International Journal of Systematic and Evolutionary Microbiology (2002), 52, 1171–1176 DOI: 10.1099/ijs.0.02185-0

Exiguobacterium undae sp. nov. and NOTE Exiguobacterium antarcticum sp. nov.

DSMZ–Deutsche Sammlung Anja Fru$ hling, Peter Schumann, Hans Hippe, Bettina Stra$ ubler von Mikroorganismen und Zellkulturen GmbH, and Erko Stackebrandt Mascheroder Weg 1b, D-38124 Braunschweig, Germany Author for correspondence: Erko Stackebrandt. Tel: j49 531 2616352. Fax: j49 531 2616418. e-mail: Erko!DSMZ.de

Four orange-pigmented strains from pond water (L1–L4) have been subjected to polyphasic taxonomic analyses. On the basis of ribotype analysis and Fourier-transform infrared spectroscopy, these strains form a genomically highly related group. 16S rDNA sequence analysis revealed 988% similarity between the 16S rDNA sequences of strains L2T and H2T, isolated previously from a microbial mat from Lake Fryxell, Antarctica. DNA–DNA reassociation values indicated the presence of two genomic clusters. While the DNA of strains L2T and L3 showed 100% DNA relatedness, strains L2T and H2T shared only 51% DNA relatedness. These two clusters differed in some phenotypic properties, e.g. utilization of melibiose, D-mannitol, adenosine 5'- monophosphate and uridine 5'-monophosphate, and in their fatty acid compositions. Based on the composition of isoprenoid quinones, peptidoglycan, polar lipids and fatty acids, these organisms are members of the genus Exiguobacterium. This is supported by 16S rDNA analyses, which revealed 97–98% similarity to Exiguobacterium acetylicum DSM 20416T and 932–938% similarity to Exiguobacterium aurantiacum DSM 6208T. E. acetylicum DSM 20416T, the closest phylogenetic neighbour, shows only 39% DNA similarity to strain L2T and 40% DNA similarity to strain H2T. Based on genomic distinctiveness and the clear differences in chemotaxonomy and physiology, two novel species are proposed, Exiguobacterium undae sp. nov. and Exiguobacterium antarcticum sp. nov.

Keywords: Exiguobacterium undae sp. nov., Exiguobacterium antarcticum sp. nov., riboprinting, FT-IR, polyphasic taxonomy

In a study applying modern, polyphasic approaches in were maintained in glycerol at k20 mC and preserved microbial taxonomy, surface water samples were taken by lyophilization. from a garden pond in Wolfenbuttel, Lower Saxony, $ Analysis of the 16S rDNA sequence of strain L2T and Germany, in June 2001 and streaked onto GS medium. phylogenetic comparison were done as described This medium (DSMZ catalogue no. 851; DSMZ, 1998) " previously (Rainey et al., 1996). The sequence of strain contained (l− )015 g glucose, 1 g yeast extract, 0 5g n n L2T is most similar (98 8% similarity) to that of the (NH ) SO ,01 g CaCO ,01 g Ca(NO ) ,005 g KCl, n % # % n $ n $ # n orange-pigmented strain H2T, originating from an 0 05 g K PO ,005 g MgSO .7H O, 0 20 g Na S.9H O n # % n % # n # # anaerobic enrichment of a microbial mat isolated from and 10 ml of a vitamin cocktail. Orange-pigmented Lake Fryxell, Antarctica (Brambilla et al., 2001). The colonies, developing within 48 h at room temperature, sequences of strains L2T and H2T were distantly related were isolated and purified on tryptone soy agar (Difco) to those of Exiguobacterium acetylicum DSM 20416T and four strains were designated strains L1–L4. Strains (97n2 and 97n9%, respectively) and Exiguobacterium T aurantiacum DSM 6208 (93n3 and 93n8%, respect- T T ...... ively). Strains L2 and H2 share 97n7% sequence Abbreviation: FT-IR, Fourier-transform infrared. similarity (Fig. 1a). The EMBL accession numbers for the 16S rDNA sequences of strains L2T (l DSM 14481T) and H2T (l DSM 14480T) are respectively AJ344151 and The genomic homogeneity of the pond isolates and AJ297437. relatives was analysed by riboprinting, using the

02185 # 2002 IUMS Printed in Great Britain 1171 A. Fru$ hling and others

T T (a) Pond strain L2T (AJ344151) L3 showed 100% relatedness to strain L2 , strain H2 92 T Lake Fryxell strain H2T (AJ297437) was more distantly related to strain L2 (51%). The 100 heterologous DNA reassociation values of strains L2T T T 100 E. acetylicum DSM 20416 (X70313) and H2 were only 39 and 40% when compared with T E. acetylicum IFO 12146T (D55730) E. acetylicum DSM 20416 . 100 E. aurantiacum DSM 6208T (X70316) Fourier-transform infrared (FT-IR) spectroscopy E. aurantiacum Z8 (AF275717) (Tindall et al., 2000) was performed in order to verify 2% the phenotypic similarity of the isolates. The data clearly support the notion that the four isolates L1–L4 T (b) Pond strain L4 form a coherent cluster, while strains H2 and the type Pond strain L1 strains of E. aurantiacum and E. acetylicum branch Pond strain L2T more distantly (Fig. 1b). Pond strain L3 Lake Fryxell strain H2T In order to determine whether the new isolates are E. aurantiacum DSM 6208T members of the genus Exiguobacterium, chemotaxo- E. acetylicum DSM 20416T nomic markers were determined. Compounds ana- lysed included cell wall amino acids (Schleifer & 12·0 9·0 6·0 3·0 0 Kandler, 1972), isoprenoid quinones (Collins et al., Euclidian distance 1977; Groth et al., 1996), polar lipids (Minnikin et al., (c) Pond strain L4 1979; Collins & Jones, 1980) and fatty acids (Miller & Pond strain L1 Berger, 1984). Membership of strains L2T and H2T of Pond strain L2T Pond strain L3 the genus Exiguobacterium was verified by the presence Lake Fryxell strain H2T of lysine as the diagnostic amino acid of the pepti- E. acetylicum DSM 20416T doglycan (Lys–Gly peptidoglycan type) (Farrow et al., E. aurantiacum DSM 6208T 1994) and MK-7, MK-8 and MK-6 as the principal quinones (76:15:2 for strain L2T, 71:20:2 for strain 60 45 30 15 0 T Spectral distance H2 ). Polar lipids are phosphatidylglycerol, diphos- phatidylglycerol, phosphatidylethanolamine, phos- ...... phatidylserine, phosphatidylinositol and an unidenti- Fig. 1. Dendrograms of relatedness among the novel isolates fied phospholipid. Phosphatidylserine and phosphati- and type strains of Exiguobacterium species. (a) Phylogeny of 16S rDNA as derived from distance-matrix analysis. Bar, 2% dylinositol were not reported in the species descriptions nucleotide deviation. Numbers at branching points are boot- of E. aurantiacum and E. acetylicum; however, strains strap percentages (1000 resamplings). (b) Sorting bacterial strains of the latter species contain phosphatidylserine (N. according to their FT-IR spectral distance. Cluster analysis was Weiss and P. Schumann, unpublished). All the isolates performed using the first derivatives of the spectra considering − − contained moderate amounts of iso (i)C"$:!, anteiso the spectral ranges 1200–900 cm 1 (3i), 900–700 cm 1 (1i), − 3000–2800 cm 1 (1i), scaling to first range, applying Ward’s (ai)C"$:!,C"':! and C"':"ω""c fatty acids (Table 1), but some differences were observed between strains L1–L4 algorithm (see Helm et al., 1991). (c) Fatty acid relationships. T The dendrogram was generated by treating the Euclidian dis- and strain H2 . The latter strain contains significantly tances of the fatty acids with the unweighted pair group method larger amounts of C"':" (c and C"):" *c fatty acids, with arithmetic average algorithm. Numerical analyses were ω ω done using the standard MIS software (Microbial ID). Statistical lacks C"):"ω(c, aiC"(:! and iC"':! fatty acids and has a procedures are described by Eerolen & Lehtonen (1988) and smaller amount of iC"(:!. E. aurantiacum shares more O’Donnell (1985). fatty acid similarity with the novel isolates than does E. acetylicum (Fig. 1c), but both type strains display significant differences: E. aurantiacum lacks iC"&:! and C"':"ω(c fatty acids, while E. acetylicum possesses automated RiboPrinter (DuPont), with EcoRI as the significant amounts of iC"&:! (Jones & Keddie, 1986) restriction enzyme. The typing profiles of strains and C"%:! fatty acids. A dendrogram of fatty acid L1–L4 were very similar, while those of the Exiguo- similarities is depicted in Fig. 1(c). bacterium strains and strain H2T were more dissimilar (Fig. 2). Cultural and physiological characteristics (Gordon, 1973; Gregersen, 1978) (API 50CHE Biolog GP DNA–DNA reassociation was performed to deter- Microplate; aminopeptidase test according to Merck mine the relatedness between strain L2T and strains L1 1.13301.0001) of the novel isolates (Tables 2 and 3) and L3, strain H2T and E. acetylicum DSM 20416T, also confirm the generic affiliation. The isolates show which appeared to be the closest relative as judged genus-specific characters such as rod-shaped mor- from 16S rDNA analysis. DNA was isolated as phology, motility, peritrichous flagellation and ab- described by Cashion et al. (1977) and Escara & sence of spore formation; they are facultatively an- Hutton (1980) and DNA–DNA reassociation, per- aerobic and catalase- and oxidase-positive. Discount- formed under optimal conditions (2iSSC at 68 mC) ing variable and weak reactions, the pond isolates (Huß et al., 1983; Jahnke, 1992), was recorded with a share the highest phenotypic similarity with strain H2T Gilford 2600 spectrophotometer. While strains L1 and (five differences), while the numbers of differences

1172 International Journal of Systematic and Evolutionary Microbiology 52 Novel Exiguobacterium species

kb 1·00 2·00 3·00 4·00 6·00 8·00 15·00 30·00 60·00

Pond strain L2T Pond strain L4 Pond strain L1 ...... E. aurantiacum DSM 6208T Fig. 2. Diversity of normalized EcoRI ribo- Lake Fryxell strain H2T type patterns found within the pond isolates, strain H2T and the type strains of the two T E. acetylicum DSM 20416 described Exiguobacterium species.

Table 1. Fatty acid composition of the novel isolates and the type strains of Exiguobacterium species ...... Only values " 1% are indicated; values & 5% are given in bold.

Fatty acid L1 L2T L3 L4 H2T E. aurantiacum E. acetylicum DSM 20416T DSM 6208T

iC"":! 2 iC"#:! 22223 3 C"#:! 12 1 iC"$:! 8 9 9 9 12 18 5 aiC"$:! 991091112 6 iC"%:! 22211 1

C"%:"ω&c 2 C"%:! 23322 3 13 iC"&:! 11 10 10 11 11 4 8 ai C"&:! 33332 1

C"':"ω""c 12 8 9 11 18 10 26 iC"':! 2222

C"':"ω&c 2 C"':"ω(c 87893 13 C"':! 12 17 16 11 13 27 10

C"(:"ω"!c 42233 1 iC"(:! 67795 6 1 aiC"(:! 1233

C"):"ω*c 5 3346 2 5 C"):"ω(c 3334 2 C"):! 4 664 55 1 Percentage of total 94 95 98 95 95 94 98 fatty acids identified

between these organisms and the type strains of the cies on the basis of genomic, phenotypic and metabolic two Exiguobacterium species, as well as those found properties. between the two type stains, are above 15. Based upon phylogenetic, chemotaxonomic and phy- Description of Exiguobacterium undae sp. nov. siological properties, the novel isolates belong to two Exiguobacterium undae (unhdae. L. fem. n. unda novel species that can be differentiated from each other waters; L. gen. n. undae of the waters). by their 16S rDNA sequences, riboprints (Fig. 2) and fatty acid patterns (Table 1; Fig. 1b) and some Surface colonies on tryptone soy agar (Difco) are metabolic properties, such as utilization of melibiose, 2–3 mm in diameter after 2 days at 25 mC, orange, -mannitol, adenosine 5h-monophosphate and uridine convex, entire and shiny. Orange pigment does not 5h-monophosphate (Tables 2 and 3). They also differ diffuse into the medium. Gram-positive, non-spore- from the two validly described Exiguobacterium spe- forming rods, varying in shape and size of rods from http://ijs.sgmjournals.org 1173 A. Fru$ hling and others

Table 2. Phenotypic properties that differentiate Exiguobacterium species and the novel isolates ...... Characteristics are scored as: k, negative; j, positive; , weak; , variable. All strains hydrolyse starch, casein, gelatin and DNA, they are motile and catalase-positive and they are negative for denitrification, nitrate reduction, urease aminopeptidase and lecithinase. Note that nitrate reduction was positive in the original description of E. aurantiacum (Collins et al., 1983). Utilization of compounds was determined with the Biolog GP Microplate panel. All strains utilized the following substrates: dextrin, amygdalin, arbutin, -fructose, gentiobiose, α--glucose, maltose, maltotriose, -mannose, 3-methyl glucose, α-ketovaleric acid, methyl β--glucoside, palatinose, -psicose, pyruvic acid, sucrose, trehalose, glycerol, adenosine, 2-deoxyadenosine, inosine, thymidine and uridine. The following substrates were not utilized by any strain: α-cyclodextrin, β-cyclodextrin, inulin, Tweens 40 and 80, -arabinose, -arabitol, -fucose, -galacturonic acid, -gluconic acid, m-inositol, α--lactose, lactulose, melizitose, -melibiose, methyl α--galactoside, methyl β--galactoside, methyl α--glucoside, methyl α--mannoside, sedoheptulosan, -rhamnose, turanose, salicin, stachyose, tagatose, xylitol, -xylose, α-hydroxybutyric acid, β-hydroxybutyric acid, p-hydroxyphenylacetic acid, α-ketoglutaric acid, alaninamide, lactamid, lactic acid methyl ester, -lactic acid, - and -malic acid, succinamic acid, succinic acid, N-acetyl -glutamic acid, -alanine, -asparagine, -glutamic acid, -pyroglutamic acid and putrescine.

Characteristic E. acetylicum E. aurantiacum Pond isolates Lake Fryxell DSM 20416T DSM 6208T L1–L4 isolate H2T

Oxidase jkjj Utilization of: Glycogen jkjj Mannan jj N-Acetyl glucosamine kjjj N-Acetyl mannosamine kk Cellobiose jkjj -Galactose kkjj -Mannitol jkjk -Raffinose kkjj -Ribose kkjj -Sorbitol jkkk Acetic acid  kjj γ-Hydroxybutyric acid kk Methylpyruvate kjkk Methylsuccinate kk Propionic acid kk\k (k)*  -Alanine jk -Alanyl glycine  k \k (k) j -Glycyl -glutamic acid kk\k (k) j -Serine  k \k (k)  2,3-Butanediol kk () j Adenosine 5h-monophosphate kkkj Thymidine 5h-monophosphate kk\j () j Uridine 5h-monophosphate kkkj Fructose 6-phosphate jkkk Glucose 1-phosphate jkkk Glucose 6-phosphate jkkk -α-Glycerol phosphate jkkk

* Reactions in parentheses refer to the type strain, L2T.

1n2i2n5–3n7 µm in the exponential phase to 1n1i1n5 dylglycerol, phosphatidylglycerol, phosphatidylserine, µm in the stationary phase. Motile (peritrichous phosphatidylinositol, phosphatidylethanolamine and flagella). Growth occurs under aerobic and anaerobic an unidentified phospholipid. Major fatty acids (" 5% conditions in both CASO and BBL medium. Growth in all strains) are iC"$:!, aiC"$:!,iC"&:!,C"':"ω""c, occurs at 20 and 41 mC but not at 45 mC; optimal C"':"ω(c,C"':! and iC"(:!; minor components are temperature for growth about 37 mC. Phenotypic listed in Table 1. Isolated from a garden pond in properties are listed in Tables 2 and 3. The pepti- Wolfenbu$ ttel, northern Germany. doglycan type is lysine–glycine. The principal isopre- T T noid quinone is MK-7; MK-6 and MK-8 occur in The type strain is strain L2 (l DSM 14481 l CIP smaller amounts. Polar lipids consist of diphosphati- 107162T). Strain L3 has been deposited as DSM 14482.

1174 International Journal of Systematic and Evolutionary Microbiology 52 Novel Exiguobacterium species

Table 3. Phenotypic properties that differentiate Exiguobacterium species and the novel isolates determined by API 50CHE ...... All strains were positive for -glucose, -fructose, N-acetylglucosamine, aesculin, salicin, maltose, sucrose, trehalose, starch, glycogen, β-gentiobiose. All strains were negative in the following reactions: erythritol, - and -arabinose, - and -xylose, adonitol, -sorbose, rhamnose, dulcitol, inositol, sorbitol, lactose, inulin, melezitose, xylitol, -turanose, -lyxose, -tagatose, -fucose, - and -arabitol, gluconate, 2-ketogluconate, 5-ketogluconate.

Test E. acetylicum E. aurantiacum Pond isolates Lake Fryxell DSM 20416T DSM 6208T L1–L4 isolate H2T

Glycerol kjjj Ribose kkjj Galactose kkjj -Mannose jkjj Mannitol jjjk Methyl α--glucoside kj (j)* k Amygdalin kj () j Arbutin kj () j Cellobiose jk (j) j Melibiose kkjk -Raffinose kk () k

* Reactions in parentheses refer to the type strain, L2T.

Description of Exiguobacterium antarcticum sp. nov. congratulate Lennart Stackebrandt (age 11) on the isolation of strains of a novel prokaryotic species in his first approach Exiguobacterium antarcticum (ant.archti.cum. N.L. to microbiology. gen. n. antarcticum of Antarctica). Surface colonies on tryptone soy agar (Difco) are References 2–3 mm in diameter after 2 days at 25 mC, bright orange, convex, entire and shiny. Orange pigment does Brambilla, E., Hippe, H., Hagelstein, A., Tindall, B. J. & Stacke- brandt, E. (2001). 16S rDNA diversity of cultured and uncultured not diffuse into the medium. Gram-positive, non- prokaryotes of a mat sample from Lake Fryxell, McMurdo Dry spore-forming rods, varying in shape and size of Valleys, Antarctica. Extremophiles 5, 22–33. rods from 1n2i3n0 µm in the exponential phase to Cashion, P., Holder-Franklin, M. A., McCully, J. & Franklin, M. 0n4i1n5 µm in the stationary phase. Motile (peritri- (1977). A rapid method for the base ratio determination of bacterial chous flagella). Growth occurs under aerobic and DNA. Anal Biochem 81, 461–466. anaerobic conditions in both CASO and BBL medium. Collins, M. D. & Jones, D. (1980). Lipids in the classification and Growth occurs at 20 and 41 mC but not at 45 mC; identification of coryneform bacteria containing peptidoglycans based J Appl Bacteriol optimal temperature for growth about 37 mC. Pheno- on 2,4-diaminobutyric acid. 48, 459–470. typic properties are listed in Tables 2 and 3. The Collins, M. D., Pirouz, T., Goodfellow, M. & Minnikin, D. E. (1977). Distribution of menaquinones in actinomycetes and corynebacteria. peptidoglycan type is lysine–glycine. The principal J Gen Microbiol 100, 221–230. isoprenoid quinone is MK-7; MK-6 and MK-8 occur Collins, M. D., Lund, B. M., Farrow, J. A. E. & Schleifer, K. H. in smaller amounts. Polar lipids consist of diphos- (1983). Chemotaxonomic study of an alkalophilic bacterium, Exiguo- phatidylglycerol, phosphatidylglycerol, phosphatidyl- bacterium aurantiacum gen. nov., sp. nov. J Gen Microbiol 129, serine, phosphatidylinositol, phosphatidylethanola- 2037–2042. mine and an unidentified phospholipid. Major fatty DSMZ (1998). Catalogue of Strains, 5th edn. Braunschweig: DSMZ. acids (" 5%) are iC"$:!, aiC"$:!,iC"&:!,C"':"ω""c,C"':! Eerolen, E. & Lehtonen, O.-P. (1988). Optimal data processing procedure for automating bacterial identification by gas-liquid chroma- and C"):"ω*c; minor components are listed in Table 1. Isolated from a microbial mat from Lake Fryxell, tography of cellular fatty acids. J Clin Microbiol 26, 1745–1753. Antarctica. Escara, J. F. & Hutton, J. R. (1980). Thermal stability and renaturation T T of DNA in dimethyl sulfoxide solutions: acceleration of the rena- The type strain is strain H2 ( l DSM 14480 l CIP turation rate. Biopolymers 19, 1315–1327. T 107163 ). Farrow, J. A. E., Wallbanks, S. & Collins, M. D. (1994). Phylogenetic interrelationships of round-spore-forming bacilli containing cell walls based on lysine and the non-spore-forming genera Caryophanon, Acknowledgements Exiguobacterium, Kurthia, and Planococcus. Int J Syst Bacteriol 44, 74–82. The excellent technical assistance and support of Maike Gordon, R. E., Haynes, W. C. & Pang, C. H. (1973). The Genus Steffen, Ina Kramer, Gabriele Gerich-Schro$ ter, Sabine Bacillus, Agricultural Handbook 427. Washington, DC: US Depart- Breymann and Ulrike Steiner is highly appreciated. We ment of Agriculture. http://ijs.sgmjournals.org 1175 A. Fru$ hling and others

Gregersen, T. (1978). Rapid method for distinction of Gram-negative Miller, L. & Berger, T. (1984). Bacterial identification by gas chroma- from Gram-positive bacteria. Eur J Appl Microbiol 5, 123–127. tography and whole cell fatty acids. Gas chromatography application Groth, I., Schumann, P., Weiss, N., Martin, K. & Rainey, F. A. note 228–41. Palo Alto, CA: Hewlett-Packard. (1996). Agrococcus jenensis gen. nov., sp. nov., a new genus of Minnikin, D. E., Collins, M. D. & Goodfellow, M. (1979). Fatty acid actinomycetes with diaminobutyric acid in the cell wall. Int J Syst and polar lipid composition in the classification of Cellulomonas, Bacteriol 46, 234–239. Oerskovia and related taxa. J Appl Bacteriol 47, 87–95. Helm, D., Labischinski, H., Schallehn, G. & Naumann, D. (1991). O’Donnell, A. G. (1985). Numerical analyses of chemotaxonomic Classification and identification of bacteria by Fourier-transform data. In Computer-assisted Bacterial Systematics, pp. 403–414. Edited infrared spectroscopy. J Gen Microbiol 137, 69–79. by M. Goodfellow, D. Jones & F. Priest. London: Academic Press. Huß, V. A. R., Festl, H. & Schleifer, K. H. (1983). Studies on the Rainey, F. A., Ward-Rainey, N., Kroppenstedt, R. M. & Stacke- spectrophotometric determination of DNA hybridization from rena- brandt, E. (1996). The genus Nocardiopsis represents a phylogenetically turation rates. Syst Appl Microbiol 4, 184–192. coherent taxon and a distinct actinomycete lineage: proposal of Jahnke, K.-D. (1992). BASIC computer program for evaluation of Nocardiopsaceae fam. nov. Int J Syst Bacteriol 46, 1088–1092. spectroscopic DNA renaturation data from GILFORD SYSTEM 2600 spectrophotometer on a PC\XT\AT type personal computer. Schleifer, K. H. & Kandler, O. (1972). Peptidoglycan types of bacterial J Microbiol Methods 15, 61–73. cell walls and their taxonomic implications. Bacteriol Rev 36, 407–477. Jones, D. & Keddie, R. M. (1986). Genus Brevibacterium Breed 1953, Tindall, B. J., Brambilla, E., Steffen, M., Neumann, R., Pukall, R., 13AL emend. Collins et al. 1980, 6. In Bergey’s Manual of Systematic Kroppenstedt, R. M. & Stackebrandt, E. (2000). Cultiveable mi- Bacteriology, vol. 2, pp. 1301–1313. Edited by P. H. A Sneath, N. S. crobial diversity: gnawing at the Gordian knot. Environ Microbiol 2, Mair, M. E. Sharpe & J. G. Holt. Baltimore: Williams & Wilkins. 310–318.

1176 International Journal of Systematic and Evolutionary Microbiology 52