The claim of methane (CH4) formation in plants has caused much controversy and debate within the scientific community over the past 4 years. Here, using both stable isotope and concentration measurements, we demonstrate that CH4 formation occurs in plant cell cultures that were grown in the dark under sterile conditions. Under non-stress conditions the plant cell cultures produced trace amounts [0.3–0.6 ng g−1 dry weight (DW) h−1] of CH4 but these could be increased by one to two orders of magnitude (up to 12 ng g−1 DW h−1) when sodium azide, a compound known to disrupt electron transport flow at the cytochrome c oxidase (complex IV) in plant mitochondria, was added to the cell cultures. The addition of other electron transport chain (ETC) inhibitors did not result in significant CH4 formation indicating that a site-specific disturbance of the ETC at complex IV causes CH4 formation in plant cells. Our study is an important first step in providing more information on non-microbial CH4 formation from living plants particularly under abiotic stress conditions that might affect the electron transport flow at the cytochrome c oxidase in plant mitochondria.
Methane (CH4), second in radiative forcing of the long-lived greenhouse gases after CO2, is the most abundant reduced organic compound in the atmosphere and plays a central role in atmospheric chemistry (Solomon et al. 2007).
Although numerous research groups have studied the biogeochemical cycle of CH4 recent investigations show that our understanding of the global CH4 budget is still incomplete (Conrad 2009). For example, CH4 released to the atmosphere from the Earth's crust through faults and fractured rocks estimated to account for 40–60 Tg year−1 (Etiope 2009) is not yet included as a geological source. Another CH4 source with high uncertainty in source strength is Northern wetlands and tundra, which are currently estimated to contribute ∼20% (Wille et al. 2008) of the annual natural emissions. The processes and environmental factors that drive CH4 emissions from tundra regions and tropical wetlands are particularly in doubt. Moreover, there are additional processes in plants that can result in the emission of CH4 through the plant canopy. The best characterized is the transport of CH4 via the canopy, which accounts for most of the flux from anoxic soils vegetated with aquatic plants (Schütz, Seiler & Conrad 1989). Such CH4 transport also exists in trees (Terazawa et al. 2007; Rice et al. 2010). While microbial CH4 production in the rhizosphere of aquatic plants appears to be a common process, production within the plant canopy is unexpected. Nevertheless, in 2006, Keppler and co-workers reported direct CH4 emissions from vegetation foliage under aerobic conditions. This controversial work was followed by a number of similar studies although several were either, unable to reproduce CH4 emissions by living plants or disagreed with the global extrapolations provided by Keppler and his co-workers (Dueck et al. 2007; Wang et al. 2008; Nisbet et al. 2009). However stable isotope labelling studies (Keppler et al. 2008; Vigano et al. 2008, 2009; Brüggemann et al. 2009) verified that CH4 is formed from both dead plant tissue and intact living plants. For example, Brüggemann et al. (2009) demonstrated that plants of Grey Poplar, grown from cell cultures under sterile conditions, released 13C-labelled CH4 after supplementation with labelled 13CO2. Although the observed CH4 emission rates were much lower than those initially reported by Keppler et al. (2006), the work of Brüggemann et al. (2009) clearly demonstrated the existence of a non-microbial CH4 source in Grey Poplar. Non-microbial CH4 release has also been shown to occur under high ultraviolet (UV) irradiation and elevated temperatures from dry and detached fresh plant material (Vigano et al. 2008). Similar results were also obtained when detached tobacco leaves and plant pectin were placed under UV radiation (280–400 nm) at ambient conditions (McLeod et al. 2008). A reaction of reactive oxygen species (ROS) with pectic polysaccharides was suggested as a possible route to CH4 formation under UV radiation (Messenger, McLeod & Fry 2009). A more detailed overview of our current understanding of CH4 formation in aerobic environments can be found in a recent concept article by Keppler et al. (2009). Interestingly, this concept article makes the suggestion that aerobic CH4 formation may occur in all eukaryotes as an integral part of cellular response towards changes in oxidative status. In this context, the disturbance of the mitochondrial electron transport chain of plants might be of particular interest to be studied for the formation of CH4 in living plants.
Plant mitochondria are key organelles in the synthesis of ATP and differ from their animal equivalent as they have specific electron transport chain (ETC) components in their cellular environment for facilitating photosynthesis (Noctor, De Paepe & Foyer 2007). The ETC in plant mitochondria consists of four multi-subunit complexes, respiratory complexes I–IV, assisted by ubiquinone and cytochrome c (Fig. 1). The mitochondrial electron transport chain is a potential source of ROS such as superoxide (O2-) and hydrogen peroxide (H2O2), with complexes I and III being their main sites of production. Complex IV receives electrons from reduced cytochrome c and passes them to an oxygen molecule (via cytochromes a and a3). The complex functionality can be inhibited by sodium azide (NaN3) which blocks electron transport (Møller 2007). Remarkably, non-microbial CH4 generation from animal cells was recently reported when endothelial rat liver cells were exposed to site-specific inhibitors (e.g. NaN3 and NaCN) of the ETC (Ghyczy et al. 2008).
To find out if a similar process could exist in plants we investigated the possibility of non-microbial CH4 formation in heterotrophic plant cell cultures. We also examined disturbance of mitochondrial functionality, by employment of ETC inhibitors such as NaN3, rotenone, salicylhydroxamic acid (SHAM) to see if it affected CH4 release from plant cell cultures. We made use of stable isotope techniques to verify in vivo formation of CH4 from heterotrophically grown cell cultures. When designing some of our experiments, we took into account the findings of Ghyczy et al. (2008) who used rat liver mitochondria and inhibitors of complex IV of the mitochondrial electron transport chain to show generation of CH4. Thus, plant cell cultures of tobacco BY-2 (Nicotiana tabacum), grape vine (Vitis vinifera) and sugar beet (Beta vulgaris L.) grown under sterile conditions were treated with inhibitors of the ETC. Headspace samples from sealed incubation vials were analysed for CH4 concentration at different time intervals throughout the experiments.
Cultivation of cells
Grape vine (Vitis vinifera)
Cell suspensions (50 mL in 250 mL Erlenmeyer flasks), incubated at 26 ± 1 °C with shaking at 160 r.p.m., were subcultured weekly by transfer of 10 mL culture to 40 mL of new medium. Following 6 d growth, these cultures were then used for the experiments. The medium used for grape vine cultures had the following composition: Gamborgs B5 including vitamins, 3.2 g L−1, sucrose 30 g L−1, casein hydrolysate (acid hydrolysed) 0.25 g L−1, kinetin 1 mg mL−1, 200 µL naphthalene acetic acid, stock 1 mg mL−1.
Sugar beet (Beta vulgaris L.)
Fifty millilitres of cell suspensions were incubated in 250 mL Erlenmeyer flasks at 26 ± 1 °C with shaking at 160 r.p.m., The cells were subcultured biweekly by transfer of 10 mL culture to 40 mL of new medium. Following 11 d growth, these cultures were then used for the experiments. The medium used for sugar beet cultures had the following composition: Murashige and Skoog (MS) Salts 4.41 g L−1, sucrose 30 g L−1, 6-benzylaminopurine 0.05 mg L−1.
Tobacco (Nicotiana tabacum cv. BY-2)
Fifty millilitres of cell suspensions were incubated in 250 mL Erlenmeyer flasks at 26 ± 1 °C with shaking at 160 r.p.m. Cells were subcultured weekly by transfer of 5 mL culture to 45 mL of new medium. Following 6 d growth these cultures were then used for the experiments. The medium used for tobacco BY-2 cultures had the following composition: MS salts 4.3 g L−1, sucrose 30 g L−1, myo-inositol 100 mg L−1, thiamine 1 mg L−1, 2,4-dichlorophenoxyacetic acid 0.2 mg L−1, KH2PO4 255 mg L−1. The pH of all media was adjusted to 5.7 using KOH. All cell cultures were grown in the dark.
Isotopic labelling of tobacco cells
When tobacco BY-2 cells were subcultured, as described above, the new medium was supplemented with 10 mg 13C sucrose (Sigma-Aldrich, St Louis, MO, USA, isotec™ 99% 13C atoms). The amount of 13C in sucrose was calculated to be in the range of 1.7 to 1.8%, corresponding to δ13C values of 539 and 631‰, respectively. The cells were then grown for the normal 6 d period prior to the start of the stable carbon isotope measurement experiments. Taking into account the dilution by the unlabelled cells with a δ13C value of ∼−25‰ (5 mL unlabelled culture to 45 mL of labelled new medium; ∼1:9) we estimated the isotope signature of the 13C labelled cells to be in the range of 477 and 565‰.
δ13C(CH4) values of the CH4 emitted from the cells were calculated as follows:
Area defines the peak area of CH4 measured by the isotope ratio mass spectrometer (IRMS) and was measured in volt second (Vs).
Treatments using electron transport chain disrupters
Cell cultures (5 mL) were transferred to sterile vials (40 mL) and different concentrations of sodium azide (NaN3), SHAM (C7H7NO3), rotenone (C23H22O6) and kanamycin (C18H36N4O11) added. Azide, SHAM and rotenone were dissolved in double distilled and sterile water. Rotenone was dissolved in dimethyl sulfoxide (DMSO). Control samples containing only cell cultures were also prepared. The vials and their contents were sealed with sterile PTFE/Silicone septa and then incubated in a growth chamber at a temperature of 26 ± 1 °C with shaking at 120 r.p.m. for varying time intervals. Results are presented as ΔCH4 (ng g−1), the difference between the CH4 headspace concentrations at the start and end of the incubation period. At the end of each experiment cultures were lyophilized for determination of cell dry matter content.
Measurement of headspace CH4 concentration by GC-FID
A sample of vial headspace (5 mL) was injected manually into a gas chromatograph (Shimadzu GC-14B, Shimadzu, Kyoto, Japan) equipped with flame ionization detector using a 10 mL Hamilton TTL gas-tight syringe. Synthetic air was used as carrier gas (c. 300 kPa). The flame ionization detector gas flow was comprised of compressed air at 40 kPa and hydrogen at 50 kPa. Methane was quantified using a reference gas containing 8.905 ppmv CH4 (diluted in synthetic air) with an analytical precision of 11 ppb.
Continuous flow isotope ratio mass spectrometry for measurement of δ13C(CH4)
Headspace gas samples were transferred from the sample vial to an evacuated 40 mL sample loop. CH4 was trapped on Hayesep D, separated from interfering compounds by gas chromatography and transferred via an open split to the IRMS (ThermoFinnigan Deltaplus XL, Thermo Finnigan, Bremen, Germany).
All 13C/12C -isotope ratios are expressed in the conventional δ notation in per mil versus V-PDB, defined as:
A tank of high purity carbon dioxide (carbon dioxide 4.5, Messer Griesheim, Frankfurt, Germany) with a known δ13C value of −23.64‰ (V-PDB) was used as the working reference gas. All δ13C values were normalized relative to V-PDB using a CH4 standard. Samples were routinely analysed three times (n = 3) and the average standard deviations of the GC/C/IRMS measurements were in the range of 0.1 to 0.3‰.
To evaluate the effect of NaN3 treatments on CH4 production (Fig. 2) and the effect of various inhibitors (Fig. 5) one-way analysis of variance (anova) and a Tukey- Honestly Significant Difference (HSD) analysis were employed. Levels of significances were defined as follows: P < 0.001 highly significant and P < 0.05 significant. Repeated measurement anova and Tukey test were used to investigate the time-dependent treatment effect on CH4 concentration (Fig. 3) and changes in isotopic signature (Fig. 4). Due to low repetition numbers (n = 3 per treatment and measurement time) the Huynh–Feldt correction (Huynh and Feldt 1970) was applied. All calculations were carried out with the software package SPSS version 17 (Chicago, IL, USA).
RESULTS AND DISCUSSION
Methane emission from cell cultures of different plants when treated with NaN3
In an initial set of experiments, tobacco cells (Nicotiana tabacum cv. BY-2) were studied as these cells are often used as a model system in molecular cell biology (Nagata, Sakamoto & Shimizu 2004). We monitored the change of CH4 concentration in the headspace of the incubation vials containing tobacco BY-2 cells when incubation under sterile conditions in the dark at 26 ± 1 °C was carried out for 19 h (Fig. 2a). The untreated cell cultures (control) showed minor CH4 uptake rates [−6.4 ± 2.6 ng CH4 g−1 dry weight (DW)], whereas treatment with NaN3 showed substantial formation of CH4 at all concentrations tested (P < 0.001). At 1 mM NaN3 cells were observed to produce 42 ± 11 ng CH4 g−1 DW, and a maximum emission of 118 ± 8 ng CH4 g−1 DW was monitored when cells were treated with 2.5 mM NaN3. Interestingly, less CH4 was formed (although not significant) when cells were exposed to higher concentrations of NaN3 (5 and 10 mM). A plausible explanation for this could be that at the higher NaN3 concentrations faster cell mortality interrupts transport of the precursor compounds to the site of CH4 formation.
In order to determine if azide-induced CH4 emission from plant cells is a general phenomenon two other plant species were examined: cells of grape vine (Vitis vinifera) and sugar beet (Beta vulgaris L.). The emission pattern for sugar beet cell cultures is shown in Fig. 2b, where it can be seen that there was slight, but not significant, CH4 uptake for the control samples (−2.9 ± 2.6 ng g−1 DW), whereas for the 1 and 5 mM NaN3 treatments CH4 emissions were similar (3.5 ± 2.2 and 5.6 ± 1.2 ng CH4 g−1 DW, respectively). CH4 emissions from cells treated at the 10 mM level were highest at 16.8 ± 13.8 ng g−1 DW, which is highly significant formation (P < 0.001) when compared with control samples.
The CH4 emission pattern for grape vine cell cultures (Fig. 2c) was quite similar to that found for tobacco BY-2 cultures. A slight, although statistically not significant, CH4 uptake was monitored for the control cells (−0.73 ± 1.14 ng g−1 DW). While at the 5 mM treatment level, emissions (13.7 ± 2.2 ng CH4 g−1 DW) were approximately three times higher than those for the 1 mM treatment (5.2 ± 2.3 ng CH4 g−1 DW), they were comparable with those at the 10 mM level (12.5 ± 3 ng CH4 g−1 DW).
The apparent observed variations for the different species of plant cell cultures might be due to differences in their tolerance, metabolic activity or uptake of this antibacterial agent. Sodium azide uptake is a function of different physiological parameters which include cell aggregates, differences in growth phase, cell wall properties, and transporters. It is also likely that NaN3 can be toxic to other cell compartments, which increase or reduce NaN3 permeability and toxicity. For example, the grape vine cells were found to grow much slower compared with the tobacco BY-2 cells and therefore might reflect a different physiological activity. Furthermore, the observed CH4 emission patterns between species could be related to differences in the formation of secondary metabolites such as phenolic compounds. Therefore, the activation of secondary metabolism to alleviate stress should also be considered to explain differences in CH4 emission patterns.
Time dependency of CH4 formation
Tobacco BY-2 cell cultures were investigated at higher time resolution over a 48 h period in order to determine the pattern of CH4 release (Fig. 3). CH4 concentrations for the control and NaN3 treatment did not alter during incubation for the first 9 h and merely reflected the initial CH4 concentration in the vials (2.02 ppmv). After a 12 h incubation period headspace concentrations in vials containing cells treated with NaN3 were approximately 5% higher than those for vials with untreated cells. Concentrations of CH4 for NaN3 treated cells further increased until termination of the experiment (48 h) to a maximum value of about 2.54 ppmv. From the change in CH4 headspace concentrations, incubation period and cell dry matter content it was possible to calculate CH4 emission rates for the incubation periods. Highest CH4 emission rates of 12 ng g−1 DW h−1 were found to occur when cells were incubated between 9 and 12 h. Between the 12 and 24 h incubations, CH4 emission rates declined slightly to a value of 9.3 ng g−1 DW h−1 but, thereafter, decreased with a value of 1.6 ng g−1 DW h−1 observed between the 24 and 48 h incubations. From these data, we assume that after incubation for 9 h NaN3 reached the threshold for exerting its effect on CH4 formation. The observed decline in emission rates after a 12 h incubation period might be related to cell mortality. Interestingly, a small increase in headspace CH4 concentration between the 12 and 24 h incubations was observed for the untreated cells, with a calculated emission rate of 0.6 ng g−1 DW h−1. This emission rate would then appear to increase considerably between the 24 and 48 h incubation periods where the large standard deviation noted between replicates after the 48 h incubation might indicate involvement of other stress factors. More detailed experiments will be necessary to further study this finding.
13C labelled tobacco cells and related stable isotope measurements of methane
As mentioned earlier only a small change was noted in CH4 concentration in the headspace above untreated cells during the initial 24 h incubation period (see Figs 2 & 3). It has previously been reported that stable carbon isotope measurements are much more sensitive in detecting CH4 emissions from plant material than by measurement of their concentration (Vigano et al. 2008; Brüggemann et al. 2009). For example, CH4 emission rates as low as 0.03 ng CH4 g−1 DW were measured from 99% 13C-labelled dried wheat (Triticum aestivum) leaves (Vigano et al. 2008). Thus, we decided to also measure the δ13C value of CH4 in the headspace above 13C isotopically labelled tobacco cells (see Methods section) without any treatment before and after incubation. Figure 4a shows that the δ13C(CH4) values for the headspace above the medium and unlabelled cells did not change over the entire incubation period and closely reflected the δ13C(CH4) of the initial value of laboratory background air (−46.1 ± 0.3‰). Similar values of δ13C(CH4) in the headspace of the vials (ranging from −45.9 to −46.3‰) were measured for all samples after incubation for 3 h and these values did not change in control samples (medium only) throughout the rest of the 24 h incubation period. In contrast there was a change in δ13C(CH4) values towards higher values (P < 0.001) in the headspace above isotopically labelled tobacco cells after both 19 and 24 h incubations (−42.6 ± 0.6 and −40.4 ± 0.8‰, respectively), verifying in vivo formation of CH4 in these cultures. Based on the changes of the δ13C(CH4) values and the isotope label of the cells emission rates of 0.3 and 0.38 ng CH4 g−1 DW h−1 were calculated for incubation times of 19 and 24 h, respectively. These emission rates are in the same range to those reported (0.16–0.7 ng g−1 DW h−1) for grey poplar shoot cultures grown under low light sterile conditions (Brüggemann et al. 2009).
In another set of experiments in which the 13C labelled tobacco cell cultures were treated with 2.5 mM NaN3 a strong and highly significant (P < 0.001) increase in the δ13C value (116 ± 39‰) of headspace CH4 was monitored after 19 h (Fig. 4b). This value further increased to 147 ± 39‰ after incubation for 24 h (±1 h). Based on the δ13C(CH4) values and increase in headspace CH4 concentration in the vials during the incubation period it is possible to correct for the background CH4 value and thus calculate the δ13C values of the CH4 emitted from the tobacco cells for both time periods. Almost identical δ13C(CH4) values of 448 ± 65‰ and 440 ± 30‰ were found for tobacco cells incubated for 19 and 24 h, respectively. These isotope numbers reflect closely the δ13C values of the 13C labelled tobacco cells (in the range of 477 to 565‰, see the Methods section) and thus clearly demonstrate in vivo formation of CH4.
Methane emission from tobacco BY-2 cells treated with other ETC inhibitors
Sodium azide is known to block electron transport from complex IV and in this study was shown to enhance CH4 release by plant cell cultures. In order to examine if other disturbances in ETC compartments could also enhance CH4 emissions SHAM and rotenone were tested on cultures of tobacco BY-2 cells. Furthermore, in order to examine if protein synthesis inhibition could have a similar effect, tobacco BY-2 cells were treated with kanamycin. CH4 emissions from tobacco BY-2 cells treated with SHAM, rotenone and kanamycin are shown in Fig. 5. SHAM is a known inhibitor of alternative oxidase (AOX) (Fig. 1), a 32–36 kDa diiron enzyme that, similar to complex IV (Møller 2007), produces water as the final product. Rotenone is routinely used as an inhibitor of complex I, a type I NADH dehydrogenase and as the first enzyme of the mitochondrial ETC catalyses the transfer of electrons from NADH to coenzyme Q (CoQ). (Møller 2007). Kanamycin is an aminoglycoside antibiotic which binds to the 30S ribosomal subunit preventing RNA translation thus inhibiting protein synthesis (Conte et al. 2009).
CH4 emissions from cultures of tobacco BY-2 cells were not significantly affected by SHAM, rotenone or kanamycin at the concentrations tested (Fig. 5), whereas addition of NaN3 resulted in a significant (P < 0.001) increase in CH4 production.
To check whether CH4 formation was related to cell viability, cells were stained with trypan blue (0.4%) after treatment and incubation for 16 and 24 h. Trypan blue only stains dead cells since live cells with intact cell membranes exclude this dye. Whereas with the NaN3 and SHAM treatments most cells had died between 16 and 24 h of incubation, cells treated rotenone and kanamycin looked predominantly viable even after 24 h (data not shown). These results would indicate that a site-specific target disturbance of the ETC at complex IV is a possible source for CH4 formation in plant cells.
CONCLUDING REMARKS AND PERSPECTIVES
In this study we have shown that CH4 formation occurs in three plants species of heterotrophically grown cell cultures. Under non-stress conditions the plant cell cultures produced small amounts of CH4 but these were increased when NaN3, a compound known to disrupt electron transport flow at the cytochrome c oxidase (complex IV) in plant mitochondria, was added to the cultures. The addition of other ETC inhibitors such as rotenone and SHAM did not result in significant CH4 formation. Furthermore, cell death is an unlikely explanation for the observed CH4 emissions since addition of SHAM did not lead to significant CH4 formation even though most cells were dead before the end of the experiment with this treatment. We suggest that CH4 formation in the mitochondria might be related to a site-specific disturbance of the ETC. This would also be in agreement with the findings of Ghyczy, Torday & Boros (2003), Ghyczy et al. (2008) who demonstrated that endothelial cells from rat liver produced CH4 when exposed to site-specific inhibitors of complex IV of the mitochondrial ETC.
Berg et al. (2010) have proposed that NaN3 can be considered as a specific tool to study endoplasmic reticulum stress in general. Furthermore, specific effects of NaN3 treatment in Arabidopsis leaves, and the possible relationship of complex IV inhibition with abiotic stress conditions (heat, salt, and heavy metal stress) in plants have been suggested by Schwarzländer, Fricker & Sweetlove (2009). Thus, our experiments should be of relevance for researchers when investigating CH4 formation in plants under environmental stress conditions.
Emission rates of up to 12 ng g−1 DW h−1 were observed for tobacco cells when treated with NaN3. This value lies at the lower end of the range of CH4 emission rates from intact living plants (12 to 370 ng g−1 DW h−1) reported by Keppler et al. (2006). Thus disturbance of the ETC in mitochondria electron transport particularly under abiotic stress might be one factor that controls CH4 formation in plants. However, other physiological factors and environmental stress factors are also suggested to be involved in its formation (e.g. Keppler et al. 2009; Messenger et al. 2009; Wang et al. 2009).
Further research is required to determine how CH4 emissions from intact plants are linked to abiotic stress factors such as temperature, drought or hypoxia. The next immediate aim of our research group is to focus on the identification of precursors in plant cells giving rise to CH4. Therefore we plan to use the information gained from this study to select a number of candidate 13C positionally labelled methyl donor compounds and to employ these with stable isotope techniques in our future studies.
The authors thank Jack Hamilton and Claudia Kammann for critical comments on the manuscript and Markus Greule for performing the IRMS measurements. We are grateful to Carl Brenninkmeijer and Dieter Scharffe for provision of the GC-FID system. We thank EON Ruhrgas for financial support. F. Keppler acknowledges support provided by the European Science Foundation (European Young Investigator Award) and the German Science Foundation (KE 884/2–1).