ARTICLE IN PRESS

0723-2020/$ - se

doi:10.1016/j.sy

�Correspondfax: +41 21 31 4

E-mail addr

Systematic and Applied Microbiology 30 (2007) 572–579

www.elsevier.de/syapm

New Afipia and Bosea strains isolated from various water sources by

amoebal co-culture

Vincent Thomas, Nicola Casson, Gilbert Greub�

Center for Research on Intracellular Bacteria (CRIB), University Hospital Center and Institute of Microbiology,

University of Lausanne, 1011 Lausanne, Switzerland

Received 16 March 2007

Abstract

It has been suspected that some species belonging to the alphaproteobacteria might cause pneumonia in humans. Itis thus of special interest to isolate new members of this phylum, and to further characterize their pathogenicity. Theamoebal co-culture method allowed the isolation of various new bacterial species during the last few years, includingfastidious alphaproteobacterial species that were isolated from complex environments. In this work, we isolated newbacterial strains from a drinking water network or from river water using amoebal co-culture with Acanthamoeba

castellanii. One Afipia sp. strain and two Bosea sp. strains presented 16SrDNA and partial rpoB gene sequencessuggesting that they could be representative of new species, and were thus further characterized using phenotypic tests.r 2007 Elsevier GmbH. All rights reserved.

Keywords: Acanthamoeba castellanii; Afipia; Bosea; Co-culture; Amoeba-resisting bacteria; River water; Drinking water; Candidatus

Introduction

Bosea and Afipia are alphaproteobacteria genera thatboth belong to the family Bradyrhizobiaceae. SeveralBosea species have been isolated from the rhizosphere ofvarious plants and have unique metabolic features:oxidation of thiosulfate [7], the ability to use opinesas sole carbon source [19], and the ability to degradeN-acyl homoserine lactones [6]. Similarly, strains relatedto Afipia felis have been isolated from Antartic soils andwere the first described bacteria being able to use bothmethanesulfonate and dimethylsulfone as sole carbonsource [17]. Some Bosea strains were also associatedwith biofilms occurring in water purification processesemploying microfiltration and reverse osmosis mem-

e front matter r 2007 Elsevier GmbH. All rights reserved.

apm.2007.06.004

ing author. Tel.:+41 21 31 44 979;

4 060.

ess: gilbert.greub@chuv.ch (G. Greub).

branes [5], and they have recently been shown to resistdecontamination treatments used for drinking waterpurification [20]. Afipia spp. accounted for as many as28% of bacterial species grown from a dental unitbiofilm [21]. Most of the species belonging to thesegenera are fastidious micro-organisms that grow betterat 30–35 1C than at 37 1C, and that are difficult to isolateon conventional axenic cultivation media. This couldpartially explain why Afipia and Bosea spp. areinfrequently isolated from environmental and/or clinicalsamples. Afipia and Bosea species are resistant todestruction by amoebae of the genus Acanthamoeba,being thus considered as amoebae-resisting bacteria(ARB). The amoebal co-culture method has beensuccessfully used to recover new bacterial species fromhospital drinking water networks [14,15]. Resistance todestruction by amoebae could be considered as avirulence trait [10] because a bacterial species able toresist the microbicidal effector mechanisms of free-living

ARTICLE IN PRESSV. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579 573

amoebae might also be able to resist destruction byhuman macrophages. Given the resistance of thesespecies to amoebae, the amoebal co-culture methodseems to be appropriate to isolate new, potentiallypathogenic species belonging to these genera. In thiswork, we describe two new Bosea strains and one newAfipia strain that were all isolated from water sourcesusing an amoebal co-culture method. According to therecommendations of Murray and Schleifer [18], we usedthe term ‘‘Candidatus’’ for the description of thesestrains, since this is only a preliminary report andadditional studies including DNA–DNA re-associationexperiments will be necessary for complete speciesdescription.

Materials and methods

Samples

From May to August 2004, 200 samples werecollected from the water network of the UniversityHospital in Lausanne, Switzerland [23]. A total of 153tap swabs, 26 water samples, and 21 shower-head swabswere collected. Additionally, two 1L Seine river watersamples were collected at the entry of the Morsang-sur-Seine (France) drinking water plant every 3 months in2005 [22]. Approx. 1 kg of sand from the sand filtrationunit was also collected at the same time points.

Amoebal microplates

Acanthamoeba castellanii strain ATCC 30010 wasgrown at 28 1C in 75 cm2 cell culture flasks (Corning)with 30mL peptone yeast-extract glucose (PYG) [9].When cells formed an homogenous monolayer, theamoebae were harvested and washed three times in50mL of Page’s acanthamoeba saline (PAS) [9] (cen-trifugations at 2000g/10min to pellet the amoebae).After the last centrifugation, the amoebae were resus-pended in PAS, and 1mL of a 5� 105 A. castellanii/mLsuspension was distributed in each well of a 24-wellCostar microplate (Corning).

Processing of samples

For water samples, 1 L samples were filtered througha 0.2 mm cellulose nitrate membrane and the membranewas re-suspended in 10mL sterile water. For sandsamples, approx. 100 g sand were transferred to sterilebottles to which sterile PBS was added for a final volumeof 200mL and vigorously vortexed for 30 s.

To recover ARB (see amoebal co-culture), 200 mLwater or PBS were spread onto amoebal microplates.Once inoculated, microplates were centrifuged at 1500g

for 30min and incubated at 32 1C. Amoebal co-cultures(F0) were subcultured on fresh amoebae on day 6 andsubcultures (F1) were incubated for 14 days at 32 1C.Amoebal co-cultures were examined daily for amoeballysis. When amoebal lysis was observed, at the time ofsubculture, and after 14 days of subculture, 100 mL co-cultures were seeded onto charcoal yeast extract (CYE)agar plates and incubated for 20 days at 32 1C. Seedingon CYE was also systematically performed from F1subcultures after 14 days of incubation.

Comparative sequence analysis

To identify the recovered bacteria, PCR amplificationand sequencing of the 16S rRNA encoding gene wasperformed directly from agar-grown bacteria resuspe-nded in sterile PBS using primers fD1 and rP2 (Table 1),producing an approximately 1500 bp fragment [24]. The16S rRNA encoding gene was sequenced for everyAfipia or Bosea species using previously describedprimers (Table 1) [1]. The discriminative partialsequence of the rpoB gene was also amplified withprimers Br3200F and Br3950R (Table 1) [11] and allsequences were compared with sequences available inthe GenBank database in October 2006, using theBLAST 2.2.2 program available on the NCBI website(www.ncbi.nlm.nih.gov). RpoB sequences were alignedwith the complete rpoB sequence of Afipia felis

AY242824 in order to select the hypervariable regioncorresponding to positions 3380–3800 [11]. The homol-ogy of the edited sequences was then analyzed by thedistance matrix program of the MEGA3 software [12].According to Khamis et al. [11], we considered thatisolates belonged to the same species when hypervari-able region of rpoB gene sequence similarity was X98%,whereas they likely represent different species whensequence similarity was p96%. With these 16S rRNAand rpoB sequences, neighbor-joining (p-distance),minimum evolution (p-distance), and parsimony (stan-dard parsimony) trees were constructed using theMEGA3 software [12].

Phenotypic tests

The phenotypes of the strains were evaluated asfollows. Morphological and tinctorial properties weredetermined by Gram and Gimenez staining. Growthwas tested at 32 1C on Columbia agar with 5% sheepblood, chocolate agar and CYE agar. Growth on CYEagar was attempted at 30 and 37 1C. Oxidase activitywas detected using a dimethyl-p-phenylenediamineoxalate disk (Pasteur Diagnostic). Catalase activitywas detected by emulsifying a colony in 3% hydrogenperoxide and by assessing the presence of microscopicbubbles. Other biochemical tests were performed by

ARTICLE IN PRESS

Table 1. Primers used in this study. PCR conditions were identical to that described in original references

Primer designation Sequence 50–30 Targeted gene Position References

fD1 AGAGTTTGATCATGGCTCAG 16S 9–28 [24]

357F TACGGGAGGCAGCAG 16S 342–357 [1]

536F CAGCAGCCGCGGTAATAC 16S 519–536 [1]

800F ATTAGATACCCTGGTAG 16S 786–802 [1]

1050F TGTCGTCAGCTCGTG 16S 1055–1069 [1]

357R CTGCTGCCTCCCGTA 16S 342–357 [1]

536R GTATTACCGCGGCTGCTG 16S 519–536 [1]

800R CTACCAGGGTATCTAAT 16S 786–802 [1]

1050R CACGAGCTGACGACA 16S 1055–1069 [1]

rP2 ACGGCTACCTTGTTACGACTT 16S 1492–1512 [24]

Br3200F TGAAGATGGTCAAGGTCTTCGT rpoB 3208–3229 [11]

Br3550F GCATGAAYGTCGGBCAGAT rpoB 3386–3405 [11]

Br3550R ATCTGVCCGACRTTCATGC rpoB 3386–3405 [11]

Br3950R GTCCGACTTSACHGTCAGCAT rpoB 3940–3960 [11]

V. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579574

inoculation of API 20NE and API 50CH strips(bioMerieux, Geneva, Switzerland), according to themanufacturer’s instructions. These strips were, respec-tively, incubated for 7 and 15 days at 32 1C. The API20NE strip tested for any reduction of nitrates, indoleproduction, urease activity, glucose acidification, argi-nine dihydrolase activity, hydrolysis of gelatin andaesculin, b-galactosidase activity and assimilation ofglucose, arabinose, mannose, mannitol, N-acetylgluco-samine, maltose, gluconate, caprate, adipate, malate,citrate and phenylacetate. The API 50CH strip tested forany acidification of glycerol, erythritol, D-arabinose,L-arabinose, ribose, D-xylose, L-xylose, adonitol, methylb-D-xyloside, galactose, glucose, fructose, mannose,sorbose, rhamnose, dulcitol, inositol, mannitol, sorbitol,methyl a-D-mannoside, methyl a-D-glucoside, N-acetyl-glucosamine, amygdalin, arbutin, aesculin, salicin,cellobiose, maltose, lactose, melibiose, sucrose, treha-lose, inulin, melezitose, raffinose, starch, glycogen,xylitol, gentiobiose, D-turanose, D-lyxose, D-tagatose,D-fucose, L-fucose, D-arabitol, L-arabitol, gluconate,2- and 5-ketoglutarate.

Resistance of new bacterial species to A. castellanii

To further characterize the putative new species, weevaluated their ability to grow in the presence of A.

castellanii. Briefly, bacteria were grown on CYE agar at32 1C for 10 days. Then, they were harvested, washed 3times in PAS buffer, 5 mm-filtered, and the suspensionwas adjusted to a McFarland value of 1. Twentymicroliters of these suspensions were used to seedamoebal microplates prepared as described above.Microplates were centrifuged for 30min at 1500g andincubated for 2 h at 32 1C. The amoebal co-cultureswere washed with PAS and further incubated for 1 week

at 32 1C in humidified atmosphere. To count thenumber of viable bacteria, co-cultures were scrapped,serially diluted in sterile PBS and seeded on CYEagar that were then incubated for 14 days at32 1C. Colony forming units (CFUs) were thencompared with those obtained in the absence ofamoebal cells.

Transmission electron microscopy

Electron microscopy was done to further characterizethe putative new bacterial species. Briefly, 10mL of 5days’ cocultures were harvested, washed in PBS, andfixed in 4% glutaraldehyde. Fixed samples were thentreated as described previously [9] and evaluated with atransmission electron microscope PHILIPS EM 201 C(Philips, Eindhoven, The Netherlands).

Results and discussion

During the two sampling cycles of the hospital waternetwork and the Seine river water, five isolates related toAfipia spp. and eight related to Bosea spp. wererecovered using the amoebal co-culture method(Table 2). All Afipia isolates were recovered from thehospital water network [23]. Four Bosea isolates wererecovered from the hospital water network [23], twofrom the Seine river and two from the drinking waterplant sand filtration unit. Among the five Afipia isolates,four were identified as Afipia birgiae using rpoB

sequencing. The last isolate, Afipia sp. CRIB-05,exhibited 93.7% sequence similarity in the rpoB

hypervariable region with the sequence of its closestrelative, Afipia massiliensis [23] (Table 2). Among theeight isolates related to Bosea spp., one hospital strain

ARTICLE IN PRESS

Table 2. Amoebae-resisting bacterial strains related to Afipia and Bosea isolated from water samples by amoebal co-culture

Strain No. of isolatesa rpoB hypervariable

region GenBank

accession no.

Closest rpoB sequence species

& accession no.

Similarity (%)

Ca. Afipia lausannensis C-RIB-5 1 (H) AY919306 Afipia massiliensis

(AY242820)

93.7

Afipia birgiaeCRIB-08 2 (H) DQ123625 Afipia birgiae (AY242821) 99.0

Afipia birgiaeCRIB-14 2 (H) b Afipia birgiae (AY242821) 100.0

Bosea eneaeCRIB-11 3 (H) b Bosea eneae (AY242841) 100.0

Bosea sp. CRIB-07 1 (H) DQ123624 Bosea eneae (AY242841) 96.5

Bosea sp. CRIB-10 1 (R) DQ440829 Bosea eneae (AY242841) 96.9

Ca. Bosea sequanensis CRIB-12 1 (R) DQ440828 Bosea thiooxidans

(AY242832)

95.3

Ca. Bosea lascolae CRIB-13 2 (S) DQ440830 Bosea thiooxidans

(AY242832)

93.6

The partial, hypervariable, rpoB gene sequence homology of the closest published species, with its name, is also included.aSample type is shown in parenthesis. H ¼ Lausanne University Hospital water network, R ¼ Seine river and S ¼ sand filtration unit of the

Morsang-sur-Seine drinking water plant.bSequences 100% identical to those of existing GenBank sequences were not deposited; Ca ¼ Candidatus.

V. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579 575

exhibited 100% rpoB similarity with that of Bosea

eneae:B. eneae CRIB-11 (isolated 3 times). Anotherhospital strain, Bosea sp. CRIB-07 (isolated once), and astrain isolated one time from the Seine river, Bosea sp.CRIB-10, shared respectively, 96.5% and 96.9% rpoB

similarity with that of B. eneae. One strain from theSeine river: Bosea sp. CRIB-12 (isolated once) and onefrom the sand filtration unit: Bosea sp. CRIB-13(isolated twice), had similarity values of their rpoB

hypervariable regions of 95.3% and 93.6%, respectively,with their closest relative: Bosea thiooxydans. Whencomparing Bosea spp. CRIB-12 and CRIB-13 to eachother, the rpoB similarity value was 94.4%.

When considering the criteria proposed by Khamiset al. [11], we could not determine if Bosea spp. CRIB-07and CRIB-10 are distinct species or if they belong toB. eneae species since their hypervariable rpoB genesequence similarity with that of B. eneae was in theambiguous 96–98% range [11]. Conversely, Afipia sp.CRIB-05, Bosea sp. CRIB-12 and Bosea sp. CRIB-13could potentially be considered as new species due to thesequence similarity values of their rpoB hypervariableregions o96% with that of their closest relative. Thus,these three strains were deposited in the Collection del’Institut Pasteur (CIP) and Deutsche Sammlung vonMikroorganismen und Zellkulturen (DSMZ) collections(see species descriptions hereafter). To get additionalinformation for these strains, we performed phyloge-netic analyses using the hypervariable sequence of therpoB gene. These analyses demonstrated that Bosea

strains CRIB-12 and CRIB-13 clustered with Bosea

thiooxydans with bootsrap values of 92%, 78% and54% using neighbor-joining (Fig. 1), minimum evolu-tion (not shown) and maximal parsimony (not shown)analyses of the hypervariable region of the rpoB gene.

For Afipia sp. CRIB-05, bootstrap values of 97%, 97%,and 58% in the neighbor-joining, minimum evolution,and maximum parsimony trees, respectively, supportedthe nodes separating this species from Afipia massiliensis.

We also sequenced the nearly complete 16S rRNAgenes and performed additional phenotypic analyses onthe new isolates. Sixteen S rRNA genes sequences ofAfipia sp. CRIB-05 (DQ123622), Bosea sp. CRIB-12(DQ440827) and Bosea sp. CRIB-13 (DQ440824) were499% identical to those of the closest species: Afipia

birgiae (AF288304, 99.6% identity), Bosea thiooxydans

(AJ250796, 99.7% identity) and Bosea minatitlanensis

(AF273081, 99.2% identity), respectively. This confirmsthat sequence similarity of the 16S rRNA gene is notsufficiently discriminative to distinguish between speciesbelonging to these genera [8,11].

Phenotypic tests showed that all strains were Gram-negative, Gimenez-positive, and oxidase-positive. Afipia

sp. CRIB-05 was able to grow within 7–10 days on CYEand on Columbia agar at 30 1C but not at 37 1C. Nogrowth was observed under anaerobic conditions. Thestrain exhibited biochemical reactions similar to those ofAfipia clevelandensis [15]: assimilation of adipate,gluconate and malate, but not of the other substrates.However, differences were also observed between Afipia

sp. CRIB-05 and A. clevelandensis: Afipia sp. CRIB-05 isable to reduce nitrates to nitrites whereas A. clevelan-

densis is not; conversely A. clevelandensis has ureaseactivity whereas Afipia sp. CRIB-05 has not. For theCRIB-05 strain, acid production was only observed byoxidation of D-Glucose (not shown). Bosea spp. CRIB-12 and CRIB-13 were both able to grow within 3–5 daysat 30 and 32 1C on CYE, Columbia and Chocolate agarin aerobic conditions in absence of CO2. No growth wasobserved under anaerobic conditions. Bosea sp. CRIB-

ARTICLE IN PRESS

Afipia

100

60

87 978

AY242831

Afipia sp SP3302452 AY599911

Afipia broomeae AY242822

Candidatus Afipia lausannensis CRIB-05 AY919306

Afipia massiliensis AY242820

Afipia birgiae AY242821

sp SP17 AY599910

Afipia clevelandensis AY242823

Afipia felis AY242824

Afipia felis genosp.A AY242825

Afipia genosp.1 AY242828

Afipia genosp.2 AY242829

Afipia genosp.3 AY242827

Bradyrhizobium liaoningense AY242831

Bradyrhizobium japonicum AY242830

Bosea eneae AY242835

Bosea eneae AY242841

Bosea eneae CRIB-11 DQ440831

Bosea sp CRIB -07 DQ123624

Bosea sp CRIB -10 DQ440829

Bosea vestrisii AY242839

Bosea vestrisii AY242840

Bosea vestrisii AY242834

Bosea massiliensis AY242836

Bosea massiliensis AY242837

Bosea thiooxidans AY242832

Candidatus Bosea sequanensis CRIB-12 DQ440828

Bosea minatitlanensis AY242833

Bosea sp 7F AY242838

Sinorhizobium meliloti AL591787

100

100

100

100

100

75

99

89

86

64

100

82

95

92

65

97

100

82

67

95

0.02

Candidatus Bosea lascolae CRIB-13 DQ440830

Afipia genosp.3 -related AY242826

Fig. 1. Phylogenetic neighbor-joining tree representing relationships of Afipia and Bosea. The tree was derived from alignment of

partial rpoB sequences. Taxonomic names and GenBank accession numbers are shown at the end of each branch. The support of

each branch, as determined from 1000 bootstrap samples, is indicated by the value at each node (in percent). Sinorhizobium meliloti

was used as outgroup.

V. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579576

12 was also able to grow at 37 1C onto CYE agar,whereas Bosea sp. CRIB-13 was not. For these twostrains, biochemical reactions were similar to thatdescribed for Bosea massiliensis 63287, except that theywere not able to reduce nitrates to nitrites [14].API50CH tests gave no positive results for these twoBosea strains. As shown in Figs. 2 and 3, the new Afipia

and Bosea species were all able to grow in co-culturewith A. castellanii ATCC 30010. For Afipia sp. CRIB-05, transmission electron microscopy after day 5 ofco-culture showed highly vacuolated amoebae thatcontained 3–5 bacteria (Fig. 2). When incubating Afipia

sp. CRIB-05 in the presence of amoebae, there was a2.2 log increase within 7 days, whereas there was only a

0.6 log increase in the absence of cells (Fig. 3). For Bosea

spp. CRIB-12 and CRIB-13, transmission electronmicroscopy after day 5 of co-culture showed vacuolesthat contained 5–12 bacteria (Fig. 2). When incubatingthese species in the presence of amoebae, the increasewas 1.8 log for Bosea sp. CRIB-12 (no increase withoutcells), and 3.0 log for Bosea sp. CRIB-13 (0.3 logincrease without cells) (Fig. 3).

The ability of these three new strains to grow inco-culture with A. castellanii make them good candi-dates for being potentially pathogenic [10]. They extendthe list of strains belonging to the genus Afipia andBosea, that both comprise species that have beenimplicated in human infections [2–4,13]. Bosea sp.

ARTICLE IN PRESS

Fig. 2. New alphaproteobacteria species Candidatus Afipia

lausannensis (A) and Candidatus Bosea sequanensis (B) in A.

castellanii after 5 days of co-culture. Highly vacuolated

amoebae contained 3–5 Candidatus Afipia lausannensis and

5–12 Candidatus Bosea sequanensis (white arrows). Electron

microscopy magnification � 10,000.

D0 D2 D4 D7

Time (in days)

Lo

g C

FU

/ml

1,0E+04

1,0E+03

1,0E+02

1,0E+01

1,0E+00

1,0E-01

Fig. 3. Growth of new Afipia and Bosea species in co-culture

with A. castellanii ATCC 30010. Bacteria were grown for 7

days in PAS buffer with (plain lines) or without (dotted lines)

A. castellanii. Data presented are mean values of 3 independent

experiments. Candidatus Afipia lausannensis in presence (~)

and in absence (}) of A. castellanii Candidatus Bosea

sequanensis in presence (m) and in absence (n) of A. castellanii

Candidatus Bosea lascolae in presence (’) and in absence (&)

of A. castellanii.

V. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579 577

CRIB-13 has been isolated from the sand filtration unitof a drinking water plant. This is the first step of thewater purification process and other Bosea strains havealso been detected by others during late stages of thisprocess [20]. Thus, it cannot be excluded that bacteria ofthis genus are able to colonize drinking water plants, asdemonstrated for mycobacteria [16]. Future studiesshould confirm or refute this hypothesis.

Description of Candidatus Afipia lausannensis

Candidatus Afipia lausannensis (lau.san.en’sis. N.L.lausannensis of Lausanne, a town in Switzerland,where this strain was isolated from the water networkof the local University Hospital). Candidatus Afipialausannensis currently includes only the type strain,

CRIB-05 that has been isolated by co-culture of ahospital water network sample with A. castellanii ATCC30010. It is a Gram-negative, Gimenez positive andoxidase positive rod that exhibit poor sugar assimila-tion. Compared to other species of the genus, Candida-

tus Afipia lausannensis has a phenotype similar to thatof Afipia clevelandensis, except that it is able to reducenitrates to nitrites and that it has no urease activity. It isable to grow on CYE agar at 30 1C, not at 37 1C.The Candidatus A. lausannensis 16S rDNA sequence(GenBank accession number DQ123622) is 96.1–99.6%identical to the 16S rDNA of Afipia spp., which thusmakes this organism a member of the genus Afipia. Thepartial hypervariable rpoB gene sequence of thisorganism (GenBank accession number AY919306) is93.7% identical to that of the closest species: Afipia

massiliensis (GenBank accession number AY242820).This strain has been deposited as ‘‘Candidatus Afipialausannensis strain CRIB-05’’ in the CIP (CIP108885T)and DSMZ (DSM18162) collections.

Description of Candidatus Bosea sequanensis

Candidatus Bosea sequanensis (sequ.a.nen’sis. M.L. n.Sequana is the latin name of the Seine river, from where thestrain has been isolated). Candidatus Bosea sequanensiscurrently includes only the type strain, CRIB-12 that hasbeen isolated by co-culture of a Seine river water samplewith A. castellanii ATCC 30010. It is a Gram-negative,Gimenez positive and oxidase positive rod exhibiting poorsugar assimilation. Compared to others species of the genus,Candidatus Bosea sequanensis has a phenotype similar tothat of B. massiliensis 63287, except that it is not able to

ARTICLE IN PRESSV. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579578

reduce nitrates to nitrites. It is able to grow on CYE agar at30 and 37 1C. The Candidatus B. sequanensis 16S rDNAsequence (GenBank accession number DQ440827) is98.7–99.7% identical to the 16S rDNA of Bosea spp. whichmakes this organism a member of the genus Bosea. Thepartial, hypervariable rpoB gene sequence of this organ-ism (GenBank accession number DQ440828) is 95.3%identical to that of the closest species: Bosea thiooxydans

(GenBank accession number AY242832). This strain hasbeen deposited as ‘‘Bosea sp. strain CRIB-12’’ in the CIP(CIP109202T) and DSMZ (DSM18161) collections.

Description of Candidatus Bosea lascolae

Candidatus Bosea lascolae (lascolae, in honor ofDr. Bernard La Scola who isolated and described severalnew Afipia [15] and Bosea [14] species). Candidatus Bosealascolae currently includes only the type strain, CRIB-13that has been isolated by co-culture of sand samples fromthe sand filtration unit of the Morsang-sur-Seine drinkingwater plant with A. castellanii ATCC 30010. It is a Gram-negative, Gimenez positive and oxidase positive rodexhibiting poor sugar assimilation. Compared to othersspecies of the genus, Candidatus Bosea lascolae has aphenotype similar to that of B. massiliensis 63287, exceptthat it is not able to reduce nitrates to nitrites. It is able togrow on CYE agar at 30 1C, not at 37 1C. The Candidatus

B. lascolae 16S rDNA sequence (GenBank accessionnumber DQ440824) is 98.3–99.3% identical to the 16SrDNA of Bosea spp. that makes this organism a memberof the genus Bosea. The partial, hypervariable rpoB genesequence of this organism (GenBank accession numberDQ440830) is 93.6% identical to that of the closestspecies: Bosea thiooxydans (GenBank accession numberAY242832). This strain has been deposited as ‘‘Bosea sp.strain CRIB-13’’ in the CIP (CIP109201T) and DSMZ(DSM18163) collections.

Acknowledgements

This work was supported by the Swiss NationalScience Foundation (SNSF) Grant no. 3200BO-105885and by SUEZ Environment (Paris). We thank thePFMU at the Medical Faculty of Geneva for assistingwith electron microscopy analysis. We also thank P.Tarr for reviewing the manuscript and J.F. Loret (SUEZEnvironnement, Paris) for providing water samples.

References

[1] T. Adekambi, M. Drancourt, Dissection of phylogenetic

relationships among 19 rapidly growing Mycobacterium

species by 16S rRNA, hsp65, sodA, recA and rpoB gene

sequencing, Int. J. Syst. Evol. Microbiol. 54 (2004)

2095–2105.

[2] P. Berger, L. Papazian, M. Drancourt, B. La Scola,

J. Auffray, D. Raoult, Amoeba-associated microorgan-

isms and diagnosis of nosocomial pneumonia, Emerg.

Infect. Dis. 12 (2006) 248–255.

[3] K.A. Birkness, V.G. George, E.H. White, D.S. Stephens,

F.D. Quinn, Intracellular growth of Afipia felis, a putative

etiologic agent of cat scratch disease, Infect. Immun. 60

(1992) 2281–2287.

[4] D.J. Brenner, D.G. Hollis, C.W. Moss, C.K. English,

G.S. Hall, J. Vincent, J. Radosevic, K.A. Birkness,

W.F. Bibb, F.D. Quinn, et al., Proposal of Afipia gen.

nov., with Afipia felis sp. nov. (formerly the cat scratch

disease bacillus), Afipia clevelandensis sp. nov. (formerly

the Cleveland Clinic Foundation strain), Afipia broomeae

sp. nov., and three unnamed genospecies, J. Clin.

Microbiol. 29 (1991) 2450–2460.

[5] C.L. Chen, W.T. Liu, M.L. Chong, M.T. Wong,

S.L. Ong, H. Seah, W.J. Ng, Community structure of

microbial biofilms associated with membrane-based water

purification processes as revealed using a polyphasic

approach, Appl. Microbiol. Biotechnol. 63 (2004)

466–473.

[6] C. d’Angelo-Picard, D. Faure, I. Penot, Y. Dessaux,

Diversity of N-acyl homoserine lactone-producing and

-degrading bacteria in soil and tobacco rhizosphere,

Environ. Microbiol. 7 (2005) 1796–1808.

[7] S.K. Das, A.K. Mishra, B.J. Tindall, F.A. Rainey,

E. Stackebrandt, Oxidation of thiosulfate by a new

bacterium, Bosea thiooxidans (strain BI-42) gen. nov.,

sp. nov.: analysis of phylogeny based on chemotaxonomy

and 16S ribosomal DNA sequencing, Int. J. Syst.

Bacteriol. 46 (1996) 981–987.

[8] M. Drancourt, C. Bollet, A. Carlioz, R. Martelin, J.P.

Gayral, D. Raoult, 16S ribosomal DNA sequence

analysis of a large collection of environmental and clinical

unidentifiable bacterial isolates, J. Clin. Microbiol. 38

(2000) 3623–3630.

[9] G. Greub, B. La Scola, D. Raoult, Amoebae-resisting

bacteria isolated from human nasal swabs by amoebal

coculture, Emerg. Infect. Dis. 10 (2004) 470–477.

[10] G. Greub, D. Raoult, Microorganisms resistant to

free-living amoebae, Clin. Microbiol. Rev. 17 (2004)

413–433.

[11] A. Khamis, P. Colson, D. Raoult, B.L. Scola, Usefulness

of rpoB gene sequencing for identification of Afipia and

Bosea species, including a strategy for choosing discrimi-

native partial sequences, Appl. Environ. Microbiol. 69

(2003) 6740–6749.

[12] S. Kumar, K. Tamura, M. Nei, MEGA3: integrated

software for molecular evolutionary genetics analysis and

sequence alignment, Brief Bioinform. 5 (2004) 150–163.

[13] B. La Scola, I. Boyadjiev, G. Greub, A. Khamis, C.

Martin, D. Raoult, Amoeba-resisting bacteria and

ventilator-associated pneumonia, Emerg. Infect. Dis. 9

(2003) 815–821.

[14] B. La Scola, M.N. Mallet, P.A. Grimont, D. Raoult,

Description of Afipia birgiae sp. nov. and Afipia

massiliensis sp. nov. and recognition of Afipia felis

ARTICLE IN PRESSV. Thomas et al. / Systematic and Applied Microbiology 30 (2007) 572–579 579

genospecies A, Int. J. Syst. Evol. Microbiol. 52 (2002)

1773–1782.

[15] B. La Scola, M.N. Mallet, P.A. Grimont, D. Raoult,

Bosea eneae sp. nov., Bosea massiliensis sp. nov. and

Bosea vestrisii sp. nov., isolated from hospital water

supplies, and emendation of the genus Bosea (Das et al.

1996), Int. J. Syst. Evol. Microbiol. 53 (2003) 15–20.

[16] C. Le Dantec, J.P. Duguet, A. Montiel, N. Dumoutier,

S. Dubrou, V. Vincent, Occurrence of mycobacteria in

water treatment lines and in water distribution systems,

Appl. Environ. Microbiol. 68 (2002) 5318–5325.

[17] S.A. Moosvi, C.C. Pacheco, I.R. McDonald, P. De

Marco, D.A. Pearce, D.P. Kelly, A.P. Wood, Isolation

and properties of methanesulfonate-degrading Afipia felis

from Antarctica and comparison with other strains of A.

felis, Environ. Microbiol. 7 (2005) 22–33.

[18] R.G. Murray, K.H. Schleifer, Taxonomic notes: a

proposal for recording the properties of putative

taxa of prokaryotes, Int. J. Syst. Bacteriol. 44 (1994)

174–176.

[19] P.M. Oger, H. Mansouri, X. Nesme, Y. Dessaux,

Engineering root exudation of Lotus toward the

production of two novel carbon compounds leads to the

selection of distinct microbial populations in the rhizo-

sphere, Microb. Ecol. 47 (2004) 96–103.

[20] J. Rapala, M. Niemela, K.A. Berg, L. Lepisto, K. Lahti,

Removal of cyanobacteria, cyanotoxins, heterotrophic

bacteria and endotoxins at an operating surface water

treatment plant, Water Sci. Technol. 54 (2006) 23–28.

[21] R. Singh, O.C. Stine, D.L. Smith, J.K. Spitznagel Jr.,

M.E. Labib, H.N. Williams, Microbial diversity of

biofilms in dental unit water systems, Appl. Environ.

Microbiol. 69 (2003) 3412–3420.

[22] V. Thomas, N. Casson, G. Greub, Criblamydia sequa-

nensis, a new intracellular Chlamydiales isolated from

Seine river water using amoebal co-culture, Environ.

Microbiol. 8 (2006) 2125–2135.

[23] V. Thomas, K. Herrera-Rimann, D.S. Blanc, G. Greub,

Biodiversity of amoebae and amoebae-resisting bacteria

in a hospital water network, Appl. Environ. Microbiol. 72

(2006) 2428–2438.

[24] W.G. Weisburg, S.M. Barns, D.A. Pelletier, D.J. Lane,

16S ribosomal DNA amplification for phylogenetic study,

J. Bacteriol. 173 (1991) 697–703.