Methylomonadaceae was the active and dominant methanotroph in Tibet lake sediments

Abstract Methane (CH4), an important greenhouse gas, significantly impacts the local and global climate. Our study focused on the composition and activity of methanotrophs residing in the lakes on the Tibetan Plateau, a hotspot for climate change research. Based on the field survey, the family Methylomonadaceae had a much higher relative abundance in freshwater lakes than in brackish and saline lakes, accounting for ~92% of total aerobic methanotrophs. Using the microcosm sediment incubation with 13CH4 followed by high throughput sequencing and metagenomic analysis, we further demonstrated that the family Methylomonadaceae was actively oxidizing CH4. Moreover, various methylotrophs, such as the genera Methylotenera and Methylophilus, were detected in the 13C-labeled DNAs, which suggested their participation in CH4-carbon sequential assimilation. The presence of CH4 metabolism, such as the tetrahydromethanopterin and the ribulose monophosphate pathways, was identified in the metagenome-assembled genomes of the family Methylomonadaceae. Furthermore, they had the potential to adapt to oxygen-deficient conditions and utilize multiple electron acceptors, such as metal oxides (Fe3+), nitrate, and nitrite, for survival in the Tibet lakes. Our findings highlighted the predominance of Methylomonadaceae and the associated microbes as active CH4 consumers, potentially regulating the CH4 emissions in the Tibet freshwater lakes. These insights contributed to understanding the plateau carbon cycle and emphasized the significance of methanotrophs in mitigating climate change.


Introduction
Methane (CH 4 ) is the second most abundant greenhouse gas in the atmosphere [1], and its concentration has increased from 1.6 ppm in 1983 to 1.9 ppm in 2023 [2].Inland water ecosystems, including lakes, are significant in CH 4 emissions, contributing to 6%-16% of total natural CH 4 emissions [3].Recent estimations suggest that the annual average release of CH 4 from lakes accounted for ∼18.6% of the global average annual CH 4 emissions [4].
The Tibetan Plateau, commonly referred to as the "Third Pole" and the "Asian Water Tower," is highly vulnerable to global warming due to its high elevation.More than half of its area exceeds 4000 m above sea level [5].This region is home to thousands of lakes, which cover a total area of ∼50 323 km 2 [6], which is roughly 57.2% of China's lake area [7].These lakes on the Tibetan Plateau have unique characteristics, such as high altitude, low annual mean temperature [8], and a range of water salinity from freshwater to hypersaline [9].In a recent study, in situ diffusive measurements of CH 4 f lux at the water-air interface of Tibet lakes were conducted [10].It was found that the diffusive CH 4 flux in freshwater lakes was 45.14 ± 58.86 μmol•m −2 •s −1 , which was 15 times higher than that observed in brackish lakes [10].These measurements provide valuable insights into the CH 4 emissions from lakes on the Tibetan Plateau.CH 4 emissions in lake sediments are inf luenced by CH 4 production and oxidation processes.The major CH 4 consumers in lakes are aerobic methanotrophs, which can utilize CH 4 as the sole carbon and energy source [11].They are capable of consuming up to ∼93% of the CH 4 produced in the deeper sediments [12].Aerobic methanotrophs in the phylum Proteobacteria can be classified into two types: Type I and Type II [13].Type I methanotrophs belong to the class Gammaproteobacteria and are further categorized into the families Methylomonadaceae (Methylococcaceae) and Methylothermaceae.Type II methanotrophs are members of the class Alphaproteobacteria and are mainly affiliated with the families Methylocystaceae and Beijerinckiaceae [14].Methanotrophs can also be found in the phylum Verrucomicrobia [15].Oswald et al. suggested that Crenothrix could be a relevant CH 4 consumer in stratified lake water [16].The particulate methane monooxygenase (pMMO) enzyme, which initiates the first step of CH 4 oxidation [17], exists in most methanotrophs, with the exception of genera Methylocella and Methyloferula.The pmoA gene, which encodes the alpha subunit of pMMO, is a widely used functional gene to detect methanotrophs [13].
The activity and distribution of methanotrophs in lakes have received significant attention due to their important role in CH 4 consumption [13,18].However, little is known about the proportion of methanotrophs within the bacterial community in the lakes on the Tibetan Plateau.Therefore, the objective of this study is to assess the relative abundance of methanotrophs in Tibet lake sediments and investigate their distribution patterns based on the large-scale sediment sampling across the Tibetan Plateau.In our previous study, methanotroph communities in Tibet lake sediments were dominated by Type I methanotrophs in freshwater lakes, specifically Methylobacter and uncultivated Type Ib methanotrophs, while Methylomicrobium was prevalent in saline lakes [19].Salinity was found to be a key factor inf luencing the composition of aerobic methanotroph communities [19].However, it remains unclear whether these methanotrophs are actively oxidizing CH 4 in these sediment environments.To address this question, we used the DNA stable-isotope probing (DNA-SIP) method, successfully identifying metabolically active microorganisms in previous studies [20][21][22].Our objective is to determine if the relatively abundant methanotrophs in Tibet lake sediments are actively involved in CH 4 oxidation.

Lake sediment sampling
From 2015 to 2020, a total of 231 surface sediment samples were collected from 98 lakes located on the Tibetan Plateau.The geographical distribution of these lakes is shown in Fig. 1.Sediment samples were collected near the maximum water depth using sediment grab samplers.The salinity of the lake water was measured using the portable multiparameter water quality meter (YSI ProQuatro).The 98 investigated lakes were categorized into three groups according to their salinity: freshwater lakes (salinity <0.1%), brackish lakes (0.1% < salinity <3.5%), and saline lakes (salinity >3.5%) [23].Sediment samples were kept in a cool box during transportation and subsequently stored at a temperature of −20 • C in the laboratory for DNA extraction.

Sequence analysis of bacterial 16S rRNA and pmoA genes
The processing of 16S rRNA data was mainly carried out on the Usearch (https://drive5.com/usearch/)and Mothur (https:// mothur.org/)platforms.The paired-end reads were merged first, followed by removing the forward and reverse primers on Usearch, cutting off the 8-bp barcodes using Python.Next, performing quality control, eliminating error sequences greater than 1.0 and leaving unique sequences.The operational taxonomic units (OTUs) were then selected based on 97% sequence similarity, and a OTUs table was created; denoising was necessary for the process.Finally, the taxonomy classification was carried out in Mothur using "classify.seqs."According to the RDP-v18 reference database, the high-quality sequences were classified based on Wang's method (cutoff = 80%).
The pmoA gene sequencing data were analyzed as previously described [24].The paired-end sequences were first merged, and the sequence quality was checked in Usearch [25].Unique sequences were selected from all sequences, and based on 90% similarity, the candidate OTU sequences of the pmoA gene were obtained in Usearch.These candidate OTU sequences were imported into ARB (http://www.arb-home.de)[26] to remove the error sequences that cannot be correctly translated into the amino acid sequence.Distance matrices were calculated in ARB based on the 156 amino acid residues of the high-quality pmoA sequence.The new OTUs were assigned using a 7% amino acid dissimilarity cutoff using the average linkage algorithm implemented in Mothur.A neighbor-joining phylogenetic tree was constructed in ARB, including the new representative OTU sequences and related reference sequences.Methanotrophic composition data were used to generate a heatmap using the R functions heatmap.2(R package gplots) [27].Finally, the phylogenetic tree and heatmap were combined in Adobe Illustrator CS6.

Co-occurrence network analysis
To investigate the potential interaction of active CH 4 -associated bacterial communities in the labeled DNA of sediment, we conducted the co-occurrence network analysis on the labeled 16S rRNA OTU in 13 C-DNA ("heavy" fraction).The sixth and seventh fractions of DNAs from 13 CH 4 incubation were combined to have sufficient data for the network analysis.The top 100 abundant OTUs were selected for network construction, and the Spearman correlations between each OTU were calculated.Only significant correlations (0.7 < correlation coefficient (ρ) < 0.9, Pvalue <.01) were used, and the networks were created using the "igraph" package in R. For visualization, Gephi software (version 0.9.2; https://gephi.org/) was used, and all these network layouts were generated using the force-based algorithm Hu Yifan.

Procedures of metagenome sequencing
Three samples, two heavy fractions of DNA of AMC sediment, and the total DNA of MPYC were used for the metagenomic sequencing.At least 1 μg DNA was used for metagenomic library construction following the manufacturer's instructions of Truseq DNA Library Prep Kits (Illumina, USA).The purified genomic DNA is sheared into smaller fragments with ∼400 bp size by Covaris, and blunt ends are generated using T4 DNA polymerase.After adding an "A" base to the 3 end of the blunt phosphorylated DNA fragments, adapters are ligated to the ends of the DNA fragments.The desired fragments were purified through gel-electrophoresis, then selectively enriched and amplified by PCR.The index tag could be introduced into the adapter at the PCR stage, followed by a library quality test.Finally, the qualified pair-end library would be used for NovaSeq 6000 sequencing (Illumina; 150 bp × 2, Shanghai BIOZERON Co., Ltd).Approximately 20 Gb of raw sequence data were generated for each sample.
The average nucleotide identity (ANI) between each pair of bins was calculated using the OAT JavaScript (http://www.ezbiocloud.net/sw/oat) for all bins.The set of medium-high quality bins obtained from each sample was dereplicated using dRep (https:// github.com/MrOlm/drep, ANI > 95%), resulting in 40 representative genome operational taxonomic units (bins with the highest dRep score, a metric that considers genome sizes, levels of completeness and contamination, strain-heterogeneity, N50 as well as how similar each genome is to all other genomes in their cluster, Supplementary 2).GTDB Toolkit (Genome Taxonomy Database, downloaded Sep 2022, version r207v2) was introduced to obtain the taxonomy information for each MAGs.All of the genes in a bin were transformed to protein sequences to generate the proteomes for each bin to reconstruct the phylogenetic tree using PhyloPhlAN (Supplementary 3) [28].
Prodigal and BLASTP predicted ORFs within MAGs against functional gene databases, including methane cycling genes databases (MCycDB, https://github.com/qichao1984/MCycDB)and nitrogen cycling genes databases (NCycDB, https://github.com/qichao1984/NCyc). Positive hits were considered to have a genome similarity of more than 70% and coverage of more than 60% of BLASTP results.The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used via the BLASTP program with an E-value cutoff of 10 −5 further to estimate the function and metabolic pathway of genes.

Statistics analysis
IBM SPSS Statistics 22 (SPSS Inc., Cary, NC) was used for data analysis.A nonparametric test (Kruskal-Wallis H) was used to test the significant difference in the relative abundance of Methylomonadaceae in total bacteria among freshwater lakes (salinity<0.1%),brackish lakes (0.1% < salinity<3.5%), and saline lakes (salinity>3.5%).

Relative abundance of Methylomonadaceae in Tibet lake sediments
The relative abundance of Methylomonadaceae in lake sediments varied along the salinity gradient (Fig. 2).Methylomonadaceae had  a significantly higher abundance than other methanotrophs in almost all lakes, accounting for 84.07% of total aerobic methanotrophs (Supplementary 4).The relative abundance of Methylomonadaceae in total bacteria ranged from 0% to 8.34%.Freshwater lakes had a significantly higher Methylomonadaceae abundance compared to brackish (P < .001)and saline lakes (P < .001).
In the CH 4 incubation experiment, we focused on sediment samples from 10 freshwater lakes: MC, MPYC, DJMC, WRC, ZGC, BGC, GRC, DTC, QGC, and AMC, where Methylomonadaceae had a relatively higher abundance than other lakes on the Tibetan Plateau (Fig. 2).Among these 10 lakes, the average relative abundance of Methylomonadaceae was about 2% in nine lakes, except MC, where Methylomonadaceae had the highest relative abundance, reaching 8.34% in bacterial communities.

CH 4 oxidation potential and dynamics of pmoA genes abundance
During the 21-day incubation, CH 4 concentrations decreased to varying extents in all the sediments (Fig. 3).In the first 7 days, the potential for CH 4 oxidation ranged from 0.216 to 7.78 ng CH 4 g −1 dry weight sediment (d.w.s) day −1 .The CH 4 concentration in DTC and AMC bottles decreased dramatically (from ∼5% to ∼1%), with the methane oxidation rates (MORs) 7.78 and 4.44 ng CH 4 g −1 d.w.s day −1 , respectively.Between the 7th and 14th days, DTC and AMC bottles consumed CH 4 dramatically.However, WRC had the highest MOR.From the 14th to the 21st days, WRC still had the highest MOR, while DJMC had the lowest MOR (0.26 ng CH 4 g −1 d.w.s day −1 ), and other lakes had moderate MORs.
The growth of methanotrophic populations during the CH 4 incubation was characterized by quantifying the pmoA genes using qPCR.The CH 4 consumption during the incubation was accompanied by the growth of methanotrophic populations, and their abundance varied among lakes (Fig. 4A).On average, the methanotrophs increased from 1.0 × 10 6 to 5.1 × 10 6 pmoA copies g −1 d.w.s after 21 days of incubation.The pmoA gene abundance in lake AMC, DJMC, DTC, GRC, MPYC, QGC, and WRC significantly increased after 21-day incubation.Some lakes showed significant increases from in situ to Day 7, such as AMC, DJMC, DTC, and WRC.Among them, the pmoA gene abundance in DJMC increased the most and reached 1.1 × 10 6 pmoA copies g −1 d.w.s.The pmoA gene abundance in the other lakes, such as GRC, MPYC, and QGC, increased significantly from Day 7 to Day 21.The pmoA gene abundance increased significantly in Lake MPYC, from 2.1 × 10 5 to 8.7 × 10 6 pmoA copies g −1 d.w.s.

Active methanotrophs in lake sediments
During the sediment incubation, the relative abundance of phylum Proteobacteria increased in lake AMC, DJMC, MC, MPYC, and WRC, especially the classes Gammaproteobacteria and Alphaproteobacteria (Fig. S1).We further identified the main methanotroph lineages from the total bacteria before and after the CH 4 incubation (Fig. 4B).Except for lakes BGC and DJMC, the relative abundance of methanotrophs all increased during the incubation.The family Methylomonadaceae had the highest relative abundance among the aerobic methanotrophs, with an average proportion of 3.94% in the total bacterial community, even reaching as high as 30.2% in MPYC sediment after 21 days of incubation (Fig. 4B).
High-throughput sequencing of 16S rRNA revealed the abundance of methanotrophs and other methylotrophs in the heavy fractions of 13 CH 4 -incubated sediment DNAs (Fig. 5).For example, in AMC lake sediments, the relative abundance of 13 CH 4 -labeled methanotrophs and other methylotrophs in heavy fraction (Layer 8 in Fig. 5) exceeded 80%.In contrast, the relative abundance of 12 CH 4 -labeled species was only abundant in the light fraction (Layers 11 and 12 in Fig. 5).Genus Methylobacter was the main active methanotroph, and there was also a high percentage of unclassified clusters in Methylomonadaceae.Nonmethanotrophic methylotrophs, including unclassified clusters in Methylophilaceae, Methylophilus, and Methylotenera, also played an essential role in CH 4 oxidation.CH 4 -oxidizing methanotrophs in Methylomonadaceae were found in 5 out of the 10 tested lakes.The other lakes (BGC, DJMC, DTC, GRC, and ZGC), with some having relatively low CH 4 oxidation potentials, did not show a high methanotroph abundance in the heavy 13 C-DNA fractions (Fig. 5).
We also used pmoA gene sequencing to classify these methanotrophs further and constructed a DNA-based phylogenetic tree using other pmoA genes as references.The pmoA sequences utilized in the classification and construction of the phylogenetic tree were sourced from the pmoA database [24], as well as NCBI blasting results with 90% identity to the target sequences.This phylogenetic tree included Type Ia, Type Ib, Type Ic, and Type II methanotrophs.Based on 7% amino acid dissimilarity, the pmoA OTU representatives were selected, and these OTUs with a relative abundance >1% were included in the phylogenetic tree, including Uniq1, Uniq2, Uniq12, Uniq24, Uniq131, and Uniq166, listed in Fig. 6.Uniq2 was the predominant OTU across nearly all lake sediment samples, with an average relative abundance of 80.9%.The amino acid dissimilarity between Uniq2 and the closest pure culture, Methylobacter sp.LW2, measured 7.09%, which was more significant than 7%, suggesting a new species or even genus in methanotrophs (refer to Fig. 2 in Knife, 2015 [13]).Uniq1 dominated the methanotrophs in the 7th and 21st-day incubation AMC samples, closely associating with Methylobacter sp.CMS7, exhibiting 6.38% amino acid and 16.56%DNA dissimilarity.Uniq166, proximal to the genus Methylomonas, showed a low average relative abundance of 1.11%.Within Type Ib methanotrophs, Uniq12 was found to be in the lake-cluster2, and Uniq131 exhibited proximity to the FWs cluster, both of which had relative abundances below 3.50%.Meanwhile, Uniq 24 was affiliated with the Methylocystis genus, exhibiting an average relative abundance of 5.47%.It was prominently observed in the 7-day incubated DJMC sample, 7-day incubated GRC sample, and 21-day incubated MPYC sample.

Co-occurrence network of methanotrophs and other bacteria
Co-occurrence networks based on the abundant 16S rRNA gene OTUs from the 13 C-DNA heavy fractions and total DNA were constructed, respectively (Fig. 7).In the 13 C-DNA network (Fig. 7), 29 OTUs were methanotrophs, and 6 OTUs were methylotrophs among the 76 nodes.Significant correlations among OTUs affiliated with B-Methylobacter OTUs were detected in Modules 1, 2, and 4. Besides, OTUs of unclassified P-Methylophilaceae enriched in Modules 3 and 5 were positively correlated with B-Methylobacter OTUs and unclassified O-Methylomonadaece OTUs.Besides the correlation between methanotrophs and methylotrophs, the correlation between methanotrophs and heterotrophs reveals a complex metabolic correlation in the 13 C-DNA network.For example, A-Arenimonas OTUs, as a denitrifying bacterium, were positively correlated with B-Methylobacter OTUs in Module 1.

Methylomonadaceae metagenome-assembled genome and its metabolic adaption
Five representatives MAGs were identified as belonging to methanotrophs, and three belonged to methylotrophs (Supplementary 2).To construct a phylogenetic tree of the family Methylomonadaceae, the MAGs (Bin_009, Bin_018, Bin_025, Bin_038, and Bin_041) were compared with reference MAGs downloaded from NCBI (Supplementary 3).The tree included various genera such as Methylomonas, Methylobacter, Methylocaldum, Methylococcus, Methylovulum, Methylomicrobium, and KS41.It is worth noting that Bin_025 was the most abundant MAGs in AMC samples, and it could not be classified at the genus level (belonging to uncultured cluster_Methylomonadaece_JABFRC01, Supplementary 3).Bin_009 affiliated with the genus Methylovulum and showed a close relationship with Methylovulum oryzae.Bin_041, the second most abundant MAG in AMC, belonged to the family Methylomonadaceae and the genus Methylobacter.Bin_041 contained the porB/D gene (Serine cycle) in the CH 4 metabolic module and the mvhA gene (Central methanogenic pathway) in the CH 4 metabolic module, which were not found in other genomes in the genus Methylobacter.Bin_018 and Bin_038 were also identified in the AMC and MPYC sediment; both belonged to the family Methylomonadaceae at the genus level KS41.
To further study the metabolism of Methylomonadaece, we analyzed the metabolic pathways of two MAGs (Bin_025 and Bin_009), which were enriched in AMC and MPYC, respectively (Supplementary 2, Fig. 8).Bin_025 was annotated as Methylomonadaece_JABFRC01, and Bin_009 was annotated as Methylomonadaece_Methylovolum.Regarding CH 4 oxidation, both Bin_025 and Bin_009 contained gene clusters (pmoCAB) encoding pMMO.No genes encoding the soluble methane monooxygenase (sMMO, mmoXYBZDC) were detected.In the process of methanol dehydrogenation to formaldehyde, these two MAGs only contain calcium-dependent (mxaFJGID) dehydrogenases and do not have enzymes with XoxF lanthanide.In the tetrahydromethanopterin (H 4 MPT) pathway that converts formaldehyde to formate, genes encoding tetrahydromethanopterin hydrolase (fae), methylenetetrahydrofolate, methenyltetrahydromethanopterin cyclohydrolase (mch), formylmethanofuran (ftr), and formylmethanofuran dehydrogenase (fwdAB) were present completely.In terms of Figure 6.Phylogenetic tree of pmoA gene sequences; the tree was constructed based on DNA distance of representative OTUs and related pmoA sequences using the Neighbor-joining method on ARB; bold Uniq highlights OTUs with a relative abundance exceeding 1% in lake sediments; a heatmap displays their relative abundance variations across different incubation times, with the transition from low to high relative abundance represented by a color shift from light to dark red.formaldehyde assimilation pathways, Bin_009 contained three pathways: the ribulose monophosphate (RuMP), the Embden-Meyerhof-Parnas (EMP), and the Enter-Doudoroff (ED) pathways.However, Bin_025 lacked essential genes (PGK, gpml, ENO), resulting in an incomplete EMP pathway.
Regarding nitrogen metabolism, Bin_025 and Bin_009 possess genes involved in denitrification and assimilation (Fig. 8).The process of nitrate conversion to nitrite (NO 3 − → NO 2 − ) can be accomplished by various enzymes, including assimilatory nitrate reductase (NAS, nasA, and nirA), respiratory nitrate reductase (NAR, narGH), periplasmic nitrate reductases (NAP, napA), as well as nitrite oxidoreductase (NXR, nxrAB).Only genes encoding NAS were found in Bin_009 and Bin_025, while genes encoding NXR were detected in Bin_025.Regarding the process of nitrite conversion to nitric oxide (NO 2 − → NO), there are two categories of nitrite Regarding hydrogen metabolism, Bin_025 and Bin_009 contain the genes hoxH, hoxY, hoxU, and hoxF, which are involved in bidirectional hydrogenase (Fig. 8).Additionally, cyc1, the ubiquinolcytochrome c reductase cytochrome c1 subunit, suggests a potential involvement in iron oxidation.These five MAGs contain genes involved in extracellular electron transfer (EET).EET is a process by which microorganisms exchange electrons with their environment, enabling them to transfer energy and perform various metabolic activities.This process involves three main pathways: multiheme c type cytochromes (MHCs), nanowires, and electron shuttles [29].First, the pilA gene encoding electrically conductive pili (e-pili) was detected in Bin_025 and Bin_009.Second, the genes encoding MHCs (CYC1, CYT1, petC) were also identified in Bin_025 and Bin_009.Lastly, the genes encoding ribof lavin (ribA, ribBA, ribD, ribE, ribF, and ribH), which serve a typical electron shuttle, were detected in all five MAGs.

Salinity had a significant effect on the CH 4 cycle in Tibet lakes
In the studied Tibet lakes, a much higher relative abundance of Methylomonadaceae was detected in the freshwater lakes compared to brackish and saline lakes (Fig. 2).Significant effects of salinity on methanotroph community composition in Tibet lakes and other lake sediments were reported previously [19].Methylomicrobium and other Type Ia salt-tolerant and halophilic methanotrophs play an active role in CH 4 cycling [19,[30][31][32].For example, Methylomicrobium became dominant with the increase in salinity, and the previously inhibited methanotrophy quickly recovered in Lake Qinghai sediment [33].In this study, we found even though methanotrophs like Methylomicrobium exist widely and were actively oxidizing CH 4 in saline lakes, their relative abundance in bacteria was generally low (<0.1%).Like methanotrophs, the relative abundance of methanogens in Tibet lake sediments was also lower in brackish water and saline lakes than in freshwater lakes [34].These findings suggest that CH 4 -cycling microorganisms, including methanogens and methanotrophs, are less abundant in brackish water and saline lakes than in freshwater lakes.
Lakes were estimated to contribute ∼18.6% of the global average annual CH 4 emissions [4].In situ measurements of lake CH 4 f lux in the Tibet lakes have shown that CH 4 emissions are much higher in freshwater lakes than in brackish and saline lakes [10].This suggests that the relative abundance and the activity of methanotrophs and methanogens were reduced according to salinization.Therefore, considering the CH 4 f lux and the relative abundance of methanotrophs and methanogens, the contribution of freshwater lakes to CH 4 cycling is higher compared to that of brackish and saline lakes.
Global warming, increased precipitation, and accelerated melting of glaciers and permafrost have led to an expansion of more than 80% of the lake area in the Qinghai-Tibet Plateau.For example, from 1979 to 2017, Lake Selincuo increased from 1667 km 2 to 2389 km 2 [35].The expansion of lakes has diluted their salinity, which seems beneficial for the survival of methanogens and methanotrophs and has the potential to stimulate CH 4 production and oxidation, thereby increasing the contribution of these brackish water and saline lakes to CH 4 cycling.

Methylomonadaceae is the active CH 4 oxidizer in Tibet lake
In the previous study, Methylomonadaceae was identified as an abundant family in lake sediments on the Tibetan Plateau [19].Using DNA-SIP, 16S rRNA amplified sequencing, and metagenome analysis, we found the potentially new genus within Methylomonadaceae that was not only abundant but also actively involved in CH 4 oxidation in these cold, high-altitude lake sediments (Figs 2, 5, and 6).This was supported by detecting abundant unique sequences Uniq1 and Uniq2 from pmoA sequencing and identifying potentially new genus Bin_025 through metagenomic binning analysis.Previous research has also explored active methanotrophs in lake sediments using DNA-SIP in various locations worldwide.These include lakes in China, England [36], India [31], Russia [37], Germany [38], and North America [21,22,39].Methylomonadaece is the dominant group responsible for CH 4 oxidation in all these studies.For example, in German lake sediments, Dumont et al. identified Methylobacter as the most active methanotroph, a genus within the family Methylomonadaece [38].In China and England, Yang et al. found that Crenothrix, also belonging to the family Methylomonadaece, was the dominant methanotroph in two lake sediments [36].In North America, He et al. identified Methylomonas, Methylobacter, and Methylosoma, all within the family Methylomonadaece, as active methanotrophs in lakes [21,22,39].In saline water lakes, abundant microorganisms, including Methylomicrobium and Methylobacter, which also belong to the family Methylomonadaece, are dominant [31,40].
Our DNA-SIP experiments found that certain MAGs within the family Methylomonadaece contained a complete CH 4 metabolic pathway, including the H 4 MPT pathway, RuMP pathway, and the tricarboxylic acid cycle.Methanotrophs exhibit remarkable metabolic f lexibility and can adapt to oxygen-deprived conditions.In anaerobic conditions, anaerobic methanotrophs can utilize various electron acceptors, such as sulfate, metal oxides like iron (Fe 3+ ) and manganese (Mn 4+ ), nitrate, nitrite, and arsenate, for the process of coupled CH 4 oxidation [40][41][42][43].In the meantime, in O 2 -limited conditions, aerobic methanotrophs from the gammaproteobacterial group dominated the methanotrophic community and exhibited activity in freshwater lakes.Their CH 4 oxidation was also stimulated by adding iron and manganese oxides [16].Under O 2 -limited conditions, iron oxides can be an alternative electron acceptor for methanotrophs [39,[44][45][46].However, since cells cannot take up solid iron oxides, EET is essential in microbial iron reduction.Genes encoding EET mentioned above were all detected in Bin_025 and Bin_009 (Fig. 8).Ribof lavin is a typical kind of electron shuttle [47].With the help of ribof lavin, the enriched methanotrophs consortium used ferric oxides as alternative electron acceptors for oxidizing CH 4 when O 2 was unavailable [45,47].
Molecular hydrogen is considered alternative energy conservation in the energetic input of H 2 , which might counter the effect of otherwise unbalanced growth conditions, such as the O 2limited environment [48].Lake sediments are O 2 -limited environments, and from the MAG results, it appears that methanotrophs have genes hoxHYUF that are potentially involved in bidirectional hydrogenase.
Methanotrophs with nitrogen metabolism genes may utilize NO 3 − as an alternative electron acceptor when O 2 is limited.It is worth mentioning that no aerobic methanotrophs that have been isolated in pure culture have been demonstrated to perform the function of complete denitrification (NO 3 − → NO 2 − → NO→N 2 O → N 2 ) [49].However, an obligate aerobic methanotrophic bacterium, Methylomonas denitrificans FJG1, has the genes encoding the nitrate reductase NAR and has been demonstrated to couple partial denitrification with CH 4 oxidation, producing nitrous oxide as a terminal product under hypoxia conditions [49].In our study, the genes encoding the reductases involved in the first (NAS, nasA, and nirA) and second (Cu-NIR, nirK) steps of denitrification (NO 3 − → NO 2 − → NO) were detected in Methylomonadaece MAGs of Tibet lakes, suggesting their potential to conduct the denitrification.Therefore, methanotrophs inf luence global change by impacting carbon and nitrogen cycling in ecosystems.

Cross-feeding among the methanotrophs and the associated bacteria
Our study revealed that lake sediments contained methylotrophs, including Methylotenera, Methylophilus, and unclassified clusters in the Methylophilaceae family, as detected in the 13 C-labeled DNAs (Fig. 7).These methylotrophs utilize methanol, an intermediate CH 4 oxidation metabolite, as a carbon source [22].This finding is consistent with previous research that observed an abundance of methylotrophs in the 13 C-labeled DNA due to their involvement in CH 4 -carbon assimilation [21,22,[50][51][52].Except for methylotrophs, other types of metabolic interactions were also identified and shown in the 13 C-labeled DNA network.This includes interactions between denitrification and CH 4 oxidation.In aerobic conditions, CH 4 oxidation coupled to denitrification can be described as aerobic CH 4 oxidation coupled to denitrification (AME-D).Denitrifying bacteria is essential in the process of reducing nitrate to nitrite.They could utilize metabolites such as formaldehyde, formate, and particularly methanol, produced by methanotrophs as substrates in aerobic CH 4 oxidation coupled with denitrification [53][54][55].During the 13 CH 4 incubation, taxa in the genus Arenimonas and family Comamonadaceae were labeled (Fig. 7).These heterotrophic denitrifiers have been observed to engage in syntrophy with aerobic methanotrophs (primarily Methylobacter, Fig. 7) under aerobic and microaerobic conditions [56].

Conclusion
Methylomonadaceae, specifically some new clusters close to the genus Methylobacter, is the dominant and active CH 4 oxidizer in the lake sediments on the Tibetan Plateau.The relative abundance of Methylomonadaceae is higher in freshwater lakes compared to brackish and saline lakes.Other methylotrophs, such as Methylotenera and Methylophilus, are also present and play a role in CH 4 carbon assimilation.There are metabolic interactions between methanotrophs and other bacteria, including denitrifying bacteria, indicating a potential coupling of CH 4 oxidation and denitrification in the sediments.The results suggest that the CH 4oxidizing microorganisms in the lake sediments are inf luenced by salinity and play a significant role in the CH 4 cycles from these lakes.

Figure 1 .
Figure 1.Geographical distribution of the 98 surveyed lakes across the Tibetan Plateau; solid dots represent 10 freshwater lakes with a notably high relative abundance of Methylomonadaceae used for DNA-SIP labeling, while hollow dots indicate other sampled lakes within the Tibetan Plateau.
CH 4 were sampled only on the seventh day.Two non-sample bottles were set to test the gas tightness of the bottles.After incubation, all subsamples were stored at −20 • C for molecular analyses.The following DNA extraction, Quantitative Polymerase Chain Reaction (qPCR), DNA-SIP fractionation, PCR amplification, and high-throughput sequencing methods are detailed in Supplementary 1.

Figure 3 .
Figure 3. Methane concentrations in bottle headspace during incubation with an initial 5% concertation.

Figure 4 .
Figure 4. (A) Log 10-converted pmoA gene copies in 10 lakes at various incubation times (in situ, Day 7, Day 21); (B) relative abundance of methanotrophs based on 16S rRNA gene at these times in the same 10 lakes.

Figure 5 .
Figure 5. Relative abundance of methylotrophs and methanotrophs based on the 16S rRNA gene in 12 C-and 13 C-DNA fractions after a 7-day incubation.
) and denitrifying bacteria (unclassified C-Comamonadaceae and A-Arenimonas OTUs) were detected.Module 2 showed the positive correlation between OTUs B-Methylobacter, unclassified O-Methylomonadaece, and U-Methylophilus.Significant positive correlations among B-Methylobacter OTUs and unclassified O-Methylomonadaece were detected in Module 4.

Figure 7 .
Figure 7. Bacterial co-occurrence network in 13 C-CH 4 -labeled DNA and total DNA of CH 4 -incubated lake sediments; abundant OTUs with > 0.15% average relative abundance were selected to construct the networks, with red and green lines indicating positive and negative correlations, respectively; node size ref lects the "degree" (number of connections), and the top eight modules were displayed.

Figure 8 .
Figure 8. Metabolic pathway of methanotrophs.Bin_009 and Bin_025 in the family Methylomonadaceae, annotated with completeness of 99.6% and 73.7%, respectively, were selected to construct the CH 4 metabolic pathway; gray and black dots represent gene detection; the absence of dots indicates no detection.