<<

International Journal of Systematic and Evolutionary Microbiology (2001), 51, 1853–1862 Printed in Great Britain

Hydrogenothermus marinus gen. nov., sp. nov., a novel thermophilic hydrogen-oxidizing bacterium, recognition of Calderobacterium hydrogenophilum as a member of the Hydrogenobacter and proposal of the reclassification of Hydrogenobacter acidophilus as Hydrogenobaculum acidophilum gen. nov., comb. nov., in the ‘Hydrogenobacter/Aquifex’

1 Institut fu$ r Allgemeine Ru$ diger Sto$ hr,1† Arne Waberski,1 Horst Vo$ lker,1 Brian J. Tindall2 Mikrobiologie, Am 1 Botanischen Garten 1–9, and Michael Thomm 24118 Kiel, Germany

2 DSMZ Braunschweig, Author for correspondence: Michael Thomm. Tel: j49 431 880 4330. Fax: j49 431 880 2194. Mascheroder Weg 1b, e-mail: mthomm!ifam.uni-kiel.de 38124 Braunschweig, Germany A novel thermophilic, hydrogen-oxidizing bacterium, VM1T, has been isolated from a marine hydrothermal area of Vulcano Island, Italy. Cells of the strain were Gram-negative rods, 2–4 µm long and 1–15 µm wide with four to seven monopolarly inserted flagella. Cells grew chemolithoautotrophically under an

atmosphere of H2/CO2 (80:20) in the presence of low concentrations of O2 (optimum 1–2%). Carbohydrates and peptide substrates were not utilized, neither for energy generation nor as a source of cellular carbon. Growth of VM1T occurred between 45 and 80 SC with an optimum at 65 SC. Growth was observed between pH 5 and 7. NaCl stimulated growth in the range 05–6% with an optimum at 2–3%. Hydrogen could not be replaced by elemental sulfur or thiosulfate as electron donors. Nitrate and sulfate were not used as electron acceptors. The major respiratory lipoquinone was a new menathioquinone. T Analysis of the fatty acids of VM1 revealed straight-chain saturated C18:0 and the unsaturated C18:1ω9c and C20:1ω9c as major components. The GMC content of the total DNA was 43 mol%. Phylogenetic analysis placed strain VM1T near the members of the genera Hydrogenobacter, Thermocrinis and Aquifex on a separate deep-branching phylogenetic lineage. Therefore, it is proposed that strain VM1T (l DSM 12046T l JCM 10974T) represents a novel species within a new genus, for which the name Hydrogenothermus marinus gen. nov., sp. nov., is proposed. In addition, it is shown that Calderobacterium hydrogenophilum should be transferred to the genus Hydrogenobacter; the name Hydrogenobacter hydrogenophilus comb. nov. (DSM 2913T l JCM 8158T)is proposed for this organism. Furthermore, on the basis of 16S rRNA sequence analysis, Hydrogenobacter acidophilus is only distantly related to Hydrogenobacter species. Owing to this finding and its growth at low pH, the name Hydrogenobaculum acidophilum gen. nov., comb. nov., is proposed for Hydrogenobacter acidophilus. The type strain is JCM 8795T (l DSM 11251T).

Keywords: hydrogen oxidation, thermophilic , Aquifex, Hydrogenobacter

...... † Present address: Institut fu$ r Meereskunde, Du$ sternbrooker Weg 20, 24105 Kiel, Germany. The GenBank accession number for the 16S rDNA sequence of strain VM1T is AJ292525.

01805 # 2001 IUMS 1853 R. Sto$ hr and others

INTRODUCTION added to the headspace of the serum bottles by use of a sterile filter after autoclaving the medium. Mass cultures of When was described (Huber et al., VM1T were grown in a 10 l titanium fermenter (Braun 1992; Burggraf et al., 1992), a new phylum of the Biotech). The fermenter was gassed with 120 ml H#,30ml −" Bacteria was found that represented the deepest CO# and 7n5 ml air min . Working with mixtures of branching of the bacterial . Aquifex pyrophilus hydrogen and can lead to highly explosive gas was characterized by its hyperthermophilic and chemo- mixtures when the hydrogen atmosphere contains more lithoautotrophic metabolism, yielding energy from the than 25% air (Aragno & Schlegel, 1992). Under normal oxidation of molecular hydrogen. Other thermophilic, conditions of fermentation, this explosive atmosphere was hydrogen-oxidizing bacteria were found to be related. not formed. To prevent hydrogen entering the room at- mospheere, the fermenter was equipped with a direct exhaust Hydrogenobacter thermophilus (Kawasumi et al., 1984) pipe out of the building and gas-tight bearings. and Calderobacterium hydrogenophilum (Kryukov et al., 1983) also could be grouped together within the In order to analyse carbon source utilization by VM1T,3g order ‘Aquificales’ (Huber et al., 1998). PIPES buffer, adjusted to pH 6n0, was added to 1 l medium as a buffer and NaHCO$ was omitted from the medium. The While Aquifex pyrophilus was isolated from a marine carbon sources meat peptone, tryptone, meat extract, yeast geothermally heated area of the Kolbeinsey Ridge, extract, lactose, -galactose, α--glucose, -ribose, -fruc- members of the genera Hydrogenobacter, Caldero- tose, sucrose, citric acid, α--maltose hydrate, starch, - bacterium and Thermocrinis (Huber et al., 1998) were xylose, -alanine, -proline, -histidine hydrochloride, isolated from freshwater habitats. The only marine glycine, methanol, ethanol, acetic acid, pyruvate, disodium species of the genus Hydrogenobacter reported so far is fumarate, -malate and ammonium formate were added ‘Hydrogenobacter halophilus’, which was isolated from individually at concentrations of 0n1%. The gas atmosphere was 99% H# (300 kPa) and 1% O#. To analyse growth of the a marine (Nishihara et al., 1990). With strain in the presence of organic carbon sources and in the respect to the optimal temperature of growth, members absence of hydrogen, the strain was cultivated under a of the genera Aquifex and Thermocrinis are hyper- N#\CO# atmosphere (80:20; 300 kPa) in the presence of 1% , showing optimal growth at 85 mC. O#. Growth on the carbon sources yeast extract, glucose, Representatives of the genera Hydrogenobacter and starch, peptone, tryptone and maltose, added individually at Calderobacterium show optimal growth around 70 mC. final concentrations of 0n1%, was assayed under a hydrogen- They do not grow at 85 mC. Here, we describe a new free atmosphere. strain of marine origin with an even lower optimal The gas atmosphere was changed to N#\CO# (80:20) to test growth temperature, of 65 mC, representing a new for fixation and the ability to use thiosulfate and phylogenetic lineage within the phylum ‘Hydro- elemental sulfur as electron donators. Oxygen was omitted genobacter\Aquifex’. when testing sulfate and nitrate (0n1% KNO$,w\v) as electron acceptors. METHODS Isolation procedure. Pure cultures were obtained by repeated transfers of serial dilution cultures. The cultures were Origin of samples. Strain VM1T was isolated from a marine checked for contamination using a phase-contrast light water sample that also contained sediment. The sample was microscope (Zeiss). The purity of the cultures was confirmed taken from a geothermally heated shallow area at Vulcano by repeated partial sequence analysis of the gene encoding beach, 3–4 m from the shore. The temperature of the sample 16S rRNA. was 83 C. The sample was collected with a 20 ml syringe m Gram staining. Gram staining was performed by using the and was transferred to a 20 ml tube containing a drop of Bacto 3-step Gram-stain procedure (Difco). resazurin solution (0n1%). The tube was sealed with a rubber stopper and reduced by adding a spatula-tip amount of Measurement of growth. Growth experiments were set up in dithionite to protect the sample from oxygen, as we planned 120 ml serum bottles that were incubated in a reciprocally initially to isolate anaerobes. shaking water bath (100 r.p.m.). Growth curves were de- termined by direct counting using a Thoma Blau Brand Culture media. Modified marine medium described by T chamber (Omnilab-Laborzentrum) with a depth of 0n02 mm ZoBell (1941) was used for the isolation of VM1 . This −" under a phase-contrast microscope (Zeiss standard 16). The medium contained (g l ): Bacto yeast extract, 1n0; Bacto doubling times were calculated from the slopes of growth peptone, 5n0; NaCl, 19n4; MgCl#.6H#O, 12n6; NaHCO$, curves of three replicates. When the pH optimum was 0n16; Na#SO%,3n24; KCl, 0n56; elemental sulfur, 2n0; determined, the pH was adjusted 1 d prior to the experiment resazurin (0n1%), 1n0 ml; trace minerals (10i) 10 ml (Balch and readjusted immediately before inoculation using uni- et al., 1979). For large-scale fermentation and subsequent versal pH paper (duotest; Macherey-Nagel). cultivation of the isolate, Bacto peptone and Bacto yeast extract were replaced by 0n3g NH%Cl, and 2n38 g Electron microscopy. A cell suspension of a well-grown CaCl#.2H#O; the amount of elemental sulfur was reduced to culture was applied to Pioloform-covered 300 mesh Cu grid, −" 0n5gl . The pH was adjusted to 7n0 with H#SO% (25%). The washed once with glass-distilled water and sputtered after medium was mixed with an UltraTurrax for 1 min, deoxy- drying with Pt\C at an angle of 40m. A culture was fixed genated under a stream of N# for 20 min and dispensed in overnight at 4 mC with 2% glutaraldehyde and 0n05% 20 ml portions in 120 ml type III borosilicate bottles ruthenium red prior to thin sectioning. After centrifugation (Pharmapack; Stute) under a N# atmosphere. Prior to (10000 g, 10 min), the cells were washed three times in sterilization for 90 min at 100 mC, the atmosphere was cacodylate buffer (0n1M,pH7n0). The pellet was post-fixed changed to H#\CO# (80:20; 300 kPa). Twenty ml of air was for 3 h at 4 mC with a mixture of equal volumes of OsO%

1854 International Journal of Systematic and Evolutionary Microbiology 51 A novel hydrogen oxidizer

(2%), ruthenium red (0n15%) and cacodylate buffer (0n2M). were recovered by cooling the hydrolysis mixture to room It was dehydrated in ascending concentrations of ethanol temperature and adding 0n5ml10%NH%HCO$ to neutralize using propylene oxide as the intermediate medium and the acid. This also caused the tert-butyl methyl ether phase embedded in Spurr’s resin. Ultrathin sections were cut using to separate from the lower, aqueous methanol phase and a Reichert-Ultracut S ultramicrotome. Sections were stained allowed recovery of the hydrolysed lipophilic material. Five- with uranyl acetate and lead citrate (Reynolds, 1963). hundred microlitres tert-butyl methyl ether was added to the Electron micrographs were taken using a Philips EM 300 aqueous methanol phase, the suspension was shaken and the electron microscope at 80 kV on Kodak electron microscope upper, tert-butyl methyl ether phase was pooled with the first film (no. 4489). tert-butyl methyl ether layer. Fatty acids, diethers and monoethers were separated on silica-gel thin layers using Extraction of respiratory lipoquinones and polar lipids. hexane\diethyl ether\acetic acid (25:25:1, by vol.) as the Respiratory lipoquinones and polar lipids were extracted solvent system. Lipid material was visualized using 5% from 100 mg freeze-dried cell material using the two-stage ethanolic dodecamolybdophosphoric acid after heating to method described by Tindall (1990a, b). Respiratory 150 mC for 30 min. The presence of vicinal hydroxyl groups quinones were extracted using methanol\hexane (Tindall, was detected using the periodate–Schiff reagent. 1990a, b) and the polar lipids were extracted by adjusting the remaining methanol\0n3% aqueous NaCl phase (containing DNA base composition. The DNA GjC content was the cell debris) to give a choroform\methanol\0n3% aque- determined by HPLC according to Mesbah et al. (1989). ous NaCl mixture (1:2:0n8, by vol.). The extraction solvent Non-methylated lambda DNA (Sigma) was used as a was stirred overnight and the cell debris was pelleted by standard. centrifugation. Polar lipids were recovered into the chloro- DNA isolation. About 5 mg lyophilized cells (a tip of a form phase by adjusting the chloroform\methanol\0n3% spatula) was resuspended in 150 µl sterile distilled water. A aqueous NaCl mixture to a ratio of 1:1:0n9 (by vol.). preincubation of 10 min at 37 mC was followed by the addition of 567 µl Tris\EDTA buffer (10 mM, pH 8n0), 30 µl Analysis of respiratory lipoquinones. Respiratory lipo- −" quinones were separated into their different classes (e.g. 10% SDS and 3 µl proteinase K (20 mg ml ; Sigma). After menaquinones and ubiquinones) by TLC on silica gel incubation for 1 h at 37 mC, 100 µl NaCl and 80 µl10% (Macherey-Nagel art. no. 805023), using hexane\tert-butyl CTAB were added and the mixture was incubated at 65 mC methyl ether (9:1, v\v) as solvent. UV-absorbing bands for 30 min. DNA was extracted by treatment with phenol\ corresponding to respiratory quinones were removed from chloroform (1:1) and twice with chloroform. It was pre- the plate and analysed further by HPLC. This step was cipitated with 2-propanol, washed with ethanol (70%) and carried out on an LDC Analytical HPLC (Thermo Sep- dissolved in 10 µl sterile distilled water. aration Products) fitted with a reverse-phase column Phylogenetic analysis. The 16S rRNA gene was amplified (2i125 mm, 3 µm, RPB"); Macherey-Nagel) using from isolated DNA by PCR. PCRs contained: 5n0 µlRP methanol\heptane (10:2, v\v) as the eluant. Respiratory buffer [1 M Tris\HCl, pH 9n0, 400 mM (NH%)#SO%,30mM lipoquinones were detected at 269 nm. MgCl#], 10 µl dNTP mix (2n5 mM each of dATP, dCTP, Analysis of polar lipids. Polar lipids were separated by two- dGTP and dTTP), 2 µl12n5 µM forward primer (5h-GAG- dimensional silica-gel TLC (Macherey-Nagel art. no. TTTGATCCTGGCTCAG-3h, positions 9–27), 2 µl 818135). The first direction was developed in chloroform\ 12n5 µM reverse primer (5h-TACGGCTACCTTGTTACG- methanol\water (65:25:4, by vol.) and the second in ACTT-3h, positions 1510–1492; Pharmacia), 0n5 µl DNA chloroform\methanol\acetic acid\water (80:12:15:4, by template (10–100 ng), 80 µl sterile distilled water and 50 µl vol.). Total lipid material and specific functional groups mineral oil. Taq DNA polymerase (2n5 U; Boehringer were detected using dodecamolybdophosphoric acid (total Mannheim) was added after a ‘hot start’. PCR was lipids), Zinzadze reagent (phosphate), ninhydrin (free amino performed in a Mastercycler (Eppendorf) using the fol- groups), periodate–Schiff (α-glycols), Dragendorff reagent lowing program: 4 min at 94 mC, 72 mC during addition of (quaternary nitrogen) and anisaldehyde\sulfuric acid the enzyme and 35 cycles of 45 s at 94 mC, 45 s at 50 mC and (glycolipids). 75 s at 72 mC. After 35 cycles, extension was continued for 10 min at 72 mC and the cycler was cooled to 4 mC. Analyses of fatty acids. Fatty acids were analysed as the methyl ester derivatives prepared from 30 mg frozen cell The PCR product was purified using the Wizard PCR Prep material. Fatty acid methyl esters were analysed by gas DNA purification system (Promega) and collected in 100 µl chromatography using a 0n2 µmi25 m non-polar capillary double-distilled water. The concentration of DNA was column and flame ionization detection. The run conditions estimated from an agarose gel stained with ethidium bro- were: injection and detector port temperature, 250 mC; inlet mide. The sequence of the PCR product was determined using the AmpliCycle TM sequencing kit (Perkin Elmer). pressure, 9 p.s.i. (62 kPa); split ratio, 50:1; injection volume, $& [ S]ATPγS was used for labelling. The 16S rRNA gene 2 µl; with a temperature program from 170 to 310 mCata −" sequences of the new isolates were aligned using   rate of 7 mC min . version 1.7 (Thompson et al., 1994) with sequences taken Analysis for the presence of compounds in addition to fatty from the Ribosomal Database Project (RDP) (Olsen et al., acids. The presence of fatty acids and other compounds (i.e. 1991) and EMBL database. Programs of the  package mono- and diethers or long chain diols) was analysed (version 3.5) (Felsenstein, 1989) were used for calculations. following hydrolysis of 10 mg dry cell material. Cell material Distance matrixes were set up using  with the from Thermodesulfobacterium commune (kindly supplied by Jukes–Cantor (Jukes & Cantor, 1969) and maximum- T. A. Langworthy) served as reference material for the likelihood option. The neighbour-joining method and the presence of non-isoprenoid mono- and diethers. Cell ma-  program generated tree estimations with a random terial was incubated for 18 h at 50 mC in 1 l methanol\tert- order input and a global rearrangement option activated. butyl methyl ether\sulfuric acid (5:5:0n2 by vol.). Fatty acid Bootstrap analysis with 1000 replicates was performed using methyl esters and other components released from the lipids the  and  programs of the same package.

International Journal of Systematic and Evolutionary Microbiology 51 1855 R. Sto$ hr and others

RESULTS (a) Enrichment and isolation In searching for marine, thermophilic, hydrogen- oxidizing bacteria, water-plus-sediment samples were collected from a shallow hydrothermal area of Vulcano Island, Italy. The enrichment culture was set up using 20 ml modified marine medium inoculated with 1 ml samples and a gas atmosphere of H#\CO#\O# (80:20: 1n2) pressurized to 2 bars (200 kPa). After incubation for 1 d at 75 mC, the culture became turbid. A series of serial dilutions led to the isolation of VM1T. Purity of the culture was checked microscopically and by se- quence analysis of 16S rRNA.

Morphological characteristics (b) Cells of VM1T were rods that were Gram-negative. The cells were 2–4 µm long and 1–1n5 µm wide. Microscope examination revealed that they were mo- tile at room temperature. Formation of endospores was not observed. Electron microscopy showed that cells contained four to seven monopolar flagella (Fig. 1a). Analysis of ultrathin sections revealed a cell wall structure typical of Gram-negative bacteria (Fig. 1b).

Physiological characterization T Strain VM1 was isolated at 75 mC. Optimal growth was observed at 65 mC, whereas no growth occurred at 35 or 85 mC (Fig. 2a). The strain grew at NaCl ...... concentrations in the range 0n5–6%. Optimal growth Fig. 1. Electron micrograph of an ice-dried, platinum-shadowed occurred at 2 and 3% (Fig. 2b). Thermophilic, cell with flagella (a) and a thin section of a cell (b) of isolate hydrogen-oxidizing bacteria are especially sensitive to VM1T. Bars, 1n0 µm. T high oxygen concentrations; VM1 tolerated O# up to 8%, although optimal growth was observed at 1–2% O# in the atmosphere (Fig. 2c). Growth occurred over a range of pH between 5 and 7 (data not shown). 20; 300 kPa) containing 1% oxygen in addition. The strain did not grow under these conditions. This Carbon source utilization and other nutritional finding indicates that hydrogen cannot be replaced as features the electron donor by these organic substances. Some thermophilic, hydrogen-oxidizing organisms can use T elemental sulfur or thiosulfate as alternate electron Strain VM1 showed good growth under lithotrophic T conditions with H# as electron donor, O# as electron donors. Growth of VM1 was dependent on elemental acceptor and CO# as the source of carbon. Therefore, sulfur in the medium, but it could not grow in the the complex components of the modified marine presence of sulfur or thiosulfate under a N#\CO# medium in which strain VM1T was isolated were atmosphere (80:20; 300 kPa) containing 1% oxygen replaced by NH%Cl as a source of cellular nitrogen. in addition. This finding suggests that sulfur is required as a source for the biosynthesis of cellular sulfur- In order to test whether the strain could use organic containing compounds, but cannot be used as the sole substances as sole carbon sources, the gas atmosphere electron donor. was changed to H#\O# (99:1) with a variety of organic carbon sources added (0n1%). None of these 25 organic Oxygen was omitted from the gas atmosphere and was compounds were utilized, which was taken as an replaced by KNO$ (0n1%) in the medium to test indication of obligately autotrophic growth. In order whether nitrate was a suitable electron acceptor. The to test whether organic substances could replace strain was not capable of using either nitrate or sulfate hydrogen as a source of energy, growth in the presence as electron acceptors. In the presence of O#, nitrate was of yeast extract, peptone, tryptone, starch, glucose and tested as a source of cellular nitrogen, but there was no maltose was assayed under a N#\CO# atmosphere (80: growth.

1856 International Journal of Systematic and Evolutionary Microbiology 51 A novel hydrogen oxidizer

(a) 180 (b) 220 160 200 140 180 160 120 140 100 120 80 100 60 80 Doubling time (min) Doubling time (min) 40 60 20 40 0 20 40 45 50 55 60 65 70 75 80 0 123 4567 Temperature (°C) NaCI concentration (%)

(c) 80

70

60

50

40 ...... Fig. 2. Growth of isolate VM1T dependent Doubling time (min) 30 on temperature (a), concentration of NaCl (b) and O2 concentration in the headspace (c). Cultures were grown at pH 6n8 and 1n9% 20 NaCl at various temperatures (a), at 65 mC 0 123 456789 and pH 6n8 (b) or at 65 mC, 1n9% NaCl and pH O2 concentration (%) 6n8 (c).

Table 1. Major components of the fatty acids of representatives of the phylum ‘Hydrogenobacter/Aquifex’ ...... Values are percentages of total fatty acids unless otherwise indicated. jjj, Major component, not quantified; k, not detected. Data for Hydrogenobacter thermophilus and ‘Hydrogenobacter halophilus’ were taken from Kawasumi et al. (1984) and Nishihara et al. (1990), respectively.

Fatty acid Hydrogenobacter ‘Hydrogenobacter Hydrogenobaculum Aquifex Hydrogenothermus thermophilus halophilus’ acidophilum pyrophilus marinus VM1T

C"):! jjj 15 23–28 22–25 22–24 C"):"ω9c – 19 16–29 4–7 15–16 C#!:"ω9c jjj* 43–44 37–43 22–28 46–51 X(Rf 0n086) – – – 16–27 –

* Position of the double bond not specified in the literature.

T VM1 was both catalase- and cytochrome-oxidase- rated C"):"ω9c and C#!:"ω9c as major components. In positive. contrast, Aquifex pyrophilus lacked C"):"ω9c as a major component but instead had another component, Analysis of fatty acids and respiratory quinones X(Rf l 0n086) (Table 1). Fatty acids were analysed as methyl ester derivates Analysis of the quinone compounds revealed one prepared from 30 mg frozen cell material. Analysis of major and one minor compound to be present, neither the major fatty acids of VM1T, Hydrogenobacter of which co-chromatographed with the two com- thermophilus and Hydrogenobacter acidophilus pounds present in Hydrogenobacter acidophilus, indi- revealed straight-chain saturated C"):! and unsatu- cating that they were novel compounds. Examination

International Journal of Systematic and Evolutionary Microbiology 51 1857 R. Sto$ hr and others

...... Second dimension Fig. 3. Two-dimensional chromatographic separation of polar lipids of Hydrogeno- bacter acidophilus (a) and VM1T (b). PL, Phospholipid; PNL, aminophospholipid; GL, First dimension glycolipid.

...... Fig. 4. Phylogenetic dendrogram based on 16S rDNA sequences showing the relationship of the novel isolate VM1T to the members of the phylum ‘Hydrogenobacter/Aquifex’. Confidence limits expressed as percentages were determined by bootstrap analysis with 1000 replicates. Only confidence limits of more than 95% are shown. of the mass spectrum of the major compound indicated reference. The molecular ion was shifted by 4 mass that it produced a fragment at m\z l 257, charac- units compared with MTK-7H%, giving a molecular teristic of the menathioquinone ring nucleus (Ishii et ion at m\z l 680. This would suggest that the com- al., 1983, 1987), MTK-7H% serving as an authentic pound also contains a heptaprenyl side chain, but that

1858 International Journal of Systematic and Evolutionary Microbiology 51 A novel hydrogen oxidizer none of the double bonds are hydrogenated. Further phylum of hyperthermophilic and thermophilic work is in progress for a complete characterization of hydrogen-oxidizers by 16S rDNA analysis. This phy- the new compound but, on the basis of evidence lum includes the genera Aquifex, Hydrogenobacter, collected to date, we infer that the major compound is Calderobacterium and Thermocrinis, which all include probably a menathioquinone with a heptaprenyl side Gram-negative bacteria with a chemolithoautotrophic chain (i.e. MTK-7). metabolism. Growth above 70 mC currently distin- guishes this group from other thermophilic, Gram- negative, hydrogen-oxidizing bacteria such as ‘Pseudo- Analysis of polar lipids monas thermophila’ and Flavobacterium thermophilum The polar lipid compositions of strain VM1T and (Aragno & Schlegel, 1992), Hydrogenophilus thermo- Hydrogenobacter acidophilus were compared. Both luteolus (Hayashi et al., 1999) and Hydrogenophilus strains were dominated by the presence of phospho- hirschii (Sto$ hr et al., 2001). The latter two belong to the lipids. Strain VM1T also contained small amounts of β-subclass of the and grow below 70 mC glycolipids. The major phospholipids present in all with a range between 50 and 68 mC. strains were an aminophospholipid (PNL) and a T phospholipid (PL1). These results are also consistent The new isolate, VM1 , exhibits, together with Hydro- genobacter acidophilus, a lower temperature optimum with the report of similar compounds in Aquifex T pyrophilus (Huber et al., 1992). PNL did not, however, of 65 mC. As an isolate of marine origin, VM1 can have an Rf value similar to that of authentic clearly be separated from most Hydrogenobacter phosphatidylethanolamine and we have therefore strains, which cannot tolerate NaCl concentrations refrained from assigning this structure to this com- of 0n3M (1n74%) or more (Nishihara et al., 1990; pound. The polar lipid composition of the strains Kristjansson et al., 1985). The only exception to date is examined differed and allowed them to be distin- ‘Hydrogenobacter halophilus’, which grows optimally guished easily on this basis (Fig. 3). No data are between 0n3 and 0n5 M NaCl (Nishihara et al., 1990). currently available for Hydrogenobacter thermophilus, However, at the time of writing, this organism was not available. ‘Hydrogenobacter halophilus’ did not grow Thermocrinis ruber or Calderobacterium hydrogeno- T philum. in 1 M NaCl, whereas VM1 tolerates 6% (1n03 M) NaCl. In contrast to ‘Hydrogenobacter halophilus’, strain VM1T is motile by means of four to seven DNA base composition flagella. In contrast to all hitherto-described represen- The mean DNA base composition was determined by tatives of the genera Hydrogenobacter, Aquifex and HPLC (Mesbah et al., 1989). The GjC content of Thermocrinis (Bonjour & Aragno, 1986; Huber et al., T strain VM1 was 43n0mol%. 1992; Shima & Suzuki, 1993; Huber et al., 1998), this organism cannot use sulfur or thiosulfate as alternate electron donors. VM1T is the only organism of this Phylogeny bacterial phylum whose growth is strictly dependent A sequence of 1433 bases of the 16S rRNA of VM1T upon the presence of hydrogen as an electron donor. was determined. Phylogenetic analysis using the Jukes Like Hydrogenbacter acidophilus, it requires elemental & Cantor equation for distance-matrix calculation and sulfur for growth. In contrast to Aquifex pyrophilus, the program  for calculation of a phylogenetic nitrate could not substitute for oxygen as an electron tree placed the strain in the same phylum as members acceptor. A comparison of the fatty acid composition of the genera Aquifex and Hydrogenobacter (Fig. 4). revealed that C"):!,C"):" and C#!:" are characteristic Within this group, the strain represents a new, deep- for Hydrogenobacter thermophilus, Hydrogenobacter acidophilus,‘Hydrogenobacter halophilus’ (Nishihara branching lineage with respective sequence similarities T of 85n9 and 87n6%toHydrogenobacter acidophilus and et al., 1990) and isolate VM1 (Table 1). Aquifex Aquifex pyrophilus. A similar topology was also found pyrophilus can be distinguished from this group be- using the maximum-likelihood option to calculate the cause it only possesses low levels of C"):"ω9c.An distance matrix and the neighbour-joining method to additional, uncharacterized, major component is calculate the phylogenetic tree. The significance of the found instead (Table 1). Mono- and diethers found in Aquifex pyrophilus (Huber et al., 1992) could not be topology was tested further by bootstrap analysis T using 1000 replicates. detected in strain VM1 . Although we included Hydrogenobacter thermophilus The major compound of the quinone systems of strain T3 and ‘Hydrogenobacter subterranea’, no Hydrogenobacter thermophilus and Hydrogenobac- information on the latter organism was found in the ter halophilus is 2-methylthio-3-VI,VII-tetrahydro- literature. heptaprenyl-1,4-naphthoquinone. In VM1T, one major and one minor compound were present, neither DISCUSSION of which co-chromatographed with the compounds present in Hydrogenobacter acidophilus. Instead, Isolate VM1T, from a shallow marine hydrothermal VM1T appears to contain a novel menathioquinone, area of Vulcano Island, Italy, was placed within the the structure of which is currently not known.

International Journal of Systematic and Evolutionary Microbiology 51 1859 R. Sto$ hr and others

Taken together, these findings indicate that VM1T Description of Hydrogenobacter hydrogenophilus represents a new genus, for which we propose the name (Kryukov et al. 1984) comb. nov. Hydrogenothermus gen. nov., with the type species Basonym: Calderobacterium hydrogenophilum. Hydrogenothermus marinus sp. nov. The species description is identical to that given by Sequences of several uncultivated bacterial rDNA Kyrukov et al. (1983). The type strain is strain INMI T T T clones related to the phylum ‘Hydrogeno- Z-829 (l DSM 2913 l JCM 8158 ). bacter\Aquifex’ have been published or deposited in the EMBL database that originate from hydro- Description of Hydrogenobaculum (ex Shima and thermal environments and one thermophilic, Suzuki 1993) gen. nov. hydrogen-oxidizing isolate, EX-H1, has been reported (Reysenbach et al., 1994, 2000). We have included the Hydrogenobaculum (Hy.dro.ge.no.bahcu.lum. Gr. n. clones whose 16S rDNA sequence had been deter- hydor water; Gr. v. genein to produce; L. neut. n. mined up to a length of at least 1433 nucleotides in our baculum small rod; N.L. neut. n. Hydrogenobaculum phylogenetic analyses. The environmental DNAs water-producing small rod). Thermocrinis EM17 and SRI-48 were closely related to Gram-negative, non-sporulating rods. Motile. Res- ruber . Most other environmental clones, VC2.1 bac27, piratory metabolism; molecular oxygen is used as the OPB13, pBB, SRI-240, SRI-40 and the isolate EX-H1, T electron acceptor, hydrogen and reduced sulfur com- clustered together with VM1 (Fig. 4). These findings pounds as electron donors and carbon dioxide as a suggest a high diversity of the novel taxon described T carbon source. Carbon dioxide is fixed via the re- here. Furthermore, relatives of VM1 that might ductive tricarboxylic acid cycle. Requires elemental represent additional genera and species seem to be sulfur or thiosulfate for growth. The optimum pH for distributed widely in geothermally heated habitats. growth is 3–4. 2-Methylthio-3-VI,VII-tetrahydro- ( In addition to the description of a new species of multiprenyl -1,4-naphthoquinone (methionaquinone) thermophilic hydrogen oxidizer, we also included in is the major component of the quinone system. the 16S rDNA analysis all members of the genera The type species is Hydrogenobaculum acidophilum. The Hydrogenobacter, Aquifex, Calderobacterium and similarity of the sequence of 16S rDNA of Hydro- Thermocrinis, the names of which have been validly genobaculum acidophilum is around 86% to that of published (Fig. 4). In addition to showing that Hydro- Hydrogenothermus marinus and 88% to that of Aqui- genobacter acidophilus is related only distantly to fex pyrophilus. Hydrogenobacter thermophilus, our results also show that Thermocrinis ruber and Calderobacterium hydro- Description of Hydrogenobaculum acidophilum genophilum are closely related to Hydrogenobacter (Shima and Suzuki 1993) comb. nov. thermophilus. While Thermocrinis ruber differs in some respect from Hydrogenobacter thermophilus, and this Basonym: Hydrogenobacter acidophilus. still justifies its inclusion in a separate genus, the close Cells are short rods (0n4–0n6i1n3–1n8 µm) that occur relationship between the type strains of Hydrogeno- singly or in linked chains containing two to four cells. bacter thermophilus and Calderobacterium hydro- Polar flagellation. are membrane- genophilum indicates that the latter species may be bound and soluble. The soluble hydrogenases do not considered to be either a strain of the species Hydro- reduce pyrimidine nucleotides. Type b, c and o cyto- genobacter thermophilus or a distinct species within the chromes are found. The optimum temperature for genus Hydrogenobacter. Shima & Suzuki (1993) have growth is 65 mC. The GjC content of the DNA is 35 presented DNA–DNA hybridization data that indi- mol% (as determined by HPLC). Straight-chain satu- cate that the type strains of the species Hydro- rated C"):! and straight-chain unsaturated C"):" and genobacter thermophilus and Calderobacterium hydro- C#!:" acids are the major components of the cellular genophilum are not members of the same species, fatty acids. despite their high degree of 16S rDNA sequence Isolated from mud samples from a solfataric field similarity. Consequently, we propose that Caldero- in Tsumagoi, Japan. The type strain is strain 3H-1T bacterium hydrogenophilum should be transferred to ( JCM 8795T DSM 11251T). the genus Hydrogenobacter as a new combination, l l Hydrogenobacter hydrogenophilus comb. nov. Description of Hydrogenothermus gen. nov. In addition, 16S rDNA sequence analyses indicate that Hydrogenobacter acidophilus belongs to a lineage that Hydrogenothermus (Hy.dro.ge.no.therhmus. Gr. n. is distant from the Hydrogenobacter and Aquifex hydor water; Gr. v. genein to produce; Gr. adj. thermos cluster (Fig. 4). Furthermore, this strain can be dis- hot; N.L. masc. n. Hydrogenothermus hot and water tinguished phenotypically from all these strains by its producer). low pH optimum for growth, pH 3–4 (Shima & Suzuki, Cells are Gram-negative rods that occur singly, in 1993). We therefore propose to classify this organism pairs and in aggregates. Cells are motile by polarly in a new genus as Hydrogenobaculum acidophilum gen. inserted flagella. No spore formation. Respiratory nov., comb. nov. metabolism with oxygen as electron acceptor. Chemo-

1860 International Journal of Systematic and Evolutionary Microbiology 51 A novel hydrogen oxidizer lithoautotrophic with hydrogen as electron donor and Bonjour, F. & Aragno, M. (1986). Growth of thermophilic, CO# as source of cellular carbon. Elemental sulfur and obligatorily chemolithoautotrophic hydrogen-oxidizing thiosulfate are not utilized as electron donors. No bacteria related to Hydrogenobacter with thiosulfate and growth factors are required. Chemo-organotrophic elemental sulfur as electron and energy source. FEMS Microbiol growth is not found. Temperature optimum about Lett 35, 11–15. 65 mC. The major quinone is a menathioquinone, Burggraf, S., Olsen, G. J., Stetter, K. O. & Woese, C. R. (1992). A probably with a heptaprenyl side chain (i.e. MTK-7). phylogenetic analysis of Aquifex pyrophilus. Syst Appl Microbiol Acyl mono- and diethers are not present. The pre- 15, 352–356. dominant fatty acids present are C"):!,C"):"ω9c and Felsenstein, J. (1989).  – phylogeny inference package C#!:"ω9c. The polar lipids comprise phospholipids, a (version 3.2). Cladistics 5, 164–166. single aminophospholipid and glycolipids. The type Hayashi, N. R., Ishida, T., Yokota, A., Kodama, T. & Igarashi, Y. species is Hydrogenothermus marinus sp. nov. (1999). Hydrogenophilus thermoluteolus gen. nov., sp. nov., a thermophilic, facultatively chemolithoautotrophic, hydrogen- Description of Hydrogenothermus marinus sp. nov. oxidizing bacterium. Int J Syst Bacteriol 49, 783–786. Huber, R., Wilharm, T., Huber, D. & 7 other authors (1992). Hydrogenothermus marinus (ma.rihnus. L. adj. marinus Aquifex pyrophilus gen. nov., sp. nov., represents a novel group of marine origin). of marine hyperthermophilic hydrogen-oxidizing bacteria. Syst Cells are motile, Gram-negative rods, 2–4 µm long and Appl Microbiol 15, 340–351. 1–1n5 µm wide, with four to seven flagella inserted. The Huber, R., Eder, W., Heldwein, S., Wanner, G., Huber, H., Rachel, strain grows chemolithoautotrophically under an at- R. & Stetter, K. O. (1998). Thermocrinis ruber gen. nov., sp. nov., mosphere of H# and CO# (80:20) with low concen- a pink-filament-forming hyperthermophilic bacterium isolated trations of O# (0n5–8%, optimum at 1–2%). No from Yellowstone National Park. Appl Environ Microbiol 64, growth on meat peptone, tryptone, meat extract, yeast 3576–3583. extract, lactose, -galactose, α--glucose, -ribose, - Ishii, M., Kawasumi, T., Igarashi, Y., Kodama, T. & Minoda, Y. fructose, sucrose, citric acid, α--maltose hydrate, (1983). 2-Methylthio-1,4-naphthoquinone, a new quinone from starch, -xylose, -alanine, -proline, -histidine an extremely thermophilic hydrogen bacterium. Agric Biol hydrochloride, glycine, methanol, ethanol, acetic Chem 47, 167–169. acid, pyruvate, disodium fumarate, -malate or Ishii, M., Kawasumi, T., Igarashi, Y., Kodama, T. & Minoda, Y. ammonium formate as a source of energy or cellular (1987). 2-Methylthio-1,4-naphthoquinone, a unique sulfur- carbon. Catalase- and cytochrome-oxidase-positive. containing quinone from a thermophilic hydrogen-oxidizing No growth on peptone, yeast extract, tryptone, starch, bacterium, Hydrogenobacter thermophilus. J Bacteriol 169, glucose or maltose as sole electron donor. Sulfur and 2380–2384. thiosulfate are not utilized as electron donors or Jukes, T. H. & Cantor, C. R. (1969). Evolution of protein sources of energy. Elemental sulfur is necessary for molecules. In Mammalian Protein Metabolism, pp. 21–132. growth. Nitrate and sulfate cannot serve as electron Edited by H. N. Munro. New York: Academic Press. acceptors. Growth occurs over the range 45 to 80 mC Kawasumi, T., Igarashi, Y., Kodama, T. & Minoda, Y. (1984). with an optimum at 65 mC at pH 5–7. NaCl stimulates Hydrogenobacter thermophilus gen. nov., sp. nov., an extremely growth between 0n5 and 6%, with an optimum at thermophilic, aerobic, hydrogen-oxidizing bacterium. Int J Syst 2–3%. Isolated from sediment of a geothermally Bacteriol 34, 5–10. heated area, 3–4 m off the beach of Vulcano, Italy. The T T Kristjansson, J. K., Ingason, A. & Alfredsson, G. A. (1985). type strain is strain VM1 (l DSM 12046 l JCM Isolation of thermophilic obligately autotrophic hydrogen T 10974 ). oxidizing bacteria, similar to Hydrogenobacter thermophilus, from Icelandic hot springs. Arch Microbiol 140, 321–325. ACKNOWLEDGEMENTS Kryukov, V. R., Savel’eva, N. D. & Pusheva, M. A. (1983). Caldero- bacterium hydrogenophilum nov. gen., nov. sp., an extreme This paper is dedicated to Professor Karl-Otto Stetter on the thermophilic hydrogen bacterium, and its activity. occasion of his 60th birthday. Mikrobiologiya 52 This work was supported in part by the Fonds der , 781–788. Chemischen Industrie. We would like to thank T. A. Mesbah, M., Premachandran, U. & Whitman, W. B. (1989). Precise Langworthy for cell material of Thermodesulfobacterium measurement of the GjC content of deoxyribonucleic acid by commune as reference material for the presence of non- high-performance liquid chromatography. Int J Syst Bacteriol isoprenoid mono- and diethers, Hans Tru$ per for valuable 39, 159–167. advice and Uta Wehmeyer for technical assistance. Nishihara, H., Igarashi, Y. & Kodama, T. (1990). A new isolate of Hydrogenobacter, an obligately chemolithotrophic, thermo- REFERENCES philic, halophilic and aerobic hydrogen-oxidizing bacterium from seaside saline hot spring. Arch Microbiol 153, 294–298. Aragno, M. & Schlegel, H. G. (1992). The mesophilic hydrogen oxidizing (Knallgas) bacteria. In The , 2nd edn, pp. Olsen, G. J., Larsen, N. & Woese, C. R. (1991). The ribosomal 3917–3933. Edited by A. Balows, H. G. Tru$ per, M. Dworkin, RNA database project. Nucleic Acids Res 19 (Suppl.), W. Harder & K.-H. Schleifer. New York: Springer. 2017–2021. Balch, W. E., Fox, G. E., Magrum, L. J., Woese, C. R. & Wolfe, R. S. Reynolds, E. S. (1963). The use of lead citrate at high pH as an (1979). : re-evaluation of a unique biological electron-opaque stain in electron microscopy. J Cell Biol 17, group. Microbiol Rev 43, 260–296. 208–212.

International Journal of Systematic and Evolutionary Microbiology 51 1861 R. Sto$ hr and others

Reysenbach, A.-L., Wickham, G. S. & Pace, N. R. (1994). Phylo- from Yellowstone National Park. Int J Syst Evol Microbiol 51, genetic analysis of the hyperthermophilic pink filament com- 481–488. munity in Octopus Spring, Yellowstone National Park. Appl Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994).  Environ Microbiol 60, 2113–2119. : improving the sensitivity of progressive multiple sequence Reysenbach, A.-L., Banta, A. B., Boone, D. R., Cary, S. C. & Luther, alignment through sequence weighting, position-specific gap G. W. (2000). Microbial essentials at hydrothermal vents. Nature penalties and weight matrix choice. Nucleic Acids Res 22, 404, 835. 4673–4680. Shima, S. & Suzuki, K. (1993). Hydrogenobacter acidophilus sp. Tindall, B. J. (1990a). A comparative study of the lipid com- nov., a thermoacidophilic, aerobic, hydrogen-oxidizing bac- position of saccharovorum from various sources. terium requiring elemental sulfur for growth. Int J Syst Bacteriol Syst Appl Microbiol 13, 128–130. 43, 703–708. Tindall, B. J. (1990b). Lipid composition of Halobacterium $ $ Stohr, R., Waberski, A., Liesack, W., Volker, H., Wehmeyer, U. & lacusprofundi. FEMS Microbiol Lett 66, 199–202. Thomm, M. (2001). Hydrogenophilus hirschii sp. nov., a novel ZoBell, C. E. (1941). Studies on marine bacteria. I. The cultural thermophilic hydrogen-oxidizing β-proteobacterium isolated requirements of heterotrophic aerobes. J Mar Res 4, 42–75.

1862 International Journal of Systematic and Evolutionary Microbiology 51