bacterial diversity in hydrothermal sediment and epsilonproteobacterial dominance in

Environmental Microbiology (2003)

5

(10), 961–976 doi:10.1046/j.1462-2920.2003.00495.x

© 2003 Society for Applied Microbiology and Blackwell Publishing Ltd

Blackwell Science, LtdOxford, UKEMIEnvironmental Microbiology 1462-2920Society for Applied Microbiology and Blackwell Publishing Ltd, 20035

10961976

Original Article

Bacterial diversity in Mid-Atlantic Ridge hydrothermal systemsP. López-

García

et al.

Received 4 March, 2003; accepted 6 June, 2003. *For correspon-dence: E-mail [email protected]; Tel. (

+

33) 1 69 15 76 08;Fax (

+

33) 1 69 15 46 97. †Present address: Département Environne-ment Profond, IFREMER, Technopole de Brest-Iroise, 29280 Plou-zané, France.

Bacterial diversity in hydrothermal sediment and epsilonproteobacterial dominance in experimental microcolonizers at the Mid-Atlantic Ridge

Purificación López-García,

1

Sébastien Duperron,

2†

Pascal Philippot,

3

Julien Foriel,

3

Jean Susini

4

and David Moreira

1

1

UMR CNRS 8079, Ecologie, Systématique et Evolution, Université Paris-Sud, bâtiment 360, 91405 Orsay Cedex, France.

2

UMR 7622 CNRS, Université Pierre et Marie Curie, 9 quai St. Bernard, 75005 Paris, France.

3

Laboratoire de Géosciences Marines, CNRS, Institut de Physique du Globe de Paris, 4 place Jussieu, 75005 Paris, France.

4

European Synchrotron Radiation Facility, ID 21, BP 220, 38043 Grenoble Cedex, France.

Summary

We report here a molecular survey based on 16SrRNA genes of the bacterial diversity found in twodeep-sea vent niches at the Mid-Atlantic Ridge:hydrothermal sediment (Rainbow site), and microcol-onizers made of three different substrates (organic-rich, iron-rich and pumice) that were exposed for15 days to a vent emission. Bacterial diversity in sed-iment samples was scattered through many bacterialdivisions. The most abundant and diverse environ-mental sequences (phylotypes) in our libraries corre-sponded to the Gammaproteobacteria, followed bythe Acidobacteria. We detected members of all thesubdivisions within the Proteobacteria. Myxobacteriallineages were the most represented within the deltasubdivision. Phylotypes ascribing to the

Cytophaga-Flavobacterium-Bacteroides

, Planctomycetales, highand low G

++++

C Gram-positives, Nitrospirae, and thecandidate division TM7 were also identified. Com-pared to this broad taxonomic coverage, micro-colonizers were almost exclusively colonized byepsilonproteobacteria, although these exhibited con-siderable morphological and phylogenetic in-group

diversity. No specificity for any of the substratestested was seen. This observation further supportsthe idea of the ecological dominance of epsilonpro-teobacteria in the fluid–seawater interface environ-ment. Because oxidation of reduced S species and/orsulphur-reduction is thought to be essential for theirenergetic metabolism in these areas, we mapped dif-ferent oxidation states of S in individual bacterial fil-aments from the iron-rich microcolonizer. For this, weused high-resolution, non-destructive synchrotronmicro-X-ray Absorption Near-Edge Spectroscopy(micro-XANES), which revealed the co-existence ofdifferent S oxidation states, from sulphide to sul-phate, at the level of individual cells. This suggeststhat these cells were metabolizing sulphur

in situ

.

Introduction

Deep-sea hydrothermal systems have attracted muchattention during the last 30 years for a number of reasons.They constitute ecosystems: (i) with a rich animal andmicrobial diversity founded basically on chemolithoau-totrophic primary production; (ii) where the most hyper-thermophilic organisms have been isolated from, and (iii)that are considered modern analogues of ancient andmore widespread hydrothermal systems (for reviews, seePrieur, 1997; Stetter, 1999; Jeanthon, 2000; Nisbet andSleep, 2001; Reysenbach and Cady, 2001; Reysenbachand Shock, 2002). Hydrothermal emissions are hot (thetemperature varies from a few degrees above seawatertemperature in diffusers to more than 300

∞

C in blacksmokers) and enriched in various metals from mantleregions as well as hydrogen sulphide and other reducedspecies. Their encounter with the cold (2

∞

C on average)and highly oxygenated deep seawater generates steepphysicochemical gradients. As a consequence, a varietyof ecological niches form in these systems that are char-acterized by different temperatures, nature of substrateand availability of electron donors and acceptors forenergy metabolism.

In addition to the study of strict hyperthermophiles fromchimney fragments and hot fluids, the prokaryotic diver-sity of warm to hot fluid-seawater-mixing regions hasbeen focus of intense research. Most microorganisms in

962

P. López-García

et al.

© 2003 Society for Applied Microbiology and Blackwell Publishing Ltd,

Environmental Microbiology

,

5

, 961–976

areas exposed to the fluid–seawater interface are thoughtto be chemolithoautotrophs or chemoorganotrophs oxi-dizing reduced forms of sulphur and iron, as was pro-posed long ago (Jannasch and Mottl, 1985). Although arelatively large bacterial diversity is found in these areas,members of the epsilon subdivision of the Proteobacteriaappear to predominate, regardless the geographic prove-nance of samples. This trend has been observed in vent-cap chambers incubated

in situ

at the Mid-Atlantic Ridge(Reysenbach

et al

., 2000; Corre

et al

., 2001), mat sam-ples from the South-east Pacific Rise (Longnecker andReysenbach, 2001), and in the epibiotic microbiota ofmany deep-sea vent animal teguments or tubes. Theseinclude the polychaetes

Alvinella pompejana

and

Paralvinella palmiformis

(Haddad

et al

., 1995; Alain

et al

.,2002a) and the vestimentiferan

Riftia pachyptila

(López-García

et al

., 2002), endemic from the Eastern Pacific,and the Atlantic shrimp

Rimicaris exoculata

(Polz andCavanaugh, 1995).

The Epsilonproteobacteria group some S-oxidizers,such as

Thiovulum

spp. which oxidize hydrogen sulphideand displays chemotaxis along oxygen gradients (Tharand Fenchel, 2001) or

Arcobacter

spp., some of whichproduce filamentous sulphur (Wirsen

et al

., 2002). In addi-tion, several epsilonproteobacterial isolates from deep-sea vents reduce sulphur (Campbell

et al

., 2001; Alain

et al

., 2002b; Miroshnichenko

et al

., 2002). Very recently,Takai

et al

. (2003) have reported the isolation of epsilon-proteobacteria from hydrothermal vents at the OkinawaTrough (Western Pacific) and the Central Indian Ridgethat cluster with environmental groups lacking cultivatedmembers so far. These are able to grow using H

2

orreduced sulphur compounds (elemental sulphur or thio-sulphate) as electron donors, and O

2

, nitrate or elementalsulphur as electron acceptors. Deep-sea vent sulphur orthiosulphate oxidizers are also frequent within the alphaand gamma subdivisions of the Proteobacteria (Jannash

et al

., 1985; Distel

et al

., 1988; Muyzer

et al

., 1995), andwithin members of the Flavobacteria (Teske

et al

., 2000).Sulphur-based metabolisms thus play a major role in thisecosystem, which is particularly substantiated by thedevelopment of mats of sulphur oxidizers sometimesassociated with apparent filamentous sulphur structures(Jannasch

et al

., 1989; Moyer

et al

., 1995; Taylor

et al

.,1999).

The prokaryotic diversity in hydrothermal sediments hasbeen considerably less studied, but it appears to be muchmore varied than that of exposed surfaces. To date, mostreports relate to Pacific areas. A recent 16S rRNA-basedmolecular survey of archaea and bacteria in Guaymasbasin hydrothermal sediment revealed a variety ofarchaea, including ANME-1 lineages that could partici-pate in anaerobic methane oxidation, and a wide diversityof lineages scattered in various bacterial divisions (Teske

et al

., 2002). This high level of prokaryotic diversity insediments is paralleled by a considerable diversity ofmicrobial eukaryotes, as deduced from 18S rRNA genesretrieved from the same site (Edgcomb

et al

., 2002). Inthe case of eukaryotes, a variety of typical photosynthe-sising lineages (e.g. green algae) was identified, what ledto the conclusion that contamination from communitiessedimenting from upper layers in the water column existedand could not be distinguished from autochthonous mem-bers. Likewise, although most bacterial lineages wereaffiliated with typical deep-sea or anaerobic bacteria, thepresence of cyanobacterial-like sequences suggests thatpart of the bacterial diversity observed in the Guaymassystem is not original from the sediment. We recentlycarried out a molecular survey of 18S rRNA genespresent in Mid-Atlantic Ridge hydrothermal sediment. Ourresults showed, in contrast, the existence of a wide diver-sity of eukaryotic lineages that were in no case related tophotosynthesisers, thus excluding the possibility of con-tamination from surface water in this site (López-García

et al

., 2003). The bacterial diversity in these samplesshould therefore correspond to autochthonous bacteria inthese deep-sea sediments as well.

Here, we present a 16S rRNA-based molecular surveyof the bacterial diversity found in this hydrothermal sedi-ment from the Rainbow site (Mid-Atlantic Ridge) presum-ably lacking contaminants from surface waters, and inthree experimental devices for microbial colonization(microcolonizers) that were exposed for two weeks to afluid emission (Mid-Atlantic Ridge). This should allow us,first, comparing Atlantic and Pacific hydrothermal sedi-ment communities, second, comparing sediment and sub-strates directly exposed to the fluid–seawater interface inthe same oceanic ridge, and finally, testing whether thereis any specificity in the colonization of the different sub-strates, namely inert mineral, organic-rich, and iron-rich,that were used in the microcolonizers. In addition, usinga synchrotron-based approach (see Philippot

et al

., 2002and Foriel

et al

., 2003 for details), we show the co-existence of different S oxidation states in individual cellsfrom microcolonizers, suggesting that they were activelymetabolizing sulphur.

Results

Bacterial diversity in hydrothermal sediment

The hydrothermal sediment from the Rainbow site stud-ied here is characterized by record enrichments in differ-ent metals and rare earth elements (Cave

et al

., 2002;Douville

et al

., 2002). Levels of Fe, Cu, Mn, V, P and Asare especially high, even when compared to those ofother hydrothermal sediments. CaCO

3

accounted for

~

32% of the sediment (dry weight), and were contributed

Bacterial diversity in Mid-Atlantic Ridge hydrothermal systems

963



,

5

, 961–976

mostly by haptophyte coccoliths and foraminifer shells(López-García

et al

., 2003). After DNA extraction, weamplified 16S rRNA genes present in the sample by poly-merase chain reaction (PCR) using different combina-tions of bacterial and prokaryotic specific primers (see

Experimental procedures

). We then constructed a totalof four different 16S rDNA libraries, from which wesequenced partially more than one hundred clones. Thisallowed us to have a first insight on the diversity of bac-terial groups represented in the sediment by

BLAST

sequence comparison with sequences in databanks. Ascan be seen in Fig. 1, we detected a variety of bacteriallineages. The most diverse and abundant clones in ourlibraries affiliated to the Gammaproteobacteria and to theAcidobacteria. The rest of phylotypes appear scatteredover a broad taxonomic distribution including the addi-tional four subdivisions of the Proteobacteria (alpha, beta,delta and epsilon), the

Cytophaga-Flavobacterium-Bacteroides

(CFB) group, the High G

+

C (Actinobacteria)and Low G

+

C (Firmicutes) Gram-positives, the Plancto-mycetales, the Nitrospirae and the candidate divisionTM7 (Fig. 1).

We then selected a number of representative clonescovering the whole diversity observed, which were com-pletely sequenced and subjected to more detailed phylo-genetic analysis (Table 1). In order to handle a reasonabletaxonomic diversity encompassing the closest environ-mental sequences to our clones and also representativesof major bacterial groups compatible with maximum like-lihood (ML) analyses, we constructed five different datasets. Four of these, including all but one sedimentsequence, are shown in Fig. 2. The fifth data set corre-sponds to the Epsilonproteobacteria (see below). In gen-eral, our clones ascribed to groups that have been already

identified in deep-sea vents or sediments, and their clos-est relatives are environmental sequences from theseenvironments, or otherwise, from anaerobic sludge ormetal-contaminated sites (Fig. 2 and Table 1). One excep-tion is clone AT-s2–18 (Fig. 2A), which clearly brancheswithin the candidate division TM7 (Hugenholtz

et al

.,2001). Members of this group had not been previouslyidentified in deep-sea vents. Among the most representedbacterial divisions were the CFB and the Acidobacteria,the latter being particularly diverse. In both cases, oursequences defined clusters with other environmental phy-lotypes (Fig. 2B).

We found representatives of all Proteobacteria sub-divisions (Fig. 2C and D and see below). Among theDeltaproteobacteria, most phylotypes branch within theMyxobacteria, gliding bacteria with complex social behav-iour and developmental cycles that are not uncommon indeep-sea vent areas (Moyer

et al

., 1995; Teske

et al

.,2002). The clone AT-s3–57 forms a well-defined clusterwith environmental sequences and with

Bdellovibrio

spp.,fast swimmer bacterial predators common in differentenvironments (Martin, 2002; Snyder

et al

., 2002). Amongthe sequences belonging to the Betaproteobacteria, weidentified a sequence (AT-s3–41) that can be ascribed tothe genus

Hydrogenophilus

(Fig. 2C), whose membersare thermophilic (Hayashi

et al

., 1999; Stohr

et al

., 2001).We also detected some phylotypes within the Alphapro-teobacteria, but most of the sediment diversity was foundwithin the Gammaproteobacteria (Fig. 2D). Some of thesephylotypes clustered with other environmental sequences,many of which (clones MERTZ) have been retrieved fromAntarctic sediments (Bowman and McCuaig, 2003). How-ever, several sequences did not have clear close relativesamong sequences in databases, particularly those of

Fig. 1.

Diversity of major bacterial groups iden-tified in Mid-Atlantic Ridge hydrothermal sedi-ment and experimental microcolonizers containing different substrates that were exposed to a fluid source. The histogram shows the proportions of clones found in different 16S rDNA libraries.

964

P. López-García

et al.



,

5

, 961–976

Tab

le 1

.

Phy

loge

netic

affi

liatio

n of

bac

teria

l clo

nes

obta

ined

from

hyd

roth

erm

al s

edim

ent a

t the

Mid

-Atla

ntic

Rid

ge (

Rai

nbow

site

) as

ded

uced

from

BLA

ST

sea

rche

s.

Clo

neS

eque

nce

leng

th (

bp)

% GC

Tota

l num

ber

of s

imila

rse

quen

ces

a

Phy

loge

netic

ascr

iptio

nC

lose

st 1

6S-r

DN

A m

atch

in d

atab

ase,

nat

ure

of h

abita

t w

here

the

y it

was

ret

rieve

d (a

cces

sion

num

ber)

%

iden

tity

AT-s

1614

5254

.35

1G

amm

apro

teob

acte

riaS

ulph

ur-o

xidi

sing

bac

teriu

m O

AII2

, sh

allo

w v

ent

(AF

1704

23)

89AT

-s26

1484

56.4

62

Gam

map

rote

obac

teria

Unc

. gam

map

rote

obac

teriu

m c

lone

ME

RT

Z_2

CM

_24,

con

tinen

tal s

helf

sedi

men

ts c

olle

cted

of

f A

ntar

ctic

a (A

F42

4060

)97

AT-s

4314

6552

.71

1G

amm

apro

teob

acte

riaU

nc. b

acte

rium

clo

ne B

PC

036,

hyd

roca

rbon

see

p se

dim

ent

(AF

1540

89)

93AT

-s68

1456

54.6

15

Gam

map

rote

obac

teria

Unc

. gam

map

rote

obac

teriu

m B

2M60

, m

arin

e se

dim

ents

(A

F22

3299

)96

AT-s

7514

6054

.88

2G

amm

apro

teob

acte

riaU

nc. g

amm

apro

teob

acte

rium

BD

6-6,

dee

p-se

a se

dim

ents

(A

B01

5576

)97

AT-s

8014

5756

.94

4G

amm

apro

teob

acte

riaS

ulph

ur-o

xidi

zing

che

moa

utot

roph

ic g

ill s

ymbi

ont

of t

he b

ival

ve

Cod

akia

cos

tata

(L2

5712

)91

AT-s

2–13

1452

53.2

43

Gam

map

rote

obac

teria

Sul

phur

-oxi

dizi

ng b

acte

rium

ND

II1.1

, sh

allo

w w

ater

hyd

roth

erm

al v

ent

(AF

1704

24)

91AT

-s2–

5914

7156

.56

1G

amm

apro

teob

acte

riaU

nc. g

amm

apro

teob

acte

rium

BD

6-6,

dee

p-se

a se

dim

ents

(A

B01

5576

)93

AT-s

3–1

1418

54.1

310

Gam

map

rote

obac

teria

Sul

phur

-oxi

dizi

ng c

hem

oaut

otro

phic

end

osym

bion

t of

the

ves

timen

tifer

an

Lam

ellib

rach

ia

colu

mna

(U

7748

1)92

AT-s

3–25

1422

55.4

15

Gam

map

rote

obac

teria

Unc

. gam

map

rote

obac

teriu

m B

D3-

6, d

eep-

sea

sedi

men

ts (

AB

0155

48)

95AT

-s3–

2614

2056

.49

1G

amm

apro

teob

acte

riaU

nc. g

amm

apro

teob

acte

rium

Sva

0071

, pe

rman

ently

col

d m

arin

e se

dim

ents

(A

J240

986)

92AT

-s3–

4814

2354

.81

2G

amm

apro

teob

acte

riaU

nc. g

amm

apro

teob

acte

rium

BD

3-6,

dee

p-se

a se

dim

ents

(A

B01

5548

)95

AT-s

3–68

1423

55.7

31

Gam

map

rote

obac

teria

Pse

udom

onas

sp.

clo

ne N

B1-

h, J

apan

Tre

nch

sedi

men

t (A

B01

3829

)91

AT-s

6315

2054

.64

1A

lpha

prot

eoba

cter

ia

Jann

asch

ia h

elgo

land

ensi

s

str

ain

Hel

10,

Nor

th S

ea (

AJ4

3815

7)92

AT-s

3–44

1381

56.8

81

Alp

hapr

oteo

bact

eria

Unc

. bac

teriu

m c

lone

MN

D8,

Gre

en B

ay f

erro

man

gano

us m

icro

nodu

le (

AF

2929

99)

90AT

-s71

1464

52.4

61

Bet

apro

teob

acte

riaU

nc. b

acte

rium

clo

ne M

B-A

2-11

5, m

etha

ne h

ydra

te-b

earin

g de

ep-s

ea s

edim

ent i

n a

Fore

arc

Bas

in (

AY09

3465

)92

AT-s

3–41

1430

59.5

91

Bet

apro

teob

acte

ria

Hyd

roge

noph

ilus

ther

mol

uteo

lus

TH

-1 (

AB

0098

28)

98AT

-s3–

3414

1355

.63

1D

elta

prot

eoba

cter

iaU

nc. d

elta

prot

eoba

cter

ium

clo

ne S

h765

B-T

zT-4

2, u

rani

um m

inin

g w

aste

pile

(A

J519

631)

92AT

-s3–

5714

4252

.54

3D

elta

prot

eoba

cter

iaU

nc. b

acte

rium

Fuk

uN9,

lake

Fuc

hsku

hle

bact

erio

plan

kton

(A

J290

009)

91AT

-s3–

6014

4355

.86

1D

elta

prot

eoba

cter

iaU

nc. b

acte

rium

clo

ne O

B3–

39,

Ant

arct

ic c

oast

al s

edim

ents

(AY

1334

45)

97AT

-s3–

6614

2155

.64

1D

elta

prot

eoba

cter

iaU

nc. d

elta

prot

eoba

cter

ium

clo

ne M

ER

TZ

_21

C

M

_284

, con

tinen

tal s

helf

sedi

men

ts c

olle

cted

of

f A

ntar

ctic

a (A

F42

4255

)97

AT-s

3–19

1382

50.9

81

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m B

D6-

6, d

eep-

sea

sedi

men

ts (

AB

0155

35)

94

a.

Seq

uenc

es 9

8–10

0% id

entic

al,

incl

udin

g pa

rtia

l seq

uenc

es.

CF

B,

Cyt

opha

ga-F

lavo

bact

eriu

m-B

acte

roid

es

; Unc

., un

cultu

red.


965



,

5

, 961–976

AT-s

214

6859

.33

2A

cido

bact

eria

Unc

.

Hol

opha

ga

sp.

clo

ne J

G37

-AG

-29,

ura

nium

min

ing

was

te p

ile (

AJ5

1936

7)92

AT-s

3614

7757

.91

Aci

doba

cter

iaU

nc. H

olop

haga

/Aci

doba

cter

ium

Sva

0725

, pe

rman

ently

col

d m

arin

e se

dim

ents

(A

J241

003)

91AT

-s54

1471

59.4

81

Aci

doba

cter

iaU

nc. b

acte

rium

clo

ne K

S85

, m

ud d

epos

its o

ff Fr

ench

Gui

ana

(AF

3282

18)

92AT

-s65

1481

57.2

12

Aci

doba

cter

iaU

nc. b

acte

rium

clo

ne T

K29

, m

arin

e sp

onge

(A

J347

029)

89AT

-s2–

5714

7459

.38

4A

cido

bact

eria

Unc

. bac

teriu

m M

ER

TZ

_2 C

M

_263

, sh

elf

sedi

men

ts c

olle

cted

off

Ant

arct

ica

(AF

4243

07)

96AT

-s3–

2314

4057

.64

1A

cido

bact

eria

Unc

. Hol

opha

ga/A

cido

bact

eriu

m S

va05

15,

perm

anen

tly c

old

mar

ine

sedi

men

ts (

AJ2

4100

4)94

AT-s

3–28

1442

57.5

33

Aci

doba

cter

iaU

nc. b

acte

rium

clo

ne K

S63

16S

, m

ud d

epos

its o

ff Fr

ench

Gui

ana

(AF

3282

01)

93AT

-s3–

2414

4058

.71

2A

cido

bact

eria

Unc

.

Hol

opha

ga

sp.

clo

ne J

G34

-KF

-153

, ur

aniu

m m

inin

g w

aste

pile

(A

J532

721)

90AT

-s3–

3714

3958

.11

Aci

doba

cter

iaU

nc. H

olop

haga

/Aci

doba

cter

ium

Sva

0725

, pe

rman

ently

col

d m

arin

e se

dim

ents

(A

J241

003)

92AT

-s3–

4214

3658

.36

1A

cido

bact

eria

Unc

. Hol

opha

ga/A

cido

bact

eriu

m S

va05

15,

perm

anen

tly c

old

mar

ine

sedi

men

ts (

AJ2

4100

4)94

AT-s

3–56

1440

58.2

12

Aci

doba

cter

iaU

nc. H

olop

haga

/Aci

doba

cter

ium

Sva

0515

, pe

rman

ently

col

d m

arin

e se

dim

ents

(A

J241

004)

95AT

-s3–

5914

4056

.78

1A

cido

bact

eria

Unc

. bac

teriu

m c

lone

TK

29,

mar

ine

spon

ge (

AJ3

4702

9)93

AT-s

9214

3653

.92

1C

FB

gro

upU

nc. b

acte

rium

clo

ne M

L121

8M-1

4, a

lkal

ine,

hyp

ersa

line

Mon

o La

ke,

Cal

iforn

ia (

AF

4525

99)

89AT

-s98

1436

52.4

4C

FB

gro

up

Fle

xiba

cter

agg

rega

ns

IF

O 1

5974

(A

B07

8038

)93

AT-s

2–33

1440

56.3

61

Hig

h G

C G

ram

-pos

itive

(Act

inob

acte

ria)

Unc

. bac

teriu

m c

lone

JT

B31

, se

dim

ents

fro

m d

eep

cold

-See

p at

the

Jap

an T

renc

h (A

B01

5267

)89

AT-s

3–3

1433

59.5

32

Hig

h G

C G

ram

-pos

itive

(Act

inob

acte

ria)

Unc

. spo

nge

sym

bion

t PA

UC

43f

(AF

1864

15)

93

AT-s

3014

8559

.55

1Lo

w G

C G

ram

-pos

itive

Unc

. bac

teriu

m w

b1_A

12,

Nul

larb

or c

aves

, A

ustr

alia

(A

F31

7743

)95

AT-s

8314

6454

.89

1N

itros

pira

Unc

. bac

teriu

m c

lone

RC

P2-

12,

fore

st w

etla

nd im

pact

ed b

y re

ject

coa

l (A

F52

3925

)

93AT

-s2–

3814

4357

.43

1P

lanc

tom

yces

Unc

. bac

teriu

m c

lone

SH

A-8

7, 1

,2-d

ichl

orop

ropa

ne d

echl

orin

atio

n co

mm

unity

(A

J306

767)

91AT

-s2–

1814

9851

.04

1T

M7

cand

idat

e di

visi

onU

nc. b

acte

rium

Noo

saA

W44

, slu

dge

from

enh

ance

d bi

olog

ical

pho

spho

rous

rem

oval

rea

ctor

(A

F26

9023

)93

Clo

neS

eque

nce

leng

th (

bp)

% GC

Tota

l num

ber

of s

imila

rse

quen

ces

a

Phy

loge

netic

ascr

iptio

nC

lose

st 1

6S-r

DN

A m

atch

in d

atab

ase,

nat

ure

of h

abita

t w

here

the

y it

was

ret

rieve

d (a

cces

sion

num

ber)

%

iden

tity

a.

Seq

uenc

es 9

8–10

0% id

entic

al,

incl

udin

g pa

rtia

l seq

uenc

es.

CF

B,

Cyt

opha

ga-F

lavo

bact

eriu

m-B

acte

roid

es

; Unc

., un

cultu

red.

966

P. López-García

et al.



,

5

, 961–976

clones AT-s3–26, AT-s2–13, AT-s16, and AT-s3–1. Theinspection of the alignment and the fact that we detectedin most cases other, closely related, sequences in ourlibraries (Table 1) suggest that these are not chimeras.Therefore, they could be representatives of novel groupsor families within this subdivision.

Bacterial diversity in microcolonizers

The three substrates that had been exposed for 15 daysto a fluid emission (organic-rich, iron-rich and inert mineral–pumice fragments) were densely colonized during thattime, as can be seen in scanning electron microscopy

Fig. 2.

Maximum-likelihood (ML) phylogenetic trees showing the position of sediment 16S rDNA bacterial clones. Trees A to D cover the diversity of different bacterial groups including closest

BLAST

matches to our clone sequences and representative neighbour species. Only bootstrap proportions above 50 are shown. Scale bars represent 10 substitutions per 100 positions for a unit branch length.


967



,

5

, 961–976

photographs (Fig. 3). The morphologies observed includedmany different types of filaments, and individual cells withdifferent diameters and lengths, often forming biofilmsembedded in a matrix. We could not observe a morpho-logical specificity of cell types on the different substrates,as a wide range of forms was always observed. The onlypossible exception could be constituted by the observationof filaments surrounded by rigid sheaths that wereobserved exclusively on the iron-rich substrate (Fig. 3C).

From the morphological diversity observed, the findingof a large microbial diversity by molecular methods wasto be expected. To test this and whether there was adifferential colonization of the different substrates, we con-structed environmental 16S rDNA libraries of DNAextracted from each one of the three substrates (see


). We partially sequenced

~

60clones, from which we selected 16 representatives thatwere sequenced completely. With the only exception ofclone AT-co1, a CFB member (Fig. 2B and Table 2), allmicrocolonizer phylotypes affiliated to the Epsilonproteo-bacteria (Figs 1 and 4; Table 2). Despite this, phylotypeswere diverse within the group, which correlates with themorphological variety observed. Our sequences branchedwithin three different clusters, from which clusters B/II andF/I (Fig. 4 and see below) are very often found in deep-sea vent environments. A proportion of the clones, repre-sented by AT-pp13, belongs to the

Arcobacter

group.Some marine

Arcobacter

spp. are autotrophic sulphideoxidizers (Wirsen et al., 2002). Several sequencesbelonged to the group B defined by Corre et al. (2001),which is equivalent to the group II by Teske et al. (2002).It includes members of Thiomicrospira, a genus includingthiosulphate oxidizers (Kuenen et al., 1992). Takai et al.(2003) have recently isolated members of this clusterwhich are able to live on a mixture of hydrogen, thiosul-phate and sulphur using nitrate or oxygen (1%) as elec-tron acceptors. Most of our sequences branched withinthe cluster F/I, which includes many deep-sea vent and

deep-sea sediment lineages as well as, although not veryclosely related to our sequences, epibionts of vent poly-chaetes and shrimp (Fig. 4). Takai et al. (2003) also iso-lated several members of this group able of growing onhydrogen-thiosulphate-sulphur mixtures using nitrate oroxygen as electron acceptors, but in this case most grewbetter with 10% oxygen in the medium. This could suggestthis lineage may occupy a more oxygen-enriched, likelycolder, part of the physico-chemical gradient at hydrother-mal vents.

In general, most clones have as closest relatives envi-ronmental sequences retrieved from deep-sea sediment,deep-sea vent environments, and epibionts on Riftia pac-hyptila tubes (R76 and R103 clones) (López-García et al.,2002). The sequences of the three substrates being inter-spersed in the epsilonproteobacterial tree, a specificity ofcolonization appears to be rejected by our data.

Imaging sulphur-metabolizing activities in bacterial filaments from microcolonizers

Many epsilonproteobacteria isolated from deep-sea ventshave sulphur oxidizing and/or reducing capabilities, as canbe deduced from their growth in laboratory conditions(Campbell et al., 2001; Alain et al., 2002b; Mirosh-nichenko et al., 2002; Takai et al., 2003). Given the avail-ability of large amounts of both, hydrogen sulphide andoxygen at the places colonized by the epsilonproteobac-teria, the oxidation of reduced sulphur states with oxygencould be hypothesised as an obvious reaction to obtainenergy from.

In order to test if sulphur oxido-reduction activities weretaking place in situ, we analysed single filaments recov-ered from the microcolonrs by micro-X-ray AbsorptionNear-Edge Structure (micro-XANES) at the EuropeanSynchrotron Radiation Facility (ESRF). Synchrotron-based analyses have been used recently to image thespatial distribution at a mm-scale of diverse potential

Fig. 3. Scanning electron microscopy of prokaryotic cell types observed in different col-onization substrates after 15 days exposure to a fluid source. A and B correspond to pumice, C to the iron-rich substrate, and D to the organic-rich substrate. Scale bars correspond to 10 mm (A, B) and 1 mm (C, D).

968 P. López-García et al.

© 2003 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 5, 961–976

Tab

le 2

.P

hylo

gene

tic a

ffilia

tion

of b

acte

rial c

lone

s fr

om s

ubst

rate

s re

cove

red

afte

r 15

day

s ex

posu

re to

a fl

uid

sour

ce a

t the

Mid

-Atla

ntic

Rid

ge (

Luck

y S

trik

e).

Clo

neS

eque

nce

leng

th (

bp)

% GC

Tota

l nu

mbe

r of

si

mila

rse

quen

cesa

Phy

loge

netic

ascr

iptio

nC

lose

st 1

6S-r

DN

A m

atch

in d

atab

ase

(acc

essi

on n

umbe

r)%

Id

entit

y

AT-c

o114

3549

.76

1C

FB

Gro

upU

nc. C

FB

bac

teriu

m B

D3-

6, d

eep-

sea

sedi

men

ts (

AB

0155

32)

89AT

-co1

114

3150

.94

1E

psilo

npro

teob

acte

riaU

nc. e

psilo

npro

teob

acte

rium

clo

ne R

76-B

47,

tube

s of

the

ves

timen

tifer

an R

iftia

pac

hypt

ila (

AF

4492

45)

97AT

-co1

214

2750

.18

3E

psilo

npro

teob

acte

riaU

nc. e

psilo

npro

teob

acte

rium

clo

ne R

76-B

49,

tube

s of

the

ves

timen

tifer

an R

iftia

pac

hypt

ila (

AF

4492

42)

96AT

-co1

314

3150

.49

9E

psilo

npro

teob

acte

riaU

nc. e

psilo

npro

teob

acte

rium

clo

ne R

103-

B19

, tu

bes

of t

he v

estim

entif

eran

Rift

ia p

achy

ptila

(A

F44

9233

)97

AT-c

o15

1437

51.4

32

Eps

ilonp

rote

obac

teria

Unc

. bac

teriu

m P

VB

_OT

U_6

, m

icro

bial

mat

fro

m h

ydro

ther

mal

ven

t sy

stem

, H

awai

i (U

1510

6)92

AT-c

o16

1428

51.2

23

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m c

lone

BD

2-1,

dee

p-se

a se

dim

ents

(A

B01

5531

)98

AT-c

o23

1430

50.5

92

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m c

lone

NK

B11

, N

anka

i Tro

ugh

sedi

men

ts a

t 38

43 m

dee

p (A

B01

3263

)98

AT-c

s314

3751

.25

1E

psilo

npro

teob

acte

riaU

nc. e

psilo

npro

teob

acte

rium

clo

ne R

103-

B19

, tu

bes

of t

he v

estim

entif

eran

Rift

ia p

achy

ptila

(A

F44

9233

)95

AT-c

s714

2950

.91

12E

psilo

npro

teob

acte

riaU

nc. e

psilo

npro

teob

acte

rium

clo

ne R

103-

B19

, tu

bes

of t

he v

estim

entif

eran

Rift

ia p

achy

ptila

(A

F44

9233

)97

AT-c

s814

3350

.84

1E

psilo

npro

teob

acte

riaU

nc. e

psilo

npro

teob

acte

rium

clo

ne R

103-

B46

, tu

bes

of t

he v

estim

entif

eran

Rift

ia p

achy

ptila

(A

F44

9238

)98

AT-c

s10

1429

50.9

512

Eps

ilonp

rote

obac

teria

Unc

. bac

teriu

m P

VB

_OT

U_3

clo

ne P

VB

_12,

mic

robi

al m

at f

rom

hyd

roth

erm

al v

ent

syst

em,

Haw

aii (

U15

104)

96AT

-cs1

514

2849

.35

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m c

lone

R76

-B78

, tu

bes

of t

he v

estim

entif

eran

Rift

ia p

achy

ptila

(A

F44

9251

)97

AT-p

p614

3451

.26

2E

psilo

npro

teob

acte

riaU

nc. b

acte

rium

PV

B_O

TU

_2 c

lone

PV

B_7

, m

icro

bial

mat

fro

m h

ydro

ther

mal

ven

t sy

stem

, H

awai

i (U

1510

0)95

AT-p

p13

1434

47.5

67

Eps

ilonp

rote

obac

teria

Unc

. bac

teriu

m f

rom

ves

timen

tifer

an t

ubew

orm

(D

8306

1)94

AT-p

p26

1422

51.0

29

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m c

lone

BD

2-5,

dee

p-se

a se

dim

ents

(A

B01

5535

)95

AT-p

p27

1428

49.0

23

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m c

lone

R76

-B78

, tu

bes

of t

he v

estim

entif

eran

Rift

ia p

achy

ptila

(A

F44

9251

)97

AT-p

p46

1433

49.9

32

Eps

ilonp

rote

obac

teria

Unc

. eps

ilonp

rote

obac

teriu

m c

lone

R76

-B12

9, t

ubes

of

the

vest

imen

tifer

an R

iftia

pac

hypt

ila (

AF

4492

52)

97

a. S

eque

nces

98–

100%

iden

tical

, in

clud

ing

part

ial s

eque

nces

.C

lone

s na

med

co

wer

e ob

tain

ed f

rom

an

orga

nic

subs

trat

e, p

p fr

om p

umic

e, a

nd c

s fr

om a

n iro

n-ric

h su

bstr

ate.

Unc

., un

cultu

red.

Bacterial diversity in Mid-Atlantic Ridge hydrothermal systems 969


biogenic markers in fossil microfilaments from the EastPacific Rise (Philippot et al., 2002; 2003). The data pre-sented here are part of a general study devoted to imagingmetabolic processes at the scale of individual microbialfilaments using high resolution and non-destructive syn-chrotron techniques. In this paper, we focus on the micro-XANES results. All other results are presented in a com-plementary paper (Foriel et al., 2003). Micro-XANES

allows studying the relative amount and distribution of thedifferent oxidation states of sulphur (S-2 to S6+) at mm-scale, being therefore appropriate to obtain chemicalredox S maps in single microbial filaments. Figure 5shows the S redox maps obtained in a portion of a singlefilament taken from our iron-rich microcolonizer. We coulddetect three main sulphur species with X-ray peaks at2471 eV, 2478 eV and 2482 eV. The marked peak at 2482

Fig. 4. Maximum likelihood phylogenetic tree showing the diversity of epsilonproteobacteria found in Mid-Atlantic Ridge samples. Clones labelled AT-co correspond to the organic-rich substrate in microcolonizers, AT-pp to those from pumice and AT-cs to those from the iron-rich substrate. The clone AT-s3–19 was obtained from hydrothermal sediment. * Definition of Groups B/F and I/II are according to Corre et al. (2001) and Teske et al. (2002) respectively. Only bootstrap proportions above 50 are shown. The scale bar represents 10 substitutions per 100 positions for a unit branch length.



eV is characteristic of sulphate. The peaks at 2471 eV and2478 eV are typical of sulphide, -SH radicals or both. Thissuggests that S-metabolizing activities were taking placewithin cells. Although the technique does not allow dis-crimination between cell surface and cytoplasmic content,the distribution of S species in cells does not appearcompletely homogeneous (Fig. 5). This could indicate thatredox reactions involving changes in S states occur pref-erentially at certain places, for instance at cellular polesor discretely in metabolically more active membraneareas.

Discussion

Bacterial diversity in Mid-Atlantic Ridge sediments

Our results show the existence of a variety of bacterialphylotypes in hydrothermal sediment and virgin sub-strates exposed for two weeks to a fluid emission at theMid Atlantic Ridge. The diversity of the metal enrichedhydrothermal sediment was comparatively larger, cover-ing a wide spectrum of bacterial divisions as is the casein Pacific hydrothermal sediment (Teske et al., 2002). Thephylogenetic diversity of bacteria was not only large, butsome phylotypes are very divergent and could representpreviously undetected lineages within bacterial divisions.This is accompanied by a large diversity, and a consider-able degree of novelty, of microbial eukaryotes on thesame Pacific and Atlantic hydrothermal sediments (Edg-comb et al., 2002; López-García et al., 2003). However,as 16S rRNA-based surveys of bacterial diversity have

been carried out for more than 15 years (Olsen et al.,1986), the chances to find out novelty at great scale arelogically decreasing. It is interesting to note the detectionof the candidate division TM7 in deep-sea hydrothermalsediments, which extends the observation of its membersto this habitat as well. Although no TM7 representativehas ever been isolated in pure culture, there is a certainamount of information about their morphology and ultra-structure after their enrichment and identification by fluo-rescent in situ hybridization in wastewater treatmentsludge. They possess typical Gram positive walls and, atleast some of the species visualized, have sheaths(Hugenholtz et al., 2001). They have been detected in avariety of habitats, including terrestrial, freshwater andmarine environments, being frequently identified in anaer-obic environments (activated sludge, peat bogs) or inmetal-contaminated soils or aquifers.

Also divergent are several gammaproteobacterial phy-lotypes detected in the Rainbow sediment without clearclose neighbours in databases (Fig. 2). Gammaproteo-bacteria are very diverse in marine plankton and sedi-ments, and might be dominant in deep-sea plankton(Fuhrman and Davis, 1997; López-García et al., 2001).They appear also diverse in deep-sea vents. Mats of thesulphur-oxidizing Beggiatoa spp. and Thiomicrospira spp.are frequent in these areas (Jannash et al., 1985; 1989;Muyzer et al., 1995), and many deep-sea vent animalendosymbionts belong to this group as well (Distel et al.,1988). Gammaproteobacteria not only accounted for thelargest proportion of clones in our sediment libraries(44%) but they were also very diverse. Being metabolically

Fig. 5. Synchrotron micro-XANES spectrum of sulphur oxidation states recorded in a microbial filament recovered from fluid-exposed microcol-onizers. The rectangle in the bright-field picture indicates the region shown in the micro-XANES analyses (left panels). The scale bar is 1 mm. The three main peaks in the graphic shown above are characteristic of sulphate (2482 eV) and sulphide or -SH radicals, or both (2471 eV and 2478 eV). Associated intensity maps show the spatial distribution of the three absorption peaks in the portion of the microbial filament scanned using an energy resolution of 2 eV (see Philippot et al., 2002; 2003; and Foriel et al., 2003).



versatile, this subdivision is ecologically very successful(Madigan et al., 1997). In addition to the Gammaproteo-bacteria, we identified members of all the remaining sub-divisions of the Proteobacteria in the Rainbow sediment(Figs 2 and 4). The Alphaproteobacteria, represented inour case by two sequences (Fig. 2C), are also versatileand frequent in marine samples, being particularly diverseand abundant in surface waters (Morris et al., 2002). Wedo not have any clue about the physiology of these micro-organisms detected by their 16S rRNA and phylogeneti-cally distant from their closest cultivated relatives.However, it is interesting to note that sulphate-producingthiosulphate-oxidizing alphaproteobacteria were fre-quently isolated from slope sediments and hydrothermalsediments in the Galapagos (Teske et al., 2000). Wedetected two phylotypes ascribing to the Betaproteobac-teria. Members of this subdivision have not been detectedin hydrothermal sediment from the Guaymas basin in thePacific (Teske et al., 2002), but they have been identifiedin cold seep sediments at the Nankai Trough (nearly4000 m depth) (Li et al., 1999) and in a vent-cap systemdisplayed in the Mid-Atlantic Ridge (Reysenbach et al.,2000). One of the betaproteobacterial sequencesbranches at the base of a group comprising the parasiticNeisseria spp. and its lifestyle is unknown, but the otherphylotype corresponds to a member of the genus Hydro-genophilus (Fig. 2C). Hydrogenophilus spp. are thermo-philic and aerobic or facultative anaerobic, growingchemolithoautotrophically oxidizing hydrogen (Hayashiet al., 1999; Stohr et al., 2001). They have been identifiedin continental geothermal regions, but this is the first timethat members of this genus are detected in deep-seavents. The presence of one Hydrogenophilus phylotypeattests for the hydrothermal influence on the sediment,although temperature measurements were not taken dur-ing sampling. Only one sequence of epsilonproteobacte-ria was detected, and it clustered with the Group F/I(Fig. 4) that was highly represented in microcolonizers.Finally, as is frequently observed in anoxic sediments,members of the Deltaproteobacteria were relatively abun-dant. Most deltaproteobacteria are sulphate reducers andstrictly anaerobic. However, in our case, we failed to iden-tify bona fide lineages of typical sulphate reducers,although they are most likely present (the diversitydetected in the sediment is still far from saturation). CloneAT-s3–57 is very distantly related to Bdellovibrio spp.,bacterial predators that are widely distributed (Martin,2002; Snyder et al., 2002), and the rest of our clonesbelonged to the Myxobacteria, gliding microorganismsdisplaying complex social and developmental cycles. Myx-obacteria are heterotrophs that were thought to be exclu-sively aerobic some time ago, but anaerobic species havebeen recently isolated (Reichenbach, 1999; Sanford et al.,2002). Myxobacteria are frequent in cold marine sedi-

ments (Ravenschlag et al., 1999), but they have also beendetected in hydrothermal sediments at Loihi seamount(Hawai) and Guaymas (Moyer et al., 1995; Teske et al.,2002).

In addition to the Proteobacteria, and several lineagesscattered in various bacterial divisions (Nitrospirae, lowGC and high GC Gram-positives, TM7 and Planctomyce-tales), we found in sediment samples a variety of Cytoph-aga-Flavobacterium-Bacteroides and, most particularly, ofthe Acidobacteria/Holophaga (Fig. 2B). Most members ofthe Acidobacteria are not yet cultivated, but have beendetected by molecular tools in activated sludge or inmarine sediments (Ludwig et al., 1997; Ravenschlaget al., 1999). Their presence has not been previouslyreported in deep-sea vents, and they have been detectedneither in Guaymas nor in the Loihi seamount area (Moyeret al., 1995; Teske et al., 2002). However, they appear tobe quantitatively very abundant in shallow submarinevents near to Milos (Greece) (Sievert et al., 2000). Thisabundance was interpreted as a consequence of anallochthonous input of organic matter to the system, sincecultivated members of the group are heterotrophs. Acido-bacteria were also found to be very abundant in soilsfollowing a geothermal heating event at Yellowstone (Nor-ris et al., 2002). Their amount increased in soils incubatedat 50∞C indicating that various members of this group arethermophilic. This opens the possibility for the Rainbowlineages being also thermophilic.

A first comparison of the bacterial diversity in Atlanticand Pacific hydrothermal sediments reveals, on the onehand, some similarities, as sediment samples show adiversity much larger than that observed in fluid–seawaterinterface but, on the other hand, there are some differ-ences. Thus, neither alpha- and beta-proteobacterial noracidobacterial sequences were retrieved from the Guay-mas sediment (Teske et al., 2002). Conversely, somebacterial divisions were not detected in our Rainbow sed-iment (e.g. candidate divisions OP11, OP5 or OP3). Thiscould reflect the fact that diversity has not been fullyexplored in both cases and/or the introduction of differentbiases during library construction. However, there couldalso be true differences owing to the particular composi-tion of the sediments and nature of the hydrothermalinfluence. Rainbow sediment exhibits a mixed diversity.Close relatives to some of our sequences have beendetected in cold seeps or cold deep-sea sediments (Liet al., 1999; Bowman and McCuaig, 2003), but others areclosely related to lineages found in deep-sea vents ortypical thermophiles such as the Hydrogenophilus-likesequences or the presumably thermophilic Acidobacteria(Sievert et al., 2000). Therefore, the Rainbow sedimentstudied here might constitute a mid-point or a transitionarea between hot hydrothermal regions and the cold,deep-sea.



Ecological importance of the Epsilonproteobacteria and S-oxidizing metabolisms at the fluid–seawater interface

The bacterial diversity that we observed in the three sub-strates exposed to an emission source for two weeks atthe Mid-Atlantic Ridge shows three things. First, the col-onization process is very fast; in only two weeks the exposedsurfaces were completely covered by microorganismsshowing a variety of morphologies (Fig. 3). Second, thereis an overwhelming predominance of diverse epsilonpro-teobacteria in the clone libraries derived from the micro-colonizers and third, there is no particular specificity ofcolonization among the different substrates assayed, asall are almost exclusively colonized by epsilonproteobac-teria. Only one clone belonging to the CFB group wasfound in the organic-rich substrate but, even if this couldbe explained by the fact these microorganisms are het-erotrophs capable of degrading complex organics, thefinding of this clone is probably not significant when com-pared to the vast diversity of epsilonproteobacteria detected.

Our results thus confirm previous observations showingthe prevalence of epsilonproteobacteria in these habitats,and that these organisms colonize all the available sur-faces regardless their nature (animal teguments, mineralprecipitates or artificial substrates). In the Rainbow sedi-ment, we identified one epsilonproteobacterial sequence,and several sequences of this group were also identifiedin the Guaymas sediment. In both cases, these phylotypeswere related to those found at the fluid–seawater inter-face, but they were a minor fraction of the overall diversity.This suggests that epsilonproteobacteria are present indeep-sea sediment and when the conditions are par-ticularly appropriate for them, as is the case in vent sur-roundings where hydrogen sulphide is easily supplied,they bloom. Our data also show that epsilonproteobacte-ria are the first colonizers occupying virgin exposed sur-faces to the fluid–seawater interface. One can speculatethat young communities on any exposed surface in thisecological niche are formed basically by fast-developingepsilonproteobacteria and that, as communities evolve,they diversify and allow the creation of heterogeneousmicroniches suitable for the subsequent colonization byother bacteria. For instance, the creation of anoxic pock-ets and the production of sulphate as one abundantsecondary metabolite (see below) would favour the settle-ment of sulphate-reducing deltaproteobacteria that wouldclose the S cycle. Deltaproteobacteria are indeedobserved in mature surfaces where epsilonproteobacteriadominate, such as tubes of big specimens of Riftia pac-hyptila, or mucous tubes of Paralvinella palmiformis (Alainet al., 2002a; López-García et al., 2002).

A prominent role of S-dependent metabolisms, andmost particularly, the oxidation of reduced forms of sul-phur in warm areas of deep-sea vents is deduced from

the availability of hydrogen sulphide and other reduced Sspecies in combination with geochemical constraints(McCollom and Shock, 1997). This is further supported bythe isolation of many microbial species that metabolizesulphur to gain energy. As mentioned above, many gam-maproteobacteria oxidize sulphur or thiosulphate, but alsoothers, including alphaproteobacterial or flavobacterialspecies from vents (Teske et al., 2000). Sulphate-reducing deltaproteobacteria would assure the completionof the S cycle in anoxic sediments and anoxic micron-iches. Sulphur-based energy reactions appear to becrucial for the metabolism of deep-sea vent epsilonpro-teobacteria, either as electron acceptor or as electrondonor. Hydrogen sulphide oxidation would be a readilyenergy source for these organisms. However, most culti-vation attempts use either hydrogen (electron donor) orsulphur (electron donor or acceptor), and therefore, thereis generally no information about the capacity to directlyoxidize hydrogen sulphide by the few strains of epsilon-proteobacteria that have been isolated from the most rep-resented groups around vents (Takai et al., 2003).Unfortunately, metabolic activities displayed in situ aredifficult to measure, and those shown in the laboratory donot necessarily reflect what microorganisms do in nature.Trying to investigate if S metabolism could be visualizedin cells having grown in deep-sea vents, we used high-resolution synchrotron-based techniques to map differentS oxidation states at a single filament-level (Philippotet al., 2002; 2003). The filament shown here (Fig. 5) wasretrieved from the iron-rich substrate of our microcoloniz-ers. As all phylotypes retrieved from this substratebelonged to this phylogenetic group, the filament mostprobably corresponded to an epsilonproteobacterium. Ourresults showing the co-occurrence in the filament of dif-ferent S species, including sulphide and sulphate, suggestthat cells were actively metabolizing sulphur (see Forielet al., 2003 for further details). Furthermore, they werelikely oxidizing sulphide to sulphate and not the opposite,as the presence of sulphate reducing bacteria (typicallydeltaproteobacteria) was not detected by 16S rDNAsequencing, and as no epsilonproteobacteria is known toreduce sulphate. Additionally, the filament morphologyfavours the exposure to the fluid–seawater interfacewhere oxygen is present (sulphate reducing bacteria arestrict anaerobes). Coupling micro-XANES analysis with insitu fluorescent hybridization using phylogenetic probescould be useful to link metabolic information to communitystructure in these environments.


Sampling and scanning electron microscopy

Samples were taken with the aid of the Remote Operated



Vehicle (ROV) Victor during the French cruise ATOS 2001 tothe Mid-Atlantic Ridge hydrothermal area, as previouslydescribed (López-García et al., 2003). A sediment core wasobtained from Rainbow hydrothermal sediment (36∞6¢N,33∞11¢W, depth 2264 m). A fraction of the sediment corre-sponding to the ~1 cm upper part was removed in a laminarflux chamber and frozen in liquid nitrogen until use. Home-made sterile microcolonizers consisting of different sub-strates placed into extensively perforated 50 ml polypropy-lene Corning tubes were deployed adjacent to a fluidemission at the Tour Eiffel chimney (Lucky Strike site,37∞17¢N, 32∞16¢W, depth 1695 m) for 15 days (see Fig. 2 inLópez-García et al., 2003), and collected in a close sterilecontainer by the ROV Victor. Two of them consisted of aninert nylon mesh embedding, respectively, a meat-basedsubstrate (autoclaved ground ham) and several pieces of 0.5-mm diameter iron wire. A third one was made of pumicefragments. The container was opened in a laminar flux cham-ber on board, and the microcolonizers were stored at 4∞C in75% ethanol, 2% NaCl. For scanning electron microscopyobservation, samples were dehydrated in increasing ethanolconcentrations (50%, 70%, 90% and 100%), critical-point-dried and gold coated. Observation was carried out with aJEOL (JSM 840 A) scanning electron microscope (SEM)operating at 17 kV at the Service de Microscopie Electron-ique de l’Institut Fédératif de Recherche Biologie Intégrative(Paris, France).

Synchrotron micro-X-ray absorption near-edge spectroscopy (micro-XANES)

Micro-XANES spectra of the different oxidation states of sul-phur (S-2 to S6+) were obtained using beamline ID21 (Susiniet al., 2002) at the European Synchrotron Radiation Facility(Grenoble, France). Analyses were performed following theprocedure described by Philippot et al. (2002; 2003) andForiel et al. (2003). For this, microbial filaments scratched outfrom different microcolonizer substrates were deposited overa Mylar® polyester film, and placed on a metallic grid.Samples were then subjected to incident beam energiesfrom 20 eV below the main absorption edge energy of sul-phate (2482 eV for S6+) to 20 eV above this value. Thestructure of the S K absorption edge was scanned in the nearedge region. The microscope used a Fresnel zone-plate asa focusing lens and delivered a microbeam of 0.8 ¥ 0.8 mm2

with a measured photon flux of about 108 photon/s (Davidet al., 2000). Recorded micro-XANES spectra were com-pared with spectra of pure standard products that were anal-ysed during the same experimental session to optimize thecorrection procedure and interpretation of the X-ray spectra.

Nucleic acid extraction

Nucleic acids were purified from small fragments of the dif-ferent microcolonizer substrates (~200–300 ml of organic andiron rich plastic mesh, and ~100 ml of pumice) and fromapproximately 50–100 ml of hydrothermal sediment. Therespective substrate volumes were separated from mothersamples in Eppendorf tubes in a laminar flux chamber. Micro-colonizer samples were rehydrated with phosphate saline

buffer (130 mM NaCl, 10 mM phosphate buffer, pH 7.7,PBS). Phosphate-buffered saline was also added to the sed-iment to a same final volume of 0.5 ml. Samples were thensubjected to six freezing/thawing cycles in liquid nitrogen tofacilitate cell lysis. Subsequently, 80 mg ml-1 proteinase K, 1%SDS, 1.4 M NaCl, 0.2 b-mercaptoethanol and 2% CTAB (finalconcentrations) were added sequentially. Lysis suspensionswere incubated overnight at 55∞C. Lysates were extractedonce with hot phenol (65∞C), once with phenol-chloroform-isoamylalcohol, and once with chloroform-isoamyl-alcohol.Nucleic acids were concentrated by ethanol precipitation.

16S ribosomal RNA gene libraries and sequencing

16S rRNA genes were amplified by PCR using different com-binations of the bacteria-specific forward primers B-27F(AGAGTTTGATCCTGGCTCAG), ANT-1 (AGAGTTTGATCATGGCTCAG) and B-63F (CAGGCCTAACACATGCAAGTC),and the prokaryote-specific reverse primer 1492R (GGTTACCTTGTTACGACTT). PCR reactions were normallyperformed under the following conditions: 30 cycles (dena-turation at 94∞C for 15 s, annealing at 55∞C for 30 s, exten-sion at 72∞C for 2 min) preceded by 2 min denaturation at94∞C, and followed by 10 min extension at 72∞C. For someamplification reactions, dimethyl sulphoxide was added to afinal concentration of 3–5%. A total of seven bacterial rDNAclone libraries (four for sediment DNA and three for microcol-onizers) were constructed using the Topo TA Cloning system(Invitrogen) following the instructions provided by the manu-facturers. After plating, positive transformants were screenedby PCR amplification of inserts using flanking vector primers.A total of 181 expected-size amplicons from these librarieswas partially sequenced (Genome Express) with the primer1492R. After preliminary phylogenetic analysis, 61 clonesrepresentative of the phylogenetic diversity found were cho-sen for complete sequencing using primer B-27F. Two ofthese clones were discarded from subsequent phylogeneticanalyses as possible chimeras after sequence checking bythe Chimera Detection program of the Ribosomal DatabaseProject II (Cole et al., 2003) and visual inspection of thesequence alignment.

Phylogenetic analysis

Closest relatives to our sequences were identified in data-bases by BLAST (Altschul et al., 1997) and retrieved fromGenBank (http://www.ncbi.nlm.nih.gov). Sequences wereautomatically aligned using the program BABA (H. Philippe,personal communication) to a 16S rRNA gene alignmentcontaining ~16 600 sequences. The multiple alignment wasthen manually edited using the program ED from the MUST

package (Philippe, 1993). A preliminary phylogenetic analy-sis of all partial sequences was done by distance methods(neighbour-joining, NJ) using the program MUST, allowing theidentification of identical or nearly identical sequences andthe selection of clones for complete sequencing. For moredetailed phylogenetic analyses of the 59 complete bacterialsequences, we selected five subsets of sequences includingthe closest relatives to our clones in databases as well ascultivated representatives. Gaps and ambiguously aligned

http://www.ncbi.nlm.nih.gov



positions were excluded from our analyses. The relativelylow number of positions used as compared to full-lengthsequences is due to the inclusion of environmentalsequences shorter than those determined in this work, butthat were closely related. The five data sets, and the numberof positions used in phylogenetic reconstruction, were asfollows: 38 sequences covering the major bacterial divisionsand including an archaeal outgroup (988 positions, Fig. 1A),46 sequences covering Planctomyces, CFB and Acido-bacteria divisions (861 positions, Fig. 1B), 41 sequencescorresponding to the a-, b- and deltaproteobacteria (929positions, Fig. 1C), 44 sequences covering the gammapro-teobacteria and including a betaproteobacterial outgroup(968 positions, Fig. 1D), and 52 sequences covering theepsilonproteobacteria and including a deltaproteobacterialoutgroup (944 positions, Fig. 1E). Maximum likelihood (ML)trees were done using TREEFINDER (Jobb, 2002) applying ageneral time reversible model of sequence evolution (GTR),taking among-site rate variation into account by using aneight-category discrete approximation of a G distribution(invariable sites are included in one of the categories). Thea parameters of the G distribution estimated from the differ-ent sequence sets were: 0.35 (set of sequences in Fig. 1A),0.26 (Fig. 1B), 0.25 (Fig. 1C), 0.19 (Fig. 1D) and 0.21(Fig. 1E). Bootstrap proportions were inferred using 1000replicates by NJ applying a G law using the estimated avalues for each data set.

Nucleotide sequence accession numbers

The sequences reported in this study were submitted toGenBank with accession numbers AY225602 (AT-s63),AY225603 (AT-s3–44), AY225604 (AT-s71), AY225605(AT-s3–41), AY225606 (AT-s3–57), AY225607 (AT-s3–34),AY225608 (AT-s3–60), AY225609 (AT-s3–66), AY225610 (AT-pp13), AY225611 (AT-co15), AY225612 (AT-co23), AY225613(AT-pp6), AY225614 (AT-cs3), AY225615 (AT-cs10),AY225616 (AT-co11), AY225617 (AT-co16), AY225618(AT-co12), AY225619 (AT-s3–19), AY225620 (AT-pp26),AY225621 (AT-pp46), AY225622 (AT-cs8), AY225623 (AT-cs15), AY225624 (AT-pp27), AY225625 (AT-cs7), AY225626(AT-co13), AY225627 (AT-s68), AY225628 (AT-s3–68),AY225629 (AT-s3–26), AY225630 (AT-s43), AY225631 (AT-s2–13), AY225632 (AT-s3–1), AY225633 (AT-s16), AY225634(AT-s26), AY225635 (AT-s80), AY225636 (AT-s2–59),AY225637 (AT-s75), AY225638 (AT-s3–48), AY225639 (AT-s3–25), AY225640 (AT-s2), AY225641 (AT-s3–24), AY225642(AT-s3–59), AY225643 (AT-s3–37), AY225644 (AT-s65),AY225645 (AT-s36), AY225646 (AT-s3–28), AY225647 (AT-s3–23), AY225648 (AT-s3–42), AY225649 (AT-s54),AY225650 (AT-s2–57), AY225651 (AT-s3–56), AY225652(AT-s2–38), AY225653 (AT-s2–18), AY225654 (AT-s83),AY225655 (AT-s2–33), AY225656 (AT-s3–3), AY225657 (AT-s30), AY225658 (AT-s92), AY225659 (AT-s98) and AY225660(AT-co1).

Acknowledgements

We are grateful to Magali Zbinden for efficient sample han-dling on board during the ATOS 2001 cruise, Hervé Philippe

for the sequence alignment program BABA, Jean Cauzid andBénédicte Ménez for late night companionship while perform-ing micro-XANES analysis at the ESRF, Françoise Gaill forsampling access, and Michèle Grasset for technical assis-tance in electron microscopy analyses. This work was sup-ported by the French CNRS-INSU Program GEOMEX.

References

Alain, K., Olagnon, M., Desbruyères, D., Pagé, A., Barbier,G., Juniper, S.K., et al. (2002a) Phylogenetic characteriza-tion of the bacterial assemblage associated with mucoussecretions of the hydrothermal vent polychaete Paralvinellapalmiformis. FEMS Microbiol Ecol 42: 463–476.

Alain, K., Querellou, J., Lesongeur, F., Pignet, P., Crassous,P., Raguenes, G., et al. (2002b) Caminibacter hydrogeni-philus gen nov., sp. nov., a novel thermophilic, hydrogen-oxidizing bacterium isolated from an East Pacific Risehydrothermal vent. Int J Syst Evol Microbiol 52: 1317–1323.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.,Zhang, Z., Miller, W., and Lipman, D.J. (1997) GappedBLAST and PSI-BLAST: a new generation of protein databasesearch programs. Nucleic Acids Res 25: 3389–3402.

Bowman, J.P., and McCuaig, R.D. (2003) Biodiversity, com-munity structural shifts, and biogeography of prokaryoteswithin Antarctic continental self sediment. Appl EnvironMicrobiol 69: 2463–2483.

Campbell, B.J., Jeanthon, C., Kostka, J.E., Luther, G.W., andCary, S.C. (2001) Growth and phylogenetic properties ofnovel bacteria belonging to the epsilon subdivision of theProteobacteria enriched from Alvinella pompejana anddeep-sea hydrothermal vents. Appl Environ Microbiol 67:4566–4572.

Cave, R.R., German, C.R., Thomson, J., and Nesbitt, R.W.(2002) Fluxes to sediments underlying the Rainbow hydro-thermal plume at 36∞14¢N on the Mid-Atlantic Ridge.Geochim Cosmochim Acta 66: 1905–1923.

Cole, J.R., Chai, B., Marsh, T.L., Farris, R.J., Wang, Q.,Kulam, S.A., et al. (2003) The Ribosomal Database Project(RDP-II): previewing a new autoaligner that allows regularupdates and the new prokaryotic taxonomy. Nucleic AcidsRes 31: 442–443.

Corre, E., Reysenbach, A.L., and Prieur, D. (2001) Epsilon-proteobacterial diversity from a deep-sea hydrothermalvent on the Mid-Atlantic Ridge. FEMS Microbiol Lett 205:329–335.

David, C., Kaulich, B., Barrett, R., Salomé, M., and Susini, J.(2000) High-resolution lenses for sub-100 nm x-ray fluo-rescence microscopy. Appl Phys Lett 77: 3851–3853.

Distel, D.L., Lane, D.J., Olsen, G.J., Giovannoni, S.J., Pace,B., Pace, N.R., et al. (1988) Sulfur-oxidizing bacterial endo-symbionts: analysis of phylogeny and specificity by 16SrRNA sequences. J Bacteriol 170: 2506–2510.

Douville, E., Charlou, J.L., Oelkers, E.H., Bienvenu, P., JoveColon, C.F., Donval, J.P., et al. (2002) The rainbow ventfluids (36∞14¢N, MAR): the influence of ultramafic rocks andphase separation on trace metal content in Mid-AtlanticRidge hydrothermal fluids. Chem Geol 184: 37–48.

Edgcomb, V.P., Kysela, D.T., Teske, A., De Vera Gomez, A.,and Sogin, M.L. (2002) Benthic eukaryotic diversity in the



Guaymas Basin hydrothermal vent environment. Proc NatlAcad Sci USA 99: 7658–7662.

Foriel, J., Philippot, P., Cauzid, J., Susini, J., Dumas, P.,Khodja, H., et al. (2003) High resolution synchrotron-basedimaging of sulfur oxidation states in individual microfossilsand contemporary microbial filaments. Geochim Cosmo-chim Acta (in press).

Fuhrman, J.A., and Davis, A.A. (1997) Widespread Archaeaand novel Bacteria from the deep sea as shown by 16SrRNA gene sequences. Mar Ecol Prog Series 150: 275–285.

Haddad, A., Camacho, F., Durand, P., and Cary, S.C. (1995)Phylogenetic characterization of the epibiotic bacteriaassociated with the hydrothermal vent polychaete Alvinellapompejana. Appl Environ Microbiol 61: 1679–1687.

Hayashi, N.R., Ishida, T., Yokota, A., Kodama, T., and Iga-rashi, Y. (1999) Hydrogenophilus thermoluteolus gen nov.,sp. nov., a thermophilic, facultatively chemolithoautotro-phic, hydrogen-oxidizing bacterium. Int J Syst Bacteriol 49:783–786.

Hugenholtz, P., Tyson, G.W., Webb, R.I., Wagner, A.M., andBlackall, L.L. (2001) Investigation of candidate divisionTM7, a recently recognized major lineage of the domainBacteria with no known pure-culture representatives. ApplEnviron Microbiol 67: 411–419.

Jannasch, H.W., and Mottl, M.J. (1985) Geomicrobiology ofdeep-sea hydrothermal vents. Science 229: 717–725.

Jannasch, H.W., Wirsen, C.O., Nelson, D.C., and Robertson,L.A. (1985) Thiomicrospira crunogena sp. nov., a colorlesssulfur-oxidizing bacterium from a deep-sea hydrothermalvent. Int J Syst Bacteriol 35: 422–424.

Jannasch, H.W., Nelson, D.C., and Wirsen, C.O. (1989) Mas-sive natural occurrence of unusually large bacteria (Beg-giatoa sp.) at a hydrothermal deep-sea vent site. Nature342: 834–836.

Jeanthon, C. (2000) Molecular ecology of hydrothermal ventmicrobial communities. Antonie Van Leeuwenhoek 77:117–133.

Jobb, G. (2002) treefinder. [WWW document]. URLhttp://www.treefinder.de

Kuenen, J.G., Robertson, L.A., and Tuovinen, O.H. (1992)The genera Thiobacillus, Thiomicrospira and Thiosphaera.In The Genera Thiobacillus, Thiomicrospira, andThiosphaera. Balows, A., Trüper, H.G., Dworkin, M.,Harder W., and Schleifer, K.H. (eds). Vol. III, New York:Springer-Verlag.

Li, L., Guezennec, J., Nichols, P., Henry, P., Yanagibayashi,M., and Kato, C. (1999) Microbial diversity in NankaiTrough sediments at a depths of 3,843 m. J Oceanogr 55:635–642.

Longnecker, K., and Reysenbach, A. (2001) Expansion of thegeographic distribution of a novel lineage of epsilon-Proteobacteria to a hydrothermal vent site on the SouthernEast Pacific Rise. FEMS Microbiol Ecol 35: 287–293.

López-García, P., López-López, A., Moreira, D., andRodríguez-Valera, F. (2001) Diversity of free-livingprokaryotes from a deep-sea site at the Antarctic PolarFront. FEMS Microbiol Ecol 36: 193–202.

López-García, P., Gaill, F., and Moreira, D. (2002) Widebacterial diversity associated with tubes of the vent wormRiftia pachyptila. Environ Microbiol 4: 204–215.

López-García, P., Philippe, H., Gaill, F., and Moreira, D.(2003) Autochthonous eukaryotic diversity in hydrothermalsediment and experimental micro-colonizers at the Mid-Atlantic Ridge. Proc Natl Acad Sci USA 100: 697–702.

Ludwig, W., Bauer, S.H., Bauer, M., Held, I., Kirchhof, G.,Schulze, R., et al. (1997) Detection and in situ identificationof representatives of a widely distributed new bacterialphylum. FEMS Microbiol Lett 153: 181–190.

Madigan, M.T., Martinko, J.M., and Parker, J. (1997) BrockBiology of Microorganisms. New Jersey: Prentice Hall.

Martin, M.O. (2002) Predatory prokaryotes: an emergingresearch opportunity. J Mol Microbiol Biotechnol 4: 467–477.

McCollom, T.M., and Shock, E.L. (1997) Geochemical con-straints on chemolithoautotrophic metabolism by microor-ganisms in seafloor hydrothermal systems. GeochimCosmochim Acta 61: 4375–4391.

Miroshnichenko, M.L., Kostrikina, N.A., L’Haridon, S.,Jeanthon, C., Hippe, H., Stackebrandt, E., and Bonch-Osmolovskaya, E.A. (2002) Nautilia lithotrophica gen nov.,sp. nov., a thermophilic sulfur-reducing epsilon-proteobac-terium isolated from a deep-sea hydrothermal vent. Int JSyst Evol Microbiol 52: 1299–1304.

Morris, R.M., Rappe, M.S., Connon, S.A., Vergin, K.L., Sie-bold, W.A., Carlson, C.A., and Giovannoni, S.J. (2002)SAR11 clade dominates ocean surface bacterioplanktoncommunities. Nature 420: 806–810.

Moyer, C.L., Dobbs, F.C., and Karl, D.M. (1995) Phylogeneticdiversity of the bacterial community from a microbial matat an active, hydrothermal vent site, Loihi Seamount,Hawaii. Appl Environ Microbiol 61: 1555–1562.

Muyzer, G., Teske, A., Wirsen, C.O., and Jannasch, H.W.(1995) Phylogenetic relationships of Thiomicrospira spe-cies and their identification in deep-sea hydrothermal ventsamples by denaturing gradient gel electrophoresis of 16SrDNA fragments. Arch Microbiol 164: 165–172.

Nisbet, E.G., and Sleep, N.H. (2001) The habitat and natureof early life. Nature 409: 1083–1091.

Norris, T.B., Wraith, J.M., Castenholz, R.W., and McDermott,T.R. (2002) Soil microbial community structure across athermal gradient following a geothermal heating event.Appl Environ Microbiol 68: 6300–6309.

Olsen, G.J., Lane, D.J., Giovannoni, S.J., Pace, N.R., andStahl, D.A. (1986) Microbial ecology and evolution: a ribo-somal RNA approach. Annu Rev Microbiol 40: 337–365.

Philippe, H. (1993) MUST, a computer package of Manage-ment Utilities for Sequences and Trees. Nucleic Acids Res21: 5264–5272.

Philippot, P., Foriel, J., Cauzid, J., Susini, J., Ménez, B., andSomogyi, A. (2002) Imaging sulfur-metabolism activities inindividual filamentous microfossils. Highlights ESRF 2002:85–87.

Philippot, P., Foriel, J., Susini, J., Khodja, H., Grassineau, N.,and Fouquet, Y. (2003) High-resolution imaging of transi-tion metal and sulfur-redox distribution in individual micro-fossils. J Phys IV 104: 381–384.

Polz, M.F., and Cavanaugh, C.M. (1995) Dominance of onebacterial phylotype at a Mid-Atlantic Ridge hydrothermalvent site. Proc Natl Acad Sci USA 92: 7232–7236.

Prieur, D. (1997) Microbiology of deep-sea hydrothermalvents. Trends Biotechnol 15: 242–244.

http://www.treefinder.de



Ravenschlag, K., Sahm, K., Pernthaler, J., and Amann, R.(1999) High bacterial diversity in permanently cold marinesediments. Appl Environ Microbiol 65: 3982–3989.

Reichenbach, H. (1999) The ecology of the myxobacteria.Environ Microbiol 1: 15–21.

Reysenbach, A.L., and Cady, S.L. (2001) Microbiology ofancient and modern hydrothermal systems. Trends Micro-biol 9: 79–86.

Reysenbach, A.L., and Shock, E. (2002) Merging genomeswith geochemistry in hydrothermal ecosystems. Science296: 1077–1082.

Reysenbach, A.L., Longnecker, K., and Kirshtein, J. (2000)Novel bacterial and archaeal lineages from an in situgrowth chamber deployed at a Mid-Atlantic Ridge hydro-thermal vent. Appl Environ Microbiol 66: 3798–3806.

Sanford, R.A., Cole, J.R., and Tiedje, J.M. (2002) Character-ization and description of Anaeromyxobacter dehalogen-ans gen nov., sp. nov., an aryl-halorespiring facultativeanaerobic myxobacterium. Appl Environ Microbiol 68:893–900.

Sievert, S.M., Kuever, J., and Muyzer, G. (2000) Identificationof 16S ribosomal DNA-defined bacterial populations at ashallow submarine hydrothermal vent near Milos Island(Greece). Appl Environ Microbiol 66: 3102–3109.

Snyder, A.R., Williams, H.N., Baer, M.L., Walker, K.E., andStine, O.C. (2002) 16S rDNA sequence analysis of envi-ronmental Bdellovibrio-and-like organisms (BALO) revealsextensive diversity. Int J Syst Evol Microbiol 52: 2089–2094.

Stetter, K.O. (1999) Extremophiles and their adaptation to hotenvironments. FEBS Lett 452: 22–25.

Stohr, R., Waberski, A., Liesack, W., Volker, H., Wehmeyer,

U., and Thomm, M. (2001) Hydrogenophilus hirschii sp.nov., a novel thermophilic hydrogen-oxidizing beta-proteobacterium isolated from Yellowstone National Park.Int J Syst Evol Microbiol 51: 481–488.

Susini, J., Salome, M., Fayard, B., Ortega, R., and Kaulich,B. (2002) The scanning X-Ray microprobe at the ESRF ‘X-Ray microscopy’ beamline. Surface Rev Lett 9: 203–211.

Takai, K., Inagaki, F., Nakagawa, S., Hirayama, H., Nunoura,T., Sako, Y., et al. (2003) Isolation and phylogeneticdiversity of members of previously uncultivated epsilon-Proteobacteria in deep-sea hydrothermal fields. FEMSMicrobiol Ecol 218: 167–174.

Taylor, C.D., Wirsen, C.O., and Gaill, F. (1999) Rapid micro-bial production of filamentous sulfur mats at hydrothermalvents. Appl Environ Microbiol 65: 2253–2255.

Teske, A., Brinkhoff, T., Muyzer, G., Moser, D.P., Rethmeier,J., and Jannasch, H.W. (2000) Diversity of thiosulfate-oxidizing bacteria from marine sediments and hydrother-mal vents. Appl Environ Microbiol 66: 3125–3133.

Teske, A., Hinrichs, K.U., Edgcomb, V., de Vera Gomez, A.,Kysela, D., Sylva, S.P., et al. (2002) Microbial diversity ofhydrothermal sediments in the Guaymas basin: evidencefor anaerobic methanotrophic communities. Appl EnvironMicrobiol 68: 1994–2007.

Thar, R., and Fenchel, T. (2001) True chemotaxis in oxygengradients of the sulfur-oxidizing bacterium Thiovulummajus. Appl Environ Microbiol 67: 3299–3303.

Wirsen, C.O., Sievert, S.M., Cavanaugh, C.M., Molyneaux,S.J., Ahmad, A., Taylor, L.T., et al. (2002) Characterizationof an autotrophic sulfide-oxidizing marine Arcobacter sp.that produces filamentous sulfur. Appl Environ Microbiol68: 316–325.

bacterial diversity in hydrothermal sediment and epsilonproteobacterial dominance in

Documents