B. Jochimsen et al.: Stetteria hydrogenophila 123Extremophiles (1998) 2:123–130 © Springer-Verlag 1998

ORIGINAL PAPER

Juan M. González · Yaeko Masuchi · Frank T. RobbJames W. Ammerman · Dennis L. MaederMiki Yanagibayashi · Jin Tamaoka · Chiaki Kato

Pyrococcus horikoshii sp. nov., a hyperthermophilic archaeon isolated froma hydrothermal vent at the Okinawa Trough

Received: December 10, 1997 / Accepted: February 4, 1998

Abstract A hyperthermophilic, anaerobic archaeon wasisolated from hydrothermal fluid samples obtained at theOkinawa Trough vents in the NE Pacific Ocean, at a depthof 1395m. The strain is obligately heterotrophic, and uti-lizes complex proteinaceous media (peptone, tryptone, oryeast extract), or a 21-amino-acid mixture supplementedwith vitamins, as growth substrates. Sulfur greatly enhancesgrowth. The cells are irregular cocci with a tuft of flagella,growing optimally at 98°C (maximum growth temperature102°C), but capable of prolonged survival at 105°C. Opti-mum growth was at pH 7 (range 5–8) and NaCl concentra-tion 2.4% (range 1%–5%). Tryptophan was required forgrowth, in contrast to the closely related strains Pyrococcusfuriosus and P. abyssi. Thin sections of the cell, viewed bytransmission electron microscopy, revealed a periplasmicspace similar in appearance to the envelope of P. furiosus.The predominant cell membrane component was tetraetherlipid, with minor amounts of diether lipids. Treatment ofthe cells by mild osmotic shock released an extract thatcontained a Zn21-dependent alkaline phosphatase. Phylo-genetic analysis of the sequences encoding 16S rRNA andglutamate dehydrogenase places the isolate with certaintywithin the genus Pyrococcus although there is relatively lowDNA–DNA hybridization (,63%) with described speciesof this genus. Based on the reported results, we propose anew species, to be named Pyrococcus horikoshii sp. nov.

Communicated by J. Wiegel

J.M. González · Y. Masuchi · F.T. Robb (*) · D.L. MaederCenter of Marine Biotechnology, Columbus Center, University ofMaryland Biotechnology Institute, 701 East Pratt Street, Baltimore,MD 21202, USATel. 181-410-234-8870; Fax 181-410-234-8896 e-mail:robb@umbi.umd.edu

J.W. AmmermanDepartment of Oceanography, TexasA&M University, College Station, TX 77843, USA

M. Yanagibayashi · J. Tamaoka · C. KatoThe Deepstar Group, Japan Marine Science and Technology Center2-15 Natsushima-cho, Yokosuka 237, Japan

Key words Archaea · Archaebacteria · Periplasmic ·Hyperthermophile · Thermococcales · Okinawa

Introduction

Since the discovery of hyperthermophilic microorganismsable to grow at temperatures close to or above 100°C(Fischer et al. 1979; Jannasch and Mottl 1982; Fiala andStetter 1986), great interest has been focused on the micro-biological and biochemical properties of these isolates(Adams 1993, 1994; Erauso et al. 1993). The existence ofcells that thrive at temperatures well above the extremesthat rapidly destroy critical purified cellular componentssuch as DNA, RNA, and enzymes implies the presence ofnovel and profound adaptations conferring stability in vivo.The hyperthermophiles do indeed have exceptionally stableproteins (Zillig et al. 1979; Adams 1993; DiRuggiero et al.1993; Adams and Kelly 1994; Jaenicke et al. 1996) andextremely active DNA repair systems (DiRuggiero et al.1997). In light of this, it is perhaps not surprising thatbiochemical features such as ADP-dependent glycolytickinases (Kengen et al. 1995) and tungsten-containingoxidoreductases (Adams and Kletzin 1994) have been dis-covered in the hyperthermophile Pyrococcus furiosus. Inaddition, based on 16S rRNA phylogenetic analysis (Woeseet al. 1990), and the sequence analysis of protein-encodinggenes, such as the H1-ATPase (Gogarten et al. 1989),glutamine synthetase (Brown et al. 1994), and tRNAsynthetases (Brown et al. 1997), the hyperthermophilicArchaea appear to occupy the deepest branches of variousphylogenetic trees. It is therefore likely that the the isolatewe describe here, in common with related extreme thermo-philes and hyperthermophiles, represents an ancestralevolutionary position.

Hyperthermophiles are normally isolated from areasundergoing vigorous hydrothermal activity, and they aresubject to varied geochemical challenges. Consequently,marine hydrothermal vents are normally colonized byhyperthermophilic strains representing diverse physiologi-

124 N. Matsuda et al.: EGF receptor and osteoblastic differentiation

cal types (reviewed by Jannasch et al. 1992). In this study,we report the isolation, by high-temperature enrichment, ofa heterotrophic strain from a deep-sea hydrothermal vent atthe Okinawa Trough in the northeastern Pacific Ocean. Wepropose that the strain be designated as a new species,Pyrococcus horikoshii.

Materials and methods

Sample collection

Hydrothermal vent samples were collected during a scien-tific cruise on board the R/V Natsushima. The manned sub-mersible Shinkai 2000 was used for sample collection. Thestrain presented herein were obtained from vent chimneysamples collected at the Okinawa Trough (27°339N,126°569E) at 1395m depth. Hot vent fluid at this locationwas at about 110°–130°C, measured by a thermistor probeinserted into the vent opening. This sampling site has beendescribed comprehensively by Fujioka et al. (1992).

Isolation and culture methods

Fluid samples were inoculated into the medium described inthe following paragraph, and incubated at 105°C overnightas an initial enrichment. Isolation was performed bygrowing successive ten-fold dilutions at 95°C into anaerobicmedia in Balch tubes, and repeatedly inoculating additionaldilution series with the final tube showing growth(end-point dilution). Aliquots from these cultures werecollected, stained with acridine orange, and observed byepifluorescence microscopy (Hobbie et al. 1977).

Unless otherwise mentioned, the medium used forgrowth and isolation contained (per liter): 24g NaCl, 4gNa2SO4, 0.7g KCl, 0.2g NaHCO3, 0.1g KBr, 0.03g H3BO3,10.6g MgCl2 ·6H2O, 1.5g CaCl2 ·2H2O, 0.025g SrCl2 ·6H2O,5g tryptone, 1g yeast extract, 1ml resazurin (0.2g/l solu-tion), 5–10g elemental sulfur. The medium was preparedunder a N2 atmosphere, and reduced by the addition of 3mlNa2S ·9H2O [25% (w/v) solution, pH 7]. The pH wasadjusted at room temperature with correction for increasedtemperature of growth.

Optimal growth conditions were determined by varyingthe NaCl concentration or pH of the medium and by incu-bating the cultures at different temperatures. The growthresponse to pH was tested from 5 to 10, with phosphatebuffer. The pH was not observed to change significantlyduring growth.

Several potential energy substrates were tested: sucrose,glucose, maltose, citrate, glycerol, cellobiose, acetate, lac-tate, pyruvate, olive oil, complex proteinaceous substrates(yeast extract, peptone, beef extract), casamino acids, a21-amino-acid mixture, and CO2. Nonproteinaceous com-pounds were tested in the presence and absence of 0.01%(w/v) yeast extract, as well as in the presence and absence ofa vitamin solution (Balch et al. 1979). In the cultures withcasamino acids and amino acid mixtures, yeast extract

was omitted. Nonproteinaceous substrates and amino acidswere supplemented at a final concentration of 0.2% (w/v).Because casamino acids (Difco, Detroit, MI, USA) lackstryptophan, asparagine, and glutamine due to acid treat-ment during production (Moore and Stein 1963), they wereadded or omitted in order to study the requirements forthese three amino acids.

The growth yield was determined in the standardmedium with and without elemental sulfur. The effect of H2

on growth was also studied by comparing the growth yieldof cultures (20ml medium/serum bottle) under an atmo-sphere of N2 or H2.

Determination of growth kinetics

Growth was determined by direct cell count using ahemocytometer and phase-contrast microscopy at a magnifi-cation of 4003. Samples were preserved with glutaraldehydeand counted within 24h. Growth rates (µ; h21) were obtainedfrom a regression line of ln N plotted against t, where N is thenumber of cells and t is the incubation time. A minimum ofthree data points was used for the regression analyses.

Electron microscopy

Samples were fixed with glutaraldehyde and platinum-shadowed using the procedures of Fiala and Stetter (1986).Thin sections were obtained according to the method de-scribed by Miroshnichenko et al. (1989). The samples wereobserved and photographed using a JEOL JEM-1210 elec-tron microscope.

DNA analysis

DNA was extracted from overnight cultures following themethod described by Charbonnier and Forterre (1995). The16S ribosomal RNA gene was amplified by polymerase chainreaction (PCR) using primers designed by Schmidt et al.(1991), and sequenced. The gdhA gene encoding glutamatedehydrogenase was amplified by using oligonucleotide prim-ers complementary to the P. furiosus gdhA gene (Vetriani etal., to be published) and purified by electrophoresis on a1.2% agarose gel. The DNA band containing the gdhA genewas excised, cloned into the TA vector (Invitrogen CA,USA), and sequenced on an ABI 373A automated se-quencer, using internal sequencing primers. The G 1 Ccontent of DNA was determined as reported by Tamaokaand Komagawa (1984) and from estimates of the meltingpoint (Marmur and Doty 1962) using a Model DU640 spec-trophotometer equipped with a high performance tempera-ture controller (Beckman Instruments, Brea, CA, USA).DNA–DNA hybridization was performed as described byKobayashi et al. (1994). The following reference specieswere used in the DNA–DNA hybridization: Pyrococcusfuriosus (DSM 3638), Pyrococcus abyssi (strain GE5),Thermococcus celer (DSM 2476). Alignment of the se-quences was carried out using the program PILEUP, in theGCG software package (Program manual for Wisconsin

B. Jochimsen et al.: Stetteria hydrogenophila 125

package, 1996). The phylogenetic tree diagrams were gener-ated by the PHYLIP suite of programs (Kuhner andFelsenstein 1994).

Osmotic shock and alkaline phosphatase assay

The procedure was derived from the method originallydescribed for osmotic shock of Escherichia coli cells (Neuand Heppel 1965). Cells were pelleted at 5000 3 g for15min, the supernatant discarded, and the pellet gently re-suspended in an osmotic shock solution containing 10% or20% sucrose, 50mM Tris buffer, and 1mM ethylene-diaminetetraacetic acid (EDTA), pH 8.0. The suspensionwas centrifuged again, the supernatant discarded, and thepellet was resuspended gently in 1/10 volume of ice-cold 0.5mM MgCl2. Finally, the suspension was cen-trifuged and the clear supernatant, termed shock fluid, wascarefully removed with a pipet. Both the shock fluid and theremaining pellet were frozen for later use in enzyme assays.

Alkaline phosphatase activity was assayed in 100mMglycine buffer, pH 9, which also included 1mM MgCl2

and 1mM ZnCl2. The substrate used was 6mM p-nitrophenyl phosphate (PNPP). Activity was assayed in arecording Beckman double-beam spectrophotometer witha temperature-controlled automated sample changer at awavelength of 410nm and a temperature of 80°C. At thistemperature there was significant hydrolysis of substrate inreagent blanks without added enzyme extract, and thesecontrol values were subtracted from enzyme assays. In atemperature stability experiment, shock fluid was incubatedat 97°C as previously described (Robb et al. 1992). Proteinwas determined by the Pierce method using the standardprotocol and bovine serum albumin (BSA) as a standard.

Lipid analysis

Core ether lipids were analyzed by the whole cell acidmethanolysis method (De Rosa and Gambacorta 1988).Lyophilized cells (100mg) were mixed with 3.5ml toluene,3.5ml of methanol, and 0.1ml of concentrated H2SO4 in areaction tube. The tube was heated at 50°C overnight, thencore lipids were extracted with 1.5ml hexane. The samplewas evaporated and the residue was redissolved in a smallamount of chloroform. The samples were analyzed by thinlayer chromatography on silica gel plates (Merck Kieselgel60; Merck, NJ, USA) developed in petroleum ether (boilingpoint 50°–80°C)–diethyl ether (85 :15, v/v). The lipids re-solved on the plates were visualized with 10% (v/v)phosphomolybdic acid in ethanol.

Nucleotide sequence accession numbers

The amplified 16S rRNA sequence reported in this paper hasbeen deposited in the DDBJ, EMBL, and GenBank nucle-otide sequence databases. The accession number of theDNA sequence corresponding to the 16S rDNA of Pyroco-ccus horikoshii is D87344. The accession number of the gdhAgene encoding glutamate dehydrogenase is AF035935.

Results

Morphology and fine cell structure

The cells appeared to be slightly irregular cocci with a diam-eter of 0.8–2µm (depending on growth conditions). Duringthe exponential phase of growth, they are often seen inpairs, and they aggregate increasingly as the culture ages.Cells in apparently dividing states are commonly seen byepifluorescence and electron microscopy. Electron micros-copy revealed a polar tuft of flagella. Thin sections showeda complex cell envelope formed by a cytoplasmic mem-brane and an external electron-dense layer enclosing aperiplasmic space (Fig. 1).

Growth conditions

The generation times of P. horikoshii incubated at differenttemperatures, and cultured at various pH and NaCl concen-trations, are shown in Fig. 2. Growth was observed from80°C to 102°C. The optimum growth temperature was 98°C.Growth did not occur at temperatures higher than 102°C;however, the strain showed prolonged survival during incu-bation at 105°C which was the condition used for enrich-

Fig. 1. Transmission electron micrograph of a thin section ofPyrococcus horikoshii. Length of bar: 0.1µm

126 N. Matsuda et al.: EGF receptor and osteoblastic differentiation

ment culture. Cultures were judged to be viable by ourability to inoculate successfully using cultures previouslyincubated for 24h at 105°C.

P. horikoshii grows optimally between pH 6 and 8, andpH 7 gave the fastest growth. No growth was obtained ata pH higher than or equal to 8.5. No growth was observedat pH lower than 5 (see Fig. 2).

NaCl concentrations from 1% to 5% allowed growth;however, there was a sharp optimum at 2.4% NaCl. Table 1summarizes these growth characteristics.

Nutritional requirements

Growth was observed on peptone, tryptone, yeast extract,beef extract, and non-hydrolyzed casein. No growth wasdetected in mineral salts medium amended with CO2 or N2

gas phases.Maximum growth rates were obtained in the presence

of elemental sulfur. However, medium lacking elementalsulfur supported growth at much lower growth rates andyields. In the presence of elemental sulfur, a H2 gas phase

Fig. 2. Influence of temperature, NaCl concentration, and pH on the growth rate of the isolate. The generation times were calculated from theslopes of exponential portions of complete growth curves (not shown)

Table 1. Comparison among Pyrococcus species: P. furiosus, P. abyssi, P. endeavori, and P. horikoshii

P. furiosus P. abyssi P. endeavori P. horikoshii

Morphology Slightly irregular cocci Slightly irregular cocci Slightly irregular cocci Slightly irregular cocciGrowth conditions

Temperature (°C)Optimum 100 96 98 98Range 70–103 67–102 80–110 ,80–102

pHOptimum 7 6.8 7 7Range 5–9 4–8.5 4–8 5–8

Salt concentration (%)Optimum 2 3 2.8 2.4Range 0.5–5 0.7–5 1–5

Substrates (C sources)Complex substrates Yes Yes Yes Yes(i.e., yeast extract,peptone, etc.)Maltose Yes Yes n/a NoStarch Yes Yes No NoPyruvate Yes Yes n/a NoCasamino Acids Very weak Yes Yes No21-Amino-acid mixture Yes Yes Yes Yes

Other requirementTryptophan No No n/a YesGrowth with S° Enhanced Enhanced Enhanced Enhanced

Doubling time (min) 37 33 60 32G 1 C content (mol %) 38 45 55 44Origin of type strain Vulcano Island North Fiji Basin Juan DeFuca Ridge Okinawa TroughDepth (m) 0 2000 2200 1395

B. Jochimsen et al.: Stetteria hydrogenophila 127

did not have any effect on growth rate and yield. This indi-cated that the formation of H2S may effectively “detoxify”H2 as proposed by Fiala and Stetter (1986). Several sub-strates were tested as sole carbon and energy sources, andalso in combination with dilute complex medium to deter-mine whether these compounds enhanced growth. The fol-lowing substrates were tested: sucrose, glucose, maltose,cellobiose, citrate, pyruvate, glycerol, lactate, acetate, andolive oil. None of the substrates tested stimulated growth,or supported growth as sole carbon sources.

The isolate grows rapidly on a mixture of 21 amino acids.However, it failed to grow on casamino acids. Sincecasamino acids (Difco) lacks the amino acids Trp, Gln, andAsn due to acid hydrolysis during production, the mediumwith casamino acids was supplemented with these threeamino acids. This supplemented casamino acids mediumsupported growth; however, the strain was unable to growin the same medium containing casamino acids but supple-mented only with Gln and Asn. Growth was also observedwhen the medium containing casamino acids was supple-mented only with Trp. This confirms the Trp requirement.These experiments were repeated with P. furiosus using thesame media, and the results confirmed that P. furiosus isprototrophic with respect to tryptophan.

DNA and phylogenetic analysis

A total of 1463 nucleotides of the 16S rRNA sequence fromthe isolate were determined. The gdhA gene encoding theglutamate dehydrogenase (GDH) was sequenced and thededuced amino acid sequence of the GDH from the isolatewas aligned with strongly homologous sequences fromrelated members of the Thermococcales. Phylogenetic trees

derived from the alignments of the 16S rRNA and theGDHs are shown in Fig. 3a and Fig. 3b respectively. Therepresentatives of the genus Pyrococcus, namely P.furiosus, and P. abyssi, were closely related to the newisolate, compared with the Thermococcus spp. (Godfroy etal. 1997; Gonzalez et al. 1995). The divergence between theGDH sequences from the new isolate and P. endeavori(ES4) is considerably less than the distance separating itfrom P. furiosus and P. abyssi. DNA–DNA hybridizationexperiments showed relatively low similarity betweenPyrococcus species and the new strain. P. furiosus has thehighest hybridization score with the new isolate (Table 2).The G 1 C content of the DNA of the isolate was deter-mined by the melting point method, 44%, and by directanalysis of the base composition, 42%.

Lipid analysis

In common with other Archaea, the isolate has ether lipidmembranes. The core lipids were separated by thin layerchromatography. During growth at high temperature, themembrane contains predominantly tetraether lipids, with asmall proportion of diether lipids, as shown in Fig. 4.

Alkaline phosphatase extraction and assay

Assays of the alkaline phosphatase and GDH in totalcell extracts or in osmotic shock preparations were linearfor 3–5 minutes at 80°C and the initial rate of reaction wasproportional to the amount of extract added. Alkalinephosphatase activity released by osmotic shock was 75% ofthe total activity found in unshocked cells after breakage.

Fig. 3. Phylogenetic trees representing the relatedness of P. horikoshiito other Archaea. a Unrooted tree of 16S rRNA sequences from re-lated Euryarcheota, using Maximum Parsimony. P. Pyrococcus; T.Thermococcus; M. jannaschii, Methanococcus jannaschii; M. fervidus,

Methanothermus fervidus. b Phylogenetic analysis of deduced aminoacid sequences of glutamate dehydrogenase (GDH), using Fitch-Margoliash analysis using T. litoralis as the outgroup. Abbreviations asin a

a b

128 N. Matsuda et al.: EGF receptor and osteoblastic differentiation

Table 2. DNA–DNA hybridization between members of the genusPyrococcus and Thermococcus celer

Species or strain % hybridization with the DNA of

JA-1 P. furiosus P. abyssi

P. horikoshii JA-1 100 50 32P. furiosus 63 100 36P. abyssi 36 45 100Thermococcus celer 25 50 36

Fig. 4. Distribution of diether and tetraether lipids in extracts frompurified membrane fractions of P. horikoshii

The activity of GDH in the osmotic shock fluid was 25% ofthe total activity, suggesting that although some cell break-age took place, alkaline phosphatase was differentiallyreleased, compared with glutamate dehydrogenase, a cyto-plasmic enzyme. These results are typical; however, thedifferential release of alkaline phosphatase was quite vari-able. In one experiment the released alkaline phosphataseactivity was 8.1-fold higher than the activity remaining withthe cells after shock treatment. The alkaline phosphataseactivity decreased exponentially during incubation at 97°Cwith a half life of 57min, as shown in Fig. 5.

Fig. 5. Exponential thermal inactivation of periplasmic alkaline phos-phatase at 96°C

Discussion

A novel hyperthermophile was isolated from hydrothermalvent fluid collected at the Okinawa Trough, in the NEPacific. The new strain is a chemoorganotroph withsimilar morphology to the previously described species ofPyrococcus. As summarized in Table 1, the new isolateshows clear physiological differences when compared withpreviously described Pyrococcus species (Fiala and Stetter1986; Erauso et al. 1993; Pledger and Baross 1991). ThePyrococcus spp. all have optimal growth temperaturesabove 90°C, whereas the existing Thermococcus speciesshow optimal growth below 90°C (Godfroy et al. 1997;Gonzalez et al. 1995).

The substrates utilized by the new strain showed thegreatest distinction, compared with P. furiosus (Fiala andStetter 1986) and P. abyssi (Erauso et al. 1993). Whereasthe latter two Pyrococcus species can utilize maltose andseveral other carbohydrates, especially cellobiose and pyru-vate, the isolate was unable to grow on nonpeptide carbonsources. In common with the other Pyrococcus species,growth was observed in minimal medium with an aminoacid mixture as the carbon and energy source. Detailedstudies have been done to determine the amino acidsrequired for growth by Pyrococcus species (Hoaki et al.1994; Watrin et al. 1995); these show that neither P. furiosusnor P. abyssi require tryptophan. The Okinawa isolate,however, definitely requires tryptophan for growth. Thearrangement of the tryptophan genes in this genus,and whether they are present, will be of great interest froman evolutionary perspective, since it was recently foundthat the tryptophan operon in P. furiosus has the samegene order as the tryptophan operons in Gramnegative bacteria (Borges et al. 1996; F. Robb and R. Weiss,unpublished).

B. Jochimsen et al.: Stetteria hydrogenophila 129

The membrane lipid composition of the new strainresembles P. abyssi, rather than P. furiosus, since the majorcomponent is the tetraether fraction, as determined usingtechniques similar to those described by Erauso et al. (1993)for the characterization of P. abyssi.

Morphological studies in Fig. 1 showed that the cellenvelope resembles that of P. furiosus, and is different fromthe cell envelope of P. abyssi. The appearance of a distinctperiplasmic space in thin sections prompted us to determinethe localization of a common extracellular enzyme, alkalinephosphatase, which is retained in the periplasmic space inGram-negative bacteria (Neu and Heppel 1965) and can bereleased by osmotic shock procedures. By using a similarprotocol to those used to release periplasmic binding pro-teins and enzymes from Gram negative bacteria, a cell-freeextract was obtained that was differentially enriched for aZn21-dependent alkaline phosphatase. This is the most ther-mostable alkaline phosphatase reported to date, and thesecond report of this enzyme in the Archaea as a whole.Alkaline phosphatase activity was purified from thehalophile Haloarcula marismortui (Goldman et al. 1990),although this may be a different class of enzyme to thealkaline phosphatase from this new hyperthermophile,since the halophile enzyme requires Ca21 rather than Zn21.These results and the appearance of an apparent spacebetween inner and outer layers of the cellular envelope(Fig. 1), indicate that a cellular compartment may existbetween the cytoplasmic membrane and the S-layer. Futurestudies on membrane uptake and exoenzyme production inthe hyperthermophilic Archaea may take advantage of thisfinding. The functions and cellular localization of manybinding proteins and catabolic enzymes, working in concertfor nutrient uptake, may be parallel in the Archaea and theBacteria. Little is known about membrane uptake systemsin the Archaea, but it was recently discovered thatthe hyperthermophiles P. furiosus and Thermococcuslitoralis have maltose uptake operons with a similar geneorder to the maltose regulons of Gram negative bacteria,including the malE gene encoding the periplasmic maltose-binding protein (DiRuggiero et al., to be published).Periplasmic enzymes and binding proteins may beimportant adaptive features in the hyperthermophiliceuryarchaeota.

The G 1 C content of DNA from the new isolate (44%)is higher than the value reported for P. furiosus, namely38% (Fiala and Stetter 1986) but it is similar to that of P.abyssi (45%) (Erauso et al. 1993). We have recently shownby extensive sequencing that a more precise value for the G1 C content of P. furiosus is actually 41% G 1 C (Borgeset al. 1996).

The gdhA gene encodes a key enzyme in nitrogenmetabolism, GDH, that has been extensively studied inthermophilic Archaea (Consalvi et al. 1991; Robb et al.1992; Kobayashi et al. 1992; DiRuggiero et al. 1993, 1996;DeVos et al. 1994). GDH ia a multimeric protein that hasbeen subject to extensive structural and molecular model-ling experiments in the hyperthermophilic Archaea. In thisstudy we used the gene sequence as a means to measuredivergence within the Thermococcales. A phylogenetic

tree, constructed using GDH sequences from P. furiosus, T.litoralis, and the new strain (Fig. 3), suggested that P.horikoshii has diverged more from P. furiosus than fromany of the other Pyrococcus species described to date, butthis divergence is less than that between Pyrococcus speciesand Thermococcus litoralis. It is interesting that the newisolate, although genetically more closely related to the Pa-cific species, P. abyssi and “P. endeavori”, is nonethelessquite similar to P. furiosus in growth physiology and mem-brane structure.

Alignment of the 16S rDNA sequence with known 16SrRNA sequences from related organisms (Pyrococcusfuriosus, Pyrococcus abyssi, Thermococcus celer) also indi-cates that the strain belongs to the genus Pyrococcus. DNA/DNA hybridization between P. furiosus, P. abyssi, and theisolate shows relatively low homology among them. Thisevidence, taken together with the nutritional characteris-tics, all indicate divergence from other members of theThermococcales, and establish that this is a new species. Wepropose to name a new species of the genus Pyrococcus:Pyrococcus horikoshii.

Description of Pyrococcus horikoshii sp. nov.

Pyrococcus horikoshii sp. nov. Gonzalez, Robb, and Kato.(ho.ri.ko.shi.i L.gen.n. horikoshii in honor of Dr. K.Horikoshi). Cells are slightly irregular cocci between 0.8and 2µm in diameter, showing a polar tuft of flagella. Thecell envelope consists of a complete S-layer enclosing aperiplasmic space around the cytoplasmic membrane. Obli-gate anaerobe. Growth occurs at temperatures between 80°and 102°C, with an optimum at 98°C. Optimum pH forgrowth is 7.0, and growth occurs from pH 5–8, withno growth observed at pH 8.5. NaCl concentration allowinggrowth ranges from 1% to 5%; optimum 2.4%. The shortestdoubling time is 32min. Elemental sulfur is not essential forgrowth but it greatly stimulates growth and results in highergrowth yield. Heterotrophic growth occurs on complex pro-teinaceous substrates (peptone, yeast extract, beef extract,tryptone, casein); on casamino acids supplemented withtryptophan in the presence of vitamins; and on a mixture of21 amino acids when vitamins are present. Tryptophan isrequired for growth. G 1 C content is 44mol%. The 16SrRNA gene sequence accession number is D87344. Inhabitsmarine hydrothermal vents in the western Pacific Ocean;specifically found in the Okinawa Trough vents at 27°339N,126°569E at a depth of 1395m.

Type strain: Pyrococcus horikoshii (JCM 9974deposited).

Acknowledgments We gratefully acknowledge the invaluableassistance of the crew of the RV Natsushima and the operators of theShinkai 2000 submersible for assistance with sampling. The support,assistance, and enthusiasm of Dr. K. Horikoshi was essential tothe collaboration that led to the isolation of this strain. JA acknowl-edges financial support from the Texas A&M Faculty DevelopmentLeave Program. FTR and JG acknowledge support by the USDeptartment of Energy, the National Science Foundation, and fromthe Lucille P. Markey Foundation. Dennis Maeder acknowledgesfinancial support from the Foundation for Research and Development,South Africa.

130 N. Matsuda et al.: EGF receptor and osteoblastic differentiation

References

Adams MW (1993) Enzymes and proteins from organisms that grownear and above 100°C. Annu Rev Microbiol 47:627–658

Adams MW (1994) Biochemical diversity among sulfur-dependent,hyperthermophilic microorganisms. FEMS Microbiol Rev 15:261–277

Adams MW, Kelly RM (1994) Thermostability and thermoactivity ofenzymes from hyperthermophilic Archaea. Bioorg Med Chem2:659–667

Adams MW, Kletzin A (1996) Oxidoreductase-type enzymes andredox proteins involved in fermentative metabolisms ofhyperthermophilic Archaea. Adv Protein Chem 48:101–180

Balch WE, Fox GE, Magrum LS, Woese CR, Wolfe RS (1979)Methanogens: reevaluation of a unique biological group. MicrobiolRev 43:260–296

Borges KM, Brummet S, Davis MF, Hujer K, Szasz J, Fuller C, ChaseJW, Robb FT (1996) A survey of the genome of the hyper-thermophile, Pyrococcus furiosus. Genome Sci Tech 1:37–46

Brown JR, Masuchi Y, Robb FT, Doolittle WF (1994) Evolutionaryrelationships of bacterial and archaeal glutamine synthetase genes. JMol Evol 38:566–576

Brown JR, Robb FT, Weiss R, Doolittle WF (1997) Evidence for theearly divergence of tryptophanyl- and tyrosyl-tRNA synthetases. JMol Evol 45:9–16

Charbonnier F, Forterre P (1995) Purification of plasmids from ther-mophilic and hyperthermophilic archaea. In: Robb FT, Place AR(eds) Archaea, a laboratory manual. Thermophiles. Cold SpringHarbor Laboratory Press, Cold Spring Harbor, New York, pp 87–90

Consalvi V, Chiaraluce R, Politi L, Vaccaro R, De Rosa M, ScandurraR (1991) Extremely thermostable glutamate dehydrogenase fromthe hyperthermophilic archaebacterium Pyrococcus furiosus. Eur JBiochem 202:1189–1196

De Rosa M, Gambacorta A (1988) The lipids of archaebacteria. ProgLipid Res 27:153–175

DiRuggiero J, Robb FT, Jagus RH, Klump H, Borges KM, Kessel M,Adams MW (1993) J Biol Chem 268:17767–17744

DiRuggiero J, Robb FT (1996) Enzymes of central nitrogen metabo-lism from hyperthermophiles: characterization, thermostability, andgenetics. Adv Protein Chem 48:311–339

DiRuggiero J, Santangelo N, Nackerdien Z, Ravel J, Robb FT (1997)Repair of extensive ionizing-radiation DNA damage at 95 degrees Cin the hyperthermophilic archaeon Pyrococcus furiosus. J Bacteriol179(14):4643–4645

Eggen RI, Geerling AC, Waldkotter K, Antranikian G, de Vos WM(1993) The glutamate dehydrogenase-encoding gene of thehyperthermophilic archaeon Pyrococcus furiosus: sequence, tran-scription, and analysis of the deduced amino acid sequence. Gene30:143–148

Erauso G, Reysenbach A-L, Godfroy A, Meunier J-R, Crump B,Partensky F, Baross JA, Marteinsson V, Barbier G, Pace NR, PrieurD (1993) Pyrococcus abyssi sp. nov., a new hyperthermophilicarchaeon isolated from a deep-sea hydrothermal vent. ArchMicrobiol 160:338–349

Fiala G, Stetter KO (1986) Pyrococcus furiosus sp. nov. represents anovel genus of marine heterotrophic archaebacteria growing opti-mally at 100°C. Arch Microbiol 145:56–61

Fujioka K, Matsuo Y, Nishimura A, Koyama M, Rodolfo KS (1992) 3.Tephras of the Izu-Bonin forearc (sites 787, 792, and 793). ProcOcean Drilling Program, College Station, Texas 126:47–74

Fischer F, Zillig W, Stetter KO, Schreiber G (1983) Chemoli-thoautotrophic metabolism of anaerobic extremely thermophilicarchaebacteria. Nature 301:511–513

Gogarten JP, Kibak H, Dittrich P, Taiz L, Bowman EJ, Bowman BJ,Manolson MF, Poole RJ, Date T, Oshima T (1989) Evolution of thevacuolar H1-ATPase: implications for the origin of eukaryotes. ProcNatl Acad Sci USA 86:6661–6665

Goldman S, Hecht K, Eisenberg H, Mevarech M (1990) ExtracellularCa21-dependent inducible alkaline phosphatase from extremelyhalophilic archaebacterium Haloarcula marismortui. J Bacteriol172:7065–7070

Godfroy A, Lesongeur F, Raguenes G, Querellou J, Antoine E,Meunier JR, Guezennec J, Barbier G (1997) Thermococcus

hydrothermalis sp. nov., a new hyperthermophilic archaeon isolatedfrom a deep-sea hydrothermal vent. Int J Syst Bacteriol 47:622–626

Gonzalez JM, Kato C, Horikoshi K (1995) Thermococcus pepto-nophilus sp. nov., a fast-growing, extremely thermophilic archae-bacterium isolated from deep-sea hydrothermal vents. ArchMicrobiol 164:159–164

Hoaki T, Nishijima M, Kato M, Adachi K, Mizobuchi S, Hanzawa N,Maruyama T (1994) Growth requirements of hyperthermophilicsulfur-dependent heterotrophic archaea isolated from a shallow sub-marine geothermal system with reference to their essential aminoacids. Appl Environ Microbiol 60:2898–2904

Hobbie JE, Dahney RJ, Jasper S (1977) Use of nucleopore filters forcounting bacteria by fluorescence microscopy. Appl EnvironMicrobiol 33:1225–1228

Jaenicke R, Schung H, Beaucamp N, Ostendorp R (1996) Structureand stability of hyperstable proteins: glycolytic enzymes fromhyperthermophilic bacterium Thermotoga maritima. Adv ProteinChem 48:181–269

Jannasch HW, Mottl J (1985) Geomicrobiology and deep-sea hydro-thermal vents. Science 229:717–721

Jannasch HW, Wirsen CO, Molyneux SJ, Langworthy TA (1992) Com-parative physiological studies on hyperthermophilic archaea isolatedfrom deep-sea hot vents with emphasis on Pyrococcus strain GB-D.Appl Environ Microbiol 58:3472–3481

Kengen SW, Tuininga JE, de Bok FA, Stams AJ, de Vos WM (1995)Purification and characterization of a novel ADP-dependent glu-cokinase from the hyperthermophilic archaeon Pyrococcus furiosus.J Biol Chem 270:30453–30457

Kobayashi T, Kwak YS, Akiba T, Kudo T, Horikoshi K (1994)Thermococcus profundus sp. nov., a new hyperthermophilicarchaeum isolated from a deep-sea hydrothermal vent. Syst ApplMicrobiol 17:232–236

Kuhner MK, Felsenstein J (1994) A simulation comparison of phylog-eny algorithms under equal and unequal evolutionary rates. MolBiol Evol 11:459–468

Marmur J, Doty P (1962) Determination of the base composition ofdeoxyribonucleic acid from its thermal denaturation temperature. JMol Biol 5:109–118

Miroshnichenko ML, Bonch-Osmolovskaya EA, Neuner A,Kostrikina NA, Chernych NA, Alevseev VA (1989) Thermococcusstetteri sp. nov., a new extremely thermophilic marine sulfur-metabolizing archaebacterium. Syst Appl Microbiol 12:257–262

Moore S, Stein WH (1963) Chromatographic determination of aminoacids by the use of automatic recording equipment. MethodsEnzymol 6:819–831

Neu HC, Heppel LA (1965) The release of enzymes from Escherichiacoli by osmotic shock and during the formation of spheroplasts. JBiol Chem 240:3685–3692

Robb FT, Park JB, Adams MW (1992) Characterization of an ex-tremely thermostable glutamate dehydrogenase: a key enzyme in theprimary metabolism of the hyperthermophilic archaebacterium,Pyrococcus furiosus. Biochim Biophys Acta 1120:267–272

Pledger RJ, Baross JA (1991) Preliminary description and nutritionalcharacterization of a chemoorganotrophic archaebacterium growingof up 110C isolated from a submarine hydrothermal vent environ-ment. J Gen Microbiol 137:203–211

Program manual for Wisconsin package (1996) Version 9, December1996, Genetics Computer Group, 575 Science Drive, Madison, WI53711, USA

Schmidt TM, DeLong EF, Pace NR (1991) Analysis of a marinepicoplankton community by 16S rRNA gene cloning and sequenc-ing. J Bacteriol 173:4371–4378

Tamaoka J, Komagawa K (1984) Determination of DNA base compo-sition by reversed-phase high-performance liquid chromatography.FEMS Microbiol Lett 25:125–128

Watrin L, Martin-Jezequel V, Prieur D (1995) Minimal amino acidrequirements of the hyperthermophilic archaeon Pyrococcus abyssi,isolated from deep-sea hydrothermal vents. Appl Environ Microbiol61:1138–1140

Woese CR, Kandler O, Wheelis ML (1990) Towards a natural systemof organisms: proposal for the domains Archaea, Bacteria, andEucarya. Proc Natl Acad Sci USA 87:4576–4579

Zillig W, Stetter KO, Janekovic D (1979) DNA-dependent RNA poly-merase from the archaebacterium Sulfolobus acidocaldarius. Eur JBiochem 96:597–604