International Journal of Systematic Bacteriology (1998), 48, 357-367 Printed in Great Britain

Methanobacterium subterraneurn sp. nov., a new alkaliphilic, eurythermic and halotolerant methanogen isolated from deep granitic groundwater

Svetlana Kotelnik~va,'.~Albert0 J. L. Macario' and Karsten Pedersen3

Author for correspondence : Svetlana Kotelnikova. Tel : + 46 3 1 773 25 65. Fax : + 46 3 1 773 25 99. e-mail : svetlana. kotelnikova @ gmm.gu. se

1 Institute of Biochemistry Deep subterranean granitic aquifers have not been explored regarding and Physiology of methanogens until now. Three autotrophic -producing were , Russian Academy of Science, isolated from deep granitic groundwater at depths of 68,409 and 420 m. These Pushchino-na-Oke, were non-motile, small, thin rods, 01-0-15 pm in diameter, and they Moscow region, RU- could use and or formate as substrates for growth 142292, Russia and . One of the isolates, denoted A8p, was studied in detail. 2 Wadsworth Center, It grew with a doubling time of 2.5 h under optimal conditions (2040 "C, Division of Molecular Medicine, New York State pH 7.8-8.8 and 0.2-1.2 M NaCI). Strain A8p is eurythermic as it grew between 3-6 Department of Health, and 45 "C. It was resistant to up to 20 mg bacitracin I-'. The G+C content was Empire State Plaza, and 54.5 mol0/o, as determined by thermal denaturation. Phylogenetic studies Department of Biomedical Sciences, School of Public, based upon 16s rRNA gene sequence comparisons placed the isolate A8p in the Health, The University at genus Phenotypic and phylogenetic characters indicate Albany, Albany, NY 12201- that the alkaliphilic, halotolerant strain A8p represents a new species. We 0509, USA propose the name Methanobacterium subterraneum for this species, and 3 GUteborg University, strain A8p (= DSM 110749 is the type strain. Institute of Cell and Molecular Biology, Department of General and Marine Microbiology, Keywords : Methanobacterium subterraneum sp. nov., , eurythermy, Box 462, Guteborg, halotolerance, methanogen 5-40530, Sweden

INTRODUCTION (lo), peat bog (49, marshy soil (21), bleach kraft mill sludge (3 l), and an oilfield (5). Methanobacterium Mesophilic, rod-shaped and hydrogen-consuming alcaliphilum, isolated from alkaline lake sediments, has methanogenic Archaea are placed in-the Methano- a pH-optimum between 8-1 and 9-1 (8). Four species, bacterium and genera. Typically, Methanobacterium formicicum, Methanobacterium species belonging to the genus Methanobrevibacter are ivanovii, Methanobacterium palustre and Methano- short rods with G + C contents lower than 32 mol YO, bacterium bryantii are autotrophic. No members of while Methanobacterium species are long rods with Methanobacterium have been described before that G + C contents higher than 32 mol %. At present the can grow below 10 "C. genus Methanobacterium is composed of seven meso- philic and six thermophilic species. On the basis of In this paper we describe the phenotypic and phylo- phylogenetic comparative analysis of 16s rRNA, it genetic characteristics of an autotrophic, halotolerant, was suggested to tranfer Methanobacterium thermo- eurythermic and alkaliphilic methanogen that is able autotrophicum, Methanobacterium thermophilum and to grow below 10°C. On the basis of these charac- Methanobacterium wolfeii to the genus Methano- teristics, we propose a new species, Methanobacterium thermobacter (7). The mesophilic species of the genus sub terraneum. Methanobacterium were isolated from sewage sludge The described here was isolated from deep granitic groundwater and is the first example of a ...... methanogen isolated from the deep subterranean Abbreviation :AODC, acridine orange direct count. biosphere (32). In this habitat, methanogens may The EMBUGenBank accession numbers for the nucleotide sequences of represent chemoautolithotrophic organisms that strains A8pT, 3067 and C2BlS are X99044, Y 12592 and X99045, respectively. initiate food chains in the oligotrophic deep sub-

~~ 00599 0 1998 IUMS 3 57 S. Kotelnikova, A. J. L. Macario and K. Pedersen

surface environment at the expense of geologically NaC1, 0.5 g K,HPO,, 0.003 g FeC1,. 7H,O, 10 ml trace produced hydrogen. element solution (44), 0.001 g rezasurin, 2.0 g NaHCO,, 0.25 g cysteine. HC1 and 0.25 g Na, .9H,O. The following carbon and energy sources were used: 2.0 g formate l-l, or METHODS hydrogen and carbon dioxide (80 :20,152 kPa). The medium did not contain buffer before sterilization, and the pH value Sources of organisms. The Aspo hard rock laboratory (HRL) of the medium, which became 7-0 after sterilization, was tunnel is located on the Baltic coast under the island of adjusted to 7.8-8.0 before inoculation with anoxic bicar- Aspo, in the vicinity of the Simpevarp nuclear power bonate (10% stock solution) and NaOH (0.1 M stock north of Oskarshamn, South-East Sweden. The host rock is solution). All experimentswere repeated at least once and all a - 1800 Ma old granodiorite belonging to the Fenno- tests were done at least in duplicate. Means of repeated tests scandian shield. The tunnel has a total length of 3600 m, is are reported and used for the graphs presented. approximately 5 x 5 m (height x width) and proceeds down with an inclination of about 14 ". It starts at the coastline Microscopy. Phase-contrast and fluorescence microscopy and continues about 1700 m under the sea floor where it were performed with an Olympus BH-2 phase-contrast !pirals down to 460 m below sea level under the island of microscope and a Carl Zeiss Axioscope equipped with UV Asp0 (34, 35). Microbiological data from boreholes in the lamps. Phase-contrasted images were obtained with an tunnel to a depth of 192 m and a length of 1420 m and in Olympus C35AD camera and an AD Exposure Control surrounding surface boreholes have been published pre- Unit. Autofluorescence of whole cells was observed with an viously (35-37). Groundwater was sampled from core- LP 420 excitation filter (13). Acridine orange direct counts drilled surface and tunnel boreholes at 10-440m below (AODCs) were done as described previously (14). Ultrathin ground. The methanogens described here were called A8p, sections of the cells were obtained by glutaraldehyde fixation 3067 and C2BIS and came from boreholes denoted (2.5 %, 2 h) followed by osmium tetroxide fixation (1 YO,2 h) KROO12A, KA3067A and HD0025A, respectively. These in 0.1 M phosphate buffer (pH 7.6) at 4 "C. The cells were boreholes correspond to tunnel lengths of 500, 3067 and embedded in Epon 8 12, then thin-sectioned and stained with 3200 m, and to depths below sea level of 68,409 and 420 m, uranyl and lead citrate. The thin sections were respectively. The samples were inoculated in the enrichment studied with an EAL 1200EX electron microscope. medium described below. Gas chromatography. Methane was determined with a Media and culturing techniques. The anaerobic technique Varian GC-3700 gas chromatograph equipped with a described by Hungate (17) was used. A total of 19 boreholes 2 m x 3.175 mm steel column packed with Porapak Q mesh were screened during 1995-1996 for the presence of 80/100 (Varian), with nitrogen as carrier gas at a flow rate of methanogens with various carbon and energy sources. 30 ml min-l, and a flame-ionization detector. The response Enrichment cultures were obtained in a medium prepared of the detector to methane was linear. The injector, column with filter-sterilized groundwater which was collected from and detector temperatures were isothermal at 100, 100 and boreholes KR0012A or KA3067A. The groundwater was 200 "C, respectively. Calibration, registration and integra- supplemented with (1-l) 10 ml trace element solution (44), tion of methane peaks were done with a Star Chroma- 1-0g yeast extract and 1 mg resazurine. The following tography Workstation, version 4.5 (Varian). carbon and energy sources were used (1-l): 3.4 g sodium Growth determination. The studied isolate A8p grew in acetate, 2.0 g formate, hydrogen and carbon dioxide aggregates and it was, therefore, not possible to use (80: 20 %, 152 kPa), 2.0 g methanol or 1.0 g trimethylamine. measurements of optical density in the cultures for biomass Five millilitre portions of this medium were distributed determinations. Instead, the linear relation between methane under -free nitrogen gas in Hungate-type gas-tight, production and biomass formation during the exponential anaerobic culture tubes (Bellco Glass 2047, 17 ml) and growth phase (38) was used to calculate specific growth rates sterilized at 121 "C for 20 min. After cooling, the following for the experiments described below. This relation was sterile, anoxic solutions were added (1-l): 5 ml vitamin confirmed in experiments where methane production during solution SL-6 (44), 1 mg , 2.0 g NaHCO,, 0-25 g growth of the isolates was found to be linearly correlated cysteine.HC1 and 0.25 g Na2S.9H20. The pH of the with the increase in cell number as determined by AODC. enrichment medium was adjusted twice, before sterilization Growth at low methane production rates (< 0.05 h-l) was and after the final additions, with 0.1 M NaOH or 0.1 M confirmed by AODC. Inoculations which did not show a cell HC1 to pH values corresponding to those of the ground- count increase and a corresponding methane production waters used for inoculation (7.25-7.5). The enrichment tubes after 20 d incubation were regarded as negative. were inoculated with 0.5 ml groundwater from the boreholes within 2 h of sampling and incubated at room temperature Susceptibility tests. The susceptibility of A8p to antibiotics for up to 5 months. The final headspace in the enrichment was determined by adding them at 1-2000mg1-l. The tubes was approximately 11 ml. The enrichment cultures minimum concentration of the antibiotic causing 15 YO which actively produced methane were subcultured by serial reduction in growth rate was considered to be inhibitory. dilutions in the presence of 0.5 g vancomycin 1-l. Pure Determination of growth requirements. Cultures of A8p cultures were obtained by mixing 1 ml of the last dilution of were transferred three times in a modified ASPM (PH 7-23), the culture which produced methane with 5 ml fresh enrich- without vitamins and organic compounds, but with 40 mM ment medium with 20 g agar 1-1 (45 "C) plus 0.5 g vanco- formate. Subsequently, an array of possible growth factors mycin l-l, and then solidified in a thin layer by rolling was added to the modified ASPM and inoculated in in butyl-rubber-stoppered, aluminium crimp-sealed tubes duplicate with the vitamin-depletedA8p culture. The control (Bellco Glass 2048, 22 ml). For the cultivation of pure did not contain any additions. Growth of the cultures was cultures, including all physiological experiments,an artificial monitored during incubation at 35 "C for 120 h. The Asp0 medium (ASPM) was used that mimicked the chemical experiments were repeated twice. Growth curves were composition of the groundwater from the KROO12B bore- obtained for each repetition of the studied compounds and hole. It contained (1-'): 0.4 g NH,Cl, 0.03 g MgCl,, 0.45 g growth rates were calculated. The differences between the

358 International Journal of Systematic Bacteriology 48 Methanobacterium subterraneum sp. nov. mean growth rate for an added compound compared to the CT-3’; (368-353) r 5’-AGK TTT CGC GCC TGC T-3’; control were evaluated with the Student’s t-distribution test (1225-1240) f 5’-ACA CGC GGG CTA CAA T-3’ (designed (1) at a significance level of P = 0.95. in our laboratory); and (339-356) f 5’-CTC CTA CGG Determination of growth parameters. Methane production GAG GCA GCA-3’ (2); (517-536) f 5’-GCC AGC MGC was monitored at 1-3 d intervals in ASPM. The mean CGC GGT AAT WC-3’; (536-519) r 5’-GWA TTA CCG specific growth rate for each incubation temperature was CGG CKG CTG-3’; (907-926) f 5’-AAA CTY AAA KGA calculated from the exponential methane-production phase. ATT GAC GG-3’; (926-909) r 5’-CCG CCA ATT CCT The effect of pH on growth of A8p was determined using TTA AGT-3’ (23) ; (1099-1 114) f 5’-GCA ACG AGC GCA different pH buffers in ASPM. They were: a mixture of ACC C-3’; (1 114-1 100) r 5’-GGG TTG CGC TCG TTG-3’ 26 mM sodium acetate and 18 mM acetic acid (pH 4.8-55), (22). Nucleotide residues other than A, T, G and C are 20mM MES (pH 5.7-6.5), 20 mM HEPES (pH 5-5-7.2), indicated by the following ambiguity codes in mixtures of 20 mM PIPES (pH 6.2-7.6), 20 mM sodium bicarbonate alternative primers: M (A or C); K (G or T); W (A or T); (PH 7.0-8.0), 30 mM Tricine (pH 7-5-8-0) and 20 mM Tris Y (T or C). Primer positions correspond to positions in the (PH 8-2-84). The culture was inoculated into five repetitions E. coli Brosius 16s rRNA gene sequence (9). The master of each buffer. The pH was measured before and after mixture contained 0.5 pl of the template (PCR product), inoculation, as well as during growth and never varied more 10 pmol fluorescent 5’-labelled primer and water to a final than 0.15 units. The mean specific growth rate for each volume of 25 p1. The mixture was distributed into four incubation pH was calculated from the exponential methane reaction tubes (6 1.11 in each tube) and 2 pl of one of the A, C, production phase. G or T reagent mixtures from the sequencing kit was added to each tube. One drop of mineral oil was added on the top The salt tolerance of A8p was determined in ASPM at 35 “C of each tube to minimize evaporation. A total of 30 cycles and pH 8.5, supplemented with NaC1. The experiment was were performed at 94°C (45 s), 55 “C (45 s) and 72°C performed in duplicate and repeated twice. The specific (1 min). The sequencing product was separated from the growth rate at each NaCl concentration was calculated from residual oil by tip dropping. Four microlitres of loading the methane production between 20 and 177 h incubation. buffer (formamide, EDTA and methyl violet) was added to Antigenic fingerprinting. Partial antigenic fingerprinting of each tube, and about 7 pI was loaded into separate wells in a the new isolates was performed using calibrated antibody gel. Gel electrophoresis was performed on an ALF DNA probes as described previously (24, 25). Sequencer (Pharmacia Biotech). Ready-mixed Long Ranger Determination of DNA composition. Standard procedures Gel solution for DNA sequencing (FMS Bioproduct- (26,43) were used for DNA extraction from the culture with Europe) and urea (ALF grade) were used in the gel solution. the following modifications. Cells were lysed by repeated Sequenced fragments were obtained with 11 different freezing and thawing followed by 2 h incubation in 50 mM primers and were aligned with the Genetics Computer Group ammonium bicarbonate, 50 mM EDTA buffer (PH 8.0) (GCG) GELASSEMBLE procedure and verified manually. 10 mg SDS ml-’ and 20 mg DTT ml-l at 60 “C. Extracted Sequence analysis. The 16s rRNA gene sequences of the DNA was dissolved in 0.1 x SSC buffer and dialysed against isolates A8p, 3067 and C2BIS were compared to sequences the same buffer. The G+C content was determined by a available in the EMBL database using the FASTA and BESTFIT thermal denaturation method (27) with a Cary Varian procedures in the GCG program package. The similarity Thermal Spectrophotometer, using the DNA of Escherichia percentages between the 16s rRNA gene sequences of the coli strain L (5 1 mol YO)as the reference. The G + C content isolates and the most closely related organisms in the was calculated using the formula G+C mol% = 2-08 x database were calculated with BESTFIT, without considering Tm- 106.4 (30). Relative standard deviation of the G+ C uncertain and unknown positions. data was 0.74 YO. A phylogenetic analysis was performed on A8p and C2BIS 165 rRNA gene amplification. DNA was extracted from the isolates A8p, C2BIS and 3067 as described above, but in the strains using the programs contained in PHYLIP version 3.5~ last step DNA was dissolved in TE buffer (10 mM Tris, package (15). Nucleotide positions that could be un- 1 mM EDTA, pH 8.0). For 16s rRNA gene amplification, ambiguously aligned for all 16s rRNA genes compared were new Archaea-specificprimers were designed which matched included in the analysis. The final data set comprised 1400 the positions (7-26) (E. coli Brosius numbering) 5’-GTC nucleotide positions, positions 38-1438 of 16 organisms. CGT TTG ATC CTG GCG GA-3’ and (1 520-1 538) 5’-AG The sequences used for the tree construction were aligned GTG ATC CAG CCG CAG A-3’. One microlitre of the using PILEUP (GCG). The distances were calculated using the extracted DNA solution was added into a mixture of 10 p1 DNADIST (18) and a tree was built running KITCH with contemporary tips. KITCH was run with a randomized input 10 x PCR buffer (Stratagene), 0.2 mM each NTP, 0.25 pM each primer, and double-distilled water in a final volume of order of data with 20 jumbles, and 11759 trees were 100 1.11. The samples were treated with 50 pg RNAase A ml-l examined during execution. The organisms used for the tree (Sigma) for 15 min at 37 “C and incubated at 95 “C for construction are listed in the legend to Fig. 3. 5 min, before addition of 1 pl Pfu DNA polymerase (Strata- gene) and 100 pl mineral oil (Sigma). A total of 34 cycles RESULTS were performed at 95 “C (30 s), 55 “C (1 min), 72 “C (2 min) followed by a final incubation at 72 “C for 10 min. Enrichment and isolation 165 rRNA gene sequencing. Repeated primer extension Groundwater from deep granitic rock aquifers (10- sequencing (RPE-sequencing) was done as described in the thermo sequenase fluorescent labelled primer cycle sequenc- 440 m below sea level) was used as the inoculant. These ing kit manual (Amersham) (3) with the following modifi- groundwaters were anoxic and oligotrophic (2.0- cations. Sequencing was performed using fluorescent dye 5’- 11.0 mg organic carbon l-l), with different salinities labelled primers: (7-26) f 5’-GTC CGT TTG ATC CTG (537-13 300 mg Cl- 1-l) and bicarbonate concentra- GCG GA-3’; (352-368) f 5’-CAG CAG GCG CGA AAM tions (53-326 mg HCO, 1-l) (29).

~ ~~ ~ International Journal ofSystematic Bacteriology 48 359 S. Kotelnikova, A. J. L. Macario and K. Pedersen

Fig. 7' Morphology of strain A8p. (ax) Electron microphotographs of thin sections of cells. (d) Phase-contrast micrograph of cells and cell aggregate. The bar indicates 50 nm in (a) and (b), 100 nm in (c) and 1 pm in (d). Abbreviations: MI cytoplasmic membrane; CW, ; CP, cytoplasm; IN, invagination of cytoplasmic membrane before cell division; 5, septum.

Active methanogenesis was observed in enrichment depth. The pH increased from 7-2-75 to 9-3-9.6 in media containing 2.0 g 1-' formate and 1.0 g tri- these cultures concurrent with growth and methane methylamine 1-' inoculated with groundwater from production in the medium with formate. All the the boreholes KROO 12A and HD0025A, correspond- enrichment cultures could grow and produced copious ing to sample dates 2 February 1995 and 25 November amounts of methane when inoculated into the medium 1995 and to 68 and 420 m below sea level, respectively. with formate or with hydrogen and carbon dioxide and Abundant methane production was also observed in they grew best at alkaline pH values (7-8-8.8). Small, an enrichment culture supplied with hydrogen and non-motile and autofluorescing rod-shaped cells were carbon dioxide (80 :20, 152 kPa) inoculated 25 observed by light microscopy in all these enrichments. November 1995 from borehole KA3067A, 409 m The cultures were purified by serial dilution in the

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~ 360 International Journal of Systematic Bacteriology 48 Methanobacterium subterraneum sp. nov.

groundwater-based media containing 2.0 g formate I-' Table 7. Effect of different growth factors on the and 0.5 g vancomycin 1-' and subsequent inoculation growth of strain A8p in roll tubes. Colonies were observed after 2 weeks. Single colonies from tubes with methane production Growth factor Concn (mg 1-') Growth rate were selected and transferred into liquid ASPM with added (h-9 2.0 g formate 1-l. The resulting cultures were morpho- logically homogeneous. Only autofluorescing cells ASPM (control) - 0.1 78 were observed. Inoculation into ASPM without car- ZnC1, 2 0.082 bon and energy sources but instead containing 20 mM CoCl, 2 0.085 and 20 mM lactate or 2 g glucose 1-1 and 1 g Na,SeO, 2 0.062 peptone l-l, showed no growth. Thus, the cultures NiCl, 2 0.095 were axenic. The cells of the studied isolates had MnC1, 20 0.107 differing autofluorescence intensities and varied Yeast extract 2000 0.101 slightly in length. The isolates were designated as Casamino acids 1000 0.109 strains A8p (68 m depth), 3067 (409 m depth) and Isobutyric acid 5 0.121 C2BIS (420 m depth). n-Butyric acid 5 0.1 13 Propionic acid 5 -* Morphology and ultrastructure * Growth not observed. Surface colonies of strain A8p were circular, granular, colourless and 0.5-1.5 mm in diameter. Cells of isolate A8p were non-motile, slightly curved, small and short (80: 20; 152 kPa) with a growth rate of 0.2-0.3 h-l and rods, 0.6-1.2 pm in length and 0.1-0.1 5 pm in diameter, did not require any growth factors or vitamins for often growing in aggregates (Fig. la, d). These aggre- growth. Strain A8p grew in ASPM (specific growth gates floated with gas bubbles when cultivated in rate, 0.2-0.4 h-') with formate (2 g 1-l) as the substrate. ASPM with formate. Ultrathin sections showed a In the absence of exogenous bicarbonate or carbon single-layered electron-dense cell wall of about 6.5 nm dioxide, methane formation from hydrogen di- in thickness (Fig. lb), typical of Gram-positive bac- minished. The addition of methanol, trimethylamine teria. The cell wall was tightly fitted to the cytoplasmic or pyruvate did not stimulate methanogenesis from membrane. The cells divided after septum formation hydrogen. In the bicarbonate-buffered medium with- that was initiated peripherally (Fig. lb, c). out hydrogen, only small amounts of methane were Colonies of strain C2BIS were circular, granular, formed (< 05%). The presence of n-propanol yellow, 1.0-2-0 mm in diameter. The cells were straight (5 mM), isobutanol (5 mM), n-butanol (5 mM), or rods with a cell wall ultrastructure typical of Gram- ethanol (5 mM) in the bicarbonate-buffered medium positive and with round ends, i.e. a mor- did not increase methane production in comparison phology similar to A8p. Under 420 nm light, bright with the control. Thus, these alcohols were not blue-green autofluorescence was observed. Colonies hydrogen donors for methanogenesis. The addition of and cells of strain 3067 had a morphology similar to acetate (30 mM), methanol (20 mM), pyruvate that of C2BIS. Because the 16s rRNA sequence of (20 mM), trimethylamine (20 mM), fructose (10 mM), isolate 3067 was similar but not identical to those of glucose (10 mM), or dimethyl sulfide (4-5 mM) did not the isolates A8p and C2BIS (which were identical, see lead to an increase in methane production. Therefore, below), phenotypic studies were continued only with these compounds were also not substrates for methano- isolate A8p. genesis by strain A8p. Tryptone (1 g 1-'), sodium acetate (1 g l-'), para- Anti biotic susceptibility aminobenzoic acid (100 mg l-l), vitamin B,, (4 mg 1-'), biotin (0.4 mg P), coenzyme M (0.1 mg l-l), CaC1, Growth of strain A8p was not inhibited by (1-'): 1 g (0.5 g 1-'), MgC1, (03 g 1-'), Na,MoO, (2-20 mg 1-'), vancomycin, 500 mg benzylpenicillin, 1 g ampicillin, Na,WO, (2 mg l-'), FeC1, (2 mg P),CuCl, (2 mg 1-') 1 g methicillin, 2 g streptomycin, 800 mg nalidixic acid, and NiCl, (0.2 mg 1-') were tested as growth factors 14 mg neomycin or 20 mg bacitracin, while its growth for strain A8p. Statistical evaluation showed that in rate was reduced one order of magnitude by (1-'): ASPM with hydrogen and carbon dioxide as energy 40-100 mg chloramphenicol, 20 mg neomycin and and carbon sources and supplemented with the 40 mg bacitracin. different growth factors listed above, strain A8p showed growth rates close to those of the controls with- Determination of growth requirements out additions, i.e. these compounds did not influence methanogenesis by strain A8p. However, growth was Strain A8p was transferred four times in ASPM with reduced by yeast extract, Casamino acids, isobutyric hydrogen as the energy source and bicarbonate as the acid, n-butyric acid, Na,SeO,, ZnCl,, CoCl,, NiC1, and carbon source. The strain was and MnCl, (Table 1). Growth in medium with vitamin autotrophic; it could use hydrogen and carbon dioxide solution SL-6 (5 ml l-') (44) was slower than that in a

~~~~~~~ ~ ~ International Journal of Systematic Bacteriology 48 361 S. Kotelnikova, A. J. L. Macario and K. Pedersen

function of the incubation temperatures and varied from 5 to 144 h. Minimal lag phases and maximal A I I specific growth rates were observed at 20-40 "C. The strain showed growth and methanogenesis at 3.6- 13 "C. Growth at 3.6,6, 13-6and 20 "C was confirmed by AODCs. At 3.6 "C, the number of cells had increased by more than two orders of magnitude, indicating active growth. In a separate experiment we verified that growth (increase in cell numbers) and methane production rates were linearly correlated I...I...I...I...I,..I...I during exponential growth at 6°C (not shown). ' 0 10 20 30 40 50 60 Growth and methane production were not detected at Incubation temperature ("C) 2 or 50 "C.The organism is, therefore, mesophilic and eurythermic. (b) 0'50t The optimum pH for growth and methane production at 35 "C in ASPM was 7.8-8-8 (Fig. 2b). This optimum was confirmed by AODCs at the end of incubation. Growth and methane production were observed at pH 9.2 and 6.5, but not above pH 9-2or below 6.5. The duration of the adaptation phase was the same at pH 7.7-9.1 and increased at pH values below 7.7 and above 9.1. The reproducibility of cell growth at pH 9.0 0.00 I- was confirmed by subculturing strain A8p in ASPM at pH9.0 twice. The level of growth and methane production observed during subculturing was repro- ducible. Thus, strain A8p is alkaliphilic. Although the growth rate of A8p was highest at 0.2 M NaCl, this strain could grow up to a salinity of 1.40 M (Fig. 2c). In ASPM with added NaCl (0.25-1.40 M) 2 0.15 A8p cells formed huge aggregates (Fig. la, d). Strain L A8p is able to both grow and produce methane in a 0.10 wide range of salinities, consequently, it is halo- 2 tolerant. 0.05

Antigenic fingerprinting 0.25 0-50 0.75 1.00 1.25 1.50 NaCl (M) The partial antigenic fingerprints of the new isolates demonstrated that they were antigenically unrelated to reference methanogens from the genera Methano- Fig. 2. Influence of temperature (a) at pH 8.5, pH (b) at 35 "C and NaCl concentration (c) at 35 "C and pH 8.5 on the growth bacterium and Methanobrevibacter. of A8p, cultivated in ASPM. Specific growth rates were calculated from methane production values and are the means of duplicate cultures. G + C content of the DNA The G + C content of strain A8p was 54.5 0.5 mol%. The thermal melting point of DNA was 77.3 +O-2 "C. vitamin-free medium. This observation was confirmed These results are means of four independent measure- after four consecutive transfers into vitamin-free and ments. vitamin-containing ASPM with hydrogen and carbon dioxide (80 :20, 152 kPa) by 5 % (v/v) inoculations. Vitamins were not essential for growth. Relative Phylogenetic analysis standard deviations for the experiments with different The 16s rRNA genes of the isolates were amplified growth factors were 2-1 2 YO. using archaea-specific primers and then sequenced. The 16s rRNA genes of the isolates A8p and C2BIS Determination of growth parameters were sequenced at positions 38-1 534 and 22-1 5 12, respectively (E. coli Brosius numbering). The 16s Methanogenesis by strain A8p was observed at 3-6- rRNA gene sequences of isolates A8p and C2BIS were 45 "C (Fig. 2a), with a broad optimum of 20-40 "C. obtained without any unambiguous or unknown bases The length of the lag phase differed greatly as a and were identified for 1474 positions. Isolate 3067

362 International Journal of Systematic Bacteriology 48 Methanobacterium subterraneum sp. nov.

IStrain A8p, strain C2BlS

...... Fig. 3. Most probable phylogenetic tree constructed using Jukes-Cantor distance matrix calculated by comparing the 165 rRNA genes (1400 bases) with a random input order of sequences. Distances were calculated by the number of substitutions per 100 bases by the Fitch-Margoliash method with contemporary tips. The resulting tree was the most probable from 11759 trees. The branch length for the given topology is a least-squares fit and is proportional to evolutionary distances. Bar, 1 % difference in nucleotide sequences, as determined by measuring the length of the horizontal lines connecting two species. For the tree construction the 165 rRNA gene sequences from the following organisms with EMBL accession numbers were used : Methanobacterium formicicum MF, M36508; Methanobacterium bryantii M.o.H., M59124; Methanobacterium thermoautotrophicum FTF, X68711; Methanobacterium thermoautotrophicum A H, X68720; Methanobacterium Marburg, X15364, Methanobacterium thermoformicicum FTF, X68713; Methanobacterium thermoformicicum 2245, X68712; bavaricum, X7 1838; , M59145; stadmanae MCB-3, M59139, Methanosaeta concilii, X16932, barkeri 227, M59144; cariaci JR1, M59130; , 2541 72. The figure shows the part of the tree with strains most closely related to strain A8p.

could not be sequenced completely (the sequenced DISCUSSION fragment was between positions 570 and 1178) due to what seemed to be blockage of the sequencing process The level of similarity among the 16s rRNA genes at position 571 by a tertiary structure. The 16s rRNA (99.3-100%) of three new strains A8p, C2BIS and gene of isolate 3067 differed by 0-7YO (13 bases) from 3067, isolated from different granitic groundwater, A8p and C2BIS (610 bases compared). Only un- implies that they can be regarded as closely related ambiguous bases were included in the comparison. organisms unless evidence of phenotypic differences The sequenced fragment of isolate 3067 included both is revealed. The rod-like morphology and similar variable and conservative regions of the 16s rRNA substrate specificity of the new isolates attest that they gene, and the differences were situated in the variable are phenotypically similar. Therefore, one organism, regions. The partial 16s rRNA gene sequence of the strain A8p, was chosen for detailed studies. isolate 3067 was not included in the phylogenetic The resolution of 16s rRNA sequencing analysis analysis. A phylogenetic tree was constructed based on between closely related organisms is generally low, but the 16s rRNA gene sequences of A8p and C2BIS in it is reliable for generic identification. The sequence comparison with organisms in the EMBL and Gen- similarities of the 16s rRNA gene of A8p showed that Bank databases (Fig. 3). The phylogenetic tree shows it was related to sequenced Methanobacterium species. that the isolates A8p and C2BIS were most closely The dissimilarity levels were 2.8-8.2 %. The levels of related to the mesophilic species of the genus Methano- dissimilarity were within the range found between bacterium. The highest similarity, 97.2 YO,was observed other species of this genus. Thus, the results of 16s with Methanobacterium formicicum. The 16s rRNA rRNA gene comparison indicated that our isolate genes of this organism differed in 38 positions over belongs to the genus Methanobacterium. Phylo- 1450 bases from A8p and C2BIS. The sequence genetically, A8p was most closely related to Methano- similarities of thel6S rRNA genes of A8p and C2BIS bacterium formicicum (97.2 YO similarity). However, to the other organisms were as follows: Methano- this evolutionary distance is sufficiently large that the bacterium wolfeii, 94.2 YO; Methanobacterium bryantii two strains most probably belong in separate species 92.5 % ; Methanobacterium thermoautotrophicum CB- (7, 12, 19, 41). The phenotypic characteristics of A8p 12, 92.2 YO; Methanobacterium thermoautotrophicum confirmed that the isolate belongs to the genus TFT and Methanobacterium thermoautotrophicum Methanobacterium and that it can be distinguished THF, 92-1 % ; Methanobacterium thermoautotrophi- from phylogenetically related taxa as shown in Table cum FTF, 92.0 YO; Methanobacterium thermoauto- 2. trophicum 2-245 and Methanobacterium thermoauto- trophicum AH, 9 1.8 YO. Cells of strain A8p were small, thin rods. The cell lnterna tional Journal of Systematic Bacteriology 48 363 S. Kotelnikova, A. J. L. Macario and K. Pedersen

Table 2. Comparison of phenotypic characteristics of mesop h i I ic Methanobacterium species

Property M. formicicum M. bryantii M. ivanovii M. alcaliphilum M. espanolae M. palustre M. utiginosum 'M.subterraneum '

Morphology Rod, 0.448 x 2-15 Rod, 051.0 x 1.5 Rod, O.W.8 x 1.2 Rod, 0,546x 2-5 Rod, 0.8 x 3-22 Rod, 0.5 x 3-5 Rod, 0.24.6 x 24Short rod, 0.14.15 x 0.61.2 Filaments, clumps Filaments, clumps Filaments Filaments Filaments Filaments Filaments Aggregates Growth H, + CO,, formate, H, + CO,, H, + CO, H, + CO, H, + CO, H, + CO,, formate H, +CO, H, + CO,, formate substrates 2-propanol+ CO,, 2-propanol+ CO,, 2-propanol+ CO,, isobutanol+ CO, isobutanol+ CO, isobutanol + CO, Autotrophy + + + - - - + + Stimulatory Acetate, cysteine Yeast extract, acetate, Acetate, cysteine Trypticase peptone, Vitamins Not reported Yeast extract None factors cysteine B, cysteine, yeast extract, Fe, W, Ni peptone Growth factors None None None Trypticase peptone, Vitamins Not reported None None yeast extract, peptone Temperature 25-50 (3745) m (37-39) 15-55 (45) 2545 (37) 15-50 (35) 2W5 (33-37) 1545 (40) 3.645 (2c-40) range, "C (optimum) pH range ND (666.8) ND (6.9-7.2) 6.5-8.5 (7'0-7'4) 7'0-9'9 (8.1-9.1) 4'67.0 (5.642) 7.0 6.0-8.5 (6.0-8.0) 6.5-9.2 (7.S8.8) (optimum) Salinity, M 025 0.26 0.19 0.012 m 0-0.3 (0.2) ND 0-1.4 (0.2-1.25) (optimum) G + C, mol YO 38.c-48.0 31.0-38.0 36.6 57.0 34.0 34.0 29.0 54.5 Reference (10) (1 1) (6) (4) (8) (31) (45) (21) This study

ND, Not done.

diameter and the thickness of the cell wall were bacterium that is able to grow at low temperatures. significantly smaller than those of previously studied A8p is also capable of growth within a wide range of methanogens in this genus (Table 2, Fig. 1). The length salinity values and at high pH values, which further of the rod-shaped cells may vary, but the cell diameter suggests that it is well-adapted to the deep granitic and thickness of the cell wall are more stable cell groundwater environment. The salinity and pH of the parameters. Thus, strain A8p differs morphologically studied groundwater at Asp0 HRL increase from 0 to from other members of the genus Methanobacterium. 0.5 M and from 7.5 to 8.5, respectively, down to a depth of 950m (40). These figures coincide with the Strain A8p had a wider temperature range than other optimum salinity and pH values obtained for strain mesophilic Methanobacterium species. It was capable A8p. The temperature, salinity and pH ranges of growth at very low temperatures (3.6-20 "C) but obtained for A8p are indicative of its habitual ad- also at mesophilic temperatures. This property prob- aptation to an active life in deep granitic aquifers and ably allows A8p to survive at different depths in support the suggestion of A8p as a new species. granitic rock aquifers, in which the temperature increases with depth from below 10 "C at ground level The adaptations to the specific environment might by 1-2 "C for every 100 m. The increase in temperature determine the unique phenotypical properties of the over the sampled depth interval, 68-420 m, was from isolated methanogen when compared to established 8.5 to 16 "C (20). The temperature in various parts of methanogens. The levels of salinity and pH tolerated this environment is stable but the groundwater in- by strain A8p were higher than those tolerated by habited by micro-organisms moves from one part to other Methanobacterium strains (Table 2). Although another through aquifers and may transport the cells its pH optimum is similar to that of Methanobacterium between granitic layers at different depths. In addition, alcaliphilum (8), strain A8p is salt-tolerant and con- the isolates were distributed over at least three different sumes formate. A8p is autotrophic like four other depths, 68, 409 and 420m below ground. The deep organisms of the genus Methanobacterium : Methano- groundwater is a unique environment which is not bacterium formicicum, Methanobacterium bryantii, limited by availability of hydrogen and carbon dioxide Methano bac ter ium ivanovii and Met hanobac terium (20), the main substrates for growth of our isolates. palustre (Table 2), but all these organisms except Thus, the varying temperature at different depths and Methanobacterium ivanovii are able to use alcohols as the absence of substrate limitation in the environment electron donors, which strain A8p could not do. Strain where strain A8p was isolated from, will favour A8p can grow on hydrogen and carbon dioxide or development of eurythermic and hydrogenotrophic formate like Methanobacterium formicicum and methanogens. None of the methanogens described in Methanobacterium palustre, but strain A8p differed the comprehensive review by Boone et al. (7) could from these species by its temperature and salinity grow below 10 "C. Methanosarcina and Methano- ranges and pH optimum. The difference in the G + C coccoides species growing under psychrophilic con- contents of strain A8p and Methanobacterium for- ditions have been isolated previously (16, 46). Conse- micicum was 16.5 mol% (Table 2). In contrast to quently, strain A8p is the first member of Methano- Methanobacterium ivanovii (4) and Methanobacterium

364 lnterna tional Journal of Systematic Bacteriology 48 Methanobacterium subterraneum sp. nov.

bryantii, strain A8p uses formate, is alkaliphilic and subterraneum, with strain A8p as the type strain. The halotolerant. Thus, strain A8p differs from known description follows. species of the Methanobacterium genus in its mor- phology, physiology, sensitivity to antibiotics, tem- perature and salinity ranges, pH optimum, substrate Description of Methanobacterium subterraneum sp. spectrum, and in 16s rRNA gene base composition nov. (Table 2). These differences support placing A8p as a Methanobacterium subterraneum (sub.ter.ra'ne.um. new species. L. adj. neut. subterraneum underground, below the The antigenic fingerprinting data demonstrated that earth/soil surface). the three isolates described here are antigenically Cells are non-motile, small and thin rods, 0.6-1-2 pm unrelated to reference methanogens and represent in length and 0-1-0.15 pm in diameter, often in novel immunotypes within the genus Methano- aggregates but not in chains. The substrates used for bacterium. Whether A8p, 3067 and C2BIS are of the growth and methane production include hydrogen and same or different immunotypes, remains to be es- carbon dioxide and formate, but not methylamines, tablished. The antigenic fingerprinting indicated no acetate, pyruvate, dimethyl sulfide, methanol or other cross-reaction of the isolates with other members of alcohols plus carbon dioxide. It grows autotrophically the family Methanobacteriaceae. The immunotypes of in mineral medium without any organic additions. the isolates could not be affiliated to any of the Growth is inhibited by yeast extract (2 g l-l), Casa- reference strains. mino acids (1 g l-'), isobutyric acid (5 mg lP), The requirement for growth factors is considered to be n-butyric acid (5 mg l-l), Na2Se0, (2 mg l-l), ZnCl, a distinguishing phenotypic property for methanogens. (2 mg l-l), CoC1, (2 mg l-'), NiC1, (20 mg 1-l) and The growth of strain A8p was inhibited by yeast MnCl, (20 mgl-l). Vitamins are not essential for extract, Casamino acids, isobutyric acid and n-butyric growth. Growth conditions : temperature 36-45 "C, acid. Low concentrations of Se4+,Zn2+, Co2+ and high pH 6.5-9.2, salinity 0.2-1-4 M NaC1. The G+C con- concentrations of Ni2+ and Mn2+ ions inhibited tent of the DNA is 54.5 & 0.5 mol % (as determined by methanogenesis (Table 1). Growth of strain A8p was thermal melting point). Isolated from granitic rock not stimulated by any growth factors and did not groundwater from the Asp0 hard rock laboratory require vitamins for growth and methanogenesis, tunnel, South-Eastern Sweden. The type strain is A8pT unlike many previously studied mesophilic species of (= DSM l1074T), and the reference strains are 3067 the genus Methanobacterium (Table 2). The nutritional and C2BIS (= DSM 11075), isolated from granitic requirements of strain A8p are in accordance with the groundwater at the depths 68, 409 and 420 m below chemistry of the environment from which this or- sea level, respectively. The 16s rRNA gene sequences ganism was isolated. The deep groundwater contains of A8pT, 3067 and C2BIS have EMBL/GenBank trace amounts of metals and it is oligotrophic with accession numbers X99044, Y 12592 and X99045, organic matter consisting mostly of fulvic and humic respectively. acids at concentrations not exceeding 11 mg 1-1 (40). Anaerobic microbial decomposition of organic ma- terial at these low concentrations cannot produce ACKNOWLEDGEMENTS inhibitory amounts of fatty acids. Thus, the nutritional The authors wish to thank Ann-Charlotte Erlandson requirements of A8p reflect its environmental ad- for thorough sequencing of the 16s rRNA genes; aptation. Lotta Hallbeck for help with the phylogenetic tree The isolates described here are the first examples of construction and Demetra Xythalis for help with methanogens from deep granitic groundwater, which antigenic fingerprinting. S. IS. thanks Karsten Pedersen can be 10000 years old at depths of 400-500 m (34). and hl; group for support during Asp0 expeditions They are representatives of life in the recently dis- and for having provided hospitality for 2 years. The covered (36) and may represent its research was supported by grants from The Swedish autotrophic part, responsible for primary production Nuclear Fuel and Waste Managment CO (SKB) and of organic matter (33, 42). Such autotrophic micro- from the Swedish Natural Science Research Council, organisms are suggested to live at the expense of grant number 8466/3 15/3 19/321. geochemically produced hydrogen, and the deep bio- sphere is then consequently postulated to be inde- REFERENCES pendent of the Sun-driven ecosystems on the Earth's surface (39, 42, 20, 28, 34). As our findings and other 1. Adler, H. A. & Roessler, E. B. (1976). Student's t-distribution. studies of the last years show, methanogenic Archaea Small sample method. In Introduction to Probability and probably play a significant role in subsurface bio- Statistics, pp. 17 1-1 9 1. San Francisco : Freeman. geochemistry (20, 28, 39). 2. Amann, R. J., Binder, B. J., Olson, R. J., Chrisholm, W., Devereux, R. & Stall, D. A. (1990). Combination of 16s Because of the phylogenetic and phenotypic charac- rRNA targeted oligonucleotide probes with flow cytometry teristics reported above, we propose that A8p is a new for analysing mixed microbial population. Appl Environ species. The proposed name is Methanobacterium Microbial 56, 1919-1925.

International Journal of Systematic Bacteriology 48 365 S. Kotelrukova, A. J. L. Macano and K. Yedersen

3. Amersham Life Science (1995). Thermo Sequenase Fluor- 21. KGnig, H. (1984). Isolation and characterization of Methano- escent Labelled Primer Cycle Sequencing Kit Manual, pp. bacterium uliginosum sp. nov. from a marshy soil. Can J 1-35. Little Chalfont: Amersham. Microbiol30, 1477-1481. 4. Belyaev, S. S., Obraztcova, A. Y., Laurinavichus, K. S. & 22. Lane, D. J. (1991). 16S/23S rRNA sequencing. In Nucleic Bezrukova, L. V. (1 986). Characteristics of rod-shaped Acid Techniques in Bacterial Systematics, pp. 115-177. methane-producing bacteria from oil pool and description Edited by E. Stackebrandt & M. Goodfellow. Chichester: of Methanobacterium ivanovii sp. nov. Microbiology Wiley. (English translation of Mikrobiologiya) 55, 821-826. 23. Lane, D. J., Pace, B., Olsen, G. J., Stahl, D. A,, Sogin, M. L. & 5. Belyaev, S.S., Obraztcova, A. Y., Laurinavichus, K. 5. & Pace, N. R. (1985). Rapid determination of 16s ribo- Bezrukova, L. V. (1986). Characterization of rod-like somalRNA sequences for phylogenetic analyses. Proc Natl methanogens isolated from oil field. Microbiology (English Acad Sci USA 82,6955-6959. translation of Mikrobiologiya) 55, 10 14-1020. 24. Macario, A. J. L. & Conway de Macario, E. (1983). Antigenic 6. Boone, D. R. (1987). Replacement of the type strain of fingeprinting of methanogenic bacteria with polyclonal Methanobacterium formicicum and reinstatement of antibody probes. Syst Appl Microbiol4,45 1-458. Methanobacterium bryantii sp. nov. nom. rev (ex Balch and 25. Macario, A. 1. L. & Conway de Macario, E. (1985). Mono- Wolf, 1981) with M.0.H. (DSM 863) as the type strain. Int clonal antibodies of predefined molecular specificity for J Syst Bacteriol37, 172-173. identification and classification of methanogens and for 7. Boone, D. R., Whitman, W. B. & Rouviere, P. (1993). Diversity probing their ecologic niches. In Monoclonal Antibodies and taxonomy of methanogens. In Methanogenesis, pp. Against Bacteria, vol. 2, pp. 213-247. Edited by A. J. L. 35-81. Edited by J. G. Ferry. London: Chapman & Hall. Macario & E. Conway de Macario. Orlando, FL :Academic 8. Boone, D. R., Worakit, S., Mathrani, 1. M. & Mah, R. A. (1986). Press. Alkaliphilic methanogens from high-pH lake sediments. 26. Marmur, J.A. (1961). A procedure for the isolation of Syst Appl Microbiol7,230-234. deoxyribonucleicacid from microorganisms. J A401 Biol3, 9'. IBrosius, J., Palmer, M. L., Kennedy, P. J. & Noller, H. F. (1978). 208-2 18. IComplete nucleotide sequence of a ribosomal RNA gene 27. Marmur, 1. & Doty, P. (1962). Determination of the base from . Proc Natl Acad Sci USA 75, 4801- composition of deoxyribonucleic acid from its thermal 4805. denaturation temperature. J Mol Biol5, 109-1 18.

10'. 1Bryant, D. R., Worakit, S., Matrani, 1. & Mah, R. A. (1987). 28. Martini, A., Budai, 1. M., Walter, L. M. & Schoell, M. (1996). Isolation and characterization of Methanobacterium for- Microbial generation of economic accumulations of meth- micicum MF. Int J Syst Bacteriol 37, 171. ane within a shallow organic-rich shale. Nature 383, 11. Bryant, M. P., Wolin, E. A., Wolin, M. J. & Wolf, R. 5. (1967). 155-158. Methanobacillus omelianski, a symbiotic association of two 29. Nilsson, A.-C. (1995). Compilaiion of groundwater chemistry species of bacteria. Arch Microbiol59,20-3 1. data from Asp0 1990-1994. Asp0 Progress Report 25-95-01, 12. Devereux, R., He, S., Doyl, C., Orkland, S., Stahl, D. & LeGall, pp. 1-200. Edited by A.-K. Nillson. Stockholm: Swedish J. (1990). Diversity and origin of Desuljivibrio species: Nuclear Fuel and Waste Management. phylogenetic definition of a family. J Bacteriol 172, 3609- 30. Owen, R. J. & Pitch, D. (1985). Current methods for es- 3619. timating DNA base composition and level DNA-DNA 13. Doddema, H. J. & Vogels, G. D. (1978). Improved identi- hybridization. In Chemical Methods in Bacterial Systema- fication of methanogenic bacteria by fluorescence micro- tics, pp. 67-93. Edited by M. Goodfellow & D. E. Minnkin. scopy. Appl Environ Microbiol36, 752-754. London : Academic Press. 14. Ekendahl, S., Arlinger, J., St%hl, F. & Pedersen, K. (1994). 31. Patel, G. B., Sprott, G. D. & Fein, J. E. (1990). Isolation and Characterization of attached bacterial populations in deep characterization of Methanobacterium espanolae sp. nov. a granitic groundwater from the Stripa research mine by 16s mesophilic, moderately acidiphilic methanogen. Int J Syst rRNA gene sequencing and scanning electron microscopy. Bacteriol40, 12-18. Microbiology 140, 1575-1 583. 32. Pedersen, K. (1993). The deep subterranean biosphere. Earth 15, Felsenstein, J. (1989). PHYLIP - Phylogeny Inference Pack- Sci Rev 34,243-260. age. Cladistics 5, 164-166. 33. Pedersen, K. (1993). Bacterial processes in nuclear waste 16, Fransmann, P. O., Springer, N., Ludwig, W., Conwey de disposal. Microbiol Eur 1, 18-23. Macario, E. & Rohde, M. (1992). A methanogenic archaeon 34. Pedersen, K. (1997). Microbial life in deep granitic rock. from an ice lake in Antarctica: Methanococcoides burtonii FEMS Microbiol Rev 20, 399-414. sp. nov. Syst Appl Microbioll5, 573-581. 35. Pedersen, K., Arlinger, J., Ekendahl, S. & Hallbeck, L. (1996). 17. Hungate, R. E. (1969). A roll tube method for the cultivation 16s rRNA gene diversity of attached and unattashed of strict anaerobes. Methoh Microbiol3b, 117-132. groundwater bacteria along the access tunnel to the Asp0 18. Jukes, T. H. & Cantor, C. R. (1969). Evolution of hard rock laboratory, Sweden. FEMS Microbiol Ecol 19, molecules. In Mammalian Protein , pp. 21-132. 249-262. Edited by N. H. Munro. New York: Academic Press. 36. Pedersen, K. & Ekendahl, 5. (1990). Distribution of bacteria 19. Kadam, P. C. & Boone, D. R. (1995). Physiological charac- in deep granitic groundwaters of Southeasthern Sweden. terization and emended description of FEMS Microbiol Ecol20, 37-52. vulcani. Int J Syst Bacteriol45, 400-402. 37. Pedersen, K. & Ekendahl, 5. (1992). Assimilation of CO, and 20. Kotelnikova, S. & Pedersen, K. (1997). Evidence for methano- introduced organic compounds by bacterial communities in genic Archaea and homoacetogenicBacteria in deep granitic ground water from Southeastern Sweden deep crystalline rock aquifers. FEMS Microbiol Rev 20, 339-349. bedrock. FEMS Microbiol Ecol22, 1-14.

766 International Journal of Svstematic Bacterioloqv 48 Methanobacterium subterraneum sp. nov.

38. Powell, G. E. (1983). Interpretations of gas kinetics of batch 42. Stevens, T. 0. & McKinley, 1. P. (1995). Lithoautotrophic cultures. Biotechnol Lett 5, 437-440. microbial ecosystems in deep aquifers. Science 270, 39. Sherwood Lollar, B., Frape, S. K., Fritz, P., Macko, S.A., 450-454. Welhan, 1. A., Blomqist, R. & Lahermo, P. W. (1993). Evidence 43. Wallance, D. M. (1987). Large and small-scale phenol extrac- for bacterially generated hydrocarbon gas in Canadian tions. Methods Enzymoll52, 33-41. Shield and Fennoscandian Shield rocks. Geochim Cosmo- 44. Wolin, E. A., Wolin, M. J. & Wolf, 1. (1963). Formation of chim 57, 5073-5085. methane by bacterial extracts. J Biol Chem 238,2882-2886. 40. Smellie, J. A. T., Laaksoharju, M. & Wiksberg, P. (1995). Aspo, 45. Zellner, G., Bleicher, K., Braun, E., Kneifel, H., Tindall, B., SE Sweden: a natural groundwater flow model derived Conway de Macario, E. & Winter, J. (1989). Characterization from hydrogeochemical observations. J Hydro1 172, 147- of new mesophilic, secondary alcohol-utilizing methanogen, 169. Methanobacterium palustre sp. nov. from a peat bog. Arch 41. Stackebrandt, E. & Goebel, B. M. (1994). Taxonomic note: a Microbioll51, 1-9. place for DNA-DNA reassociation and 16s rRNA analysis 46. Zhilina, T. M. & Zavarzin, G. A. (1991). Methane production in the present species definition in bacteriology. Int J Syst by pure culture of Methanosarcina at low temperature. Dokl Bacteriol44, 842-849. Akad Nauk SSSR 317, 1242-1245.

~~ ~ In ternationa I JournaI of Systematic Bacteriology 48 367