Evidence for nitrite-dependent anaerobic methane oxidation as a previously overlooked microbial methane sink in wetlands Bao-lan Hua, Li-dong Shena, Xu Liana, Qun Zhub, Shuai Liua, Qian Huanga, Zhan-fei Hea, Sha Genga, Dong-qing Chengb, Li-ping Loua, Xiang-yang Xua, Ping Zhenga, and Yun-feng Hea,1 a Department of Environmental Engineering, Zhejiang University, Hangzhou 310058, China; and bDepartment of Life Science, Zhejiang Chinese Medical University, Hangzhou 310053, China

The process of nitrite-dependent anaerobic methane oxidation (n-damo) was recently discovered and shown to be mediated by “Candidatus Methylomirabilis oxyfera” (M. oxyfera). Here, evidence for n-damo in three different freshwater wetlands located in southeastern China was obtained using stable isotope measurements, quantitative PCR assays, and 16S rRNA and particulate methane monooxygenase gene clone library analyses. Stable isotope experiments confirmed the occurrence of n-damo in the examined wetlands, and the potential n-damo rates ranged from 0.31 to 5.43 nmol CO2 per gram of dry soil per day at different depths of soil cores. A combined analysis of 16S rRNA and particulate methane monooxygenase genes demonstrated that M. oxyfera-like bacteria were mainly present in the deep soil with a maximum abundance of 3.2 × 107 gene copies per gram of dry soil. It is estimated that ∼0.51 g of CH4 m−2 per year could be linked to the n-damo process in the examined wetlands based on the measured potential n-damo rates. This study presents previously unidentified confirmation that the n-damo process is a previously overlooked microbial methane sink in wetlands, and n-damo has the potential to be a globally important methane sink due to increasing nitrogen pollution. activity

| methane cycle | overlooked methane sink

different electron acceptors, including nitrate, nitrite, sulfate, and oxygen. It was found that the addition of nitrate led to a significant increase in 14CO2 formation, suggesting the occurrence of n-damo in this lake (10). Wetlands are the world’s largest natural source of methane, accounting for ∼20–40% of total methane emissions (5, 15). The n-damo is predicted to occur near the oxic–anoxic interface, having low concentrations of oxidizable substrates other than methane, as well as low level of sulfate and high level of nitrate (16). Such conditions prevail in anoxic wetlands (17). To date, however, there have been no reports of direct evidence for the occurrence of n-damo in wetlands. The primary objective of this study was to determine whether n-damo acts as a previously overlooked methane sink in wetlands. To achieve this objective, three different types of freshwater wetlands, including the Xiazhuhu wetland (natural wetland), the Xixi wetland (urban wetland), and a paddy field (man-made wetland), were studied in southeastern China. The potential n-damo rates in these three wetlands were determined using 13 CH4 isotope-labeling experiments. The distribution and diversity of M. oxyfera-like bacteria were studied based on 16S rRNA and particulate methane monooxygenase (pmoA) gene clone library analyses, and the abundance of these bacteria was quantified by quantitative (q) PCR.

N

itrite-dependent anaerobic methane oxidation (n-damo) is a recently discovered process that was reported to be performed by “Candidatus Methylomirabilis oxyfera” (M. oxyfera) (1, 2). The n-damo process constitutes a unique link between the two major global nutrient cycles of carbon and nitrogen (1). Furthermore, it may act as an important and overlooked sink of the greenhouse gas methane (3). Methane is greater than 25-fold more effective at trapping heat than is carbon dioxide (CO2) on a per-molecule basis and is responsible for 20% of global warming (4, 5). The n-damo process may alleviate the greenhouse effect by converting methane to CO2. Although several enrichment cultures of M. oxyfera-like bacteria have been obtained from freshwater sediments (1, 6–8) and peatlands (9), the distribution of these bacteria in environments is not well understood. Two recent studies reported the presence of M. oxyfera-like bacteria in two freshwater lakes, Lake Constance (10) in Germany and Lake Biwa in Japan (11). Wang et al. (12) and Zhu et al. (9) reported the distribution of M. oxyfera-like bacteria in a paddy field and in a minerotrophic peatland, respectively. In addition, Shen et al. (13, 14) recently reported the distribution of M. oxyfera-like bacteria in the sediments of Qiantang River (China) and Jiaojiang Estuary (China). However, molecular evidence for the presence of M. oxyfera-like bacteria is not a definitive indication of the occurrence of n-damo process. To date, direct evidence to support the occurrence of n-damo in the environment was lacking. Deutzmann and Schink (10) used radiotracer experiments (14CH4) to detect n-damo activity in Lake Constance. The lake sediments were evaluated by tracking 14 CO 2 formation in the presence of www.pnas.org/cgi/doi/10.1073/pnas.1318393111

Significance Given the current pressing need to more fully understand the methane cycle on Earth, in particular, unidentified sinks for methane, identifying and quantifying novel sinks for methane is fundamental importance. Here, we provide previously unidentified direct evidence for the nitrite-dependent anaerobic methane oxidation (n-damo) process as a previously overlooked microbial methane sink in wetlands by stable isotope measurements, quantitative PCR assays, and 16S rRNA and particulate methane monooxygenase gene clone library analyses. It is estimated that n-damo could consume 4.1–6.1 Tg of CH4 m−2 per year in wetlands under anaerobic conditions, which is roughly 2–6% of current worldwide CH4 flux estimates for wetlands. Given the worldwide increase in nitrogen pollution, this methane sink may become more important in the future. Author contributions: X.-y.X., P.Z., and Y.-f.H. designed research; B.-l.H., L.-d.S., X.L., Q.Z., S.L., Q.H., Z.-f.H., and S.G. performed research; B.-l.H., L.-d.S., X.L., Q.Z., S.L., Q.H., D.-q.C., and L.-p.L. analyzed data; and B.-l.H., L.-d.S., and Y.-f.H. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The sequences reported in this paper have been deposited in the GenBank database [accession nos. KC905781–KC905856 (Candidatus Methylomirabilis oxyfera 16S rRNA) and KC905857–KC905908 (Candidatus Methylomirabilis oxyfera pmoA)]. 1

To whom correspondence should be addressed. E-mail: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1318393111/-/DCSupplemental.

PNAS | March 25, 2014 | vol. 111 | no. 12 | 4495–4500

ECOLOGY

Edited by Donald E. Canfield, University of Southern Denmark, Odense, Denmark, and approved February 11, 2014 (received for review October 2, 2013)

Results Physicochemical Characteristics of the Wetlands. The vertical profiles of the redox potential, pH, temperature, ammonium, nitrite, nitrate, total nitrogen, organic carbon, and methane of the soil cores at 10-cm intervals are shown in Fig. S1. The redox potential showed a decreasing trend from the surface layer to the deep layer, which decreased from 62.3 to 19.0, 43.9–23.8, and 69.2– 20.4 mV, respectively, in soil cores collected form the Xiazhuhu wetland, the Xixi wetland, and the paddy field, suggesting the anoxic conditions of the soil cores. The soil nitrite concentration of soil cores collected from the Xiazhuhu wetland peaked at 10to 20-cm depth and then decreased with depth from 2.18 to 0.12 mg·kg−1 N and rapid decrease of nitrite concentrations were observed at depths of 20–30, 50–60, and 90–100 cm (Fig. S1). The soil nitrite concentration of soil cores collected from the paddy field also peaked at 10- to 20-cm depth and then decreased with depth from 1.12 to 0.16 mg·kg−1 N, and rapid decreases of nitrite concentrations were observed at depths of 10–30, 50–60, and 90–100 cm (Fig. S1). The soil nitrite concentration of soil cores collected from the Xixi wetland decreased with depth from 1.72 to 0.15 mg·kg−1 N, and rapid decreases of nitrite concentrations were observed at depths of 20–40, 50–60, and 90–100 cm (Fig. S1). For the methane, the electron donor of n-damo, showed increasing trends with depth (Fig. S1). The concentration of methane in soil gas collected from the Xiazhuhu wetland increased with depth from 3.1 × 103 to 5.5 × 104 mg·m−3, and rapid increases of methane concentrations were observed at depths of 50–60 and 90–100 cm. The methane concentration of soil cores collected from the Xixi wetland increased with depth from 8.9 to 3.2 × 104 mg·m−3, and rapid increases of methane concentrations were also observed at depths of 50–60 and 90–100 cm. The methane concentration of soil cores collected from the paddy field increased with depth from 8.7 × 102 to 1.0 × 105 mg·m−3, and rapid increases of methane concentrations were observed at depths of 10–30, 50– 60, and 90–100 cm. It was hypothesized that the methane at the layers of 20–30, 50–60, and 90–100 cm has the potential to be oxidized anaerobically via nitrite because of the observed opposing gradients of methane and nitrite. The sulfate, which is the most common electron acceptor for anaerobic methane oxidation (AOM) in marine environments (4), was below the detection limit (∼2 mg·kg−1 S) in all core samples, whereas the turbidimetric method used for determination of soil sulfate is not very sensitive. Therefore, the possibility that the below-ground methane could also be oxidized via sulfate cannot be excluded. Based on the above observations, a total of nine core samples collected from depths of 20–30, 50–60, and 90–100 cm of each wetland were selected for further molecular analyses and activity tests. Phylogenetic Analysis of M. oxyfera-Like 16S rRNA Genes. Phylogenetic analyses showed that the recovered 16S rRNA gene sequences were grouped into eight separate clusters, which were assigned to two groups of M. oxyfera-like bacteria, groups A and B (Fig. 1), according to Ettwig et al. (6). Sequences of cluster I, which were recovered from the Xixi wetland, showed a high similarity to the 16S rRNA gene of M. oxyfera at 97.2–97.8% identity. Sequences of cluster II, which were recovered from the paddy field, were less similar to the 16S rRNA gene of M. oxyfera, showing 95.6–96.7% identity. Sequences of cluster III, which were recovered from the Xiazhuhu wetland, showed 94.7–95.4% similarity to the 16S rRNA gene of M. oxyfera. Sequences of clusters IV and V, which were recovered from the Xixi wetland, showed 92.6–92.8% and 91.8–92.4% identities to the 16S rRNA gene of M. oxyfera, respectively. Sequences of clusters VI and VII, which were recovered from the paddy field, showed 91.2–91.6% and 90.4–90.8% identities to the 16S rRNA gene of M. oxyfera, respectively. Sequences of cluster VIII, which were recovered from the Xiazhuhu wetland, showed only 88.2–90.1% identity to the 16S rRNA gene of M. oxyfera. Furthermore, the most abundant clone sequences at depths of 50–60 and 90–100 cm were affiliated 4496 | www.pnas.org/cgi/doi/10.1073/pnas.1318393111

Fig. 1. Neighbor-joining phylogenetic tree showing the phylogenetic affiliations of M. oxyfera-like 16S rRNA gene sequences in soil cores collected from the three wetlands. Bootstrap values were 1,000 replicates, and the scale bar represents 2% of the sequence divergence. The identifiers XZ30, XZ60, and XZ100 represent samples collected from depths of 20–30, 50–60, and 90–100 cm, respectively, at the Xiazhuhu wetland. The identifiers XX30, XX60, and XX100 correspond to samples collected from depths of 20–30, 50– 60, and 90–100 cm, respectively, at the Xixi wetland. The identifiers PF30, PF60, and PF100 indicate samples collected from depths of 20–30, 50–60, and 90–100 cm, respectively, at the paddy field.

with group A, whereas the majority of sequences recovered from depth of 20–30 cm were affiliated with group B (Fig. 1). Phylogenetic Analysis of M. oxyfera-Like pmoA Genes. Phylogenetic analyses showed that the retrieved pmoA gene sequences were grouped into three distinct clusters (Fig. 2), which was consistent with the detection of three distinct clusters of 16S rRNA genes in group A (Fig. 1). Sequences of cluster I that were recovered from the Xixi wetland showed 90.5–91.3% similarity to the pmoA gene of M. oxyfera. Sequences of cluster II, which were recovered from

Fig. 2. Neighbor-joining phylogenetic tree showing the phylogenetic affiliations of M. oxyfera-like pmoA gene sequences in soil cores collected from the three wetlands. Bootstrap values were 1,000 replicates, and the scale bar represents 5% of the sequence divergence.

Hu et al.

Diversity of M. oxyfera-Like 16S rRNA and pmoA Genes. Similar levels of M. oxyfera-like 16S rRNA gene diversity were observed in the Xiazhuhu and Xixi wetlands (Table S1 and Fig. S2). DOTUR analysis indicated that a total of five operational taxonomic units (OTUs) were observed in the Xiazhuhu and Xixi wetlands, respectively, based on a threshold of 3% difference in the recovered 16S rRNA gene sequences. A higher diversity of M. oxyfera-like 16S rRNA genes was observed in the paddy field having nine OTUs. The diversity of M. oxyfera-like pmoA genes was lower than the 16S rRNA gene diversity. DOTUR analysis indicated that only one OTU was detected in the three wetlands based on a threshold of 7% difference in the recovered pmoA gene sequences (Table S1 and Fig. S2). Abundance of M. oxyfera-Like Bacteria. The copy numbers of M. oxyfera-like 16S rRNA genes in the nine core samples were determined using qPCR, as previously described (6). The copy numbers of M. oxyfera-like 16S rRNA genes varied from 3.0 ± 0.5 × 106 to 3.2 ± 0.7 × 107, 1.7 ± 0.2 × 106 to 1.0 ± 0.1 × 107, and 1.5 ± 0.2 × 106 to 4.5 ± 0.3 × 106 copies per gram of dry soil in the Xiazhuhu wetland, the Xixi wetland, and the paddy field, respectively. The copy numbers of M. oxyfera-like 16S rRNA genes varied at different depths of soil cores collected from the three wetlands, with the highest copy number detected at a depth of 50–60 cm (Fig. 3A). Potential Rates of n-damo Process. Stable isotope tracer experiments were conducted on the nine samples to determine the potential rates of n-damo in the examined wetlands. The results showed that for the slurries amended with only 13CH4, no significant accumulation of 13CO2 was observed at any soil core depth (except for the 20- to 30-cm soil core depth of the Xixi wetland and the paddy field) (Fig. 4), indicating that the levels of ambient dissolved oxygen and NOx− in most incubations were so low after preincubation that can’t be used by methane-oxidizing bacteria. When both 13CH4 and NO2− were present, 13CO2 accumulated at each analyzed depth for the majority of soil cores (Fig. 4), suggesting the occurrence of n-damo. For those slurries

A

B

Fig. 3. The abundance of M. oxyfera-like bacteria (A) and the potential ndamo rates (B) of soil cores collected from the three wetlands.

Hu et al.

amended with 13CH4 and SO42−, however, 13CO2 values remained near background levels in each of soil incubation (Fig. 4), indicating a lack of discernible activity of sulfate-dependent AOM in the examined soil samples. The potential n-damo activities ranged from 0.31 ± 0.06–5.43 ± 0.19, 0.68 ± 0.13–4.92 ± 0.04, and 1.68 ± 0.03–2.04 ± 0.06 nmol of CO2 per gram of dry soil per day in the Xiazhuhu wetland, the Xixi wetland, and the paddy field, respectively. It was found that the potential n-damo activity in the deeper layers (50–60 and 90–100 cm) was much higher than in the upper layer (20–30 cm) (Fig. 3B). The highest potential n-damo rates were detected at a depth of 50–60 cm for the Xiazhuhu and Xixi wetlands and at 90–100 cm for the paddy field. In contrast, the lowest potential n-damo activity was observed at 20- to 30-cm depth for the Xiazhuhu and Xixi wetlands, and the potential ndamo activity was below the detection limit at a depth of 20–30 cm for the paddy field. Discussion In this study, the potential role of n-damo as a previously overlooked methane sink was studied in three different types of freshwater wetlands. The presence of M. oxyfera-like bacteria was confirmed by 16S rRNA and pmoA gene clone library analyses. qPCR results further confirmed the presence of M. oxyfera-like bacteria throughout the examined soil cores, with abundance varying from 1.5 × 106 to 3.2 × 107 16S rRNA gene copies per gram of dry soil. Incubations using 13CH4 tracer indicated the occurrence of n-damo in the majority of the examined soil cores, and the potential n-damo rates ranged between 0.31 and 5.43 nmol CO2 per gram of dry soil per day. Taken together, these results demonstrate that the n-damo process represents a previously overlooked microbial methane sink in wetlands. The distribution and diversity of M. oxyfera-like bacteria have been studied in a limited number of freshwater lakes, rivers, and wetlands. For example, based on the detection of M. oxyfera-like 16S rRNA genes, three and five OTUs have been observed in profundal and littoral sediments of Lake Constance (10), respectively, and six OTUs have been found in profundal sediments of Lake Biwa (11). A relatively higher diversity of M. oxyfera-like 16S rRNA genes has been reported in freshwater river sediments, with a total of 15 OTUs (13). M. oxyfera-like pmoA gene sequences have been detected in Lake Constance (10), Lake Biwa (11), Qiantang River (13), Jiaojiang Estuary (14), and the deep layer of paddy soil (12) and peatland (7). DOTUR analysis of pmoA genes deposited in GenBank showed that pmoA genes recovered from these habitats were very closely related to one another (with only one OTU being observed for each habitat), except for the Qiantang River (13) and Jiaojiang Estuary (14), for which a total of 13 and 16 OTUs were observed, respectively. In this study, a total of five, five, and nine OTUs of 16S rRNA genes from M. oxyfera-like bacteria were observed in the Xiazhuhu wetland, the Xixi wetland, and the paddy field, respectively, and only one OTU of pmoA genes was observed in the three wetlands (Table S1 and Fig. S2). Thus, the diversity of M. oxyfera-like 16S rRNA genes and pmoA genes in the examined wetlands was similar to the diversity observed in the reported lake sediments and wetlands but lower than the diversity observed in river sediments and estuarine sediments. Previous studies demonstrated that group A of M. oxyfera-like bacteria was the dominant bacteria responsible for carrying out the n-damo process (1, 6–10, 18). The most abundant clone sequences of the 16S rRNA genes obtained from the deeper layers (50–60 and 90–100 cm) of the examined soil cores were affiliated with group A, whereas most sequences detected in the upper layer (20–30 cm) were affiliated with group B (Fig. 1). Furthermore, pmoA genes of M. oxyfera-like bacteria could only be recovered from the deeper layers (50–60 and 90–100 cm) of the examined soil cores, whereas they could not be recovered from the upper layer (20–30 cm; Fig. 2). These results are in agreement with those obtained from other studies, in which group A M. oxyfera-like bacteria and pmoA genes were both present in deep profundal sediment of lakes (10, 11) and the PNAS | March 25, 2014 | vol. 111 | no. 12 | 4497

ECOLOGY

the paddy field, showed 88.4–89.5% identity to the pmoA gene of M. oxyfera. Sequences of cluster III that were recovered from the Xiazhuhu wetland showed 85.6–86.4% identity to the pmoA gene of M. oxyfera. Moreover, the pmoA gene sequences were only detected within the deeper layers (50–60 and 90–100 cm) of the soil cores, and pmoA gene sequences were not detected in the upper layer (20–30 cm) (Fig. 2).

Fig. 4. The formation of

13

CO2 from

13

CH4 in incubations of soils collected from different depths of the three wetlands.

deep layer of paddy fields (12) and minerotrophic peatlands (9), all of which are characterized by stable environmental conditions and the coexistence of methane and nitrate/nitrite. qPCR showed that M. oxyfera-like bacteria were present throughout the examined soil cores, and their 16S rRNA gene abundance varied from 1.5 × 106 to 3.2 × 107 copies per gram of dry soil, which is similar to values reported for minerotrophic peatland and river sediments (106–107 copies per gram of dry soil) (9, 13). In the present study, a higher abundance of M. oxyfera-like bacteria was observed in the deeper layers (50–60 and 90–100 cm) (Fig. 3A). These results are similar to those observed in peatland, where a higher abundance of M. oxyfera-like bacteria was found in the deep layer (80–85 cm) relative to shallow layers (9). In addition to the presence of M. oxyfera-like bacteria, stable isotope experiments confirmed the occurrence of n-damo in the 4498 | www.pnas.org/cgi/doi/10.1073/pnas.1318393111

examined wetlands. In incubations amended with only 13CH4, it was observed that a very small amount of 13CO2 seemed to be accumulated at a depth of 20–30 cm of the Xixi wetland (0–20 h) and the paddy field (0–10 h) (Fig. 4). This may indicate that the preincubation did not remove all of the ambient nitrite at a depth of 20–30 cm of these two wetlands, which a relatively higher nitrite concentration than was observed in situ (Fig. S1). Therefore, it is likely that the residue nitrite after preincubation stimulates the production of 13CO2 in incubations amended with only 13CH4 at a depth of 20–30 cm of the Xixi wetland and the paddy filed. In addition, it was found that significant amount of 13CO2 accumulated at each analyzed depth (especially at depths of 50–60 and 90–100 cm) for the majority of soil cores in incubations amended with 13CH4 and NO2− compared with the incubations amended with only 13CH4 or 13CH4 and SO42−. This Hu et al.

Hu et al.

important microbial methane sink is often exclusively attributed to aerobic methane oxidation (21, 22). In the present study, the occurrence of n-damo was confirmed in three wetlands, thereby altering our understanding of the mechanisms for reducing methane emissions from wetlands. Considering that the nitrite concentration at a depth of 20–30 cm (0.49–2.18 mg·kg−1 N) was similar to the nitrite concentration in incubations (0.67–2.16 mg·kg−1 N), the average potential n-damo rate (0.33 nmol of CO2 per gram of dry soil per day) obtained from depth of 20–30 cm of the three wetlands was used to estimate the potential importance of n-damo process as a methane sink in wetlands. Based on the average n-damo rate (0.33 nmol CO2 per gram of dry soil per day) and the average density of soil [2.65 g·cm−3 (23)], we estimate a total of 0.51 g·m−2·y−1 CH4 could be oxidized to CO2 via n-damo in wetlands. According to this rate, n-damo has the potential to consume ∼4.1–6.1 Tg of CH4 on average each year, assuming that the total area of the world’s wetlands is 8–12 million square kilometers (24, 25). This is roughly 2–6% of total current CH4 flux estimates for wetlands [100–200 Tg·m−2·y−1 CH4 (26)]. However, the estimate of the environmental importance of n-damo process as a methane sink in wetlands still appears to be associated with a very large uncertainty because this process is likely to be highly variable in space and time owing to spatial and temporal heterogeneity in substrate availability and redox potential in wetlands. Thus, more studies that mimic the in situ conditions as closely as possible are required to determine the quantitative importance of n-damo as a methane sink in wetlands. Worldwide, anthropogenic nitrogen inputs (e.g., inorganic nitrogenous fertilizer additions) are increasing rapidly in freshwater habitats, and thus nitrite and nitrate are becoming the major electron acceptors under anoxic conditions. Therefore, AOM coupled to nitrite reduction could be more likely to occur than AOM coupled to sulfate reduction in freshwater habitats. Moreover, nitrogen input to marine ecosystems via river runoff has been increasing, providing electron acceptors (nitrite/nitrate) for AOM other than sulfate which is thought to be the most common electron acceptor for AOM in anoxic marine environments (27, 28). Therefore, the n-damo process may be globally important and has the potential to be an important methane sink in natural ecosystems due to increasing nitrogen pollution. Conclusions In this study, we provide previously unidentified direct evidence for the n-damo process as a previously overlooked methane sink in wetlands. Based on the measured potential n-damo rates, it is estimated that n-damo has the potential to consume 4.1–6.1 Tg of CH4 on average each year in wetlands under anaerobic conditions, which is roughly 2–6% of total current CH4 flux estimates for wetlands. Given the worldwide increase in nitrogen pollution, this methane sink may become more important in the future. Materials and Methods Site Description and Sample Collection. The Xiazhuhu wetland is located in Zhejiang Province and is the largest natural wetland in southeastern China, having a total area of 36.5 km2. The Xixi wetland, located in Hangzhou in Zhejiang Province, is a rare urban wetland with a total area of 11.5 km2. It is the first and only wetland in China combining urban life, farming, and culture. The paddy field selected for this study is also located in Hangzhou and represents a typical agricultural region of subtropical southeastern China. In each wetland, at least five soil cores were collected in September 2012. All soil cores were collected using a stainless steel ring sampler (5 cm in diameter and 100 cm in length). Analytical Methods. The soil pH, temperature, and redox potential of the intact soil were measured in situ using an IQ150 pH meter (IQ Scientific Instruments). Ammonium, nitrite, and nitrate were extracted from the soil using 2 M KCl as previously described (29). The soil organic carbon content was determined by the K2Cr2O7 oxidation method, and the total nitrogen content was determined using the FOSS Kjeltec2300 analyzer (FOSS Group). Soil sulfate was extracted using calcium phosphate [Ca(H 2 PO 4 ) 2 ] and

PNAS | March 25, 2014 | vol. 111 | no. 12 | 4499

ECOLOGY

suggests that the oxidation of methane in the examined soil cores is driven by nitrite (not sulfate) under anoxic conditions, and the n-damo is limited by nitrite under in situ conditions. Previously, direct evidence for the occurrence of n-damo in natural environments had been lacking. Recently, Deutzmann and Schink (10) reported the occurrence of n-damo in Lake Constance using radiotracer experiments, with a potential activity of 1.8–3.6 nmol of CO2 per milliliter of sediment per day. However, the slurries were incubated for weeks to months, indicating that microbial growth should be taken into consideration. Besides, because a high concentration of nitrate (2 mM) was added for determining the potential n-damo activity in this reported study, the occurrence of the recently reported AOM coupled to nitrate reduction cannot be ruled out (19). In addition, Zhu et al. (9) investigated the potential importance of n-damo in a minerotrophic peatland. After 3 mo of incubation using 13CH4 and NO2− amendments, the deep soil core layer (80–100 cm) clearly showed n-damo activity, producing 9.0 nmol of CO2 per gram of soil per day. However, methane oxidation was not detected within the first 2 wk of incubation, indicating that the increase of n-damo activity was attributable to the enrichment and growth of M. oxyfera-like bacteria. In the present study, n-damo activity was confirmed in the majority of the examined soil core samples within 20 h of incubation (Fig. 4). Higher potential n-damo rates were observed in the deep layers of examined wetlands (Fig. 3B), where a high abundance of group A M. oxyfera-like bacteria was observed relative to the surface layer (Fig. 3A). A recent study has demonstrated that the addition of oxygen to M. oxyfera-like enrichment cultures results in an instant decrease in the methane and nitrite conversion rates (20). Therefore, the low abundance of M. oxyfera-like bacteria and the low potential n-damo rates observed in the upper soil layer may be caused by the possible penetration of oxygen into this layer, which has a negative effect on the abundance and activity of these anaerobes. The incubation experiments that mimicked the in situ soil nitrite and methane concentrations were conducted in the current study. The in situ concentrations of NO2− were 0.12–2.18, 0.15–1.72, and 0.16–1.12 mg·kg−1 N in soil cores collected from the Xiazhuhu wetland, the Xixi wetland, and the paddy field, respectively (Fig. S1). The final concentration of NO2− in our slurries was ∼0.67–2.16 mg·kg−1 N. This concentration was in the same range to the in situ concentration of NO2− measured in the examined wetlands. On the other hand, the methane concentration in the headspace of our slurries was 6.5 × 104 mg·m−3. The in situ methane concentrations in soil gas ranged from 3.1 × 103 to 5.5 × 104, 8.9–3.2 × 104 and 8.7 × 102 to 1.0 × 105 mg·m−3 at different depths of soil cores collected from the Xiazhuhu wetland, the Xixi wetland, and the paddy field, respectively (Fig. S1). Thus, the methane concentration used in incubation experiments was similar to the in situ methane concentration measured at different depths of soil cores. Coexistence of high concentrations of soil nitrite (nitrate) and methane were observed in the examined soil cores (especially at the upper 60-cm depth; Fig. S1). However, the redox potential (19.4–60.2 mV) of the examined soil cores indicated the anoxic conditions of the soil cores. Under the anoxic conditions, the soil nitrite (nitrate) could be reduced via various biological processes (such as denitrification and anaerobic ammonium oxidation). Thus, the observed profiles of soil nitrite (nitrate) and methane in the current study may suggest a transient situation of the examined soil cores. Furthermore, the soil ammonium was extracted from the soil using 2 M KCl, and this may lead to the oxidation of ammonium to nitrite (nitrate) after 1-h extraction. In addition, it was found that the potential n-damo rates mainly peaked at a depth of 50–60 cm of the three wetlands. However, the in situ nitrite concentration at this layer (0.16–0.70 mg·kg−1 N) was substantially lower than the nitrite concentration in incubations (0.67–2.16 mg·kg−1 N). Therefore, the potential ndamo rates obtained from the incubation experiments may still overestimate the in situ n-damo rates. It is estimated that ∼50% of the methane produced in wetlands is consumed before it reaches the atmosphere, and this

determined by barium sulfate (BaSO4) turbidimetric method at 420 nm (30). The below-ground gas samples were gathered through soil gas samplers. The polyvinyl chloride (PVC) tube connected with a rubber tube used as soil gas sampler (Fig. S3). The PVC tubes were placed horizontally at 10-cm intervals. The end of each PVC tube was covered with nylon stocking as separator to prevent the soil from blocking the holes, and the end of each rubber tube was sealed with rubber septa. The gas samples were collected from each rubber tube through rubber septa using 100-mL polypropylene syringes. The first 20 mL of collected gas from the tube was rejected, and the remaining gas was injected into polyethylene-coated aluminum bags for further methane concentration analyses; 50-μL gas samples were withdrawn from the polyethylene-coated aluminum bags with a gas-tight glass syringe (Agilent) and injected into an Agilent 6890N gas chromatograph (Agilent) equipped with a Porapak Q column and a flame ionization detector. The oven temperature was set at 100 °C, and the injection and detector temperatures were both set at 130 °C. Isotope Tracer Experiments. Soil samples of known weight were transferred to He-flushed, 75-mL glass vials together with He-purged deionized water. The soil slurries were preincubated under anaerobic conditions for at least 30 h to remove residual NOx− and oxygen as closely as possible. The slurries were subsequently split into three treatment groups receiving different amendments: 13CH4 (13C at 99.9%) (treatment 1); 13CH4 + NO2− (treatment 2); and 13 CH4 + SO42− (treatment 3). The final concentrations of NO2− in treatment 2 and SO42− in treatment 3 were 0.67–2.16 mg·kg−1 N and 1.53–4.80 mg·kg−1 S, respectively, by injecting 100 μL of He-purged stock solution through the septa of each vial. Three independent experiments were performed for each treatment. Immediately after the preincubation step, 2 mL of headspace gas in each vial was removed and replaced with an equal volume of 13CH4, resulting in a final concentration of 6.5 × 104 mg·m−3 methane in the headspace of each vial. The production of 13CO2 was measured directly from the headspace of each vial using a continuous flow isotope ratio mass spectrometer (Agilent 7890/5975C inert MSD; Agilent) as previously described (6). The potential rates of n-damo were calculated by the linear regression of the concentration of produced 13CO2 in the headspace of the vial over time. The 1. Raghoebarsing AA, et al. (2006) A microbial consortium couples anaerobic methane oxidation to denitrification. Nature 440(7086):918–921. 2. Ettwig KF, et al. (2010) Nitrite-driven anaerobic methane oxidation by oxygenic bacteria. Nature 464(7288):543–548. 3. Shen LD, et al. (2012) Microbiology, ecology, and application of the nitrite-dependent anaerobic methane oxidation process. Front Microbiol 3:269. 4. Knittel K, Boetius A (2009) Anaerobic oxidation of methane: Progress with an unknown process. Annu Rev Microbiol 63:311–334. 5. Denman KL, et al. (2007) Couplings between changes in the climate system and biogeochemistry. Climate change 2007: The physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, eds Solomon S, et al. (Cambridge Univ Press, Cambridge, UK). 6. Ettwig KF, van Alen T, van de Pas-Schoonen KT, Jetten MS, Strous M (2009) Enrichment and molecular detection of denitrifying methanotrophic bacteria of the NC10 phylum. Appl Environ Microbiol 75(11):3656–3662. 7. Hu S, et al. (2009) Enrichment of denitrifying anaerobic methane oxidizing microorganisms. Environ Microbiol Rep 1(5):377–384. 8. Kampman C, et al. (2012) Enrichment of denitrifying methanotrophic bacteria for application after direct low-temperature anaerobic sewage treatment. J Hazard Mater 227-228:164–171. 9. Zhu BL, et al. (2012) Anaerobic oxidization of methane in a minerotrophic peatland: Enrichment of nitrite-dependent methane-oxidizing bacteria. Appl Environ Microbiol 78(24):8657–8665. 10. Deutzmann JS, Schink B (2011) Anaerobic oxidation of methane in sediments of Lake Constance, an oligotrophic freshwater lake. Appl Environ Microbiol 77(13):4429–4436. 11. Kojima H, et al. (2012) Distribution of putative denitrifying methane oxidizing bacteria in sediment of a freshwater lake, Lake Biwa. Syst Appl Microbiol 35(4):233–238. 12. Wang Y, et al. (2012) Co-occurrence and distribution of nitrite-dependent anaerobic ammonium and methane-oxidizing bacteria in a paddy soil. FEMS Microbiol Lett 336(2):79–88. 13. Shen LD, et al. (2014) Distribution and diversity of nitrite-dependent anaerobic methane-oxidising bacteria in the sediments of the Qiantang River. Microb Ecol 67(2): 341–349. 14. Shen LD, et al. (2014) Molecular evidence for nitrite-dependent anaerobic methaneoxidising bacteria in the Jiaojiang Estuary of the East Sea (China). Appl Microbiol Biotechnol, 10.1007/s00253-014-5556-3. 15. Bastviken D, Tranvik LJ, Downing JA, Crill PM, Enrich-Prast A (2011) Freshwater methane emissions offset the continental carbon sink. Science 331(6013):50.

4500 | www.pnas.org/cgi/doi/10.1073/pnas.1318393111

coefficients of determination (R2) for linear regression of the 13CO2 concentration change over time were greater than 0.90 for most datasets. DNA Extraction and PCR Amplification. Considering the importance of biological replicates as reported by Prosser (31), at least two or three soil cores in each wetland were subject to molecular analyses in the current study. It was found that the community structures of M. oxyfera-like bacteria were similar at the same depth between different soil cores in each wetland. Thus, the molecular data of one representative soil core in each wetland was presented. Soil DNA was extracted using a Power Soil DNA kit (Mo Bio Laboratories) according to the manufacturer’s instructions. Extracted DNA was then examined using 1.0% agarose gel electrophoresis. The 16S rRNA and pmoA genes of M. oxyfera-like bacteria were amplified using nested PCR protocols as previously described (17, 32). Detailed information on the primers used is provided in Table S2. Cloning, Sequencing, and Phylogenetic Analysis. The PCR products were cloned using the pMD19-T vector (TaKaRa Bio) according to the manufacturer’s instructions. Phylogenetic analysis of the sequences was conducted using Mega 5 software with the neighbor-joining method, and the robustness of tree topology was tested by bootstrap analysis (1,000 replicates). qPCR. The copy numbers of the 16S rRNA genes of M. oxyfera-like bacteria in the collected samples were determined by qPCR, as previously described (6). The standard curve was constructed from a series of 10-fold dilutions of a known copy number of plasmid DNA (Fig. S4). Statistical Analyses. The OTU cutoff values of 3% and 7% were applied to determine M. oxyfera-like 16S rRNA and pmoA genetic diversity, respectively, using the DOTUR program. ACKNOWLEDGMENTS. This work was supported by Natural Science Foundation Grants 51108408 and 40081198 and the Shanghai Tongji Gao Tingyao Environmental Science and Technology Development Foundation.

16. Thauer RK, Shima S (2006) Biogeochemistry: Methane and microbes. Nature 440(7086): 878–879. 17. Zhu G, Jetten MS, Kuschk P, Ettwig KF, Yin C (2010) Potential roles of anaerobic ammonium and methane oxidation in the nitrogen cycle of wetland ecosystems. Appl Microbiol Biotechnol 86(4):1043–1055. 18. Luesken FA, et al. (2011) Diversity and enrichment of nitrite-dependent anaerobic methane oxidizing bacteria from wastewater sludge. Appl Microbiol Biotechnol 92(4):845–854. 19. Haroon MF, et al. (2013) Anaerobic oxidation of methane coupled to nitrate reduction in a novel archaeal lineage. Nature 500(7464):567–570. 20. Luesken FA, et al. (2012) Effect of oxygen on the anaerobic methanotroph ‘Candidatus Methylomirabilis oxyfera’: Kinetic and transcriptional analysis. Environ Microbiol 14(4):1024–1034. 21. Reeburgh WS (2007) Oceanic methane biogeochemistry. Chem Rev 107(2):486–513. 22. Bodelier PLE (2011) Interactions between nitrogenous fertilizers and methane cycling in wetland and upland soils. Curr Opin Environ Sustainability 3(5):379–388. 23. Boyd CE (1995) Bottom Soils, Sediment, and Pond Aquaculture (Chapman & Hall, New York), p 348. 24. Ramsar Secretariat, Climate change and wetlands: Impacts, adaptation, and mitigation. Wetlands: Water, Life and Culture, Eighth Meeting of the Conference of the Contracting Parties to the Convention on Wetlands (Ramsar, Iran, 1971), Valencia, Spain, November 18–26, 2002. 25. Lehner B, Döll P (2004) Development and validation of a global database of lakes, reservoirs and wetlands. J Hydrol (Amst) 296(1):1–22. 26. Dlugokencky EJ, Nisbet EG, Fisher R, Lowry D (2011) Global atmospheric methane: Budget, changes and dangers. Phil Trans R Soc A 369(1943):2058–2072. 27. Orphan VJ, House CH, Hinrichs KU, McKeegan KD, DeLong EF (2002) Multiple archaeal groups mediate methane oxidation in anoxic cold seep sediments. Proc Natl Acad Sci USA 99(11):7663–7668. 28. Holler T, et al. (2011) Carbon and sulfur back flux during anaerobic microbial oxidation of methane and coupled sulfate reduction. Proc Natl Acad Sci USA 108(52): E1484–E1490. 29. Shen LD, et al. (2013) Broad distribution of diverse anaerobic ammonium-oxidizing bacteria in chinese agricultural soils. Appl Environ Microbiol 79(19):6167–6172. 30. Chesnin L, Yien CH (1950) Turbidimetric determination of available sulfates. Soil Sci Soc Am Proc 5:149–151. 31. Prosser JI (2010) Replicate or lie. Environ Microbiol 12(7):1806–1810. 32. Luesken FA, et al. (2011) pmoA primers for detection of anaerobic methanotrophs. Appl Environ Microbiol 77(11):3877–3880.

Hu et al.

Evidence for nitrite-dependent anaerobic methane oxidation as a previously overlooked microbial methane sink in wetlands.

The process of nitrite-dependent anaerobic methane oxidation (n-damo) was recently discovered and shown to be mediated by "Candidatus Methylomirabilis...
1MB Sizes 1 Downloads 3 Views