Aceticlastic strain Mx was enriched from a sample of the top 5–10 cm deep anaerobic (sulfidic) sediment overlaid by the bottom brines taken from the soda lake Tanatar-6 in Kulunda Steppe (Altai, Russia; N51°37′/E79°48′) in July 2011. The total salinity, carbonate alkalinity, and pH of the brines were 160 g l⁻¹, 1.7 M and 10.04, respectively (Sorokin et al., 2015). In the laboratory, the 20 ml of the sediment:brine slurry (1:5, v/v) was homogenized by shaking, and the coarse sandy sediment fraction was removed by a series of low-speed centrifugations. The remaining fine sediment particles were further collected by a high-speed centrifugation, and the pellet was resuspended in 2 ml anaerobic culture medium (see below) and used as an inoculum for the enrichment cultivation. A sodium carbonate/bicarbonate-based buffer with pH 9.5 containing 0.5 M of total Na⁺ as carbonates, 0.1 M NaCl, and 1 g l⁻¹ K₂HPO₄ was used as a mineral base medium for the enrichment. After sterilization, the medium was supplemented with 4 mM NH₄Cl, 1 mM MgCl₂, acidic trace metals (1 ml l⁻¹) and vitamins (1 ml l⁻¹) (Pfennig and Lippert, 1966), basic Se/W solution (1 ml l⁻¹) (Plugge, 2005), and yeast extract (0.01 g l⁻¹). Sodium acetate was used as a substrate for methanogenesis at concentrations of 20–100 mM (added from a 2 M sterile stock solution). Eighty ml medium was dispensed into 115 ml sterile serum bottles, supplemented with 0.5 mM Na₂S (from 1 M filter-sterilized solution), and oxygen was removed by three cycles of evacuation and flushing with sterile argon gas. Final reduction of the medium was achieved by adding a drop of 10% dithionite solution in 1 M NaHCO₃. The primary enrichments (three parallel incubations) were inoculated with 1 ml of the fine sediment slurry (see above) and incubated at 28°C statically with periodic measurements of methane concentrations in the gas phase. When the methane concentration reached 30–40% in the gas phase and visible turbidity appeared in the undisturbed liquid phase, the sediment-free cultures were transferred twice 1:100 under the same conditions, and the last successful transfer was used to inoculate a 1:10 dilution series under the same conditions in 15 ml Hungate tubes with 10 ml medium. The dilution series were repeated multiple times in combination with a preliminary low-speed centrifugation step, which allowed enrichment for filamentous Methanothrix-like cells. Numerous attempts have also been made to obtain colonial growth on plates with agar-shake techniques incubated in 2.5 l anaerobic jars (Oxoid) under argon gas phase and in the presence of an O₂-scavenging catalyzer, but they were not successful.

Strain M04Ac was isolated from a sample of mud collected from the active griffin of terrestrial mud volcano Gnilaya Gora, Taman Peninsula, Krasnodar region, Russia. Coordinates of the sampling point were 45.251°N, 37.436°E. Samples were collected in May 2017, from the upper 20 cm of mud, pH 8.5, temperature 21°C, 15.7 mM Cl⁻, 5.3 mM SO 42-. Samples were taken anoxically in tightly stoppered plastic bottles and transported to the laboratory. Enrichment and isolation were performed in the liquid medium of the following composition (per liter distilled water): KH₂PO₄, 0.33 g; NH₄Cl, 0.33 g; KCl, 0.33 g; MgCl₂·6H₂O, 0.33 g; CaCl₂·6H₂O, 0.033 g; NaCl, 10.00 g; NaHCO₃, 2.00 g; Na₂S·9H₂O, 0.5 g; 1 ml trace element solution (Slobodkin et al., 2012), 1 ml vitamin solution (Wolin et al., 1963), and 1 ml resazurin solution (0.001g l⁻¹). The medium was prepared by boiling and cooling it under N₂ (100%) flow. Afterwards, NaHCO₃, vitamins and Na₂S·9H₂O were added. The medium was dispensed in 10 ml aliquots into 17 ml Hungate tubes, and the head space was filled with 100% N₂ (high-purity grade). The tubes were autoclaved at 1 atm, 121°C for 20 min. The pH of sterile medium was adjusted to 8.0–8.5 at 25°C with 10% sterile anaerobic NaOH solution using anaerobic technique. Sodium acetate from sterile anoxic stock solution was added before inoculation as the growth substrate to a final concentration of 20 mM. Growth of the strains during isolation process was monitored by a phase-contrast microscope (Olympus CX-43 and Zeiss Axioplan Imaging 2 microscope).

Growth of both cultures during physiological experiments was estimated by OD₆₀₀ after vigorous homogenization. To determine cell morphology of the strain M04Ac, transmission electron microscopy was performed using JEM-1400 electron microscope (JEOL, Japan).

All the cultivation experiments were done in triplicate. Temperature (from 10 to 60°C) and pH growth ranges (from 5.0 to 11.0) of the strain M04Ac were determined using the same medium as for the strain purification. For pH experiments, the following Good's buffers (Sigma-Aldrich) were used (30 g l⁻¹): MES (pH 6 and 6.5), HEPES (pH 7 and 7.5), Tricine (pH 8.0 and 8.5), CAPSO (pH 9.0 and 9.5), and CAPS (pH 10–11). The NaCl requirement for growth was determined at optimal pH in the medium of the same mineral composition but lacking NaCl. Varying amounts of NaCl (0–10.0%, w/v) were added directly into 50-ml serum bottles before sterilization.

The salinity profile of the strain Mx was studied by using sodium carbonate/bicarbonate buffer with pH 9.5 and a total Na⁺ concentration ranging from 0 to 2 M. The pH range was determined by using HEPES/NaHCO₃ buffer for the pH range from 6 to 8 and a NaHCO₃/Na₂CO₃ buffer for the range from pH 8.5–11. Both profiles were obtained in growing cultures and in washed (resting) cells, harvested from 0.5 l culture grown at optimal pH–Na⁺ conditions at the end of exponential growth phase. It has to be stressed that the final pH values deviated substantially from the initial values, especially at both extremes and, therefore, the final pHs were measured and are indicated in the results. All growth and cell incubation experiments were done in triplicate.

Methane was analyzed by GC using Chromateck Crystall 5000 (Ufa, Russia); column Hayesep 80–100 mesh, 2 m x 3 mm at 40°C; FID detector at 200°C; and argon as the carrier gas at 25 ml min⁻¹. Concentrations of acetate were determined in filtered supernatants after neutralization with 1 M HCl to pH 7. The HPLC conditions were as follows: Animex HPX-87H column at 60°C, eluent 5 mM H₂SO₄ at 0.6 ml min⁻¹, and UV and RI detectors.

Ectoine was extracted from freeze-dried cells of strain Mx with 80% ethanol enforced by bead beating, followed by centrifugation and filtration through a 0.22 μm filter. The concentration of ectoine was measured with a Shimadzu Prominence-i LC-2030C plus HPLC (Shimadzu, Japan) equipped with a UV detector at 210 nm and a Polaris, NH2, 180 Å, 3 μm, 4.6 × 150 mm column (Agilent Technologies, USA) with a mobile phase of acetonitrile/H₂O 75/25 (%) at a flow rate of 0.6 ml min⁻¹. Oven temperature was set at 30°C. Ectoine and hydroxyectoine standards (Sigma Aldrich, Germany) were used for quantification.

DNA from mud samples as well as from enrichment cultures was isolated using FastDNA Spin Kit for Soil according to the manufacturer's protocol (MP Biomedicals, Santa Ana, California, USA). Whole-genome sequencing and metagenomic bioinformatics analysis were used to assess the purity of cultures at the final stages of their purification according to methods described in Slobodkin et al. (2023). 16S rRNA gene sequencing was performed as described previously (Slobodkina et al., 2022). The GenBank accession numbers for 16S rRNA gene sequences of strains Mx and M04Ac were KP205578 and OQ918309, respectively. The 16S rRNA gene sequences of the strains were compared with other sequences in GenBank (Benson et al., 1999) by using the BLAST program (Altschul et al., 1990) and by means of the EzBioCloud server (Yoon et al., 2017) (http://www.ezbiocloud.net) to identify their closest relatives. Sequences were aligned by MAFFT v7.427 (G-INS-i strategy) (Nakamura et al., 2018) for 16S rRNA gene-based phylogenetic analyses. A shotgun WGS library preparation and sequencing for both strains were done in BioSpark Ltd., Moscow, Russia using KAPA HyperPlus Library Preparation Kit (KAPA Biosystems, UK) according to the manufacturer's protocol and NovaSeq 6000 system (Illumina, San Diego, CA, USA) with the reagent kit, which can read 100 nucleotides from each end. Raw reads were processed with Trimmomatic (Bolger et al., 2014) for adapter removal and quality filtering. Assembly of reads was performed by Unicycler v0.4.8 (Wick et al., 2017). These Whole Genome Shotgun projects have been deposited at DDBJ/ENA/GenBank under the accession JARFPL000000000 for M04Ac and JARFPK000000000 for Mx. For the satellite bacteria, Whole Genome Shotgun projects have been deposited at DDBJ/ENA/GenBank under the BioProject ID PRJNA1010173. Metagenomic bioinformatics approaches were used to separate genomes in binary cultures as described in Slobodkin et al. (2023). Metawrap v1.3 (Uritskiy et al., 2018) quant bins workflow was used to determine the relative abundance of the microorganisms in binary cultures by analyzing coverage using Salmon (Patro et al., 2017). Gene search and annotation were performed by the Prokaryotic Genome Annotation Pipeline (PGAP) (Tatusova et al., 2016).

For phylogenetic reconstructions, 122 conserved single-copy archaeal protein-coding marker genes (Parks et al., 2020) as well as 16S rRNA gene were used. The trees were built using the IQ-TREE 2 program (Minh et al., 2020) with fast model selection via ModelFinder (Kalyaanamoorthy et al., 2017) and ultrafast approximation for phylogenetic bootstrap (Hoang et al., 2018), as well as approximate likelihood-ratio test for branches (Anisimova and Gascuel, 2006). The average amino acid identity (AAI) between the selected genomes was calculated using the EzAAI v1.1 (Kim et al., 2021). The average nucleotide identity (ANI) between the selected genomes was calculated using the pyani v0.2.12 (Pritchard et al., 2016). in silico DDH were calculated using the Genome-to-Genome distance calculator (GGDC) (Meier-Kolthoff et al., 2013). The relative evolutionary divergence (RED) values were calculated using GTDB-Tk v2 (Chaumeil et al., 2022).

Only a single culture out of multiple attempts whereby acetate served as substrate for methanogenesis in soda lake sediments resulted in an enrichment of an aceticlastic methanogen, while in all other cases acetate was consumed by syntrophic associations of the acetate-fermenting Ca. Syntrophonatronum acetioxidans partnered with hydrogenotrophic Methanocalculus alkaliphilus (at moderate salinities) or Mc. natronophilus (at salinities above 2 M Na⁺) (Sorokin et al., 2014, 2015, 2016). In contrast to the syntrophic pathway dependent on the bacterial partner, in a single positive culture (at 0.6 M total Na⁺ and pH 9.5) enriched from the soda lake Tanatar-6 in Kulunda Steppe, acetate-dependent methanogenesis was not inhibited by the addition of streptomycin and kanamycin, at 0.1 g l⁻¹ each. In fact, antibiotics helped to stabilize and increase activity of the aceticlastic organism, whose morphology was very similar to that of Methanothrix with a domination of multicellular filaments. Several rounds of dilution to extinction resulted in a highly enriched culture, but it still contained a low number of small bacterial rod-shaped cells apparently resistant to the used antibiotics. Using whole-genome sequencing and metagenomic bioinformatics analysis, we assessed the phylogenetic affiliation and relative abundance of these bacteria. The culture was contaminated with only one type of bacteria, which belonged to the Bacteroidales and was represented by only 1% of the community microorganisms. An analysis of the phylogenetic position of this bacterium based on 120 conserved single-copy bacterial protein-coding marker genes is shown in Supplementary Figure 1A. According to GTDB, it belongs to family-level lineage UBA7960 and to the genus-level lineage SKTA01 that consists only of MAGs sequenced from Kulunda hypersaline soda lake sediments. There were no other contaminating microorganisms in the culture. Preincubation with rifampicin (0.1 g l⁻¹) for 2–3 days with further dilution series in the absence of antibiotics eliminated the bacterial contamination, but the resulting pure culture was not transferrable and, hence, could not be deposited in a culture collection. The strain was designated as Mx.

An enrichment culture was initiated by 10% (w/v) inoculation of a mud sample into anaerobic sterile medium with 3,4-dimethoxybenzoic acid (10 mM), which after 3 months of cultivation was replaced by acetate (20 mM) as a substrate and ampicillin (0.5 g l⁻¹) for suppression of bacterial growth. After a month of incubation on acetate at 30°C in the dark, the growth of long, non-motile filaments and production of methane was observed. After three subsequent 10% (v/v) transfers, serial 10-fold dilutions in the same liquid medium with the addition of yeast extract (0.05 g l⁻¹) significantly enhanced the growth of the culture. The dilution series were repeated several times in combination with a preliminary low-speed centrifugation step (at 2 500 rpm for 7 min) which enriched pellet in filamentous Methanothrix-like cells. Incubation with rifampicin (0.1 g l⁻¹) for 2–3 days with further dilution series was ineffective as significant growth inhibition of the target organism occurred. Attempts to obtain separate colonies were also unsuccessful either with 1% Gelrite gellan gum or with 1.5% agar as the solidifying agent. When isolating Methanothrix-like cells, the satellite bacterium belonging to the Acholeplasmataceae family (Supplementary Figure 1B) was isolated, whole-genome sequenced, and tested for antibiotic sensitivity. Further use of a mixture of lincomycin and streptomycin (0.05 g l⁻¹ each) resulted in a highly enriched culture. Still, by using whole-genome sequencing and metagenomic bioinformatics analysis, we detected the same contamination represented by Acholeplasmataceae bacterium isolated in the previous step; its relative abundance was 2.5%. There were no other contaminating microorganisms in the culture. The resultant highly enriched culture was used for subsequent phenotypic characterization, since further attempts to separate Methanothrix-like cells from the bacterial component led to a sharp decrease in the growth rate and, ultimately, to the death of the culture, similar to the purified Mx culture.

The 16S rRNA gene sequence identity between two novel strains was 98.03%; between M. harundinacea 6Ac and strains M04Ac and Mx, it was 98.30 and 98.64%, respectively. The deduced amino acid sequence identities of mcrA, the gene encoding the α-subunit of methyl-coenzyme M reductase, between M. harundinacea 6Ac and two novel strains were 88.73% for the strain Mx and 88.37% for the strain M04Ac, while between two strains it corresponded to 95.53%. Classification of M. harundinacea, M04Ac and Mx strains into three separate species was strongly supported by ANI values (87.30% between two new strains; 85.65% between M. harundinacea 6Ac and strain Mx; 85.70% between M. harundinacea 6Ac and strain M04Ac). In silico DDH estimation between two new strains was 29.4%, between M. harundinacea 6Ac and strain Mx−25.0%, and between M. harundinacea 6Ac and strain M04Ac−25.5%. AAI value between two new strains was 83.72%, between M. harundinacea 6Ac and strain Mx−79.21%, and between M. harundinacea 6Ac and strain M04Ac−80.18%. Thus, considering all these parameters, the three studied strains are different species of the same genus. The similarity values of strains M04Ac and Mx to the three validly described representatives of the family Methanotrichaceae, based on their 16S rRNA and mcrA gene sequences, as well as ANI and AAI analysis, are summarized in Supplementary Table 1.

According to the results of our phylogenomic analysis based on 122 conserved single-copy archaeal protein-coding marker genes (Figure 1A), strains M04Ac and Mx clustered together with Methanothrix harundinacea and a number of other metagenome-assembled genomes (MAGs) to form a separate lineage, which was significantly separated from the clusters formed by Methanothrix soehngenii and Methanothrix thermoacetophila. This separate genus-level group is proposed here to form a new genus, Methanocrinis (see below). Our data correlate well with the Genome Taxonomy Data Base (GTDB, Parks et al., 2022), where each validly described Methanothrix species forms its own genus-level lineage. In fact, the relative evolutionary divergence (RED) values for these three genera as they are shown in the Figure 1A are 0.872, 0.840, and 0.897 for Methanothrix (A), Methanothrix (B), and Methanocrinis, respectively, whereas, the median value of RED for genus delineation in the GTDB 214 is 0.909. Based on the phylogenomic tree, it is also possible to analyze the origin of the MAGs belonging to these three lineages. While the ecotopes inhabited by all three genera are quite diverse, only Methanocrinis group contains MAGs from marine ecotopes, which supports our assumption that the genus Methanocrinis contains salt-tolerant species.

The identity values between the 16S rRNA gene sequences of the two new strains and M. harundinacea on the one hand and M. soehngenii and M. thermoacetophila on the other hand strongly confirm that these are different genera of microorganisms. These values fall in the range of 91.54–92.55%. Phylogenetic analysis based on the 16S rRNA gene also confirms the conclusion that each of the currently validly described Methanothrix species represents a separate branch at a relatively deep level and that the branch represented by M. harundinacea is most distantly related (Figure 1B). However, the phylogenetic status of the “Methanosaeta pelagica,” which was isolated from a marine habitat (Mori et al., 2012), remains unresolved. This microorganism undoubtedly gravitates toward the proposed genus Methanocrinis, but its genome sequence is necessary to resolve its affiliation with this genus.

By outlining the membership of the genus Methanocrinis based on the 16S rRNA gene (Figure 1B), we were able to investigate its environmental distribution based on the data available in public databases. Out of 39,243 sequences that belong to Methanosaetaceae in the SILVA SSU 138.1 database, we selected 5,276 sequences that belonged to the genus Methanocrinis using the BLASTn program. To avoid counting replicates, sequences with a unique combination of “isolation source” and “work title” entries were selected. Thus, a representative selection of 228 16S rRNA gene sequences was obtained (Supplementary Table 2). Among them, sequences found in organic-rich, anthropogenic ecotopes were most common. These include ecotopes representing various kinds of bioreactors (33.5%) and waste disposal systems (9.7%). The second most frequent ecotopes were associated with hydrocarbons, e.g., oil fields and natural gas fields (25.7%). Freshwater ecotopes accounted for 8.7%, coastal ecotopes, such as estuary and mangroves, accounted for 7.3%, and seawater ecotopes accounted for 4.4%. Sequences from various subterranean ecotopes, such as mines and groundwaters, accounted for 5.8%, whereas sequences from mud volcanos accounted for 3.4%. All other types of ecotopes were encountered in single cases.

The individual cells of both strains were straight rods with flat ends with cell sizes 1.7–6.5 x 0.9–1.5 and 1.9–4.8 x 0.6–1.0 μm for strain M04Ac and Mx, respectively. They also formed flexible chains of cells within sheaths of variable lengths (Figure 2). The satellite of strain Mx belonging to the Bacteroidales group was represented by small sporadic rods (Figure 2A). The satellite culture from Acholeplasmataceae family, associated with the strain M04Ac, had a size of about 0.4 μm and was represented by small cocci that were poorly distinguishable in a phase-contrast microscope (Figure 2B). The cells of novel strains were non-motile. However, on electron micrographs of M04Ac cells, long, thin surface appendages of an unknown nature were noticed (Figure 2C). Such fimbria-like structures have never been reported before in any members of the Methanotrichaceae family.

Both strains appeared to be typical obligately aceticlastic methanogens similar to other known Methanothrix species. From the tested classical methanogenic substrates including H₂, formate, CO, MeOH, and acetate, only the latter supported growth and methane formation. The resting cells grown with acetate also formed methane only from acetate. For the strain Mx, addition of yeast extract at low concentrations (0.01–0.02 g l⁻¹) slightly decreased the usually long lag phase after fresh transfer, but there was no obligate dependence on it. In contrast, low concentrations of yeast extract (0.05 g l⁻¹) highly stimulated growth of strain M04Ac, but the percentage of the bacterial component also increased significantly. The addition of vitamins did not have significant influence on growth of either strains.

Growth rate of the aceticlastic strains was extremely slow, typical for this group of methanogens (explained usually by the lowest energy yield of the aceticlastic methanogenesis), although strain Mx was obviously even more slowed down by the extreme conditions. Since the growth was not exponential, only a rough estimation of the doubling time was possible, and it was around 25–30 d. The molar conversion of acetate to methane was close to the theoretical 1:1 (Supplementary Figure 2A). For strain M04Ac, the specific growth rate in acetate-containing basal medium was 0,011 h⁻¹ (doubling time 63 h) (Supplementary Figure 2B). However, it should be taken into account that the enrichment culture grew much faster than the isolate. Upon further purification, growth of strain M04Ac eventually slowed down to 4–5 months, and culture finally became not-transferrable.

The temperature range for growth of strain M04Ac was 20–45°C, with an optimum at 37°C. No growth was detected at 50°C or above, as well as at 18°C or below after incubation for 4 months. The pH range for growth was 7.5–10.0, with an optimum at pH 9.0. No growth was observed at pH values below 7.0 or above 10.0 (Figure 3A). Thus, strain M04Ac isolated from terrestrial mud volcano was alkalitolerant, similar to the freshwater Methanotrichaceae members, and it did not require high sodium concentrations for growth above trace amounts supplied with basic medium. Growth was observed at NaCl concentrations of 0–5.0% (w/v) with an optimum at 0.5–1.0% NaCl (w/v).

According to the pH-salt profiling, strain Mx represents a moderately salt-tolerant obligate alkaliphile, growing optimally at 0.2–0.3 M total Na⁺ (as sodium carbonates) and within the pH range from 7.8 to 10.2 with an optimum between 9.3 and 9.5. Also typical for the soda lake isolates, the alkalitolerance of methanogenic activity of the Mx resting cells was substantially extended in comparison to the growth optimum in culture (Figures 3B, 4). This means that the organism can still remain metabolically active during periods of elevated alkalinity even without being able to grow.

The draft genome of strain M04Ac was assembled into 96 contigs with genome size of 2,444,195 bp and the DNA G + C content of 58.31%. The genome was predicted to contain 2,423 protein-coding sequences and 53 RNA genes, including three complete rRNAs (1 5S, 1 16S, and 1 23S) and 48 tRNAs. The nomenclatural type is GCA_029167205.1^Ts according to SeqCode requirements (Hedlund et al., 2022; Whitman et al., 2022). The draft genome of strain Mx was assembled into 100 contigs with genome size of 2,412,901 bp and the DNA G + C content of 58.18%. The genome contained 2,406 protein-coding sequences and 48 RNAs genes, including three complete rRNAs (1 5S, 1 16S, and 1 23S) and 43 tRNAs. Detailed genomes statistics can be found in Supplementary Table 3. The nomenclatural type is GCA_029167045.1^Ts according to SeqCode requirements (Hedlund et al., 2022; Whitman et al., 2022). In general, the genomes of M04Ac and Mx strains are similar in size and gene content. We have analyzed the genomic data in relation to methanogenic metabolism and halo-alkaline adaptations of the novel strains.