Scale bars indicate 5 μm. (B and C) Statistics for carbon (B) (log10 [13C atom%]) and nitrogen (C) isotopic composition (log10 [15N atom%]) for type II methanotrophic bacteria and other eubacteria individuals. Asterisks indicate significant differences between type II methanotrophic bacteria and other eubacteria in unpaired two-sample Student’s t test (P < 0.01). (D) Carbon (log10 [13C atom%]) and nitrogen isotopic composition (log10 [15N atom%]) for type II methanotrophic bacteria and other eubacteria individuals presented as a scatterplot. The linear regression indicates a significant positive correlation between 13C and 15N enrichment for type II methanotrophic bacteria at all time points (P < 0.01). Ad. Rindicates Adjusted R-Squared.

Concentrations of 13C and 15N increased with high variability on a logarithmic scale at 23 and 42 h. Notably, a large fluctuation in the 15N enrichment of type II methanotrophs occurred among the cells at 42 h, and a slight saturation of 13C was noted (Fig. 2B and C). This finding indicates that at 42 h, individual type II methanotrophic cells enabled the accumulation of either 13C alone or both 13C and 15N in the root tissues (Fig. S5), which may be allowed by the creation of heterogeneous microenvironmental niches of type II methanotrophs, including in the vascular cylinders and epidermal cell layers of root tissues (6, 9). In addition, inter- or intraspecific variation of oxygen sensitivity in methanotrophs, mainly type II, has been reported (Table S4). It is also possible that different methanotrophic species or strains in the root system could differ in their sensitivity to oxygen. Given that the metabolic specialization of heterogeneous nitrogen fixation can occur at the single-cell level in diazotroph cyanobacteria (20), some type II methanotrophic cells also may transform into low- or nonnitrogen-fixing mode to save energy for creating an anoxic microenvironment.

Representative correlated FISH and NanoSIMS images indicating variations in 15N accumulation within type II methanotrophic cells (dotted lines) in the root tissue at 42 h after isotope incubation and 13C enrichment in both cells. (A) FISH panel of type II methanotrophs and other eubacteria hybridized with Ma450 (green, Alexa 488) and EUB338 mix (red, Cy3) probes. (B to E) NanoSIMS mapping images of carbon isotope ratio (13C/12C), nitrogen isotope ratio (12C15N/12C14N), 12C14N counts, and secondary electron (SE) micrographs. Scale bars represent 5 μm. Download FIG S5, JPG file, 2.4 MB.

Variation of oxygen sensitivity in type I and II methanotrophs. Download Table S4, XLSX file, 0.01 MB.

Interestingly, this varied pattern of 15N enrichment in single cells differed from a marked increase in 15N and 13C concentrations observed up to 42 h in bulk root tissues (Fig. 1C). This outcome suggests the potential influence of several potential factors, such as the accumulation from other nitrogen fixers including type I methanotrophs and/or a release of fixed nitrogen (ammonium or organic nitrogen) from type II methanotrophs at the single-cell level into the root system. In fact, peatland methanotrophs can provide not only carbon but also nitrogen to peat mosses, suggesting carbon and nitrogen accumulation in the field (12). Although further work is need, our findings expand our knowledge of the intact carbon and nitrogen cycle at the single-bacterial-cell level, particularly in the paddy rice root system.

Because type II methanotrophs in intact root tissues accumulated stable isotopes from both 13CH4 and 15N2 gases at the single-cell level (Fig. 2D), root-associated type II methanotrophs might have simultaneously performed methane oxidation and assimilation and methane-dependent nitrogen fixation in vivo in the root tissues of paddy rice. Given that nitrogen fixation heterogeneously varied at the single-cell level, we hypothesize that type II methanotrophic cells contribute nitrogen flux to root systems after nitrogen fixation or affect root systems’ nitrogen accumulation through the creation of microenvironmental niches. Our findings provide insights into potential in situ interactions that occur between methanotrophy and diazotrophy in terrestrial carbon and nitrogen cycles (12, 21) as well as in agricultural settings (9, 18, 22).