Making plant methane formation visible—Insights from application of 13C‐labeled dimethyl sulfoxide

Abstract Methane (CH4) formation by vegetation has been studied intensively over the last 15 years. However, reported CH4 emissions vary by several orders of magnitude, thus making global estimates difficult. Moreover, the mechanism(s) for CH4 formation by plants is (are) largely unknown. Here, we introduce a new approach for making CH4 formation by plants clearly visible. By application of 13C‐labeled dimethyl sulfoxide (DMSO) onto the leaves of tobacco plants (Nicotiana tabacum) and Chinese silver grass (Miscanthus sinensis) the effect of light and dark conditions on CH4 formation of this pathway was examined by monitoring stable carbon isotope ratios of headspace CH4 (δ13C‐CH4 values). Both plant species showed increasing headspace δ13C‐CH4 values while exposed to light. Higher light intensities increased CH4 formation rates in N. tabacum but decreased rates for M. sinensis. In the dark no formation of CH4 could be detected for N. tabacum, while M. sinensis still produced ~50% of CH4 compared to that during light exposure. Our findings suggest that CH4 formation is clearly dependent on light conditions and plant species and thus indicate that DMSO is a potential precursor of vegetative CH4. The novel isotope approach has great potential to investigate, at high temporal resolution, physiological, and environmental factors that control pathway‐specific CH4 emissions from plants.

Recently CH 4 produced in plants was suggested to act as a signaling and regulatory molecule playing a physiological role (for more detailed information see review by Wang et al., 2020). It has also been thought that CH 4 interacting within the metabolism of vegetation plays a role in plant stress response, growth, development, and rooting (Cui et al., 2015;Kou et al., 2018;Qi et al., 2017). Abiotic stress in plants leads to oxidative damage, due to rapid overproduction of ROS (Møller et al., 2007). In order to counterbalance the harmful effects of ROS, plants possess antioxidant enzymes as defense systems. The activity of these antioxidant enzymes has been shown to positively correlate with CH 4 , potentially boosting the plants' defense systems (Cui et al., 2015Foyer & Noctor, 2011;Noctor et al., 2012;Uzilday et al., 2014;Zhu et al., 2016).
So far, several pathways of CH 4 formation in living and dead plants have been identified: (1) UV-light produced reactive oxygen species (ROS) reacting with methyl groups of plant compounds for example pectin (Bruhn et al., 2009;Keppler et al., 2009;McLeod et al., 2008;Messenger et al., 2009), (2) plant mitochondria in which the electron transport chain was interrupted (Wishkerman et al., 2011), (3) the amino acid methionine in the presence of hydrogen peroxide (H 2 O 2 ) and iron salts form methyl radicals, which further react to form CH 4 or oxidized products such as methanol (Althoff et al., 2014;Benzing et al., 2017;Lenhart, Althoff, et al., 2015), (4) amino acid methyl groups reacting with ROS (Althoff et al., 2014;Martel & Qaderi, 2019). In recent years great progress in the understanding of CH 4 formation by plants has been made with the identification of organic precursor compounds and the likely important role of methyl radical formation (Althoff et al., 2014;Messenger et al., 2009). Recently Ernst et al. (2022) proposed a reaction mechanism for CH 4 formation that might occur on a cellular level across all living organisms. Thereby, CH 4 is formed by the interaction of ROS with free iron and methylated sulfur and nitrogen compounds in living cells. Furthermore, the authors showed that increasing the level of oxidative stress enhanced CH 4 production in all of the investigated organisms. However, it is clear that there are many physiological and environmental variables that likely control emissions. Particularly, under field-like conditions, an enormous experimental and analytical effort has to be made to determine CH 4 emission rates from specific organisms to the atmosphere but also to identify the environmental factors that regulate their emissions (Wang et al., 2020). Consequently, the detection and quantification of plant CH 4 emission rates remain challenging tasks not only because of the lack of knowledge about mechanism(s) and effects of environmental factors on CH 4 formation, but also due to the need for highly sensitive and appropriate experimental facilities, as changes in CH 4 mixing ratios during incubation experiments are often only in the ppbv range (see e.g., Brüggemann et al., 2009;Keppler et al., 2006). In order to overcome many of the analytical problems and to gain deeper insights into the complex processes involved in plant CH 4 emissions, stable isotope labelling techniques are now commonly used (Brüggemann et al., 2009;Keppler et al., 2006;Lenhart, Weber, et al., 2015, Weber, et al., 2015. Hence, in the recent past, stable carbon isotope labeled compounds, such as 13 C-CO 2 and 13 C-methionine, have already been used as a tool to demonstrate CH 4 emissions by vegetation. When conducting stable carbon isotope labelling experiments, methionine was established as a precursor of CH 4 release by plants and emission rates were shown to be controlled by physical stress factors (Lenhart, Althoff, et al., 2015).
However, the addition of methionine to plant systems is accomplished through assimilation via the roots and thus the actual amount of absorbed methionine is hard to assess. In contrast, dimethyl sulfoxide (DMSO) can be applied to the plant leaves and due to its physiochemical properties, it penetrates the leaf surface and enters the plant system almost instantly (Anchordoguy et al., 1992;Gurtovenko & Anwar, 2007). DMSO has been used as an indicator for the production of hydroxyl radicals (·OH) in biological systems by the formation of CH 4 (Klein et al., 1981;Repine et al., 1981). Here, a methyl radical is cleaved from DMSO during reaction with ·OH (or other ROS) eventually leading to the formation of CH 4 by abstraction of hydrogen from other available organic molecules. In plants DMSO leads to increased H 2 O 2 concentrations (Mannan et al., 2010) and was linked to plant parts experiencing oxidative stress, such as yellowing and senescing leaves, indicating its antioxidant effects (Husband & Kiene, 2007).
In this study, we tested an isotopic approach to gain further insight into the various environmental factors that control CH 4 emissions from plants. Particular focus was put on the role of light-dark cycles and light intensity on CH 4 formation from plants and if these factors can be made visible by applying isotopically labeled DMSO as a CH 4 precursor. Therefore, we applied 13 C-labeled DMSO onto the leaves of tobacco plants (Nicotiana tabacum) and Chinese silver grass (Miscanthus sinensis), which had been cultivated under sterile conditions, and monitored both CH 4 mixing ratios and δ 13 C-CH 4 values during static and dynamic chamber incubation experiments. For high-resolution measurements of CH 4 mixing ratios and δ 13 C-CH 4 values during dynamic experiments, we employed Cavity Ring-Down Spectroscopy (CRDS), while for static incubation experiments, gas samples were withdrawn and measured by isotope ratio mass spectrometry (IRMS).

| Selected plants
Nicotiana tabacum and Miscanthus sinensis were cultured under sterile conditions for our investigations. Both plant species were chosen for this study as there was already considerable in-house practical experience in the sterile incubation of these species. Moreover, these are plants with a distant phylogenetic relationship (dicotyledonous vs monocotyledonous plants). They differ significantly not only morphologically and anatomically, but also physiologically, for example, in their type of photosynthesis, as N. tabacum is a C 3 -plant and M. sinensis is a C 4plant. This physiological disparity causes different responses to varying parameters such as CO 2 concentration, light intensity, or temperature.

| Culture of plants under sterile conditions and application of DMSO
Cultivation of N. tabacum and M. sinensis was achieved following the approach of Lenhart et al. (2019). Both cultures were then incubated with a 16 h daylight phase (light intensity: 150 μmol m −2 s −1 ; temperature: 23°C) and an 8 h dark phase (temperature: 23°C). The constant temperature of 23°C throughout our experiments was chosen to prevent the potential influence of temperature changes on our results, making light intensity the only changing variable. Measurements on the cultures were conducted once the shoots grew to ca. 10 cm (see Table S1 for plant characteristics). By this stage, the cuttings of N. tabacum were fully rooted, while the M. sinensis shoots did not develop roots under these conditions. In order to determine the optimal amount of DMSO to apply to the plant leaves so that no visible impairment to the plant was observed pre-tests were run with varying DMSO concentrations. Therefore, aqueous solutions of DMSO with concentrations between 5% and 20% were applied to the plant leaves.
The plants treated with 5% DMSO solution showed no visible impact, however higher concentrated DMSO solutions led to visible yellowing and damage to the plants (see Figure S1a-c). Hence, an aqueous solution of 5% DMSO was used for our investigations. For incubation experiments of plant cultures, 0.8 ml of a 13 C-labeled DMSO solution was applied onto the leaves of the plants using a calibrated pump dispenser.
The solution consisted of an aquatic solution of 5% DMSO, whereby 6% of the added DMSO was 13 C-labeled and 0.25% of the surfactant Silwet®. Considerable care was taken to ensure that the DMSO solution was uniformly applied to the plant leaves.

| Dynamic incubation experiments
The experimental setup used for dynamic incubations is illustrated in via an IQS plug connector (6 mm inner diameter) (see Figure S2b).
The CRDS enabled simultaneous measurement of headspace CH 4 mixing ratios and δ 13 C-CH 4 values of the headspace air inside the glass flasks with an inflow rate of 23 ml min −1 . In order to avoid lowpressure conditions during the measurements, which might influence the well-being and metabolism of the plants, ambient air was allowed to flow into the glass flasks after first passing through a gas washing bottle (volume: 25 ml; deionized water volume: 15 ml). This in order to correct headspace δ 13 C-CH 4 change rates for the dilution with laboratory air during dynamic incubation (see Methods S1, Figure S3). Hence, δ 13 C-CH 4 values corrected by the dilution model are reported.

| Static incubation experiments
The experimental setup used for static incubations is illustrated in Figure 1  and transferred to evacuated 12 ml exetainers (Labco). At every sampling time, 40 ml were removed for CH 4 mixing ratio measurements and another 40 ml for δ 13 C-CH 4 stable isotope measurements. Directly after sampling air with an equivalent volume of withdrawn headspace gas was added to the flasks to avoid pressure changes. The mixing ratio of CH 4 and δ 13 C-CH 4 values were corrected to account for this dilution.

| Definition of δ values
Stable carbon isotope values are given in the conventional "δ (delta) notation" in per mil versus Vienna Peedee Belemnite (V-PDB). The δ notation is defined as the relative difference in the isotope ratio of a substance compared to the standard substance Vienna Peedee Belemnite (see Equation 1). δ 13 C-CH 4 values are expressed in ‰ versus V-PDB throughout the whole of this manuscript.

| Analysis of δ 13 C-CH 4 values and CH 4 mixing ratios in dynamic experiments
Headspace CH 4 mixing ratios and δ 13 C-CH 4 values during the dynamic incubation experiment were measured using a Picarro G2201-i CRDS. As described above the CRDS was directly connected to the incubation glass flasks with a flow rate of 23 ml min −1 and measurements were made with a very high temporal resolution of 1 Hz. Before measurement, the headspace gas was passed through a Nafion drying tube (Nafion MD110, PermaPure LLC) in order to remove water from the sample and thus avoid optical interferences during the measurement of CH 4 .
Thus, we determined dry mole fractions of CH 4 which we simply express as mixing ratio. The guaranteed CH 4 mixing ratio precision of the manufacturer was 5 ppbv + 0.05% of reading, while the given precision for δ 13 C-CH 4 values is below 0.8‰. In order to rule out instrument drift during measurements with the CRDS compressed air (stored in 1 L Tedlar gas sampling bags) was measured before and after the incubation experiment. The measured drift controls showed no difference for the CH 4 mixing ratios and only minor differences (0.3%) for δ 13 C-CH 4 values. Hartmann (2018), showed that there is a difference between δ 13 C-CH 4 values measured with our CRDS and those with a continuous flow mass spectrometry system (CF-IRMS, see Section 2.7 below). Therefore, reported δ 13 C-CH 4 values were corrected according to Hartmann (2018).

| Analysis of δ 13 C-CH 4 values and CH 4 mixing ratios in static experiments
Headspace samples collected from the static incubation experiments were analyzed for their CH 4 mixing ratios by a gas chromatograph (GC, Bruker Greenhous Gas Analyzer 450-GC) equipped with a flame ionization detector (FID). Before entering the analytical system, the gas sample was passed through a chemical trap filled with Drierite® to remove water. For calibration of the GC-FID five reference gases (Deuste Gas Solutions GmbH), ranging from 1 ppmv (parts per million by volume) to 21 ppmv were used. Peaks were integrated using Galaxie software (Varian Inc.).

| RE SULTS
We first present the results for headspace CH 4 mixing ratio measurements and then show the stable carbon isotope values of headspace CH 4 separately for the dynamic and static incubation experiments of the plant species N. tabacum and M. sinensis after the application of 13 C-labeled DSMO. For both the dynamic and static incubations we first show results of the effect of light-dark cycles and thereafter the influence of light intensity on the change of headspace δ 13 C-CH 4 values in the incubation flasks. Finally, we compare the headspace δ 13 C-CH 4 change rates for each treatment, both incubation types and the differences between the two plant species.

| Changes in CH 4 mixing ratio during incubation in static and dynamic chambers
The measured total range of observed changes in headspace CH 4 mixing ratios (change rate provided in ppbv h −1 ) during incubation of N. tabacum and M. sinensis samples (with added 13 C-labeled DSMO) and controls (medium with added 13 C-labeled DSMO) ranged from −9 ± 11 to 32 ± 33 ppbv h −1 (Figure 2). For comparison, Figure 2 shows the results of the calculated CH 4 change rates derived from the observed changes in headspace δ 13 C-CH 4 values of the plants when 13 C-labeled DMSO was applied (see isotopic sections below,  For static incubations, the observed change rate in CH 4 mixing ratios was in the range of the SD of the triplicate measurements, but as SDs were high (up to 2 ppbv h −1 ) no clear trend could be detected for any of the treatments (see Figure S4). The calculated formation rate of CH 4 derived from DMSO was in the range of 0.04 ± 0.02 for controls, 0.08 ± 0.01 and 1.8 ± 0. In both the dynamic and static incubation experiments the influence of dark-light cycles and light intensity on CH 4 formation were difficult to clearly establish by measurement of mixing ratios because changes lay within the range of the observed SDs. Therefore, the influence of these factors on the release of CH 4 by the plant species were further assessed by the release of 13 C-CH 4 after the application of 13 C-labeled DMSO, as changes in the resulting δ 13 C-CH 4 values might be much better resolved than those monitored by mixing ratio measurements. (

| Effect of light-dark cycles
In

| Static incubation experiment--Effect of light
In addition to the dynamic light-dark experiments, the effect of exposure to light on headspace δ 13 C-CH 4 values was also investi-  (Brüggemann et al., 2009;Keppler et al., 2008;Lenhart, Althoff, et al., 2015). However, the isotope labelling approach revealed considerable effects of light and light intensity on the patterns of released 13 C-CH 4 and thus on CH 4 formation by both plant species. Therefore, this technique allows to better resolve the influences of environmental and physiological factors in and between different plant species. Please note, that the applied isotope approach has only limited use for determining total non-methanogenic CH 4 emission rates by vegetation, as the observed changes in δ 13 C-CH 4 values are limited to only one potential precursor compound (DMSO). However, vegetative CH 4 formation via other precursor compounds and pathways occurs simultaneously and is not monitored by this method. The recently proposed common mechanism of CH 4 formation across all living organisms by Ernst et al. (2022) supports the application of 13 Clabeled methylated compounds for tracking the formation of CH 4 . Thereby, CH 4 is produced via the reaction of ROS, free iron, and methylated compounds within the cells of organisms.

| Effect of dark-light cycles
The calculated increase of headspace δ 13 C-CH 4 change rates showed that 13 C-CH 4 formation from 13 C-labeled DMSO when applied to leaves of N. tabacum and M. sinensis plants was related to light exposure. Moreover, the results from both dynamic and static isotope labelling incubation experiments showed similar patterns for the formation of CH 4 when they are expressed as δ 13 C-CH 4 change rates. (Figure 5). During exposure to light, both plant species emitted 13 C-enriched CH 4 as indicated by higher δ 13 C-CH 4 change rates.
Interestingly, N. tabacum showed no change in δ 13 C-CH 4 values during incubation in the dark, while δ 13 C-CH 4 change rates by M. sinensis still accounted for about 50% compared to those incubated under light exposure. Furthermore, in all cases, the change in δ 13 C-CH 4 values was observed to occur almost immediately after the plants were exposed to light (see Figures 3 and 4). sinensis still emitted 13 C-CH 4 at a ~ 50% lower rate relative to that when under light (see Table S2 for formation rates of CH 4 from 13 Clabeled DMSO for each incubation flask). Our results are in good agreement with observations in which CH 4 emissions by canola, sunflower, and chrysanthemum were not only linked to the exposure of blue light but also blue light intensity (Martel et al., 2020;Martel & Qaderi, 2017. High light radiation might lead to stress reactions in plants, due to an overload of the photoreceptors. Consequently, increased ROS levels might cause higher CH 4 formation rates due to the reaction with methylated compounds (Ernst et al., 2022). Furthermore, higher CH 4 emissions under light irradiation were also linked to higher concentrations of amino acid methyl groups (e.g., methionine), which in previous studies were found to be precursors of CH 4 by the reaction with ROS (Althoff et al., 2014;Ernst et al., 2022;Han et al., 2017;Lenhart, Weber, et al., 2015).
Simultaneously, ROS are accumulated in plants as a consequence of irradiance stress or adjustment to changing light conditions, due to an imbalance between light energy and the cells' capacity to use it (Mullineaux et al., 2018). Interestingly, photosynthesis was recently mentioned as the main driver of CH 4 formation in cyanobacteria (Bižić et al., 2020).
Additionally, the different responses to light and dark conditions on the formation of CH 4 by N. tabacum and M. sinensis, clearly shown by the formation of 13 C-CH 4 by M. sinensis in the dark while N. tabacum shows no CH 4 formation, might indicate that there is possibly more than one process of CH 4 formation in plants depending on their physiology and metabolism (e.g., N. tabacum as a C 3 plant and M. sinensis as a C 4 plant). However, at present, this observation remains difficult to explain and will require further investigation.
Even though the general patterns of 13 C-CH 4 formation were similar for both plants and both incubation types, the measured headspace δ 13 C-CH 4 change rates quantitatively differed between the two types of incubation. This might likely be caused by the differences between the dynamic and static incubations. While the dynamic incubation is a flow-through system that is constantly in exchange with ambient air, during the static incubation the headspace is not exchanged. This difference might lead at least to partial closing of stomata or larger oxidative or abiotic stress reactions in the plant species incubated under dynamic conditions and therefore explain quantitative differences between the plant species. In addition, the changes in the partial pressures of CO 2 and O 2 in the static incubations are larger. Nevertheless, even though a certain degree of uncertainty remains as to the cause of the differences between the observed headspace δ 13 C-CH 4 change rates for both incubations, the patterns observed in our experiments are comparable and therefore allow for some mechanistic interpretation of the results.

| Influence of light intensities
Different light intensities were also observed to influence plant CH 4 formation by N. tabacum and M. sinensis. Please note, that light intensities used for this study (100 and 250 μmol m −2 s −1 ) were much lower when compared to sunlight (ca. 2000 μmol m −2 s −1 ). During higher light intensity incubations, a trend toward higher headspace 13 C-CH 4 emissions by N. tabacum was detected while, contrastingly, M. sinensis tended toward slightly lower 13 C-CH 4 emissions.
Even though we do not know the causal sequence of these different responses to light intensity for either plant species, a likely factor might be the differences in ROS production during the acclimation process to different abiotic stresses (here, light intensity), as ROS are not only toxic by-products of stress metabolism but also function as signal transduction molecules in plants (Choudhury et al., 2017). A further role could be played by the different photosynthesis types of N. tabacum (C 3 -plant) and M. sinensis (C 4 -plant). C 3 plants generally have a lower limit of light saturation compared to C 4 plants, which under higher light intensities might lead to higher oxidative stress, hence more excess energy and an increased formation of ROS and therefore explain higher 13 C-CH 4 formation in N. tabacum compared to M. sinensis (Ernst et al., 2022;Martel & Qaderi, 2017;Messenger et al., 2009 (Hohenberger et al., 2012), leading to the formation of methyl radicals from the thiomethyl group and ultimately CH 4 (Althoff et al., 2014;Benzing et al., 2017;Ernst et al., 2022). DMSO has been detected in several plants such as the sea daisy Wollastonia biflora, saltmarsh grasses, and sugarcane (Dacey et al., 1987;Husband & Kiene, 2007;Otte et al., 2004;Paquet et al., 1994) and is formed by the oxidation of dimethlysulfoniumpropionate (DMSP) and/or DMS, both compounds which can be produced within plant cells (Husband & Kiene, 2007). In plants, DMSO has been shown to exhibit cryoprotective, radioprotective, osmoprotective as well as antioxidant abilities (Husband & Kiene, 2007;Lee & De Mora, 1999;Sunda et al., 2002). The antioxidant properties of DMSO lead to the formation of CH 4 through its reaction with ROS which produces a methyl radical from the organosulfur compound (Ernst et al., 2022;Herscu-Kluska et al., 2008;Repine et al., 1981). Furthermore, higher DMSO concentrations have been linked to plant parts that showed signs of oxidative stress such as yellowing and senescing (Husband & Kiene, 2007), thus suggesting that CH 4 formation is a response to oxidative stress and reaction with ROS is a likely mechanism in vegetation.

| CON CLUS ION
This study presents an isotope labelling method for making CH 4 formation in plants under variable environmental conditions clearly visible. This was achieved by the application of 13 C-labeled DMSO onto the leaves of the plant species N. tabacum and M. sinensis and the measurement of headspace δ 13 C-CH 4 values with a CRDS and an IRMS formed during dynamic and static incubations. Furthermore, we demonstrate that DMSO, a compound known to be ubiquitous in the environment and to occur in plants, might be a potential precursor of vegetative CH 4 emissions. The formation of CH 4 in this study was highly dependent on exposure to light, light intensity, and plant species. Both investigated plant species produced considerably more CH 4 when they were exposed to light compared to in the dark. However, while N. tabacum showed no CH 4 formation when incubated in the dark, M. sinensis still showed about 50% of CH 4 formation when exposed to light. The effect of light intensity was smaller compared to the effect of light exposure and also differed between the two plant species. While N. tabacum showed slightly more CH 4 formation at a higher light intensity, M. sinensis produced less CH 4 under these conditions. Even though in recent years many studies contributed to the understanding of CH 4 formation in aerobic environments and plants, the exact underlying mechanism(s), as well as the influence of environmental factors, are still hardly known.
The method of using dynamic incubations in combination with 13 Clabeled compounds and in-situ measurements with a CRDS enables the visualization of changes in mechanism or pathway-specific plant CH 4 formation patterns almost in real-time. Thus, this new approach has great potential not only for investigating the mechanism(s) of CH 4 formation, but also for any underlying physiological processes that cannot be made visible alone through measurements of CH 4 mixing ratios. Furthermore, our study indicates the highly complex nature of non-methanogenic emissions from vegetation and the multitude of environmental and physiological factors that control emissions. It also further highlights the still profoundly uncertain entity of CH 4 emissions by vegetation, especially with regard to global estimations and demonstrates the need for more research on the exact mechanism(s) of CH 4 formation in vegetation.