Hydrogenotrophic Methanogenesis and Autotrophic Growth of Methanosarcina thermophila

Although Methanosarcinales are versatile concerning their methanogenic substrates, the ability of Methanosarcina thermophila to use carbon dioxide (CO2) for catabolic and anabolic metabolism was not proven until now. Here, we show that M. thermophila used CO2 to perform hydrogenotrophic methanogenesis in the presence as well as in the absence of methanol. During incubation with hydrogen, the methanogen utilized the substrates methanol and CO2 consecutively, resulting in a biphasic methane production. Growth exclusively from CO2 occurred slowly but reproducibly with concomitant production of biomass, verified by DNA quantification. Besides verification through multiple transfers into fresh medium, the identity of the culture was confirmed by 16s RNA sequencing, and the incorporation of carbon atoms from 13CO2 into 13CH4 molecules was measured to validate the obtained data. New insights into the physiology of M. thermophila can serve as reference for genomic analyses to link genes with metabolic features in uncultured organisms.


Introduction
Biogenic methane (CH 4 ) is produced by methanogenic archaea, using three main substrates: acetate, CO 2 , and substances containing a methyl group [1] (Table 1). Among all methanogenic archaea, only the order Methanosarcinales includes members able to metabolize all three substrates [1]. Acetoclastic methanogenesis is exclusively performed by the genera Methanosarcina and Methanosaeta, both members of the Methanosarcinales, which differ in their substrate specificity and their affinity to acetate [1,2]. Methylotrophic methanogenesis can be hydrogen-dependent or hydrogen-independent and is limited to Methanosarcinales, Methanomassiliicoccales, and one species of Methanobacteriales [1,3]. Furthermore, genome analyses suggest hydrogendependent methylotrophic methanogens in the new phylum Verstraetearcheota [4]. Hydrogen-dependent species use hydrogen (H 2 ) to reduce the methyl group to CH 4 [1,5]. Hydrogen-independent methanogenesis involves the reduction of methyl groups with electrons deriving from the oxidation of further methyl groups, so that for each three CH 4 molecules, one molecule of CO 2 is produced [1,5].
Contrary to the two preceding pathways, hydrogenotrophic methanogenesis, the reduction of CO 2 with H 2 to CH 4 , can be performed by nearly all methanogens. Among them, obligate CO 2 -reducing methanogens and microorganisms able to use a broad range of substrates can be distinguished. They differ in some of the involved enzymes and the mode of energy conservation [6]. Organisms thought to be unable to perform hydrogenotrophic methanogenesis are found solely within the Methanosarcinales. It was shown for instance that the mesophilic methanogen Methanosarcina acetivorans is unable to use CO 2 for methanogenesis [1].
The organism Methanosarcina thermophila was firstly described under the name TM-1 by Zinder and Mah in 1979 [7]. It was isolated from a thermophilic anaerobic sludge digester and is able to metabolize acetate, methanol, methylamine, and trimethylamine [7]. In the last few years, M. thermophila was repeatedly detected in various biogas fermenters with molecular methods, which indicates that it might play a central role in active communities of anaerobic digesters [8][9][10]. The methanogen is thought to be crucial to overcome process disturbances due to high acetate levels in biogas reactors [11,12] and to be outstandingly resilient encountering changing temperatures during anaerobic digestion [13]. The observations in literature about the ability of M. thermophila to use CO 2 as a methanogenic substrate and a carbon source range from no methanogenesis or growth [7,14] to weak growth [15] on CO 2 , but no concrete data is published concerning this topic.
In the past years, sequencing approaches revealed new distinct groups of archaea that were classified as potential methanogens due to specific genes linked to methanogenesis [4,16,17]. The physiological characterization of cultivable methanogens is crucial to validate the correlation between molecular data and functional traits. Therefore, we investigated the consumption of H 2 and CO 2 by M. thermophila cultivated either with methanol as co-substrate or without organic substrates. Further, we determined the rate of CH 4 production, acetate excretion, and DNA yield during the autotrophic incubation of M. thermophila.

Media and Incubation
Conditions. The mineral medium contained per liter 0.35 g K₂HPO₄, 0.23 g KH₂PO₄, 0.244 g MgSO₄, 0.25 g CaCl₂ * 2H₂O, 2.25 g NaCl, 0.002 g FeSO₄ * 7 H₂O, 2.49 g NH₄Cl, 1 mL resazurine solution (0.115% w/v) as redox indicator, 1 mL trace mineral solution (SL-10 DSMZ medium 320), 20 mL NaHCO₃ solution (10% w/v), and 975 mL distilled water. The medium was flushed with a N 2 /CO 2 mixture (70 : 30) and simultaneously cooled down to approximately 5°C to enable additional CO 2 to dissolve. After the pH was adjusted to 6.8, 50 mL of medium was anaerobically aliquoted in 250 mL serum bottles, which were flushed with either a N 2 /CO 2 (70 : 30) or a H 2 /CO 2 (80 : 20) gas mixture to guarantee anaerobic conditions. Subsequently, the bottles were sealed and autoclaved. The sterile medium was amended with 0.2 mL Na 2 S * 9 H 2 O solution (23.1% w/v), 0.2 mL cysteine-HCl solution (7.5% w/v), and 0.5 mL vitamin solution (VL-141 DSMZ) per bottle. Due to earlier protocols, 2 mL erythromycin solution (0.1% w/v) was added per bottle to avoid bacterial infections right before the inoculation. This precautional measure proved to be unnecessary, as no contaminations of the culture appeared, when it was inoculated in a rich medium containing no erythromycin at the end of the investigation. Furthermore, 0.25 mL pure methanol were amended if necessary. To raise the partial pressure of the substrate gases, headspaces were upgraded initially with 100 mL extra filter sterilized gas. The Na 2 S and the cysteine-HCl solutions were autoclaved; the vitamin solution, the erythromycin solution, and the methanol were filter sterilized. The samples were inoculated with Methanosarcina thermophila TM-1 (DSM strain 1825, obtained from DSMZ-German Collection of Microorganisms and Cell Cultures, Germany) via a syringe and incubated at 50°C ± 0.5°C and 70 rpm in a closed batch system.

Gas and Chemical Analysis.
To quantify gas amounts, the overpressure in the headspaces of the bottles was measured with a digital precision monometer (GDH 200-13, Greisinger electronic, Germany) and normalized with the ambient pressure (data from ZAMG (Zentralanstalt für Meterologie und Geodynamik, Austria)). The gas composition was determined with a Shimadzu GC2010 as described in [18], using a TCD (thermal conductivity detector). The samples were taken and immediately injected with 1 mL syringes. The pH value was monitored to ensure stable incubation conditions. It was measured with a glass electrode and was invariable in all experiments. For the analysis of acetate concentrations, 1 mL samples were centrifuged for 10 min at 20.000 ×g to remove solid components. The supernatants were filtrated through a 0.2 μm RC (Phenomenex, Germany) filter and analyzed via HPLC on a Shimadzu Prominence system as described before [19]. To observe the incorporation of carbon atoms from CO 2 molecules into CH 4 molecules, 10 mL 13 CO 2 (36% (v/v), diluted in carbon-free air (Messer, Austria)), was added to the headspace of the serum bottles. The proportion of 13 C in CO 2 and CH 4 gas was determined with a Picarro G2201-i Analyzer (USA).

DNA-Based Analysis.
To quantify the dsDNA content in the culture fluid, genomic DNA was extracted from the pellet of 1 mL culture fluid using a NucleoSpin® Soil Kit (MACHERY-NAGEL, Germany). Extraction was performed according to the manufacturer protocol, using SL1 in the first lysis step. The DNA content in the extracts was measured with a Quantus™ Fluorometer (Promega, USA, Cat number E6150). To ensure the identity of the culture and to exclude an infection with another hydrogenotrophic microorganism, DNA from a well growing sample was extracted at day 21. Genomic DNA was amplified by PCR, using the archaeal primers 109f [20] and 1492r [21]. The PCR mix contained per reaction volume of 50 μL: 19.4 μL PCR grade water, 26.4 μL Red Taq DNA Polymerase 2x Master Mix (VWR, USA, Cat. number 733-2547), 1.1 μL of each primer, and 2 μL template. The reaction was executed in a FlexCycler (Analytik Jena, Germany) with 10 min at 95°C for initial denaturation, followed by 35 cycles of 30 s at 95°C, 30 s at 52°C, and 45 s at 72°C. The PCR product was sequenced by Eurofins Genomics (Germany), and the resulting nucleotide sequences were analyzed with NCBI BLAST. 2.4. Statistics. The statistical analyses were performed using STATISTICA 12 (StatSoft®). After testing the data for normality and homogeneity of variance, significant differences between groups were calculated by one-way or multivariate ANOVA (analysis of variance). To assess relationships between variables, a Pearson correlation was used. The alpha level used throughout was 0.05 for significant and 0.01 for highly significant results.

Results
3.1. Growth on Methanol and CO 2 . In a first approach, Methanosarcina thermophila was grown on a mineral medium containing methanol and H 2 /CO 2 in the headspace ( Figure 1). The headspace of two inoculated samples was replaced by a sterile N 2 /CO 2 mixture, serving as H 2 -free controls to quantify the gas fluxes generated during the degradation of methanol ( Figure 1). A not inoculated negative control, containing H 2 /CO 2 in the headspace (data not shown), resulted in no CH 4 production, and the H 2 and CO 2 contents stayed unchanged over the whole incubation period of 23 days. The presence of H 2 in the bottles had a positive effect on the cumulative CH 4 production and a negative effect on the net CO 2 production after 23 days.
To quantify gas fluxes occurring separately from the methanol degradation, the net gas turnover in the H 2 -free controls was subtracted from the net gas turnover in the H 2 -containing bottles. Referring to Figure 1, the results showed that H 2 variants consumed 4.21 mmol H 2 and 0.82 mmol CO 2 as well as produced 0.66 mmol CH 4 more than the H 2 -free controls.

3.2.
Growth on H 2 /CO 2 . In a next step, a mineral medium, containing solely CO 2 as carbon source and H 2 as electron acceptor, was inoculated with 0.1 mL sediment of an active culture of M. thermophila, grown on a methanol-acetate medium. The small inoculation volume was chosen to prevent the transfer of potential organic carbon sources. In the first generation of such setup, three of nine samples produced CH 4 during 38 days of incubation (data not shown). One of the samples actively producing CH 4 of the first generation was frozen and subsequently utilized to inoculate (0.1 mL) the second generation of M. thermophila grown on H 2 /CO 2 . In this trial, three out of five samples produced between 1.4 and 1.7 mmol CH 4 within 56 days of incubation, with lag phases ranging from zero to 21 days. The other two samples and the negative controls, bottles containing either no inoculum or no H 2 , did not yield any CH 4 . The theoretical potential CH 4 production (disregarding anabolism), calculated according to the available CO 2 and H 2 content at the beginning of the incubation, would have been 2.50 and 2.33 mmol per bottle, respectively (Table 1). Therefore, the actual measured CH 4 production could mathematically derive from the reduction of CO 2 and accounts for approximately 65% of the potential CH 4 production. The sequencing results of an aliquot of culture fluid from day 21 showed 99.69% identity of the sample with the ordered Methanosarcina thermophila strain DSM 1825 (NCBI accession number: AB973357.1). From the next experiment, all incubation bottles were inoculated with 1 mL of an active CO 2 culture to ensure a higher rate of successful cultivations than achieved with 0.1 mL transfer volume. Indeed, in generation three, all six samples showed visible growth. In three of six parallels, M. thermophila was incubated in a medium lacking cysteine   and erythromycin to detect possible CH 4 production, resulting from the utilization of those two medium components as methanogenic substrate. The presence or lack of cysteine and erythromycin had no significant effect on the cumulative CH 4 production or the cumulative CO 2 and H 2 consumption until the end of the incubation (multivariate ANOVA: p = 0 58). The average CH 4 yield was 1.53 ± 0.03 mmol, the average H 2 consumption 5.53 ± 0.25 mmol, and the average CO 2 consumption −0.55 ± 0.14 mmol in all six bottles after 35 days. At this point of the incubation, the pressure in the bottles was already negative, as for every produced molecule of CH 4 five substrate molecules are consumed (Table 1). For this reason, gas measurements at later time points were less trustworthy and therefore not taken into account for data analysis, although CH 4 concentration in the headspace continued to increase. Hydrogenotrophic methanogenesis in three parallels of the third generation (with erythromycin and cysteine) was further characterized concerning DNA content and concentration of acetate in the medium (Figure 2). The concentration of acetate reached up to 0.90 mM, which is the equivalent of 0.05 mmol/bottle.

Carbon Flow and Methanogenic Performance.
To validate whether the carbon of the produced CH 4 molecules derived from CO 2 molecules, 13 C-labelled CO 2 was added to two of three parallels of the fourth generation. The addition of 10 mL CO 2 with 37% 13 C resulted in an average 13 CO 2 concentration of 5.22% in the headspace of the two samples. After 3 weeks and an average CH 4 production of 0.75 ± 0.12 mmol, the 13 C content of the produced CH 4 was approximately 3.62% and thus in the same range as the 13 C content of the remaining CO 2 (approximately 3.46%) in the labeled bottles. The 13 C proportions of CH 4 (1.07%) and CO 2 (1.02%) in the bottle without labelled CO 2 were, however, distinctively lower and within the natural range. Thus, it can be concluded that the labeled carbon atoms were transferred from the CO 2 pool to the CH 4 pool. During the fifth generation, the sampling intervals of three parallels were shortened to quantify the rate of hydrogenotrophic methanogenesis performed by M. thermophila. From day 3 onwards, CH 4 production showed a rather linear (R 2 = 0 97, p < 0 01) than exponential pattern, with an average rate of CH 4 production of 0.04 mmol/day (0.11 mmol/ day/L initial H 2 /CO 2 ) ( Figure 3). Further, there was a strong linear correlation between the production of CH 4 and the consumption of H 2 and CO 2 , respectively ( Figure 4). To complete the investigations, autotrophically grown cells were microscopically compared with cells grown on methanol and acetate. As also confirmed by sequencing data, there were no signs for contaminations in the culture grown on CO 2 . The comparison of heterotrophically and autotrophically cultivated organisms showed decreased fluorescence in CO 2 cultures, indicating a lower level of the molecule F 420 and therefore a lower methanogenic activity in those cells, corresponding to the different CH 4 production rates on methanol and H 2 /CO 2 (Figure 1).

Discussion
The present study on autotrophic growth by Methanosarcina thermophila started with the investigation of CO 2 and H 2 as co-substrates of methanol. The collected data from gas measurements showed a biphasic CH 4 production of M. thermophila, with a second lag phase, occurring during the shift from consumption of the preferred substrate methanol to consumption of CO 2 (Figure 1). Interestingly, previous studies investigating Methanosarcina bakeri strain 227 and strain MS by Ferguson and Mah [22] as well as Hutten et al. [23] did not observe a biphasic growth pattern.  In the present study, the observed CO 2 production, during the degradation of methanol, was consistent with the stoichiometry of the hydrogen-independent methylotrophic methanogenesis, with every fourth methanol molecule being oxidized to CO 2 [1]. This pathway of methanol degradation was also suggested for the genus Methanosarcina by Zinder [24]. After the depletion of methanol, CH 4 production continued, although slower, and was accompanied by decreasing H 2 and CO 2 levels. Therefore, it could be shown that M. thermophila is able to perform hydrogenotrophic methanogenesis in a methanol-CO 2 medium (Figure 1). Reduction of CO 2 in the presence of methanol was already uniformly observed by Zinder and Mah [7] as well as Mladenovska and Ahring [14]. Their findings, however, deviate from each other concerning the CO 2 consumption after the depletion of methanol. Zinder and Mah [7] stated that metabolism of H 2 stopped as soon as methanol was depleted, whereas Mladenovska and Ahring [14] found ongoing methanogenesis after methanol was exhausted. As mixotrophically grown cells transferred into a new H 2 /CO 2 medium did not show any growth or CH 4 production during their experiments, Mladenovska and Ahring [14] further stated the hypothesis that methanol seems to be critical for cell formation, which was clearly not true for the culture used in the present experiments. The ability or inability of M. thermophila to produce CH 4 from CO 2 as a sole methanogenic substrate is mentioned in various articles, but there are only two publications in which the topic was experimentally investigated. Zinder and Mah [7] did not succeed to grow M. thermophila autotrophically during their initial isolation and characterization of the organism in 1979 and stated further that they found no clear explanation for this fact. In 1985, Zinder et al. [15] stated that growth of M. thermophila "may occur slowly on H 2 -CO 2 ," but the corresponding data were not published and only distributed to other authors via personal communication [25]. Therefore, the present study was conducted to provide the first concrete data on the autotrophic growth of M. thermophila ( Figure 2). Several measures were taken to assure that the CH 4 actually was produced by M. thermophila and derived from CO 2 . The possibility of CH 4 production from organic carbon in the inoculation material was eliminated by multiple transfers of small volumes into fresh medium. The carbon-containing medium components, erythromycin and cysteine, were also excluded as methanogenic substrates. Further, the identity and purity of the methanogen culture were confirmed via microscopy and DNA sequencing. Minor differences in the sequences are due to ambiguities in the sequencing.
During the incubation of M. thermophila in the absence of organic methanogenic substrates, CH 4 production as well as H 2 and CO 2 consumption largely corresponded to the stoichiometric model in which four molecules of H 2 and one molecule of CO 2 are used to produce one molecule of CH 4 ( Figure 4). Furthermore, the actual transfer of labeled carbon atoms from the CO 2 to the CH 4 pool via hydrogenotrophic methanogenesis could be shown. The fact that M. thermophila produced and excreted acetate, although it was grown under oligotrophic conditions and acetate being the preferred substrate compared with H 2 /CO 2 , was unexpected ( Figure 2). Similar observations were made, however, by Westermann et al. [26], demonstrating that Methanosarcina barkeri released acetate up to millimolar concentrations into the surrounding media, as did Methanosarcina mazei, although in smaller quantities. A possible explanation for these findings is that acetate is produced in the course of assimilation of CO 2 into cell carbon via intermediates including activated acetic acid or acetyl coenzyme A [27] and subsequently leaks the cell by passive diffusion [28]. The reuptake of lost acetate is limited by the minimum threshold for acetate utilization by Methanosarcina spp., which is known to be in the range of 0.2 to 1.2 mM [29]. This could explain the continuously increasing acetate concentration during the autotrophic methanogenesis by M. thermophila and may provide an indication that the organism is integrating carbon from CO 2 into the biomass. Apart from this, the present data further supports the evidence that M. thermophila is not only producing CH 4 from CO 2 and H 2 but is also generating biomass autotrophically. As the specific growth morphology of the Methanosarcinales prevented the direct quantification of the cell number, the production of biomass, although at a low level, was determined by quantifying the DNA content in the culture fluid ( Figure 2). Contrary to the findings of Zinder and Mah [7] for methanogenesis from acetate and methanol, CH 4 production from H 2 /CO 2 was rather linear than exponential and much slower than growth on acetate or methanol. However, linear methane production was also observed for Methanosarcina bakeri showing a CH 4 production rate of 0.23 mmol/day/L initial H 2 /CO 2 under similar incubation conditions, with the determined rates being twice as high compared with this study [23]. Low methane production rates from H 2 /CO 2 might have been attributed to the high molar volume of gases limiting substrate addition, the diffusion of gases into the nutrition medium, and the challenging adaptation to a new type of methanogenic substrate. Further, authors investigating hydrogenotrophic methanogenesis by Methanosarcina spp. found higher growth rates in complex media than in mineral media [22,30]. The role of M. thermophila as hydrogenotrophic methanogen in biogas production can only be estimated from the obtained data, as the applied H 2 partial pressure was much higher than in a bioreactor. Most acetogenic reactions require a H 2 partial pressure below 10 −4 bar to be thermodynamically favorable [31]. According to Lovley and Ferry [32], M. thermophila produced and consumed H 2 to maintain H 2 partial pressures between 0.67 and 1.6 mbar during growth on acetate or methanol, indicating that the threshold for hydrogen uptake is rather low. Furthermore, Maestrojuan and Boone found that Methanosarcina vacuolata produced only 30-40% of the expected methane in a mineral medium containing H 2 /CO 2 , probably due to decreasing substrate concentrations shifting thermodynamics [30].

Conclusions
Methanosarcina thermophila showed a biphasic CH 4 production growing mixotrophically on methanol and H 2 /CO 2 , switching from primarily methylotrophic methanogenesis to hydrogenotrophic methanogenesis as soon as methanol was depleted. Furthermore, it could be shown that M. thermophila is, contrary to the common opinion, able to perform hydrogenotrophic methanogenesis independently from other methanogenic substrates and to build up biomass autotrophically. Achieved CH 4 production rates were lower than those commonly found during methanogenesis from the preferred substrates acetate or methanol, but although carbon supply during incubations was restricted by the available volume of the headspace, M. thermophila successfully built up visible amounts of biomass. Further, the comprehensive physiological characterization of organisms is the foundation of functional genome analyses. Experimental data on the metabolic abilities of cultured methanogens are crucial to draw conclusions on the metabolic capabilities of uncultured archaea. We hope that the present study will help future investigations to refine this linkage.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper.