New insight to the role of microbes in the methane exchange in trees : evidence from metagenomic sequencing

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ New insight to the role of microbes in the methane exchange in trees : evidence from metagenomic sequencing © 2021 The Authors New Phytologist © 2021 New Phytologist Foundation Published version Putkinen, Anuliina; Siljanen, Henri M.P.; Laihonen, Antti; Paasisalo, Inga; Porkka, Kaija; Tiirola, Marja; Haikarainen, Iikka; Tenhovirta, Salla; Pihlatie, Mari Putkinen, A., Siljanen, H. M., Laihonen, A., Paasisalo, I., Porkka, K., Tiirola, M., Haikarainen, I., Tenhovirta, S., & Pihlatie, M. (2021). New insight to the role of microbes in the methane exchange in trees : evidence from metagenomic sequencing. New Phytologist, 231(2), 524-536. https://doi.org/10.1111/nph.17365 2021 Viewpoints New insight to the role of microbes in the methane exchange in trees: evidence from metagenomic sequencing Summary Methane (CH 4 ) exchange in tree stems and canopies and the processes involved are among the least understood components of the global CH 4 cycle. Recent studies have focused on quantifying tree stems as sources of CH 4 and understanding abiotic CH 4 emissions in plant canopies, with the role of microbial in situ CH 4 formation receiving less attention. Moreover, despite initial reports revealing CH 4 consumption, studies have not adequately evaluated the potential of microbial CH 4 oxidation within trees. In this paper, we discuss the current level of understanding on these processes. Further, we demonstrate the potential of novel metagenomic tools in revealing the involvement of microbes in the CH 4 exchange of plants, and particularly in boreal trees. We detected CH 4 -producing methanogens and novel monooxygenases, potentially involved in CH 4 consumption, in coniferous plants. In addition, our field flux measurements from Norway spruce (Picea abies) canopies demonstrate both net CH 4 emissions and uptake, giving further evidence that both production and consumption are relevant to the net CH 4 exchange. Our findings, together with the emerging diversity of novel CH 4 -producing microbial groups, strongly suggest microbial analyses should be integrated in the studies aiming to reveal the processes and drivers behind plant CH 4 exchange. Introduction The first evidence on aerobic methane (CH 4 ) emissions by terrestrial vegetation was provided by Keppler et al. (2006), estimating that plants –including woody and grass species –are a large source of CH 4 . Since then, numerous studies (e.g. Keppler et al., 2008; Wang et al., 2008; Br€uggemann et al., 2009; Bruhn et al., 2009, 2014; Martel & Qaderi, 2017, 2019) have confirmed aerobic CH 4 emissions from terrestrial plants. During the past decade, tree stems from tropical to boreal forests and trees growing under varying hydrological conditions have been found to emit CH 4 through multiple mechanisms behind the emissions (Carmichael et al., 2014; Barba et al., 2019). Although the CH 4 emissions from tree stems and from aerobic production in plant canopies are widely recognized, neither of these sources are yet included in the global CH 4 budget (Saunois et al., 2020). Overall, discussion on aerobic plant CH 4 production has mainly concentrated on plant physiology, which was recently reviewed by L. Li et al. (2020), whereas a more general view of the current understanding of tree-derived CH 4 fluxes, magnitudes, processes, and drivers has been presented by Carmichael et al. (2014), Covey & Megonigal (2019) and Barba et al. (2019). Potential microbial CH 4 production within the aboveground tree habitat remains less studied in comparison with other mechanisms. So far, the presence of the most-well known CH 4 producers –the methanogenic archaea –has been reported only from broadleaf tree stems: first, based on basic cultivation methods (Zeikus & Ward, 1974; Zeikus & Henning, 1975), and recently based on molecular biology (Yip et al., 2019; H-L. Li et al., 2020). Tree-canopy-derived CH 4 emissions are considered to be formed mostly by abiotic/plant physiological processes (Bruhn et al., 2014; Lenhart et al., 2015a), whereas the potential role of microbial CH 4 production has been overlooked –atleastpartlydue to theassumption that theanaerobic methanogens would not thrive within the oxygen-producing canopy habitat. Recently discovered, aerobic CH 4 -producing microbial groups –such as fungi (Lenhart et al., 2012) and cyanobacteria (Bizicet al., 2020) –have not yet been thoroughly considered as sources of CH 4 in living tree stems or canopies. Atmospheric hydroxyl (OH) radicals are recognized as the main sink for atmospheric CH 4 , whereas the largest biological sink is microbiological CH 4 oxidation that occurs mostly in soils (Kirschke et al., 2013). CH 4 consumption in plants has been observed both in the field and in laboratory studies (Kirschbaum & Walcroft, 2008; Sundqvist et al., 2012; Zhang et al., 2014; Halmeenm€aki et al., 2017; Ste zpniewska et al., 2018). Research has mainly concentrated on CH 4 -rich environments, such as peatlands, where the importance of Sphagnum moss-associated methanotrophic bacteria is well recognized (e.g. Larmola et al., 2010). Although boreal tree shoots have been shown capable of in situ CH 4 consumption (Sundqvist et al., 2012), not much research has been done to reveal themechanisms behind this process. Identification of within-tree methanotrophs (Doronina et al., 2004; Van Aken et al., 2004; Iguchi et al., 2012) points to microbial CH 4 oxidation, but a possibility for a nonmicrobial sink cannot be ruled out either. In this viewpoint, we discuss the magnitude and current processlevel understanding of CH 4 exchange of trees, with an emphasis on the role of microbes in the so far considered nonmicrobial CH 4 production in plant tissues. We propose that the lack of microbial observations is not caused by the absence of these populations, but at least partly due to undeveloped methods with poor detection limits. To demonstrate the potential of novel molecular biology tools, we provide two types of metagenomics data from coniferous tree tissues: (1) functional genes detected through an extensive 524 New Phytologist (2021) 231: 524–536 Ó2021 The Authors New Phytologist Ó2021 New Phytologist Foundation www.newphytologist.com This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Forum screening of public metagenome entries (National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA)) published so far; and (2) functional gene detection through a novel probe-targeted metagenome sequencing method. In addition, we present CH 4 flux data from Norway spruce (Picea abies) shoots indicating the occurrence of both CH 4 production and consumption in the tree canopy. Through our findings from the tree canopies, we highlight the need for extending the discussions of aerobic CH 4 production from abiotic and plant physiological processes to plant–microbe interactions. Furthermore, we discuss how modern tools could advance our knowledge on the CH 4 cycling microbes from detection of presence to the understanding of active processes, and to the characterization of novel microbial groups involved in the tree CH 4 exchange. Current understanding of the methane production in trees Trees as sources of methane Tree stem flux measurements indicate that trees across vegetation zones and growing habitats are mostly sources of CH 4 (Carmichael et al., 2014; Barba et al., 2019). The stem CH 4 emissions are proposed to form either through internal CH 4 formation within tree tissues (e.g. Wang et al., 2016) or transport of soil-derived, microbially formed CH 4 emitted through tree stems (e.g. Rusch & Rennenberg, 1998), or a combination of these two as reviewed recently by Barba et al. (2019). The number of field-scale studies on canopy CH 4 exchange remain low despite the numerous laboratory studies reporting aerobic, presumably nomicrobial CH 4 production in plant leaves first presented by Keppler et al. (2006). Existing field evidence on canopy CH 4 exchange indicates both emissions and uptake of CH 4 and calls for further studies (Sundqvist et al., 2012; Machacova et al., 2016; Pangala et al., 2017; this study). Aerobic methane production through nonmicrobial mechanisms Increasing evidence suggests that aerobic CH 4 formation in plant leaves may be an integral part of cellular responses to changing redox conditions in all eukaryotes, and that this common biochemical CH 4 source may exist in all eukaryotes –plants, animals, fungi, and algae (Keppler et al., 2009; Liu et al., 2015). Several studies also suggest that aerobic nonmicrobial CH 4 formation occurs through reactive oxygen species (ROS) generation and a subsequent release of CH 4 from precursor compounds, such as pectic methyl groups, methionine, or other substrates (e.g. Keppler et al., 2008; McLeod et al., 2008; Vigano et al., 2008; Bruhn et al., 2009; Lenhart et al., 2015a). Although plant ROS are produced during aerobic respiration and photosynthesis as a normal by-product of aerobic plant metabolism, ROS production can be induced by different environmental stressors (Huang et al., 2019). Similarly, environmental stressors like ultraviolet (UV) radiation and elevated temperature (e.g. McLeod et al., 2008; Vigano et al., 2008; Bruhn et al., 2009, 2014; Qaderi & Reid, 2009), physical injury of the plant –for example, leaf damage caused by cutting and hypoxia (Wang et al., 2009) –water stress (Qaderi & Reid, 2009), and low light levels (Martel & Qaderi, 2017) have been observed to stimulate aerobicnonmicrobial CH 4 production inplants. Still,the biochemical pathways behind stress-induced CH 4 formation and its potential occurrence in natural conditions remain unknown. Microbes as potential methane producers in trees Different parts of the trees (leaves, stem, bark, roots) serve as unique habitats for a variety of microbial communities that can live either as epiphytes on the plant surface, or as endophytes inside the plant tissues –used here to include also pathogens, as rationalized by Griffin & Carson (2018). Together, different microbes, including bacteria, archaea, and fungi, form the tree microbiome (Terhonen et al., 2019). Studies on tree microbiomes have largely concentrated on fungi and less on bacteria or archaea (Griffin & Carson, 2018; Harrison & Griffin, 2020) and tree-stem-associated prokaryotes especially are still poorly characterized (Baldrian, 2017). Geographically, endophyte studies have focused on the tropical and temperate regions, leaving boreal and alpine ecosystems poorly examined (Harrison & Griffin, 2020). Many of the tree-associated microbes are important to the host plants via promoting plant growth and increasing resistance to stress and pathogens, whereas some of them can negatively affect plant growth (Frank, 2018; Terhonen et al., 2019; Chaudhry et al., 2020). Owing to varying conditions caused by both host metabolic processes and abiotic stress factors, such as drought and UV radiation, tree foliage microbiomes especially are highly dynamic systems (Chaudhry et al., 2020). In addition, colonization patterns affect the microbiome composition: whereas some endophytes can stay with their host the whole plant life cycle (vertical transmission), most are estimated to originate from the environment through horizontal transmission either from the soil or through the air (Frank et al., 2017). On first thought, trees and other plant tissues seem to be mostly aerobic environments and, as such, unsuited habitats for anaerobic organisms, like methanogenic archaea (Kirschke et al., 2013). As recently reviewed by Covey & Megonical (2019), however, anoxia can prevail inside both healthy and infected tree stems, leading to CH 4 production (Fig. 1b). In both situations, anoxia could be created through oxygen-consuming metabolic processes of the treeassociated endophytic microbes, such as fungi, and also through active stem and root respiration (Teskey et al., 2008), helping to maintain favorable conditions for methanogenesis. In particular, fungal-mediated decay of heartwood (i.e. heart rot disease) has been suggested as an important driver of CH 4 emissions from living trees (Covey et al., 2012). Moreover, anaerobic methanogens have also been detected in the roots of Norway spruce, Scots pine (Pinus sylvestris), silver birch (Betula pendula), and black alder (Alnus glutinosa) (Bomberg & Timonen, 2009; Bomberg et al., 2011; Fig. 1c). As in the stems, root-associated archaea are predicted to benefit from the O 2 consumption of other microbes, and also ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation New Phytologist (2021) 231: 524–536 www.newphytologist.com New Phytologist Viewpoints Forum 525 directly from the carbon (C) compounds exuded from the roots (Bomberg et al., 2011). Local anaerobic microenvironments could exist also in the needles and leaves of trees, where the activity of endophytes, and thus O 2 consumption, is enhanced by the fresh, photosynthesisderived C compounds (Fig. 1a). This type of interaction has been reported at least from gramineous plants (Minamisawa et al., 2004), where anaerobic nitrogen (N)-fixing clostridia are supported by other, nondiazotrophic endophytes. Although direct canopy-derived evidence is still lacking, anaerobic N fixation occurs also in coniferous needles (Moyes et al., 2016), demonstrating the potential for other anaerobic processes as well. Although oxygen effectively inhibits archaeal CH 4 production (Fetzer et al., 1993; Yuan et al., 2009), at least some methanogens can still tolerate oxic conditions –as previously observed, for example, in upland soils (Peters & Conrad, 1996; Angel et al., 2012). Lyu & Lu (2018) evaluated the mechanisms behind this tolerance in their recent meta-analysis of both genomic and environmental data. They found strong evidence of two distinct methanogen clusters, with one of them harboring expanded category of oxygen tolerance features, including ways to combat ROS-derived oxidative stress. This split into clusters was largely in line with the classification into phylogenetic methanogen orders and their evolutionary history in relation to atmospheric O 2 levels. Moreover, global analysis of methanogens detected in oxic habitats gave further evidence that these particular methanogens have the potential to survive in the presence of oxygen (Lyu & Lu, 2018) –and thus possibly even within the canopy habitat. In addition to methanogenic archaea, tree-derived CH 4 could be produced by other microbial groups, better suited for a life in aerobic conditions. First, saprophytic fungi produce CH 4 at least in nonliving wood material (Lenhart et al., 2012) . The fungal CH 4 was shown to derive from methionine (Lenhart et al., 2012), a precursor compound linked with plant stress-induced aerobic CH 4 production (Lenhart et al., 2015a). On the other hand, Lenhart et al. (2012) also suggested that the fungal CH 4 production can be connected to chloromethane (CH 3 Cl) (a) FOLIAGE (b) TRUNK (c) ROOTS CH4-producing saprotrophic fungi CH4-producing cyanobacteria Methanogenic archaea Methanotrophic bacteria Strong direct evidence CH4PRODUCTION CH4OXIDATION Potential locations Indirect/weak evidence Sapwood Heartwood Heartwood infected by pathogenic fungi 10 cm 1 mm 1 cm 1 mm Fig. 1 Locations of different methane (CH 4 )-producing and the so far known CH 4 -consuming microbes, the methanotrophic bacteria, present within different tree compartments: (a) foliage,(b) trunk, (c) roots, as determined based on the references and new resultspresented in this viewpoint. The locations are marked with dots colored based on the level of scientific evidence: strong evidence indicates more than one study and/or detected in several tree species; indirect/weak evidence indicates only one study or one tree species, or detected activity (e.g. flux measurements); potential locations indicates potentially suitable conditions for the given microbial group.Nitrogenase-related CH 4 production wasnot included in the figure owingto the still low understanding of itspotential occurrence in the tree habitat. Red arrows, CH 4 production; blue arrows, CH 4 consumption. New Phytologist (2021) 231: 524–536 www.newphytologist.com ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation Viewpoints Forum New Phytologist 526 formation and the type of substrates available, which might limit this process to the wood-decay fungi. As some of the wooddecaying fungi can also infect living trees (Asiegbu et al., 2005), and since needles harbor complex fungal microbiomes (Pirttil€a &W€ali, 2009), their role in the tree stem and canopy CH 4 exchange may be significant, but this remains to be resolved (Fig. 1b). Another recently discovered CH 4 -producing group are cyanobacteria, which were linked to this process both under oxic and anoxic conditions (Bizicet al., 2020). CH 4 production was suggested to occur through general cell metabolism, such as photoautotrophic C fixation, and mechanisms that are dependent on photosynthetic products during light, and on storage compounds during dark. Since photosynthesis-performing chloroplasts in plants are known to have evolved from cyanobacteria through endosymbiosis (Raven & Allen, 2003), these bacteria could be connected to CH 4 production of land plants as well (Fig. 1a,b). This relation further underlines the decadal discussion on mechanistic understanding of aerobic CH 4 formation in plants (e.g. Keppler et al., 2009; Liu et al., 2015). So far, the most direct link between cyanobacteria and tree-related CH 4 emissions are the cyanobacteria-containing cryptogamic covers, such as lichens, which can grow on tree stems and have shown small CH 4 emissions in laboratory incubations (Lenhart et al., 2015b; Fig. 1b). Finally, CH 4 is also produced during the process of N fixation when it involves the iron nitrogenase, and to a lesser extent the vanadium nitrogenase (Zheng et al., 2018). These enzymes are found in various species representing both archaea and bacteria (McRose et al., 2017). This finding is interesting owing to recent indications that endophytic diazotrophs are essential for coniferous trees growing on nutrient-poor soil (Moyes et al., 2016; Puri et al., 2020) and also considering cryptogamic covers, where cyanobacteria utilize these enzymes (Bellenger et al., 2020). Furthermore, another nitrogenase-type enzyme system, found in various microbial groups, was recently reported to produce CH 4 from dimethyl sulfide (North et al., 2020), produced, for example, by various bacteria in terrestrial environments (Carrion et al., 2015, 2017). Although the link between all of these nonarchaeal groups and CH 4 production in the living trees is uncertain, the aforementioned findings suggest that CH 4 formation in terrestrial ecosystems is a far more widespread trait than previously thought and also warrants their evaluation in the tree CH 4 studies. Methanogenic microbes in canopies of coniferous trees Compared with potential aerobic CH 4 production in plants, the role of methanogenic archaea is often left undetermined in current CH 4 -exchange studies focusing on the photosynthesizing plant parts. This situation stems at least partly from practical reasons, as we have lacked methods with adequate resolution to identify microbial populations behind locally relatively small, but globally significant emissions, and methods that allow linking previously unrelated organisms with particular functions, like CH 4 -production. PCR-based methods, such as amplicon sequencing, are a standard tool in microbial ecology. Yet, coverage limitations often make them unfit for the analysis of rare, poorly characterized endophytes. Accordingly, only two PCR-based studies (and only two molecular analyses in general) have been published on tree-dwelling methanogens, and only from tree stems (Yip et al., 2019; H-L. Li et al., 2020). The recent developments in high-throughput sequencing techniques have led to the rise of metagenomic methods, which can potentially revolutionize the analysis of various microbiomes, such as tree endophytes. Compared with PCR, metagenomic approaches entail a far wider perspective: whole microbiomes within plant tissues can be sequenced, and with the right analytical tools even genes from novel taxa can be revealed. We evaluated two different metagenomic sequencing approaches in revealing potential CH 4 -producing microbes in trees, with a focus on boreal tree canopies. First, we conducted an extensive meta-analysis of methanogenic functional genes in already published data-entries in the SRA (https://www.ncbi.nlm. nih.gov/sra) related to pine and spruce tissues (a detailed description is given in Supporting Information Methods S1), similarly as previously for the methanogens in the SRA-data from peatlands (Br€auer et al., 2020). In brief, gene fragments of CH 4 - producing archaea (mcrA, coding for the methyl-coenzyme M reductase) were searched with HMMER (hidden Markov model search of gene structures) from the published SRA database (NCBI), and phylogenetics of them were analyzed against obtained cultured and candidate divisions of functional genes. Second, we utilized a novel ‘probe-targeted capture’ method (Aalto et al., 2020; Siljanen et al., 2021) to analyze genes related to CH 4 cycling from Norway spruce needles collected from eastern Finland (Kuopio). The same method was used recently for the detection of N-cycling microbes in plant biomass (Aalto et al., 2020). Here, capture reaction was carried out with 12 190 unique probes, which were designed based on the currently known mcrA gene diversity in the public databases (Siljanen et al., 2021). Before the analysis, spruce branches were incubated in aerobic conditions in a medium of sodium acetate containing diluted nitrate mineral salts for 14 d with 100 ppm CH 4 in the headspace to enhance the detection of both methanogens and methanotrophs (results for the methanotrophs are reported later in this paper). Interestingly, both of our approaches revealed known methanogen species within the spruce canopies (Figs 1a, 2, S1; Table 1). An SRA database search revealed signs of methanogenic mcrA genes in both pine and spruce trees (Fig. 2; Tables S1, S2). However, although SRA sequences gave indications of a wide diversity (orders Methanosarcinales, Methanomicrobiales, and Methanobacteriales for spruce-derived entries and Methanomicrobiales for pine-derived entries), the number of quality-checked sequences was small (Table 1). Our captured metagenome analysis included Norway spruce needles only from one location (Kuopio, Finland) and, as a smaller sample set, was expected to express lower mcrA diversity than the global database search. Still, by providing much higher length sequences (average 250 bp vs 100 bp of the SRAs) specifically enriched by mcrA-targeting probes, this method was able to give more reliable evidence on the presence of methanogenic archaea ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation New Phytologist (2021) 231: 524–536 www.newphytologist.com New Phytologist Viewpoints Forum 527 within the conifer habitat (Fig. 2; Table 1). The vast majority of the sequences belonged to orders Methanomicrobiales and Methanosarcinales within the class Methanomicrobia. Within the former order, most sequences were related to the genus Methanoregula (1799 out of 4066 sequences from spruce 1 matched with Candidatus Methanoregula boonei with a likelihood-weight ratio LWR >0.95; scale for LWR: 0–1). Within Methanosarcinales, spruce mcrA sequences grouped with Methanothrix (formerly Methanosaeta) species (Methanothrix soehngenii linked with LWR >0.95 with 303/4066 sequences from spruce 1 and 101/1865 from spruce 2). All of these well-described genera/species are common inhabitants in, for example, the waterlogged layers of peatlands, where they perform anaerobic methanogenesis as the last step of organic matter degradation (Br€auer et al., 2020). The largest groups found, Methanoregula and Methanothrix, produce CH 4 by reducing carbon dioxide (CO 2 ) with hydrogen (H 2 ), or by splitting of acetate to CH 4 and CO 2 , respectively. Accordingly, the detection of the Methanothrix L77117 1 771823-773477 Methanocaldococcus jannaschii DSM 2661 CP000477 1 590544-588887 Methanothrix thermophila PT JX141395 1 1-464 Methanobacterium formicicum KOR-1 mrtA CP000300 1 2545676-2543963 Methanococcoides burtonii DSM 6242 KM041257 1 2-488 Methanofollis liminatans A mcrA CP009515 1 197568-195858 Methanosarcina lacustris Z-7289 AB479391 1 2-721 Methanoregula formicicum SMSP mcrA CP003117 1 500591-502252 Methanothrixh arundinacea 6Ac AF313803 1 2-471 Methanothrix soehngenii VeAc9 mcrA KM041254 1 2-504 Methanothermobacter marburgensis mcrA CP002737 1 1238720-1240374 Methanotorris igneus Kol 5 CP002069 1 861025-859314 Methanohalobium evestigatum Z-7303 AF414041 1 1-437 Methanofollis liminatans DSM 4140 mcrA EU715818 1 744-2 Methanolobus zinderi SD1 mcrA KT387805 1 1141805 Uncultured Bathyarchaeota cloneCX10 BA1 24 9 CP003167 1 861950-863646 Methanoregula formicicum SMSP CP017921 1 61885-63601 Methanohalophilus halophilus Z-7982 DQ229161 1 1-519 Methanogenium boonei mcrA AB703644 1 1-1170 Methanofollis ethanolicus mcrA NBRC 104120 mcrA JQ511369 1 1-425 Methanocalculus alkaliphilus AMF2 mcrA CP000559 1 1595689-1593992 Methanocorpusculum labreanum Z KT387806 1 1661817 Uncultured Bathyarchaeota cloneCX10 BA2 13 151 CP002057 1 1273204-1274859 Methanococcus voltae A3 CP000780 1 552905-551251 Ca. Methanoregula boonei 6A8 CP009517 1 3100739-3099032 Methanosarcina barkeri 3 AB496719 1 1-746 Methanolinea mesophila mcrA CP010070 1 210620-212263 Ca. Methanoplasma termitum MpT1 CP003167 1 2552112-2553766 Methanoregula formicicum SMSP CP002009 1 260010-258356 Methanocaldococcus infernus ME HE964772 2 1548527-1546837 Methanoculleus bourgensis MS2T CP001696 1 393468-392637 Methanocaldococcus fervens AG86 AY386125 1 2-1123 Methanobacterium aarhusense mcrA AB288270 1 2-742 Methanoculleus chikugoensis mcrA U22244 1 2-489 Methanolubus tindarius mcrA CP002117 1 2417258-2415575 Methanoplanus petrolearius DSM 11571 HQ896500 1 455-1 Methanomassiliicoccus luminyensis B10 mcrA CP004049 1 363749-365391 Ca. Methanomethylophilus alvus Mx1201 CP002278 1 713931-715572 Methanothermus fervidus DSM 2088 AF414037 1 1-416 Methanothrix soehngenii DSM 3671 mcrA AB703641 1 1-1169 Methanolinea tarda NBRC 102358 mcrA AP011532 1 507632-509298 Methanocella paludicola SANAE DNA EF026570 1 3-668 Methermicoccus shengliensis ZC-1 mcrA CP002278 1 756135-754476 Methanothermus fervidus DSM 2088 CP002737 1 455360-457008 Methanotorris igneus Kol 5 AB542746 1 1-1144 Methanobacterium alcaliphilum NBRC 105226 mcrA AE009439 1 614620-616275 Methanopyrus kandleri AV19 LT608329 1 1413251-1411597 Methanothermobacter wolfeii SIV6 CP001710 1 1394951-1393301 Methanothermobacter marburgensis str Marburg CP002565 1 676986-675313 Methanothrix soehngenii GP-6 KM259864 1 1-442 Methanosalsum natronophilum AME9 McrA CP001710 1 1425701-1424056 Methanothermobacter marburgensis Marburg AB542760 1 2-1139 Methanobacterium palustre NBRC 105230 mrtA CP002551 1 338949-340594 Methanobacterium lacus AL-21 CP009516 1 105661-107367 Methanosarcina horonobensis HB-1 CP003362 1 1659823-1661533 Methanomethylovorans hollandica DSM 15978 CP009509 1 92002-93709 Methanosarcina mazei WWM610 EU919432 1 1-467 Methanobrevibacter woesei GS mcrA U22245 1 2-489 Methanolobus vulcani mcrA AB842184 1 3-1141 Methanobacterium alcaliphilum NBRC 109449 mrtA AB703638 1 1-1147 Methanothermobacter wolfeii NBRC 100332 mcrA CP005934 1 1549873-1548225 Ca. Methanomassiliicoccus intestinalis Issoire-Mx1 AF414044 1 1-438 Methanomicrobium mobile DSM 1539 mcrA AB679169 1 2-1148 Methanothrix pelagica mcrA AM114193 2 1035015-1033351 Methanocella arvoryzae MRE50 CP002551 1 2316117-2317758 Methanobacterium lacus AL-21 CP014265 1 1666298-1664653 Methanobrevibacter olleyae YLM1 Tree scale: 1 Methanosarcinales Methanomicrobiales Methanocellales Bathyarchaeota Thermoplasmata Methanopyrales Methanobacteriales, mrtA Methanococcales, mcrA Methanococcales, mrtA, Methanobacteriales, mcrA Fig. 2 Phylogenetic placements of the methanogenic mcrA gene (coding for the alpha subunit of the methyl-coenzyme M reductase (MCR))sequences retrievedthrough theSequenceReadArchive(SRA) databasesearchandcapturedmetagenomicsequencingfromtwopooledspruce needle samples(spruces1 and 2) among the known methanogen and Bathyarchaeotal mcrAand mrtA (coding for the isozyme of the MCR) sequences inferred using the iTOL-tree (Letunic & Bork,2019) with RAXML (Stamatakis, 2014). In the SRA-database search, 4822 pine and 1215spruce SRA files in total were screened with HMMER for mcrA genes (details in the Supporting Information Methods S1). Only the best phylogenetic placement is shown for each query sequence, and only sequences with likelihood-weight ratios >0.2 are included.Size of the placement icons reflects the relative number of sequences for a given placement within each sample. As an exception, the smallest size icon is used for positions with one to five sequence placements. The total number of sequences for each sample typeis listed in Table 1. Original reference tree with 189 sequences together with bootstrap values is in Fig. S1. Ca., Candidatus. New Phytologist (2021) 231: 524–536 www.newphytologist.com ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation Viewpoints Forum New Phytologist 528 sequences may have been enhanced by the spruce incubation treatment with acetate in the growth medium. However, our results still reflect the taxa present within untreated needles. In soils, both acetoclastic and hydrogenotrophic methanogenic pathways are sustained by the activity of other microbiota, such as syntrophic microbes (Br€auer et al., 2020). This is likely the case also in the needle habitat and needs to be investigated through a wider analysis of the whole microbiome and related microbe–microbe interactions. It should be noted that plant metabolic processes could also serve as a source of substrates for the microbial methanogenesis. For example, acetate is continuously recycled within the plant cells (Zhang et al., 2017) and could, thus, be available for the acetoclastic methanogens such as the Methanothrix species. Considering the largely oxygenic conditions within spruce needles, detection of Methanomicrobiales and Methanosarcinales is fitting: they belong to the specific cluster of methanogens suggested to contain enhanced O 2 tolerance mechanisms (Lyu & Lu, 2018). For them to be actually active in CH 4 production, at least temporally anoxic microhabitats are needed –potentially involving the oxygen-consuming activity of other endophytes. Methane consumption within tree stems and canopies: an unrecognized methane sink? Evidence of methane consumption by trees from flux measurements Despite numerous CH 4 flux studies on tree stems, consumption of CH 4 in stems has been rarely reported (Barba et al., 2019; Welch et al., 2019; Moldaschl et al., 2021). As the net CH 4 exchange is the sum of both production and consumption processes, it remains unclear whether consumption exists but is mostly overcome by a higher CH 4 production rate. The few existing studies on tree canopy CH 4 exchange show that tree canopies can act as both sources and sinks of CH 4 (Sundqvist et al., 2012; Machacova et al., 2016; Halmeenm€aki et al., 2017; Pangala et al., 2017; this study). Based on field measurements, canopy CH 4 consumption has, to our knowledge, been reported only by Sundqvist et al. (2012). All the tree species they measured –coniferous trees: Norway spruce and Scots pine; and broadleaf trees: birch (Betula pubescens) and rowan (Sorbus aucuparia)–were observed to mostly consume CH 4 with an average CH 4 consumption rate of 11.2 µgh 1 m 2 leaf area (LA) at lower branches of the trees during autumn period. Sundqvist et al. (2012) estimated that, with the uptake rate they measured, the tree canopy CH 4 sink could be of similar strength to the soil sink. Here, we present results from two Norway spruce field campaigns, which further demonstrate the functioning of shoots as both sinks and sources of atmospheric CH 4 (Fig. 3; details in Methods S1). The average CH 4 exchange rate was 0.3 ng g 1 DW h 1 for mature spruce trees (Skogaryd, Sweden), and 1.4 ng g 1 DW h 1 for 2 to 3-yr-old tree saplings (Helsinki, Finland). The scale and variation of the fluxes was clearly higher in the young samplings, possibly reflecting their growth phase and dynamic conditions during the spring period. However, our fluxes from both mature spruce shoots and from the saplings were markedly smaller than those measured by Sundqvist et al. (2012). As interpreted from Sundqvist et al. (2012, Fig. 1), their CH 4 fluxes from mature spruce shoots ranged from c.40 to 32 µgh 1 m 2 LA, which scales to a range of c.200 to 160 ng g 1 DW h 1 (Hager & Sterba, 1985). This high variability between the studies underlines the need for more flux measurements from tree canopies, and the need to consider both CH 4 production and consumption when evaluating the role of trees in the forest CH 4 balance. Most importantly, to uncover the drivers of these processes, potential involvement of microbes should be studied using simultaneous collection of tree tissue samples. Table 1 Number of quality-controlled (likelihood-weight ratio LWR >0.20) sequences acquired/retrievedin this study together with their LWR values from the phylogenetic placement analysis (details in Supporting Information Methods S1). Sample and analysis type Marker gene Host tree species (plant part used in the sequencing) No. of sequences LWR range LWR mean  SD Spruce 1, capture sequencing mcrA Picea abies (needles) 4066 0.26–1.00 0.83 0.21 mmoX 3782 0.20–1.00 0.67 0.20 Psm_mmoX 1 0.55 — Spruce 2, capture sequencing mcrA Picea abies (needles) 1856 0.20–1.00 0.56 0.22 mmoX 270 0.21–0.99 0.69 0.24 Psm_mmoX nd nd nd SRA database, pine mcrA Pseudotsuga menziesii (megagametophyte) 2 0.22–0.29 0.26 0.05 mmoX Pinus taeda (needles), Pseudozuga menziesii (needles) 4 0.24–0.43 0.35 0.08 Psm_mmoX Pinus sylvestris (needles), Pinus canariensis (cambial cells), Pinus contorta (foliage), Pseudozuga menziesii (megagametophyte) 4 0.22–0.97 0.56 0.35 SRA database, spruce mcrA Picea abies (megagametophyte) 6 0.21–0.36 0.28 0.06 mmoX Picea abies (megagametophyte) 2 0.27–0.62 0.45 0.24 Psm_mmoX Picea abies (megagametophyte) 50 0.02–1 0.57 0.28 In the Sequence Read Archive (SRA)-database search, 4822 pine and 1215 spruce SRA files in total were screened with HMMER for mcrA,mmoX, and pmoA genes. Average lengths for capture and SRA sequences were 250 bp and 100 bp, respectively. No traditional pmoA gene fragments were detected. nd, not detected. Psm_mmoX,Pseudomonas sp.-related novel mmoX. ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation New Phytologist (2021) 231: 524–536 www.newphytologist.com New Phytologist Viewpoints Forum 529 Evidence of methane consumption by trees from detection of tree-associated methanotrophs Methanotrophs vary in their preference for the concentration of CH 4 and can thus be divided into high (atmospheric CH 4 ) and low-affinity oxidizers –although some of them oxidize CH 4 both in high and low concentrations (Chowdhury & Dick, 2013; Ho et al., 2019). High-affinity methanotrophs are responsible for the CH 4 sink of the upland soils, whereas low-affinity populations thrive, for example, in waterlogged soils with high in situ CH 4 production (Knief et al., 2003; Chowdhury & Dick, 2013). Trees could potentially provide microhabitats for both types of methanotrophy. In the canopies, CH 4 concentrations are likely closer to atmospheric concentrations, and thus they might harbor high-affinity-type oxidizers as both epiand endophytes. Our results, however, point to the presence of methanogenic activity in the needles, which might support low-affinity oxidizers. In line with this, Iguchi et al. (2012) reported isolation of common methanotrophs, Methylomonas sp. and Methylocystis sp., from Norway spruce and Pinus parviflora needles (Fig. 1a). Although their CH 4 oxidation capacity was not tested, the isolates were obtained using a high CH 4 concentration (20%). Likewise, Doronina et al. (2004) were able to isolate a Methylocystis-related strain from the needles of Picea pungens. All the isolates mentioned also grew with methanol (CH 3 OH), and thus represented the so-called facultative methanotrophs, potentially supported by CH 3 OH formation in the plant physiological processes (Dorokhov et al., 2018). More recently, similar facultative methanotrophs were detected through 16S ribosomal RNA (rRNA) gene sequencing from the needles of Pinus radiata (Rua et al., 2016) and Norway spruce (Haas et al., 2018). Tree stems have been shown to occasionally hold very high CH 4 concentrations (Covey et al., 2012) and might, thus, serve as a habitat for low-affinity oxidizers, similar to, for example, Sphagnum mosses in peatlands (e.g. Larmola et al., 2010; Putkinen et al., 2012). This is supported by the detection of both 16S rRNA genes related to common alpha and gammaproteobacterial methanotrophs and methanogens (as discussed earlier in the paper) in the stems of Populus deltoides (Yip et al., 2019; Fig. 1b). In addition to living stems, methanotrophs have been found in fallen logs infected by fungi (M€akip€a€aet al., 2018) – another tree habitat where in situ CH 4 production likely takes place (Covey et al., 2012), and, in unquantifiable amounts, in Scots pine roots (Halmeenm€aki et al., 2017, Fig. 1c). Based on these findings, and the CH 4 consumption detected in the field measurements, we suggest that methanotrophs are present in the conifer habitat and that their in-depth characterization is possible with the improved metagenomics tools now available. For this purpose, we used the same two metagenomic approaches as with the methanogens (details in Methods S1; Br€auer et al., 2020). First, the SRA database was searched for methanotrophic functional genes pmoA and mmoX, coding for the particulate and soluble forms of methane monooxygenase (MMO), respectively. Second, the same genes were targeted with the capture enrichment approach (Aalto et al., 2020) to detect methanotrophs in spruce shoots, which were first incubated to enhance the detection of CH 4 -cycling microbes (as described in the section Methanogenic microbes in canopies of coniferous trees; Dunfield et al., 2003; Dedysh et al., 2005). Capture reaction included 640 unique probes for mmoX and 19 900 probes for pmoA (Siljanen et al., 2021). Both analyses revealed similar patterns: monooxygenase (MO) genes were found, but except for two pine-derived SRAsequences, similar to alphaproteobacterial mmoX genes, they were not related to pmoA or mmoX genes of known methanotrophs (Figs 4, S2; Tables 1, S1, S2). Almost all other SRA sequences with proper likelihood weight values (Table 1) were from the same project targeting the genome of the host tree, P. abies, with sequenced DNA deriving from the spruce megagametophyte (Nystedt et al., 2013) –likely reflecting the lack of microbiome targeting analyses in general. Except for one actinobacterial propane MO (PMO) match, these SRA sequences were related to novel Pseudomonas sp.-related MO genes, which, in addition to butane, have been linked with CH 4 oxidation (Cooley et al., 2009). By contrast, P. abies-derived sequences, captured with mmoX probes, all grouped either with actinobacterial or proteobacterial PMOs. As with the mcrA analysis, the quality of the short SRA-database-retrieved sequences was lower than the ones produced in the capture sequencing (Table 1). No similarities to ‘traditional’ pmoA genes were found with the capture approach. Taken together, our analysis revealed only minor indications of currently known, ‘traditional’ methanotrophs in the analyzed conifers. However, the novel MOs, detected both in the SRA and in the captured metagenomics data, might have the potential to consume CH 4 in the tree canopies. In general, understanding of the alkane/alkene monooxygenases is far from complete and their functioning within the trees has not been examined. Recent analysis indicates that PMO and MMO enzymes share a common ancestor but have evolved in different directions. Consequently, only MMO and butane MO (BMO) seem to be capable of breaking the C–H bond of CH 4 (Osborne & Haritos, 2019). PMOs and BMOs can primarily break the molecule at the secondary C, which is estimated to require a maximum cleavage energy of 400 kJ mol 1 . Breaking of the C–HbondofCH 4, with an estimated cleavage energy requirement of 431 kJ mol 1 , would at least be a lot less energetically efficient by the PMOs than by MMOs. Yet, we cannot rule out the possibility of PMOs or BMOs oxidizing CH 4 as a co-substrate or unspecifically, as previously shown with BMO from Pseudomonas butanovorans (Cooley et al., 2009). Evidently, we need a deeper understanding of the tree-associated CH 4 consumption mechanisms and microbial communities involved. Excluding the two alphaproteobacterial mmoX SRAfragments, our metagenomic approaches could not detect methanotrophs related to previous needle isolates (Doronina et al., 2004; Iguchi et al., 2012). This likely reflects the well-known challenge to cultivate single strains from complex environmental communities: the aforementioned facultative methanotroph isolates likely represent strains adapted to higher CH 4 concentrations (low-affinity oxidizers). They have proven easier to grow in the laboratory conditions than high-affinity methanotrophs –the first strain able New Phytologist (2021) 231: 524–536 www.newphytologist.com ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation Viewpoints Forum New Phytologist 530 to grow in atmospheric CH 4 concentration was isolated by Tveit et al. (2019). Sequence reads gained in our study could represent so far uncultivated CH 4 oxidizers adapted to low/trace level concentrations of CH 4 (induced by our incubation with CH 4 at 100 ppm). It should be noted that, owing to the presence of acetate in the growth media, our incubation might have favored facultative methanotrophs adapted to the use of this alternative C source. Future directions for moving beyond descriptive studies Currently, we still need more research even on the presence of CH 4 - producing and consuming microbes in the aboveground tree habitat. As reviewed in this paper, the few existing studies on this topic have been largely based on either cultivation, which is biased towards distinct species thriving in the laboratory, or on the sequencing of universal 16S rRNA genes, which lacks information on specific functions and the sensitivity for the rare species. Modern metagenomic tools have the potential for more detailed characterization of tree microbiomes, giving insights to both taxonomy and function. Still, based on the project descriptions behind the retrieved SRA-database entries, metagenomic sequencing is still mostly targeting the host tree genomes more than the associated microbiomes. In addition, owing to the large genome size of the host tree compared with the epiand endophytes, microbiome sequencing through the regular shotgun approach is hindered by a low signal-to-noise ratio (Schneider et al., 2021). In that sense, targeted capture metagenomics shows greater potential to uncover even rare microbial genes among the plant-cell DNA, as we showed here for the spruce shoots. Though the metagenomic tools can generate a vast amount of genomic data, linking unknown DNA fragments to given functions and species is limited by the low amount of annotated reference sequences/genomes in the databases (Kaul et al., 2016; Schneider et al., 2021). To solve this, traditional cultivation approaches are still needed to complement the sequencing methods. Successful isolation of the relevant microbial strains would allow evaluation of the role of putative enzymes in the CH 4 cycle, such as Pseudomonasrelated MOs. With pure cultures, full bacterial/archaeal genomes can be acquired, allowing the analysis of not only CH 4 metabolism but also other traits related to, for example, survival in the plant habitat and interactions with the host (Frank, 2018). Genomes of uncultivated organisms can be derived also through single-cell methods (Rinke et al., 2014), and by building them from metagenomic data (i.e. metagenome assembled genomes; Parks et al., 2017). As an alternative, novel genes/enzymes can be connected to particular functions with the help of metagenomicsbased functional screening approaches (Ngara & Zhang, 2018) and by the use of isotope applications, such as nanoscale secondary ion mass spectrometry and stable isotope labeling, or their combinations (Musat et al., 2016). 08/06/18 10/06/18 30/07/18 31/07/18 01/08/18 04/08/18 −15 −10 −5 0 5 10 15 ng CH4h−1 gDW −1 Jun Jul / Aug (a) (b) 01/04/20 08/04/20 15/04/20 22/04/20 29/04/20 06/05/20 13/05/20 20/05/20 27/05/20 03/06/20 10/06/20 −60 −40 −20 0 20 Date ng CH4h−1 gDW −1 Fig. 3 Spruce shoot methane(CH 4 ) fluxes (median, quartilesand interquartile ranges) measuredat (a) Skogaryd Research Catchmentspruce forest, Sweden, in June–August 2018, and (b) Helsinki yard saplings, Finland, in April–June 2020. Shoot fluxes were measured using manually operated transparent shoot chambers, as in Machacova et al. (2016), connected to an online CH 4 /CO 2 greenhouse gas analyzer (UGGA; ABB - Los Gatos Research, San Jose, CA, USA). In total, there were34 separate shoot flux measurementswith mature trees in Skogaryd and89 separate shoot flux measurementswith spruce saplings inHelsinki. Details of the measurement setup and data processing are given in the Supporting Information Methods S1. Note the different y-axis scales in the two graphs. ©2021 The Authors New Phytologist ©2021 New Phytologist Foundation New Phytologist (2021) 231: 524–536 www.newphytologist.com New Phytologist Viewpoints Forum 531