International Journal of Systematic and Evolutionary Microbiology (2015), 65, Shizuoka-shi, Shizuoka , Japan

Transcription

1 International Journal of Systematic and Evolutionary Microbiology (2015), 65, DOI /ijsem Novel environmental species isolated from the plaster wall surface of mural paintings in the Takamatsuzuka tumulus: Bordetella muralis sp. nov., Bordetella tumulicola sp. nov. and Bordetella tumbae sp. nov. Nozomi Tazato, 1 Yutaka Handa, 1 Miyuki Nishijima, 1 Rika Kigawa, 2 Chie Sano 2 and Junta Sugiyama 3 Correspondence Junta Sugiyama jsugiyam@tecsrg.co.jp 1 Technical Department, TechnoSuruga Laboratory Co. Ltd, 330 Nagasaki, Shimizu-ku, Shizuoka-shi, Shizuoka , Japan 2 Independent Administrative Institution, National Research Institute for Cultural Properties, Tokyo, Ueno-Koen, Taito-ku, Tokyo , Japan 3 TechnoSuruga Laboratory Co., Ltd, Chiba Branch Office, No. 4 Sanko Bldg., Hasama-cho, Funabashi-shi, Chiba , Japan Ten strains of Gram-stain-negative, non-spore-forming, non-motile coccobacilli were isolated from the plaster wall surface of 1300-year-old mural paintings inside the stone chamber of the Takamatsuzuka tumulus in Asuka village (Asuka-mura), Nara Prefecture, Japan. Based on 16S rrna gene sequence analysis of the isolates, they belonged to the proteobacterial genus Bordetella (class Betaproteobacteria) and could be separated into three groups representing novel lineages within the genus Bordetella. Three isolates were selected, one from each group, and identified carefully using a polyphasic approach. The isolates were characterized by the presence of Q-8 as their major ubiquinone system and C 16 : 0 ( %), summed feature 3(C 16 : 1 v7c and/or C 16 : 1 v6c; %) and C 17 : 0 cyclo ( %) as the predominant fatty acids. The major hydroxy fatty acids were C 12 : 0 2-OH and C 14 : 0 2-OH. The DNA G+C content was mol%. DNA DNA hybridization tests confirmed that the isolates represented three separate novel species, for which the names Bordetella muralis sp. nov. (type strain T b T 5JCM T 5NCIMB T ), Bordetella tumulicola sp. nov. (type strain T b T 5JCM T 5NCIMB T ) and Bordetella tumbae sp. nov. (type strain T b T 5JCM T 5NCIMB T ) are proposed. These results support previous evidence that members of the genus Bordetella exist in the environment and may be ubiquitous in soil and/or water. The Takamatsuzuka tumulus, a circular burial mound measuring only 20 m across and located in Asuka village (Asuka-mura), Nara Prefecture, Japan, is thought to have been built in the late 7th to early 8th century and was designated a special historic site by the Agency for Cultural Affairs (ACA), Japan, in This tumulus (tomb) is well known in Japan because of the polychrome mural paintings Abbreviation: PNPG, 2-nitrophenyl b-d-galactopyranoside. The GenBank/EMBL/DDBJ accession numbers for the 16S rrna gene sequences of strains T b T, T b T and T b T are LC053647, LC and LC053656, respectively. Two supplementary tables and a supplementary figure are available with the online Supplementary Material. drawn directly on the thin plaster walls of the stone chamber interior (approx. 2.7 m deep, 1.0 m wide and 1.1 m tall). In particular, a painting of a group of four women on the west wall is commonly called the Asuka beauties (Asuka Bijin). All of the paintings were designated national treasures by the government in In accordance with specialist advice, the mural paintings in the stone chamber were kept under almost the same conditions as before the excavation, a low temperature (mean around 15 8C) and a high relative humidity (almost 100 %) (Ishizaki et al., 2004; Miura et al., 2005), and have never been open to the public (Kigawa et al., 2004). Nevertheless, the colourful mural paintings covering the side and ceiling walls inside the stone chamber have been damaged by biodeterioration (Kigawa et al., 2009; Ishizaki & Kigawa, 2011) G 2015 IUMS Printed in Great Britain

2 Three novel environmental Bordetella species The ACA tried all possible countermeasures to preserve the paintings and protect them against invasive micro-organisms. In September 2004, when the temperature increased above 20 8C, extensive white fungal mycelium, which covered part of the paintings, and mites were discovered (Kigawa et al., 2009). In June 2005, the ACA decided to relocate the chamber, and the relocation started in April Before the dismantling operation began, the stone chamber was cooled, as a precautionary measure, to reduce fungal/microbial activity (September 2005 April 2007) (Ishizaki et al., 2006). Thereafter, the stone chamber interior was kept at a temperature of approximately 10 8C and a relative humidity of 97 % (Inuzuka & Ishizaki, 2007). By mid-2007, all of the painted stone walls had been relocated to an outside facility in Asuka village for necessary restoration and to save them from further deterioration (Ishizaki & Kigawa, 2011). To elucidate the cause of biodeterioration of the mural paintings, we have been conducting microbiological surveys inside the Takamatsuzuka tumulus stone chamber since 2004 (Sugiyama et al., 2009). In this survey, we isolated 10 strains of the genus Bordetella from the plaster wall surface of the mural paintings as one of the predominant genera of bacteria in samples collected mainly during our 2006 sampling. As of March 2015, eight species of the genus Bordetella have been reported. Of those, seven species, including the type species Bordetella pertussis, are mammalian and avian parasites and pathogens (e.g. Weyant et al., 1996; Sanden & Weyant, 2005; Weiss, 2006). The remaining species in the genus, Bordetella petrii, was isolated from an anaerobic, dechlorinating bioreactor culture enriched from a river sediment, and it may be non-pathogenic (von Wintzingerode et al., 2001). However, since its initial isolation, B. petrii has been isolated not only from other environmental samples (e.g. Sfanos et al., 2005; Wang et al., 2007; Chowdhury et al., 2007), but also from many clinical samples (e.g. Fry et al., 2005; Stark et al., 2007; Le Coustumier et al., 2011). The whole-genome sequence of B. petrii DSM T was determined and compared with those of other virulent species of Bordetella (Gross et al., 2008). The results of genomic DNA analysis of B. petrii suggested that the species might have pathogenicity similar to that of host-restricted obligately pathogenic species of Bordetella, eventhoughit is able to thrive under environmental conditions (Gross et al., 2008). Our 10 isolates were closely related to B. petrii at the molecular and phylogenetic level, and they were also derived from environmental samples. Consequently, this is the second full description of members of the genus Bordetella that were isolated from the environment, following B. petrii (von Wintzingerode et al., 2001). In this article, we describe the taxonomic characterization of the isolates and propose three novel species in the genus Bordetella (family Alcaligenaceae, order Burkholderiales, class Betaproteobacteria, phylumproteobacteria). Because the Takamatsuzuka tumulus is a special historic site and, particularly, the mural paintings are designated national treasures, sampling was performed carefully under restricted conditions (Sugiyama et al., 2009). A total of nine samples collected from Takamatsuzuka tumulus between February 2006 and April 2007 were used for bacterial isolation (Table S1, available in the online Supplementary Material). Seven of the samples collected from the surface of mural paintings were cottonswab samples of mouldy spots and biofilms. One of the remaining two samples was a soil sample collected from a floor stone on the east side of the stone chamber interior, whereas another was collected as joint plaster fragments of a stone from the north wall when work to dismantle the stone chamber had just started in April Bacteria were isolated from the cotton-swab samples by the smear plate method on nutrient agar (CM3; Oxoid) plates. The other two samples were suspended in small amounts of sterile distilled water (2 3 ml) and spread over CM3 plates with cotton swabs. The plates were kept under aerobic conditions at 30 8C for 3 days. The colonies that appeared were transferred to new plates for the isolation of pure cultures. After purification of the bacterial colonies, isolated bacteria were grouped by morphological characteristics (colony shape, colour, viscosity, cell shape, Gram staining, spore formation and motility/gliding motility). 16S rrna gene sequence determination was then performed for representative isolates of each group. For preservation, cells of the isolates were grown on CM3 plates at 30 8C for 2 days, suspended in sterile distilled water supplemented with 10 % (v/v) glycerol and stored at 280 8C. The 10 isolates used in this study are detailed in Table S1. The type strains of the eight species of the genus Bordetella were used as references: these were B. pertussis DSM 5571 T (type species of the genus), B. avium DSM T, B. bronchiseptica NBRC T, B. hinzii JCM T, B. holmesii DSM T, B. parapertussis DSM T, B. petrii DSM T and B. trematum DSM T. These strains were cultured according to the methods recommended by the respective culture collections. The almost-complete 16S rrna gene sequences (1456 bp) of all 10 isolates were determined according to previously described methods (Tazato et al., 2012). Phylogenetic trees were reconstructed using the MEGA software, version 6.0 (Tamura et al., 2013), and two tree reconstruction algorithms, the neighbour-joining (Kimura, 1980; Saitou & Nei, 1987) and maximum-likelihood (Tamura & Nei, 1993; Nei & Kumar, 2000; Tamura et al., 2013) methods. The reliability of the clustering results was assessed by bootstrap analysis (Felsenstein, 1985). The two algorithms gave similar tree topologies and robust bootstrap values (Fig. 1). Comparison of the 16S rrna gene sequences by using the BLAST tool (Altschul et al., 1997) in DDBJ ( indicated that the isolates diverged phylogenetically within the genus Bordetella. The 10 isolates were separated into three groups by 16S rrna gene sequence similarity. The sequences of T b T, T b and T b were identical and most closely related to those of B. petrii DSM T (98.6 %

3 N. Tazato and others Bordetella tumulicola T b T (LC053650) Bordetella tumulicola T b (LC053651) Bordetella tumulicola T b (LC053652) Bordetella tumulicola T b (LC053653) Bordetella tumulicola T b (LC053655) Bordetella tumulicola T b (LC053654) Bordetella tumbae T b T (LC053656) 74 Bordetella muralis T b T (LC053647) 99 Bordetella muralis T b (LC053648) 96 Bordetella muralis T b (LC053649) Bordetella petrii DSM T (AM902716) `Bordetella ansorpii SMC-8986 (AY594190) Bordetella hinzii LMG T (AF177667) Bordetella bronchiseptica ATCC T (U04948) 81 Bordetella parapertussis ATCC T (U04949) 99 Bordetella holmesii CDCF 5101 T (U04820) 95 Bordetella pertussis ATCC 9797 T (U04950) Bordetella avium ATCC T (AF177666) 90 Bordetella trematum DSM T (AJ277798) Achromobacter ruhlandii ATCC T (AB010840) 99 Achromobacter xylosoxidans IAM T (D88005) 100 Alcaligenes aquatilis LMG T (JX986974) Alcaligenes faecalis NBRC T (AB680368) 58 Oligella urethralis CIP T (AF133538) Taylorella equigenitalis NCTC T (AF408197) Burkholderia cepacia ATCC T (AF097530) Fig. 1. Neighbour-joining tree based on nearly complete 16S rrna gene sequences (1375 nt) of the novel isolates and the type strains of related species. Bootstrap values (percentages of 1000 replications).50 % are shown at branches. Filled circles indicate that the corresponding nodes were recovered in a tree generated by the maximum-likelihood method with bootstrap values.70 %. Burkholderia cepacia ATCC T was used as an outgroup. Bar, 1 change per 100 nucleotide positions. similarity), B. avium ATCC T (98.2 %) and B. hinzii LMG T (98.1 %). In addition, six isolates, T b T, T b, T b, T b, T b and T b, had identical sequences and were closely related to B. petrii DSM T (98.8 %), B. hinzii LMG T (98.7 %), B. avium ATCC T (98.5 %) and B. trematum DSM T (98.5 %). Finally, the sequence of isolate T b T was closely related to those of B. petrii DSM T (98.5 %), B. hinzii LMG T (98.1 %) and B. avium ATCC T (98.1 %). The levels of 16S rrna gene sequence similarity between the novel isolates and representatives of the other genera in the family Alcaligenaceae, including the other species of the genus Bordetella, were below 97.9 %. The degree of sequence similarity among the three groups was %. The neighbour-joining phylogenetic tree depicted in Fig. 1 suggested that strains T b T (and two other isolates), T b T (and five other isolates) and T b T belonged to the same cluster but formed an independent subgroup that was tightly clustered away from the type strains of the previously known species of the genus Bordetella. In addition, the three groups of isolates were divided phylogenetically into three novel lineages and were placed in the same group as B. petrii DSM T, the first member of this genus isolated from the environment (von Wintzingerode et al., 2001), 4832 International Journal of Systematic and Evolutionary Microbiology 65

4 Three novel environmental Bordetella species with 74 % bootstrap support (Fig. 1). The 16S rrna gene sequence of T b T showed greater sequence similarity (.99 %) to unidentified isolates (arsenic-resistant bacterium R-8, GenBank accession no. JX130378; pyridinedegrading bacterium PDD-9, JQ394931). These isolates were derived from environmental samples, suggesting that members of the genus Bordetella might be ubiquitous in environments such as water and soil. Three isolates, T b T, T b T and T b T, one from each of the three different groups, were selected for subsequent experiments. The isolates were cultured at 30 8C on CM3 agar. For anaerobic cultivation, oxygen was absorbed using an Anaero Pack gas system (Anaero Pack disposable; Mitsubishi Gas Chemical). Observation of cell morphology, characterization of conventional phenotypic features and determination of the growth range, as well as optimal ph and temperature, were performed according to the procedures described by Barrow & Feltham (1993) and Tindall et al. (2007). API 20NE and API ZYM kits (biomérieux) were used for assessing biochemical features and enzyme activities according to the manufacturer s instructions, except that the incubation time in the API ZYM tests was extended to 24 h to obtain reliable results. For electron microscopy (JEM-2000FX; JEOL) observations, cells were grown at 37 8C for24honcm3andthennegatively stained with uranyl acetate. The three selected isolates were Gram-stain-negative, catalase- and oxidase-positive, non-spore-forming and non-motile. The three isolates were coccobacilli with no flagellum (Fig. S1). The isolates possessed a respiratory type of metabolism, growing aerobically and incapable of growth under anaerobic conditions. Colonies of the three isolates grown on CM3 plates were opaque and pale yellow. The isolates grew at ph , with an optimum of ph Strains T b T and T b T grew at C, with an optimum of C, whereas strain T b T was able to grow at C, with an optimum of C. None of the isolates produced a water-soluble brown pigment on heart infusion (HI) agar containing tyrosine (Weyant et al., 1996), and acid production from D-glucose was not observed. All three isolates were positive for reduction of nitrate to nitrite. Differentiation among the three isolates could be achieved by growth on Simmons citrate agar, urease activity, alkalinization of litmus milk, utilization of D-xylose, adipate, citrate, L-malate and phenylacetate and alkaline phosphatase activity. The characteristics that differentiate the three novel species proposed herein from each other and from their closest phylogenetic neighbours are summarized in Table 1. Other phenotypic features of the isolates are given in the species descriptions. Members of the genus Bordetella are generally known not to utilize carbohydrates such as glucose for growth (Weiss, 2006). Our three representative isolates showed similar phenotypic characteristics. However, they were distinguished from each other and from their phylogenetically closest relative, B. petrii, on the basis of the following phenotypic features: growth on Simmons citrate agar, urease activity, alkalinization of litmus milk, assimilation of D-xylose, adipate, citrate, L-malate and phenylacetate and alkaline phosphatase activity (Table 1). In addition, B. petrii was also different from the three novel isolates in the following characteristics: growth temperature, assimilation of D-glucose and growth on MacConkey agar. Our three isolates were able to grow weakly at 10 8C, whereas the closest relative B. petrii revealed no growth. As for other Bordetella species, available data on the growth at 10 8C are lacking. Our isolates were derived from environmental samples, while the other species of Bordetella, except B. petrii, were isolated from animals and humans as pathogens. The isoprenoid quinone content was determined in cells grown aerobically with shaking in nutrient broth at 30 8C for 24 h. Extraction and measurement of quinones were performed as described by Nishijima et al. (1997) using HPLC (Waters 600 series HPLC system). The three isolates possessed ubiquinone 8 (Q-8) only; no other ubiquinones or menaquinones were detected. Such quinone profiles are common among strains belonging to the family Alcaligenaceae (Busse & Auling, 2005; Austin, 2014). For cellular fatty acid composition determinations, cells of the three isolates and the reference type strains, except B. pertussis DSM 5571 T, were grown aerobically on trypticase soy agar (Becton Dickinson) plates at 37 8C for 24 h. B. pertussis DSM 5571 T, which is a fastidious bacterium, was grown on horse-blood-containing medium (per litre: 100 ml potato extract, 5.5 g NaCl, 10 g glycerol, 15 g agar, 750 ml distilled water, 150 ml horse blood, ph 6.7) at 37 8C for 24 h under aerobic conditions. Fatty acid methyl esters were prepared and analysed by GC on a 7890A GC system coupled with a 7683B Series injector (Agilent Technologies), according to the instructions for the Microbial Identification System (version 6.0), and identified using the TSBA6 database (MIDI). The major fatty acids of the three isolates were C 16 : 0 ( %), summed feature 3 (C 16 : 1 v7c and/or C 16 : 1 v6c; %) and C 17 : 0 cyclo ( %). The major hydroxy fatty acids were C 12 : 0 2-OH and C 14 : 0 2-OH (Table 2). The cellular fatty acid profiles of our isolates were consistent with the description of the family Alcaligenaceae provided by Busse & Auling (2005) and Austin (2014). Some differences in cellular fatty acid compositions were found among the three novel isolates (Table 2). DNA extraction, purification and determination of the G+C content of the three isolates were performed as described previously (Tazato et al., 2012). The G+C content of DNA of the three isolates was mol%. The isolates were distinguished by their genomic DNA G+C contents from B. petrii, B. hinzii, B. trematum, B. parapertussis, B. pertussis and B. bronchiseptica (66 70 mol%) (Vandamme et al., 1995, 1996). In particular, a difference in DNA G+C content of more than 6 mol% was observed between our three isolates and B. petrii and B. hinzii, which are the phylogenetically closest species. The genomic DNA G+C contents of the three isolates

5 N. Tazato and others Table 1. Differential characteristics of the three novel species and species of the genus Bordetella Strains/species: 1, B. muralis sp. nov. T b T ;2,B. tumulicola sp. nov. T b T ;3,B. tumbae sp. nov. T b T ;4,B. petrii (unless indicated, data from von Wintzingerode et al., 2001); 5, B. hinzii (Vandamme et al., 1995, 1996); 6, B. avium (Kersters et al., 1984; Vandamme et al., 1996; Weyant et al., 1996); 7, B. trematum (Vandamme et al., 1996); 8, B. holmesii (Weyant et al., 1995; Vandamme et al., 1996); 9, B. bronchiseptica (Vandamme et al., 1996; Weyant et al., 1996); 10, B. parapertussis (Vandamme et al., 1996; Weyant et al., 1996); 11, B. pertussis (Vandamme et al., 1996; Weyant et al., 1996). +, Positive;, negative; W, weakly positive; V, strain-dependent; ND, no data available. All taxa are negative for acid production from D-glucose and assimilation of D-glucose, lactose, sucrose, maltose and D-mannitol. Characteristic Motility Oxidase Growth on/at: MacConkey agar * Simmons citrate agar * V C W W W 2* ND ND ND ND ND ND ND 25 8C * V + V C * V V 2 Reduction of nitrate to V nitrite (API 20NE) Urease activity Brown pigment on HI agar with tyrosine Litmus milk alkalinization *D ND Assimilation of: L-Arabinose * 2 2 ND ND ND ND ND D-Xylose * D-Gluconate Caprate V 2 2 Adipate Citrate * ND + ND ND ND ND ND L-Malate V 2 2 Phenylacetate Enzyme activity Alkaline phosphatase * + V + + V 2 2 Esterase lipase (C8) * + V + + V + + Lipase (C14) * 2 2 V Chymotrypsin * 2 V V Valine arylamidase * + V Naphthol-AS-BIphosphohydrolase DNA G+C content (mol%) Isolation source Mural painting * * Mural painting Mural painting River sediment Chicken Turkey Chronic otitis media infection Human blood Mammalian respiratory tract Human respiratory tract Human respiratory tract *Data obtained in this study with strain DSM T. DShowed an acid curd reaction International Journal of Systematic and Evolutionary Microbiology 65

6 Three novel environmental Bordetella species Table 2. Cellular fatty acid compositions of the isolates and the type strains of species of the genus Bordetella Strains: 1, B. muralis sp. nov. T b T ;2,B. tumulicola sp. nov. T b T ;3,B. tumbae sp. nov. T b T ;4,B. petrii DSM T ;5, B. hinzii JCM T ;6,B. avium DSM T ;7,B. trematum DSM T ;8,B. holmesii DSM T ;9,B. bronchiseptica NBRC T ; 10, B. parapertussis DSM T ; 11, B. pertussis DSM 5571 T. Values are percentages of total fatty acids; fatty acids that represented less than 1 % of the total in all strains are not included.,,1 % or not detected. All data were obtained in this study by under the same culture conditions [trypticase soy agar (except B. pertussis DSM 5571 T ; see text) at 37 8C for 24 h] and the same analytical methods (MIDI system). Fatty acid Saturated C 12 : C 14 : C 16 : C 17 : C 17 : 0 cyclo C 18 : Unsaturated C 15 : 1 v6c 2.6 C 17 : 1 v5c 4.7 C 17 : 1 v7c 1.1 C 18 : 1 v9c 2.9 C 20 : 2 v6,9c 3.8 Branched iso-c 16 : iso-c 19 : iso-c 20 : anteiso-c 15 : anteiso-c 17 : anteiso-c 17 : 1 v9c 4.8 Hydroxylated C 10 : 0 3-OH C 12 : 0 2-OH C 12 : 0 3-OH 1.8 C 14 : 0 2-OH C 15 : 0 2-OH 1.3 C 16 : 0 2-OH 1.0 Summed features* *Summed features represent groups of two or more components that could not be separated by GLC with the MIDI system. Summed feature 2 comprises C 12 : 0 ALDE?, unknown , iso-c 16 : 1 I and/or C 14 : 0 3-OH. Summed feature 3 comprises C 16 : 1 v7c and/or C 16 : 1 v6c. Summed feature 5 comprises C 18 : 2 v6,9c and/or anteiso-c 18 : 0. Summed feature 8 comprises C 18 : 1 v7c and/or C 18 : 1 v6c. and the two remaining species of the genus Bordetella (60 63 mol%) (Kersters et al., 1984; Weyant et al., 1995) were more similar, but these values are lower than those of the above-mentioned six species. For DNA DNA hybridization experiments, genomic DNA of the three novel isolates, T b T, T b T and T b T, and of the most closely related type strain, B. petrii DSM T, was hybridized in microdilution wells using the fluorometric direct-binding method (Ezaki et al., 1989). The levels of DNA DNA relatedness measured were,42 % for all tested combinations (Table S2). These values clearly indicated that the three isolates, T b T, T b T and T b T, represented three novel genospecies within the genus Bordetella, based on the 70 % pragmatic and operational guideline for species delineation (Wayne et al., 1987). Seven species in the genus Bordetella, including the type species, are known as pathogenic bacteria for humans and animals (Weyant et al., 1996). Only one species, B. petrii, was isolated from the environment (river sediment), and was thought to be a non-pathogenic bacterium in the original description (von Wintzingerode et al.,

7 N. Tazato and others 2001). All of our isolates were found to be closely related to B. petrii, and they were also derived from environmental samples that were exposed to high humidity and low temperature (,15 8C). Generally, pathogenic members of the genus Bordetella are able to grow at temperatures around 37 8C and are fastidious, probably due to host parasite interactions. However, our isolates were able to grow at 10 8C and under non-fastidious nutrient conditions; they were isolated from the plaster wall surface of mural paintings that were thought to have provided poor nutrient support for bacterial growth and were kept at temperatures of approximately 10 8C. We did not obtain isolates of Bordetella from further surveys during dismantlement of the stone chamber. The 10 isolates belonging to the genus Bordetella were obtained from samples collected in 2006 and in April 2007 when the dismantlement work had just started, while the chamber was cooled to,10 8C. A similar survey was conducted in the Kitora tumulus, which is located near the Takamatsuzuka tumulus and also had mural paintings inside the stone chamber. However, we were unable to isolate strains of Bordetella from Kitora tumulus samples, although several other species, such as Gluconacetobacter asukensis, were commonly isolated from both tumuli (Sugiyama et al., 2009; Tazato et al., 2012; Nishijima et al., 2013). In the Lascaux cave in France, Palaeolithic paintings were surveyed for clinical bacteria by a culture-independent method, and few Bordetella-related clones were detected (Bastian et al., 2009). Partial sequences of the 16S rrna genes of clones were thought to be derived from the closely related Bordetella ansorpii, which isolated from a clinical sample by Ko et al. (2005) and proposed as a novel species of the genus Bordetella, but its name is still without standing in nomenclature. In conclusion, our results obtained using a polyphasic taxonomic approach support the recognition of three novel species of the genus Bordetella (family Alcaligenaceae, order Burkholderiales and class Betaproteobacteria), for which we propose the names Bordetella muralis sp. nov., Bordetella tumulicola sp. nov. and Bordetella tumbae sp. nov. This is the second record of members of the genus Bordetella existing in the environment, and thus they may be ubiquitous in soil and/or water. Description of Bordetella muralis sp. nov. Bordetella muralis (mu.ra9lis. L. fem. adj. muralis of a wall, referring to the isolation of the type strain from wall plaster). Cells are Gram-stain-negative coccobacilli measuring mm, non-motile, strictly aerobic, catalase- and oxidase-positive. Colonies are smooth, convex, entire, opaque, glistening and pale yellow, 2 3 mm in diameter on nutrient agar plates after 48 h of incubation at 30 uc. No growth factor is required. Grows on MacConkey agar and Simmons citrate agar. Litmus milk does not turn alkaline. A water-soluble brown pigment is not produced. Growth occurs at ph 5 10 (optimum ph 7 8) and uc (optimum uc). Acid is not produced from D-glucose. Positive reactions for reduction of nitrate to nitrite and utilization of L-arabinose, adipate, DL-malate, citrate and phenylacetate. Gives negative reactions for indole production, arginine dihydrolase activity, urease activity, aesculin hydrolysis, gelatin hydrolysis and b-galactosidase activity [2-nitrophenyl b-d-galactopyranoside (PNPG)]. Can use D-xylose as a sole carbon and energy source. The following substrates are not used: D-glucose, D-mannose, lactose, sucrose, maltose, N-acetyl-D-glucosamine, D-mannitol, D-gluconate and n- caprate. Enzyme activities (by API ZYM) of esterase (C4), leucine arylamidase, acid phosphatase and naphthol- AS-BI-phosphohydrolase are positive. Tests for alkaline phosphatase, esterase lipase (C8), lipase (C14), valine arylamidase, cystine arylamidase, trypsin, a-chymotrypsin, a-galactosidase, b-galactosidase, b-glucuronidase, a-glucosidase,b-glucosidase, N-acetyl-b-glucosaminidase, a-mannosidase and a-fucosidase activities are negative. The ubiquinone is Q-8 and the major fatty acids are C 16 : 0, summed feature 3 (C 16 : 1 v7c and/or C 16 : 1 v6c) and C 17 : 0 cyclo. The major hydroxy fatty acids are C 12 : 0 2-OH and C 14 : 0 2-OH. The type strain, T b T (5JCM T 5NCIMB T ), was isolated from a cotton-swab sample collected from the surface of mural paintings on the west wall of the four Asuka beauties (Asuka Bijin) inside the stone chamber of Takamatsuzuka tumulus in Asuka village, Nara Prefecture, Japan. The DNA G+C content of the type strain is 59.8 mol%. Description of Bordetella tumulicola sp. nov. Bordetella tumulicola [tu.mu.li9co.la. L. n. tumulus tumulus or tomb; L. suff.-cola (from L. n. incola) dweller; N.L. n. tumulicola tumulus dweller]. Cells are Gram-stain-negative coccobacilli measuring mm and non-motile, strictly aerobic, catalaseand oxidase-positive. Colonies are smooth, convex, entire, opaque, glistening and pale yellow, 1 2 mm in diameter on nutrient agar plates after 48 h of incubation at 30 uc. No growth factor is required. It grows on MacConkey agar but not Simmons citrate agar. Litmus milk turns alkaline. A water-soluble brown pigment is not produced. Growth occurs at ph 5 10 (optimum 7 8) and uc (optimum uc). Acid is not produced from D-glucose. Positive reactions for reduction of nitrate to nitrite and urease activity. Negative reactions for indole production, arginine dihydrolase activity, aesculin hydrolysis, gelatin hydrolysis and b-galactosidase activity (PNPG). No tested carbon sources (D-glucose, L-arabinose, D-xylose, sucrose, lactose, D-mannose, D-mannitol, N-acetyl-D-glucosamine, maltose, gluconate, n-caprate, adipate, DL-malate, citrate and phenylacetate) are utilized. Enzyme activities (by API ZYM) of alkaline phosphatase, esterase (C4), leucine arylamidase, acid phosphatase and naphthol-as-bi-phosphohydrolase 4836 International Journal of Systematic and Evolutionary Microbiology 65

8 Three novel environmental Bordetella species are positive. Tests for esterase lipase (C8), lipase (C14), valine arylamidase, cystine arylamidase, trypsin, a-chymotrypsin, a-galactosidase, b-galactosidase, b-glucuronidase, a-glucosidase, b-glucosidase, N-acetyl-b-glucosaminidase, a-mannosidase and a-fucosidase activities are negative. The ubiquinone is Q-8 and the major fatty acids are C 16 : 0, C 17 : 0 cyclo and summed feature 3 (C 16 : 1 v7c and/or C 16 : 1 v6c). The major hydroxy fatty acid is C 14 : 0 2-OH. The type strain, T b T (5JCM T 5NCIMB T ), was isolated from a black stain on the forehead of one of the Asuka beauties (Asuka Bijin) on the mural painting surface on the west wall inside the stone chamber of Takamatsuzuka tumulus in Asuka village, Nara Prefecture, Japan. The DNA G+C content of the type strain is 59.6 mol%. Description of Bordetella tumbae sp. nov. Bordetella tumbae (tum9bae. L. gen. fem. n. tumbae of a tomb, referring to the isolation source of the type strain). Cells are Gram-stain-negative coccobacilli measuring mm, non-motile, strictly aerobic, catalaseand oxidase-positive. Colonies are smooth, convex, entire, opaque, glistening and pale yellow, 2 3 mm in diameter on nutrient agar plates after 48 h incubation at 30 uc. No growth factor is required. Grows on MacConkey agar and Simmons citrate agar. Litmus milk does not turn alkaline. A water-soluble brown pigment is not produced. Growth occurs at ph 5 10 (optimum 7 8) and uc (optimum uc). Acid is not produced from D-glucose. Positive reactions for reduction of nitrate to nitrite, urease activity, utilization of L-arabinose, adipate, DL-malate, citrate and phenylacetate. Gives negative reactions for indole production, arginine dihydrolase activity, aesculin hydrolysis, gelatin hydrolysis and b-galactosidase activity (PNPG). The following substrates are not used as sole carbon and energy sources: D-glucose, D-mannose, D-xylose, lactose, sucrose, maltose, N-acetyl-D-glucosamine, D-mannitol, D-gluconate and n-caprate. Enzyme activities (by API ZYM) of alkaline phosphatase, esterase (C4), leucine arylamidase, acid phosphatase and naphthol-as-bi-phosphohydrolase are positive. Tests for esterase lipase (C8), lipase (C14), valine arylamidase, cystine arylamidase, trypsin, a-chymotrypsin, a-galactosidase, b-galactosidase, b-glucuronidase, a-glucosidase, b- glucosidase, N-acetyl-b-glucosaminidase, a-mannosidase and a-fucosidase activities are negative. The ubiquinone is Q-8 and the major fatty acids are C 16 : 0, summed feature 3 (C 16 : 1 v7c and/or C 16 : 1 v6c) and C 17 : 0 cyclo. The major hydroxy fatty acid is C 12 : 0 2-OH. The type strain, T b T (5JCM T 5NCIMB T ), was isolated from a black stain on the right shoulder of a woman on the mural paintings on the west wall inside the stone chamber of Takamatsuzuka tumulus in Asuka village, Nara Prefecture, Japan. The DNA G+C content of the type strain is 60.0 mol%. Acknowledgements This study was permitted by the Agency for Cultural Affairs, Japan. Part of this study was supported financially by Grants-in-Aid for Scientific Research (A) from the Japan Society for the Promotion of Science (JSPS) (no to C. S., ). We appreciate the curators and staff of JCM (Tsukuba, Japan) and NCIMB (Aberdeen, UK) for their prompt response in relation to the deposition of type and other strains. We are also grateful for the help of one of the anonymous reviewers who provided insightful comments and useful suggestions that improved the manuscript. References Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped blast and psi-blast: a new generation of protein database search programs. Nucleic Acids Res 25, Austin, B. (2014). The family Alcaligenaceae. InThe Prokaryotes, 4th edn, vol. 4, pp Edited by E. Rosenberg, E. F. DeLong, S. Lory, E. Stackebrandt & F. Thompson. Berlin: Springer. Barrow, G. I. & Feltham, R. K. A. (1993). Cowan and Steel s Manual for the Identification of Medical Bacteria, 3rd edn. Cambridge: Cambridge University Press. Bastian, F., Alabouvette, C. & Saiz-Jimenez, C. (2009). Bacteria and free-living amoeba in the Lascaux Cave. Res Microbiol 160, Busse, H.-J. & Auling, G. (2005). Family Alcaligenaceae De Ley, Segers, Kersters, Mannheim and Lievens 1986, 412 VP. In Bergey s Manual of Systematic Bacteriology, 2nd edn, vol. 2C, pp Edited by D. J. Brenner, N. R. Krieg, J. T. Staley & G. M. Garrity. New York: Springer. Chowdhury, S. P., Schmid, M., Hartmann, A. & Tripathi, A. K. (2007). Identification of diazotrophs in the culturable bacterial community associated with roots of Lasiurus sindicus, a perennial grass of Thar Desert, India. Microb Ecol 54, Ezaki, T., Hashimoto, Y. & Yabuuchi, E. (1989). Fluorometric deoxyribonucleic acid-deoxyribonucleic acid hybridization in microdilution wells as an alternative to membrane filter hybridization in which radioisotopes are used to determine genetic relatedness among bacterial strains. Int J Syst Bacteriol 39, Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, Fry, N. K., Duncan, J., Malnick, H., Warner, M., Smith, A. J., Jackson, M. S. & Ayoub, A. (2005). Bordetella petrii clinical isolate. Emerg Infect Dis 11, Gross, R., Guzman, C. A., Sebaihia, M., MartinsdosSantos, V. A. P., Pieper, D. H., Koebnik, R., Lechner, M., Bartels, D., Buhrmester, J. & other authors (2008). The missing link: Bordetella petrii is endowed with both the metabolic versatility of environmental bacteria and virulence traits of pathogenic Bordetellae. BMC Genomics 9, 449. Inuzuka, M. & Ishizaki, T. (2007). [Environmental measurement for dismantling the stone chamber of Takamatsuzuka tumulus]. Sci Conserv 46, , (in Japanese). Ishizaki, T. & Kigawa, R. (2011). Conservation of the mural paintings of the Takamatsuzuka and Kitora tumuli in Japan. In Lascaux and Preservation Issues in Subterranean Environments, Proceedings of the Symposium, pp Edited by N. Coye. Paris: Fondation Maison des sciences de l homme. Ishizaki, T., Miura, S., Inuzuka, M. & Khalil, M. (2006). [Examination and choice of cooling methods for the mound of Takamatsuzuka tumulus as protective measures against fungi in the stone chamber]. Sci Conserv 45, (in Japanese)

9 N. Tazato and others Ishizaki, T., Sano, C. & Miura, S. (2004). [Study of temperature and humidity in the stone chamber of Takamatsuzuka tumulus and water content profile in the surrounding mound]. Sci Conserv 43, 87 94, (in Japanese). Kersters, K., Hinz, K. H., Hertle, A., Segers, P., Lievens, A., Siegmann, O. & De Ley, J. (1984). Bordetella avium sp. nov., isolated from the respiratory tracts of turkeys and other birds. Int J Syst Bacteriol 34, Kigawa, R., Sano, C. & Miura, S. (2004). [Past and present situation of microorganisms in Takamatsuzuka tumulus]. Sci Conserv 43, 79 85, (in Japanese). Kigawa, R., Sano, C., Ishizaki, T., Miura, S. & Sugiyama, J. (2009). Biological issues in the conservation of mural paintings of Takamatsuzuka and Kitora tumuli in Japan. Study of environmental conditions surrounding cultural properties and their protective measures. In Study of Environmental Conditions Surrounding Cultural Properties and Their Protective Measures, pp Edited by C. Sano. Tokyo: National Institute of Cultural Properties. Kimura, M. (1980). A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J Mol Evol 16, Ko, K. S., Peck, K. R., Oh, W. S., Lee, N. Y., Lee, J. H. & Song, J. H. (2005). New species of Bordetella, Bordetella ansorpii sp. nov., isolated from the purulent exudate of an epidermal cyst. J Clin Microbiol 43, Le Coustumier, A., Njamkepo, E., Cattoir, V., Guillot, S. & Guiso, N. (2011). Bordetella petrii infection with long-lasting persistence in human. Emerg Infect Dis 17, Miura, S., Ishizaki, T. & Akamatsu, T. (2005). [Temperature changes in 30 years at Takamatsuzuka tumulus]. Sci Conserv 44, , (in Japanese). Nei, M. & Kumar, S. (2000). Molecular Evolution and Phylogenetics., New York: Oxford University Press. Nishijima, M., Araki-Sakai, M. & Sano, H. (1997). Identification of isoprenoid quinones by frit-fab liquid chromatography-mass spectrometry for the chemotaxonomy of microorganisms. J Microbiol Methods 28, Nishijima, M., Tazato, N., Handa, Y., Tomita, J., Kigawa, R., Sano, C. & Sugiyama, J. (2013). Gluconacetobacter tumulisoli sp. nov. Gluconacetobacter takamatsuzukensis sp. nov. and Gluconacetobacter aggeris sp. nov., isolated from Takamatsuzuka tumulus samples before and during the dismantling work in Int J Syst Evol Microbiol 63, Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, Sanden, G. N. & Weyant, R. S. (2005). Genus Bordetella Moreno-Lópes 1952, 178 AL. In Bergey s Manual of Systematic Bacteriology, 2nd edn, vol. 2C, pp Edited by D. J. Brenner, N. R. Krieg, J. T. Staley & G. M. Garrity. New York: Springer. Sfanos, K., Harmody, D., Dang, P., Ledger, A., Pomponi, S., McCarthy, P. & Lopez, J. (2005). A molecular systematic survey of cultured microbial associates of deep-water marine invertebrates. Syst Appl Microbiol 28, Stark, D., Riley, L. A., Harkness, J. & Marriott, D. (2007). Bordetella petrii from a clinical sample in Australia: isolation and molecular identification. J Med Microbiol 56, Sugiyama, J., Kiyuna, T., An, K.-D., Nagatsuka, Y., Handa, Y., Tazato, N., Hata-Tomida, J., Nishijima, M., Koide, T. & other authors (2009). Microbiological survey of the stone chambers of Takamatsuzuka and Kitora tumuli, Nara Prefecture, Japan: a milestone in elucidating the cause of biodeterioration of mural paintings. In Study of Environmental Conditions Surrounding Cultural Properties and Their Protective Measures, pp Edited by C. Sano. Tokyo: National Research Institute for Cultural Properties, Tokyo. Tamura, K. & Nei, M. (1993). Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evol 10, Tamura, K., Stecher, G., Peterson, D., Filipski, A. & Kumar, S. (2013). MEGA6: molecular evolutionary genetics analysis version 6.0. Mol Biol Evol 30, Tazato, N., Nishijima, M., Handa, Y., Kigawa, R., Sano, C. & Sugiyama, J. (2012). Gluconacetobacter tumulicola sp. nov. and Gluconacetobacter asukensis sp. nov., isolated from the stone chamber interior of the Kitora Tumulus. Int J Syst Evol Microbiol 62, Tindall, B. J., Sikorski, J., Smibert, R. M. & Krieg, N. R. (2007). Phenotypic characterization and the principles of comparative systematics. In Methods for General and Molecular Microbiology, 3rd edn, pp Edited by C. A. Reddy, T. J. Beveridge, J. A. Breznak, G. Marzluf, T. M. Schmidt & L. R. Snyder. Washington, DC: American Society for Microbiology. Vandamme, P., Hommez, J., Vancanneyt, M., Monsieurs, M., Hoste, B., Cookson, B., Wirsing von König, C. H., Kersters, K. & Blackall, P. J. (1995). Bordetella hinzii sp. nov., isolated from poultry and humans. Int J Syst Bacteriol 45, Vandamme, P., Heyndrickx, M., Vancanneyt, M., Hoste, B., De Vos, P., Falsen, E., Kersters, K. & Hinz, K. H. (1996). Bordetella trematum sp. nov., isolated from wounds and ear infections in humans, and reassessment of Alcaligenes denitrificans Rüger and Tan Int J Syst Bacteriol 46, von Wintzingerode, F., Schattke, A., Siddiqui, R. A., Rösick, U., Göbel, U. B. & Gross, R. (2001). Bordetella petrii sp. nov., isolated from an anaerobic bioreactor, and emended description of the genus Bordetella. Int J Syst Evol Microbiol 51, Wang, F., Grundmann, S., Schmid, M., Dörfler, U., Roherer, S., Munch, J. C., Hartmann, A., Jiang, X. & Schroll, R. (2007). Isolation and characterization of 1,2,4-trichlorobenzene mineralizing Bordetella sp. and its bioremediation potential in soil. Chemosphere 67, Wayne, L. G., Brenner, D. J., Colwell, R. R., Grimont, P. A. D., Kandler, O., Krichevsky, L., Moore, L. H., Moore, W. C., Murray, R. G. E. & other authors (1987). Report of the ad hoc committee on reconciliation of approaches to bacterial systematics. Int J Syst Bacteriol 37, Weiss, A. (2006). The genus Bordetella. In The Prokaryotes, pp Edited by M. Dworkin, S. Falkow, E. Rosenberg, K.-H. Schleifer & E. Stackebrandt. 5, 3rd edn., New York: Springer. Weyant, R. S., Hollis, D. G., Weaver, R. E., Amin, M. F. M., Steigerwalt, A. G., O Connor, S. P., Whitney, A. M., Daneshvar, M. I., Moss, C. W. & Brenner, D. J. (1995). Bordetella holmesii sp. nov., a new gramnegative species associated with septicemia. J Clin Microbiol 33,1 7. Weyant, R. S., Moss, C. W., Weaver, R. E., Hollis, D. G., Jordan, J. G., Cook, E. C. & Daneshvar, M. I. (1996). Identification of Unusual Pathogenic Gram-negative Aerobic and Facultatively Anaerobic Bacteria, 2nd edn. Baltimore: Williams & Wilkins International Journal of Systematic and Evolutionary Microbiology 65