Mariana Forearc Serpentinite Mud Volcanoes Harbor Novel Communities of Extremophilic Archaea

12 Pages·2013·0.62 MB·English

GeomicrobiologyJournal(2013)30,430–441
Copyright(cid:2)C Taylor&FrancisGroup,LLC
ISSN:0149-0451print /1521-0529online
DOI:10.1080/01490451.2012.705226
Mariana Forearc Serpentinite Mud Volcanoes Harbor Novel
Communities of Extremophilic Archaea
ANDREAC.CURTIS1,C.GEOFFERYWHEAT2,PATRICIAFRYER3,andCRAIGL.MOYER1∗
1BiologyDepartment,WesternWashingtonUniversity,Bellingham,Washington,USA
2GlobalUnderseaResearchUnit,UniversityofAlaskaFairbanks,Fairbanks,Alaska,USA
3HawaiiInstituteofGeophysicsandPlanetology,SchoolofOceanandEarthScienceandTechnology,UniversityofHawaii,
Honolulu,Hawaii,USA
ReceivedMarch2011,AcceptedJune2012
ExtremophilicarchaealcommunitieslivinginserpentinizedmudsinfluencedbypH12.5deep-slabderivedfluidsweredetectedand
theirrichnessandrelatednessassessedfromacrosssevenserpentinitemudvolcanoeslocatedalongtheMarianaforearc.Inaddition,
samplesfromtwonearsurfacecoresections(HolesDandE)atODPSite1200fromSouthChamorroweresubjectedtoSSUrDNA
clonelibraryandphylogeneticanalysisresultinginthediscoveryofseveralnoveloperationaltaxonomicunits(OTUs).Fivedominant
OTUsofArchaeafromHole1200DandsixdominantOTUsofArchaeafromHole1200Eweredeterminedbygroupshavingthree
or more clones. Terminal-restriction fragment length polymorphism (T-RFLP) analysis revealed all of the dominant OTUs were
detected within both clone libraries. Cluster analysis of the T-RFLP data revealed archaeal community structures from sites on
BigBlueandBlueMoontobeanalogoustotheSouthChamorroHole1200Esite.Theseuniquearchaealcommunityfingerprints
resulted from an abundance of potential methane-oxidizing and sulfate-reducing phylotypes. This study used deep-sea sediment
coringtechniquesacrosssevendifferentmudvolcanoesalongtheentireMarianaforearcsystem.Thediscoveryanddetectionofboth
novel Euryarchaeota and Marine Benthic Group B Crenarcheaota phylotypes could be efficacious archaeal indicator populations
involvedwithanaerobicmethaneoxidation(AMO)andsulfatereductionfueledbydeepsubsurfaceserpentinizationreactions.
Keywords:Archaea,communitystructure,molecularecology,subsurfacemicrobiology
Introduction theoverridingplateintoahydratedcompositeofserpentinite,
muds and fluids that extrude from fault-derived conduits re-
TheMarianaconvergentplatemargin,anonaccretionarysub- sulting in giant (reaching sizes up to 50 km in diameter and
duction zone (Uyeda 1982), lies along the eastern boundary 2kminheight)mudvolcanoes(Fryeretal.1999;Mottletal.
ofthePhilippineplateinthewesternPacific(Figure1).Since 2003,2004).Thesemudvolcanoesformazoneofseamounts
theEocene(∼45Ma)thePacificplatehasbeensubductingbe- from 30 to 120 km behind the trench axis and often have
neaththePhilippineplate,ultimatelygivingrisetotheMari- active springs that deliver water to the seafloor that is typi-
anaforearc,aregionofserpentinitemudvolcanoesformedbe- callyfresherthanseawaterandhighlyalkalinewithpH∼12.5
tweenthetrenchandvolcanicarc(Fryeretal.1999;Maekawa (Fryeretal.2006;Hulmeetal.2010;Mottletal.2003,2004;
etal.1993).Dehydrationreactions,inducedbyincreasingpres- Wheatetal.2008).
sure and temperature within the down-going slab, transform In 1996, springs were discovered on the summit of South
ChamorroSeamount(thesouthernmostmudvolcanoonthe
forearc) 85 km arcward of the Mariana Trench (Fryer and
WearegratefultothecaptainandcrewofboththeR/VJOIDES
Mottl 1997). In 2001, as part of Ocean Drilling Program
ResolutionandR/VThomasG.Thompsonfortheirassistancein
(ODP) Leg 195, advanced piston core (APC) samples were
collecting samples as well as the Ocean Drilling Program. This
recoveredandwhole-roundcoresectionswereacquiredfrom
projectwasfundedinpartbyWesternWashingtonUniversity’s
Site 1200 near one of the active springs to investigate the
Office of Research and Sponsored Programs, U.S. Science Sup-
structure and diversity of microbial populations that exist in
port Program (USSSP) and the Consortium for Ocean Leader-
ship,andbyNationalScienceFoundationawardOCE-0727086 the shallow subsurface (Shipboard Scientific Party 2002). In
(toCM). 2003,weobtainedpiston,triggerandgravitycores(inaddition
∗Address correspondence to Craig L. Moyer, Biology Depart- tousingtheROVJasonIItocollectpush-coresamples)from
ment MS#9160, Western Washington University, 516 High seven mud volcanoes (Fryer et al. 2006). The sampled mud
Street,Bellingham,WA98225,USA;Email:[email protected]. volcanoes include (from north to south) Pacman, Blip, Nip,
wwu.edu Quaker,BigBlue,BlueMoonandSouthChamorro(Figure1).
MarianaMudVolcanoesHarborExtremophilicArchaea 431
continuestoremaintheleastunderstood(Orcuttetal.2011;
Whitmanetal.1998).Bioticandabioticchemicalreactionsas
wellasgeologicalprocessesassociatedwithmicrobialpopula-
tionsinthedeepsubsurfacebiosphereareinterlinked,butlit-
tleisknownofthecomplexityofthoselinkages(Gold1992).
Seamounts are also now being recognized as environments
hostingenhancedmicrobialdiversityandbiomass(Emerson
andMoyer2010).Previousstudieshaveelucidatedtheimpor-
tance of hydrogen-driven microbial communities; hydrogen
is a by-product of the serpentinization process (Takai et al.
2004a).
FluidsemanatingfrommudvolcanoesalongtheMariana
forearcareultimatelyderivedfromthesubductingslab(yield-
ing an ultrahigh pH) and to date, the Izu-Bonin-Mariana
forearc represents the only known example of slab-derived
rocks being returned to the surface in an active subduction
zone;howeveringeologictimesuchprocesseswerecommon
(Fryer et al. 2006; Maekawa et al. 1993). As such, this sub-
surface system not only offers the possibility of uncovering
an unique microbial ecosystem based on chemolithotrophic
primary producers, but additionally yields the potential for
findingpossibleanalogsforextraterrestriallife(Schulteetal.
2006) and better understanding origins of life on Earth
(MartinandRussell2007;Teskeetal.2003).
The goal of this study was to detect and discover ex-
tremophilicarchaealpopulations,anddeterminetherichness
and relatedness of these populations within their respective
communities. Archaeal communities from seven seamounts
wereassessedwithT-RFLP,whiletwositesatopthesouthern-
most seamount, South Chamorro, were additionally inves-
tigated using clone library analysis (Figure 1). T-RFLP is
Fig.1.BathymetricmapofthesouthernMarianaarcandforearc,
a rapid and reproducible SSU rDNA fingerprinting tech-
ranging from northernmost mud volcano (Conical Seamount)
tosouthernmost(SouthChamorroSeamount)nearesttoGuam nique that has the potential for reliably reflecting changes
(afterFryeretal.1999)(colorfigureavailableonline). in community structure and diversity allowing for compar-
isonsofrichnessandevennessamongsamples(Hartmannand
Widmer2008;Marsh2005).
Estimates for the chemical composition of fluids at depth Althoughabacterium(Marinobacteralkaliphilus)waspre-
atthesesitesrevealsaconsistentvariationinthefluidcompo- viouslyisolatedfromsubseafloormuds(Takaietal.2005),no
sitionwithdistancefromthetrenchaxis,whichisaproxyfor bacteria were detected using PCR-based molecular methods
thedepthtothesubductingslab(Hulmeetal.2010).Ingen- from any of the samples in this study. In a previous study
eral,theseupwellingfluidscontainelevatedlevelsofdissolved (designed to gain insights into microbial biomass) samples
sulfatethatareenrichedinsulfide,whichisattributedtoboth fromSouthChamorro(ODPHole1200E)wereexaminedvia
anaerobic methane oxidation (AMO) and sulfate reduction phospholipid fatty acid (PLFA) analysis which showed two
inamicrobialhabitatoverwhelminglydominatedbyArchaea subsurfacepeaksthatrangedfrom∼500to1000timesmore
(Hulmeetal.2010;Mottletal.2003,2004;Wheatetal.2008). archaeal biomass than bacterial in a setting with essentially
The Mariana forearc’s mud volcanoes thus provide a unique no sedimentary organic carbon input (Mottl et al., 2003).
view into the process of devolatization of a subducting slab Geochemical and lithological properties of muds and sedi-
andtheserpentinizationprocesseswithintheoverridingplate ments (e.g., mineral composition, pore water geochemistry
as a mechanism for fueling subsurface communities of ex- andmethaneconcentration)havebeenshowntoaffectmicro-
tremophilicArchaea. bialhabitatsindeepmarinesettings(Inagakietal.2006).Ther-
Fluids from the subducting slab rise at rates of at least modynamic modeling of ultramafic-hosted microbial habi-
tens of cm/yr (maximum measured is 36 cm/yr at Big Blue tats indicates that nearly twice as much chemical energy is
Seamount),fasterthantherockmatrix, basedonsystematic availabletochemolithotrophsinanultramaficsettingthanin
variations in pore water chemical profiles and an advection- basalt-hostedhabitats(McCollom2007).Theinvestigationof
diffusionmodel,providingmethane,hydrogen,andsulfatefor samples from Mariana serpentinite seamounts using molec-
microbialconsumption(Fryeretal.1999;Hulmeetal.2010). ular biological techniques provides an excellent opportunity
Thedeepsubsurfacebiosphereretainsthelargestbiomass formicrobialexplorationtostudyphylogeneticdiversityand
of microorganisms compared to any other habitat on Earth ecologicalsignificanceinanextremesubsurfaceenvironment
and the extent of microbial diversity in the subsurface representingthehighestpHlimitforlife(Takaietal.2005).
432 Curtisetal.
MaterialsandMethods unsuccessful at isolating sufficient gDNA at deeper depths
fromODPcoresamples.
SampleCollectionandSiteDescriptions
MudsampleswerecollectedfromthesouthernMarianafore- PCRAmplificationofSSUrDNAforCloneLibrary
arc on the summit of South Chamorro Seamount (Figure 1)
PCR primers for the initial amplification of archaeal SSU
aboardthedrill-shipR/VJoidesResolution,duringODPLeg
rDNAs templates were analogous to those used by DeLong
195(MarchthroughMay,2001).CollectionsoccurredatODP
Site1200,HolesD&E,atadepthof∼3kmbelowsealevel. (1992);primingsiteswereidentical.The21Fprimer[5(cid:6)-TTC
YGGTTGATCCYGCCRGA]wassynthesizedwhereY=
An advanced hydraulic piston corer (APC) was used to col-
lect whole-round samples (n = 18 [Hole 1200D] and n = 21 pyrimidineanalog“P”andR=purineanalog“K”(GlenRe-
search,Sterling,VA,USA).The958Rprimer[5(cid:6)-YCCGGC
[Hole1200E]),whichwereprocessedimmediatelyaftersplit-
GTTGANTCCAATT]wassynthesizedwhereY=pyrim-
ting of the core liner. Core samples were extruded and then
idine analog “P” and N = an equal mixture of analogs “P”
brokenfromthecoreratherthancuttominimizethepossibil-
and“K”atasingleposition.ThePCRparameterswere:2min
ityofcontamination.Aseptichandlingwasconductedforall
denaturationat94◦Cfollowedby25to30cyclesof94◦Cfor
samplesandprocessedwithin30minunderlaminarflow.
Subsampleswerecollectedandstoredat4◦CinsterilePBS 1 min, 56◦C for 1.5 min and 72◦C for 3 min. Cycles were
followed by a final hold at 72◦C for 7 min. PCR conditions
for fluorescent and light microscopy. The remaining samples
werequickfrozeninliquidnitrogenandmaintainedat−80◦C included: 50 ng template gDNA, 1 U JumpStart Taq DNA
polymerase(Sigma-Aldrich,St.Louis,MO,USA),5µl10X
for land-based molecular analyses. Core samples were addi-
PCRreactionbuffer,2.5mMMgCl ,1µMeachforwardand
tionallysubjectedtotracertestsduringcollectiontoassesspo- 2
reverseprimer,200µMeachdNTP,10µgBSAandmolecular
tential seawater intrusion during drilling. A perfluorocarbon
gradewatertoatotalvolumeof50µl.Negativecontrolreac-
tracerwasusedduringdrillingofODPHole1200D,whereas
tionswereperformedwitheachroundofamplification.Am-
fluorescentmicrosphereswereusedentirelyduringODPHole
plified products were examined against a 1 kb ladder DNA
1200EsamplingandincoresectionsfivethroughnineofHole
size standard using 1% agarose gel electrophoresis to deter-
1200D(e.g.,Smithetal.2000).Moredetailedsitedescriptions
minesizeandefficiency.
andsampleprocessingproceduresareaddressedinMottletal.
(2003).
Coresampleswerealsocollectedfromsevenseamountsum-
ScreeningofCloneLibraries
mits (from south to north – South Chamorro, Blue Moon,
BigBlue,Quaker,Nip,Blip,andPacman;Figure1)alongthe For each of the two clone libraries, five replicate PCRs were
MarianaforearcduringR/VThomasG.ThompsoncruiseTN generatedandreactionswereconcentratedanddesaltedwith
154 (March through May, 2003). Piston, trigger and gravity Microcon50ultrafiltrationunits(Millipore)andelutedwith
core samples were obtained and processed immediately fol- 30µlmoleculargradewater.ThePCRampliconswerecloned
lowing the splitting of core liners on deck (max core depth with a TA cloning kit following manufacturer’s instructions
∼6m).PushcoresampleswereobtainedbytheROVJasonII (Invitrogen, Carlsbad, CA, USA). All putative clones were
and processed immediately upon surfacing (max core depth streaked for isolation and the inserts assayed for correct size
∼30cm).Samplesofwholecoresections(0to10cmbsf)were usingPCRwithM13FandM13Rprimers(Moyer2001).All
collectedandasepticallyprocessedwithin30minofretrieval. reactionsweresizedwitha1kbladderDNAsizestandardby
Moredetailedsitedescriptionsandsampleprocessingproce- 1%agarosegelelectrophoresis.
duresofthesesamecoresamplesareaddressedinHulmeetal.
(2010). Deeper core samples were tested in all cases (during
OTUDeterminationandRarefaction
both expeditions), but none were found to contain sufficient
biomasstowarrantfurthernucleicacid-basedstudy. Partialsequencedatafromallcloneswerequalitycontrolled
andalignedusingtheprogramBioEdit(Hall1999).Sequences
weretrimmedtoexcludethearchaeal21Fprimerandthere-
GenomicDNAExtractionandPurification after include ∼300 bp of the 5(cid:6)-end of the SSU rRNA gene.
DNAextractionsfromtheuppermostcoresamplesfromODP TheprogramDOTURwasusedtodeterminethenumberof
Holes 1200D and 1200E (0 to 10 cmbsf) were accomplished uniqueoperationaltaxonomicunits(OTUs)foreachcloneli-
using the method described by Zhou et al. (1996) with ap- braryusinga97%similaritycutoff(SchlossandHandelsman
proximately 10 g of sample ground until nearly powder-like 2005).
withamortarandpestlethatwaspartiallysubmergedinliq-
uid nitrogen (without the use of bead-beating, otherwise the
SSUrDNASequencing
extraction process was the same as described below for T-
RFLP). DNA extracts were visualized in a 1% agarose gel Representative SSU rDNA clones from OTUs containing
andconcentratedwithMicrocon50ultrafiltrationunits(Mil- three or more clones were fully sequenced (Moyer 2001) us-
lipore, Bedford, MA, USA) to ∼100 µl in volume. Extracts ing an ABI 3100 automated sequencer (Applied Biosystems,
were then quantified by UV spectrophotometry at 260 nm. Foster City, CA, USA). Internal sequencing primers were
Unfortunately,thismethod(aswellasmultipleothers)proved the same as those of Moyer et al. (1998). SSU rDNA gene
MarianaMudVolcanoesHarborExtremophilicArchaea 433
Table1.SummaryofdominantarchaealOTUsfromODPHole1200D
Lineage,Group SimilarityValues1 NumberofClones2 Representativeout
Crenarchaeota,MarineGroupI 0.92–0.98 19 DSAOTU1
Crenarchaeota,MarineGroupI 0.92–0.98 12 DSAOTU2
Crenarchaeota,MarineGroupI 0.91–0.98 7 DSAOTU3
Crenarchaeota,MarineGroupI 0.93–0.97 7 DSAOTU4
Crenarchaeota,MarineGroupI 0.93–0.97 6 DSAOTU5
1Rangeofpair-wisecomparisonsacrossentiregroup.
2TotalnumberofSSUrDNAclonesrepresentingthelibrarywas62.
sequences were contiguously assembled using BioNumerics NucleotideSequenceAccessionNumbers
(AppliedMaths,Saint-Martens-Latem,Belgium)andchecked
The SSU rDNA sequences representing the OTUs used in
forchimerasusingboththeBellerophonserver(Huberetal.
this analysis have been submitted to GenBank and assigned
2004)andMallard,whichallowscomparisonstoanypotential
accessionnumbersEF414497throughEF414507.
parentsequence(Ashelfordetal.2006).
PhylogeneticAnalysis GenomicDNAExtractionandPurificationforT-RFLP
SequenceswerecomparedtothoseinGenBankusingBLAST DNAextractionsofcoresamples(0to10cmbsf)fromallseven
(Altschul et al. 1990; Benson et al. 2000). SSU rDNA se- forearc seamounts were performed with modifications using
quences with the closest similarity to each OTU as well as thecombinedprotocolsofMilleretal.(1999)andZhouetal.
otherrepresentativeswereusedinphylogeneticcomparisons. (1996).Approximately5gofsampleand5gofacid-washed
All sequences were imported into ARB and aligned to the zirconium-silica beads were placed into a 50 ml centrifuge
SSU Jan04 database using the ARB fast aligner (Ludwig tube. To this tube, 13.5 ml DNA-extraction buffer (100 mM
et al. 2004). Phylogenetic analyses were restricted to regions Tris-HCl, 100 mM Na EDTA, 100 mM sodium phosphate,
2
of moderately to highly conserved nucleotide positions that 1.5MNaCland1%CTAB[pH8.0]),1mgproteinaseKand
were unambiguously aligned for all sequences. Phylogenetic 3.75mgachromopeptidasewereadded.Sampleswereplaced
placements were calculated using fastDNAml version 1.2.2 at37◦Candshakenhorizontallyat225rpmfor30min.Next,
(Olsenetal.1994)usingthegeneraltwo-parametermodelof 300 µl of 20% SDS was added and samples were incubated
evolution(KishinoandHasegawa1989)andallowingforthe ina65◦Cwaterbathfor2hwithgentleinversionevery15to
global swapping of branches. The search for an optimal tree 20min.
was repeated within these parameters until the best log like- Each mixture was then transferred to a 30 ml sterile con-
lihood tree was calculated in at least three independent tree tainer and placed on a bead-beater at maximum speed for
calculations. 3 min (BioSpec Products, Bartlesville, OK, USA). The mix-
Eachphylogenetictreewasbootstrapped500timesallow- turewasnexttransferredproportionallyintotwo50mltubes
ingfortheglobalswappingofbranches.Thesearchforeach andcentrifugedat6,000×gfor10minatroomtemperature.
bootstrapwasrepeateduntilthebestloglikelihoodscorewas Thesupernatantwasremovedandplacedintotwofresh50-ml
calculatedforatleasttwoindependentbootstrapcalculations. tubesandgDNAwasextractedanadditionaltimewith5ml
UsingtheprogramBioEdit,similarityvaluesweredetermined ofDNA-extractionbufferfollowedbyincubationat65◦Cfor
fromanidentitymatrixforallgroupmembersusingthe5(cid:6)-end 15minandcentrifugationat12,000×gfor10min.
∼300bpdata.Apair-wisecomparisonacrosseachgroupwas All supernatants were combined, mixed with an equal
possible by determining the proportion of identical residues volume (∼30 ml) of phenol-chloroform-isoamylalcohol
betweenalignedsequences(Tables1and2). (24:24:1v/v)andcentrifugedat6,000×gfor10minatroom
Table2.SummaryofdominantarchaealOTUsfromODPHole1200E
Lineage,Group SimilarityValues1 NumberofClones2 Representativeout
Crenarchaeota,MarineGroupI 0.92–0.98 19 ESAOTU1
Euryarchaeota,Methanobacteria 0.79–0.99 13 ESAOTU2
Crenarchaeota,MarineBenthicGroupB 0.82–0.97 7 ESAOTU3
Euryarchaeota,Methanosarcinales 0.85–0.99 4 ESAOTU4
Euryarchaeota,Methanobacteria 0.80–0.99 3 ESAOTU5
Crenarchaeota,MarineBenthicGroupB 0.81–0.97 3 ESAOTU6
1Rangeofpairwisecomparisonsacrossentiregroup.
2TotalnumberofSSUrDNAclonesrepresentingthelibrarywas60.
434 Curtisetal.
temperature. The aqueous phase was transferred to two Results
50 ml tubes and linear acrylamide (20 µg/ml final) was
added to each. The gDNA was precipitated with0.6 volume CloneLibraryAnalysis
isopropanolfor1hatroomtemperatureandwaspelletedvia
ArchaealSSUrRNAgenesequencesweredeterminedfrom62
centrifugationat12,000×gfor20minatroomtemperature.
clonesobtainedfromSouthChamorroSeamount,ODPHole
The supernatant was decanted and the remaining pellet
1200D.UsingtheDOTURprogram,atotalof13Hole1200D
was rinsed with 70% EtOH and dissolved in 200 µl of TE
subsurface Archaea or DSA OTUs were identified based on
(10 mM Tris, 0.1 mM EDTA, pH 8) buffer. All extracts
≥97% sequence similarity and 5 dominant DSA OTUs were
were further cleaned using Qiaex II (Qiagen, Valencia, CA,
chosen as representatives of the community based on their
USA)withelutionin50µlofTEandthenquantifiedbyUV
group containing three or more clones (Table 1). Archaeal
spectrophotometryat260nm.
SSU rDNAs were determined from 60 clones obtained from
South Chamorro Seamount, ODP Hole 1200E. Again using
T-RFLPPreparation the DOTUR program, a total of 20 Hole 1200E surface Ar-
chaea or ESA OTUs were identified and 6 dominant ESA
Three replicate archaeal PCRs were performed, each using
OTUswerechosenasrepresentativesofthecommunity,based
50 ng of gDNA, with the same priming sites as described
onthesamethreeormoreclonecriteria(Table2).
above for clone library generation. The forward primer was
5(cid:6)-end-labeledwith6-FAMwiththeamplificationconditions To compare observed taxa richness between the libraries
with respect to the absolute richness within the communi-
and subsequent treatment with eight tetrameric restriction
ties, a sample-size independent method was required. Rar-
endonucleases as described by Davis and Moyer (2008).
efaction estimates the number of OTUs in a community of
All reactions were desalted using Sephadex superfine G-75
finitesizebyrandomsamplingwithoutreplacement.Thees-
(Amersham Biosciences, Uppsala, Sweden) and dehydrated.
timatednumberofOTUswereplottedagainstthenumberof
Reactionswereresuspendedin15µlformamidewith0.33µl
ROX-500internalsizestandard,denaturedat95◦Cfor5min clonesanalyzedforeachclonelibrary.EstimatesofOTUrich-
ness (number of OTUs predicted) yielded ODP Hole 1200E
andseparatedbycapillaryelectrophoresisusinganABI3100
having approximately 33% greater richness than ODP Hole
genetic analyzer with a 50-cm capillary array and POP-6
1200D(Figure2).
(Applied Biosystems). Each reaction was separated and
visualizedatleasttwicetoensurereproducibility.
PhylogeneticAnalysis
T-RFLPNormalizationandAnalysis
Each OTU from ODP Holes 1200D and 1200E had a rep-
Terminal-restrictionfragments(T-RFs)weresizedagainstthe resentative clone that was fully sequenced and phylogenetic
ROX-500internalsizestandardusingGeneMapperv.3.7(Ap-
pliedBiosystems).Electropherogramswereimportedintothe
programBioNumerics(AppliedMaths,Sint-Martens-Latem, 60
Belgium). Community fingerprints were compared in BioN-
umerics using average Pearson product moment correlation Archaea Hole 1200E
50
(Ha¨ne et al. 1993) and unweighted pair group method with s Archaea Hole 1200D
arithmetic mean (UPGMA) cluster analysis of all eight re- OTU Maximum Richness
striction digests using the relative fluorescent proportions of of 40
eachelectropherogram.Thecopheneticcorrelationcoefficient er
b
wascalculatedtoassesstherobustnessoftheclusteranalysis m 30
u
N
groupings. Peak detection was limited to peaks between 50 d
e
and 500 base pairs in size and with height at least 3% of the at 20
m
maximumvalueofthefingerprint(DavisandMoyer2008). sti
E
10
DirectCellCounts
0
All samples underwent an examination of total cell counts
0 10 20 30 40 50 60
in an attempt to visualize the distribution and abundance of
Number of Clones Screened
subsurface microorganisms. Shipboard-prepared subsamples
fixedin2.5%glutaraldehydewerevigorouslyvortexedthen10
Fig. 2. Rarefaction curves comparing the estimated population
to25µlwasremovedwithawide-borepipettipandappliedto
richnessoftwoclonelibrariesfromSouthChamorroSeamount,
pre-blackened,25mmdiameter,0.2µmporesize,polycarbon- ODP Site 1200 Holes D and E. Estimates indicate an increase
ate filters (Shipboard Scientific Party 2002). Duplicate filters of ∼33% in OTU richness in the 1200E archaeal community
werestainedwith5µMofSyto13(MolecularBioprobes,Eu- sampledoverthatof1200D.Standarddeviationconfirmssignif-
gene,OR,USA)and10fieldsfromeachfilterwereexamined icantdifferencebetweenthetwosamplesites.Curveofslope=
usingepifluorescencemicroscopy.Cellenumerationincluded 1representstheoreticalmaximumrichness.Errorbarsrepresent
normalizationtothefilteredvolume. standarderrorestimates.
MarianaMudVolcanoesHarborExtremophilicArchaea 435
ESA OTU 3
ESA OTU 6 B
100 Black Sea microbial mat BS-SR-E2 (AJ578142) p
5073 Haakon MosbHy aMakuodn V Molocsabnyo MHuMdM VVolPcaongo-3 H7 M(AMJ5V7P9o3g1-84)1 (AJ579321) Grou
78 9954 MHAaRa khoynd rMotohsebrmy aMl uvde nVt oVlcCa2n.o1 HAMrc3M1V (APoFg0-63858 (2A2J)579317) hic
Guaymas Basin hydrothermal sediments C1-R043 (AF419642) nt
60 Black Sea microbial mat BS-SR-G11 (AJ578146) Be
Black Sea microbial mat BS-SR-B9 (AJ578139) e
Black Sea microbial mat BS-SR-E3 (AJ578143) n
93 Cascadia Margin cold seep Hyd24-Arch07b (AJ578114) ari
Haakon Mosby Mud Volcano HMMVPog-30 (AJ579313) M
51
NW Atlantic Continental Rise CRA8-27cm (AF119128)
African deep gold mine fluids SAGMA-A (AB050205)
Aegean Sea sediments 19b-5 (AJ294854)
Snake River Plain aquifer biofilm ARCHEUKDBF14 (DQ190076)
83
92 Cascadia Margin methane hydrate fluids ODPB-A6 (AF121094)
NE Atlantic ocean water C48 (U71118)
DSA OTU 3 I
100 7571 MMayrioajnina KTrneonlcl hh yadrcrhoatheeornm naol .v 1e5n t( DpM87C315A0)103 (AB019725) oup
8190 ESA OTU 1 Gr
77
Seattle Aquarium clone RB01 (DQ085101) e
89 Cenarchaeum symbiosum (U51469) rin
Loihi hydrothermal vent mat PVA OTU 4 (U46680) a
M
West Philippine Basin sediments pPCA4.21 (AB049032)
DSA OTU 2
82
West Pacific warm pool sediments MBWPA35 (AJ870319)
DSA OTU 1
54 DSA OTU 5
DSA OTU 4
100 Pyrobaculum arsenaticum (AJ277124)
88 Pyrobaculum neutrophilum (X81886)
Thermocladium modestius (AB005296)
81 Desulfurococcus mobilis (M36474) ei
Thermosphaera aggregans (X99556) ot
63 Sulfophobococcus zilligii (X98064) pr
o
50 100 Thermodiscus maritimus (X99554) m
Uncultured Desulfurococcus sp. SUBT-9 (AF361213) r
e
Sulfolobus acidocaldarius (D14876) h
99 Sulfolobus solfataricus (X90478) T
100 Acidianus ambivalens (X90484)
80 Acidianus infernus (X89852)
65 Sulfolobus metallicus (D85519)
Aquifex pyrophilus
Fig. 3. Phylogenetic tree showing relationships of DSA OTU and ESA OTU phylotypes within the Marine Benthic Group B and
MarineGroupI(Crenarchaeota)asdeterminedbymaximumlikelihoodanalysisofalignedSSUrDNAsequences.Numbersatnodes
represent bootstrap values based on 500 bootstrap resamplings. The outgroup is represented by Aquifex pyrophilus. The scale bar
represents0.10fixedmutationspernucleotideposition.Bootstrapvaluesareshownforfrequenciesatoraboveathresholdof50%.
affiliationsweredetermined(i.e.,aphylotype).ForODPHole The most dominant ESA OTU (1) at ODP Hole 1200E
1200D, all dominant DSA OTUs resided within the Ma- resideswithintheMarineGroupICrenarchaeota,comprising
rine Group I Crenarchaeota (Figure 3). DSA OTU 1 com- 32% of the clone library (Figure 3). Three of the remaining
prised 31% of the clone library. The closest match to known fivedominantESAOTUs(i.e.,34%oftheclonelibrary)were
databasesequencesforthisOTUbelongstoanunculturedcre- affiliatedwiththeEuryarchaeotaandwerefoundwithineither
narchaeote,cloneMBWPA35(AJ870319),originallydetected the Methanobacteria or Methanosarcinales (Figure 4). ESA
fromsedimentinthewesternPacificnearChina(Wangetal. OTUs 3 and 6 were found most closely related to Marine
2005). DSA OTU 5 comprised 10% of the clone library and Benthic Group B Crenarchaeota (OTU 3, representing 12%
clusteredclosestwiththenearestrelativetoDSAOTU1.DSA of the library, and OTU 6, representing 5% of the library,
OTU4comprised12%oftheclonelibraryandgroupedmost respectively). ESA OTU 1 had a closest match belonging to
closelywiththesamephylotypeclosesttoDSAOTU1,clone an uncultured marine archaeon (pMC1A103) detected from
MBWPA35.DSAOTU2comprised19%oftheclonelibrary sediment obtained from a clear smoker chimney at Myojin
and had a closest match to an uncultured archaeon labeled Knoll(TakaiandHorikoshi1999).
pPCA4.21(AB049032)obtainedfromasubseafloorenviron- ESA OTU 2 comprised 22% of the clone library, and
mentoffthecoastofJapan(Inagakietal.2001).DSAOTU3 hadaclosestmatchtoanunculturedarchaeonCRA13-11cm
comprised12%oftheclonelibrarywithaclosestmatchtoan (AF119124)collectedfromdeep-seasedimentwithintheAt-
unidentified archaeon (D87350) detected from the Mariana lanticOcean(Vetrianietal.1999).ESAOTU5represented5%
Trenchatadepthof10,898m(Katoetal.1997). oftheclonelibraryandwasmostcloselyrelatedtothenearest
436 Curtisetal.
99 Methanocaldococcus jannaschii (NC_000909) cci
o
c
71 Methanocaldococcus fervens (AF056938) o
n
a
100 Methanococcus voltae (M59290) eth
M
Methanococcus vannielii (M36507)
100 Thermoplasma acidophilum (NC_002578) mo- mata
Picrophilus oshimae (X84901) Therplas
Methanococcoides euhalobius (X98192)
Methanosarcina semesiae (AJ012742)
100
64
Methanosarcina lacustera (AF432127)
Methanosarcina siciliae (U89773)
78
Methanosarcina barkeri (AJ002476) es
al
98 Methanohalophilus oregonensis (U20152) n
100 Methanolobus vulcani (U20155) rci
a
NE Japan Sea cold seep sediments AJS72-22 (AB239074) os
90 n
Methanococcoides burtonii (X65537) a
100 h
100 ESA OTU 4 et
M
61 Methanococcoides alaskense (AY941802)
AMO bioreactor AMOS4A-4113-H07 (AY323224)
99 100 Methanosaeta concilii (X16932)
Methanosaeta concilii strain FE (M59146)
Methanoplanus petrolearius (U76631)
es
100 Methanocorpusculum labreanum (AY260436) bial
o
100 Methanofollis liminatans (AF095271) r
c
mi
Methanofollis aquaemaris (AY186542) o
n
100 Methanoculleus bourgensis (AF095269) ha
et
Methanoculleus palmaeoli (Y16382) M
77 Myojin Knoll hydrothermal vent pMC1A4 (AB019754)
78 Southern Mariana Trough sediments Papm3A53 (AB213103)
100
Eastern Mediterranean mud mound Urania-2A-24 (AY627502)
ESA OTU 5 a
68 ri
e
100 NW Atlantic Ocean continental rise CRA13-11cm (AF119124) ct
ESA OTU 2 ba
69 o
n
Southern Mariana Trough sediments Papm3A05 (AB213087) a
h
Methanobacterium bryantii (M59124) et
82 M
94 Methanobacterium formicicum (AF028689)
100
Methanobacterium palustre (AF093061)
Methanobacterium oryzae (AF028690)
Aquifex pyrophilus (M83548)
Fig. 4. Phylogenetic tree showing relationships of ESA OTU phylotypes within the Methanobacteria and Methanosarcinales (Eur-
yarchaeota)asdeterminedbymaximumlikelihoodanalysisofalignedSSUrDNAsequences.Numbersatnodesrepresentbootstrap
valuesbasedon500bootstrapresamplings.TheoutgroupisrepresentedbyAquifexpyrophilus.Thescalebarrepresents0.10fixed
mutationspernucleotideposition.Bootstrapvaluesareshownforfrequenciesatoraboveathresholdof50%.
relative to ESA OTU 2, both among the Euryarchaeota or- clone CRA8–27 cm (AF119128), detected from the NW At-
der Methanobacteria (Figure 4). ESA OTU 4 comprised lanticOcean(Knitteletal.2005).
7% of the clone library and was most closely related to
a methylotrophic methanogen, Methanococcoides alaskense
T-RFLPAnalysis
(AY941802) detected from anoxic marine sediment in Skan
Bay,Alaska(Singhetal.2005)amongtheEuryarchaeotaor- Selected electropherograms (treatments with HaeIII and
der Methanosarcinales (Figure 4). ESA OTUs 3 and 6 were HhaI) from ODP Hole 1200D show T-RFs that match ei-
both most closely related to a deep-sea sediment archaeon, ther all or a majority of the DSA OTUs detected within the
MarianaMudVolcanoesHarborExtremophilicArchaea 437
Fig.5.T-RFLPelectropherogramof(A)HaeIIIand(B)HhaIdigestsfromODPHole1200DatSouthChamorroSeamount.Arrows
indicateT-RFsmatchingDSAOTUs,allfromwithintheMarineGroupI(Crenarchaeota).
Hole1200Dclonelibraryasdeterminedbyinsilicodigestion epifluorescent microscopy, ranged from 9.5×106 cells/ml
ofMarineGroupICrenarchaeotaphylotypes(Figure5).Se- (at South Chamorro and Big Blue) to 1.1×108 cells/ml (at
lectedelectropherograms(treatmentswithHaeIIIandHhaI) Pacman),fallingwithintherangeofapproximatelyoneorder
fromODPHole1200EalsoshowT-RFsthatwhencombined ofmagnitudefromthelowesttothehighestcellcountsfound
match all of the ESA OTUs within the Hole 1200E clone li- acrossthesevenseamountsexamined(datanotshown).
brary,whichcontainedbothCrenarchaeotaandEuryarchaeota
phylotypes(Figure6).ToensuremaximaldetectionofOTUs, Discussion
atotalofeightrestrictionendonucleaseswereusedwhenan-
alyzing all samples. Similar T-RF patterns that matched the The goal of this study was to discover and detect the most
dominantphylotypes(foundinthetwoODPSite1200clone abundant microbial populations, as well as study their rich-
libraries)werealsodetectedusingtheremainingsixrestriction nessandrelatedness,withintheserpentinizedmudvolcanoes
enzymes: AluI, BstUI, MspI, RsaI, HinfI, and Sau96I (data on the Mariana forearc using SSU rDNA clone library and
notshown). T-RFLPanalyses.Inapreviousstudy,phospholipidfattyacid
Clusteranalysisofsamples(collectedwithmultiplecoring or PLFA results from ODP Hole 1200E indicated an excep-
techniques) from seven different forearc seamount locations tionallyhighsubsurfacearchaealbiomass(500to1000times
revealed archaeal communities similar to the ODP Hole higherthanbacterialbiomass)collocatedinzonesofdissolved
1200Elibrary(i.e.,ESAOTUs)whichcontainednovelphylo- sulfidemaximaandsulfateminima.Thiswastakentoindicate
types(EuryarchaeotafromwithineithertheMethanobacteria, thepresenceofanarchaealdominatedcommunitypotentially
Methanosarcinales or the Crenarchaeota Marine Benthic capable of both AMO and sulfate reduction (Mottl et al.,
GroupB).Theseuniquearchaealcommunitiesweredetected 2003).
in samples from the summits of South Chamorro (Hole Under the highly productive waters off Peru, subsurface
1200E), Big Blue and Blue Moon Seamounts (Figure 7). archaeal consortia using AMO have been shown to occur in
Total cell counts from these samples, as determined by marine sediment habitats where methane is consumed at the
438 Curtisetal.
Fig.6.T-RFLPelectropherogramof(A)HaeIIIand(B)HhaIdigestsfromODPHole1200EatSouthChamorroSeamount.Arrows
in (A) indicate T-RFs matching ESA OTU 3 and 6 from within the Marine Benthic Group B (Crenarchaeota) and ESA OTU 2
from within the Methanobacteria (Euryarchaeota). Arrows in (B) indicate T-RFs matching ESA OTU 2, 4 and 5 from within the
MethanobacteriaandMethanosarcinales(Euryarchaeota)andESAOTU1fromwithintheMarineGroupI(Crenarchaeota).
expense of sulfate (Biddle et al. 2006). The highly alkaline benefitofanalyzingarchaealcommunitiesfromnovelecosys-
(up to pH ∼12.5) deep-sourced fluids from serpentinized tems with unique geochemical parameters spread across a
mud volcanoes of the Mariana forearc provide the oppor- widegeographicrange(Figure1).
tunity to investigate archaeal dominated communities and Rarefaction curves comparing the estimated popula-
their ecological significance within the subsurface biosphere. tion richness of two clone libraries from South Chamorro
Theincorporationofsevenseamountsinthisstudyoffersthe Seamount(ODPHoles1200Dand1200E)indicatedthatwhile
Fig.7.UPGMA/Pearsonproduct-momentclusteranalysisofT-RFLParchaealcommunityfingerprintsspanningsevenserpentinized
mudvolcanoesalongtheMarianaforearcusingeightrestrictiondigesttreatmentsoneachof10samples.Uniquearchaealcommunity
fingerprints resulted from an abundance of putative methane-oxidizing and sulfate-reducing phylotypes clustered from three sites:
South Chamorro (ODP Hole 1200E), Big Blue and Blue Moon Seamounts. The numbers at nodes are the cophenetic correlation
coefficients.
MarianaMudVolcanoesHarborExtremophilicArchaea 439
organismaldiversitywaslow,saturationofeithercommunity T-RFLPdatafromHole1200Eallowedfordiscrimination
hadnotyetbeenreached(Figure2).Thesignificantincrease amongeachoftheMBGBCrenarchaeotaandEuryarchaeota
in richness of the Hole 1200E clone library (∼33% greater populationsdetected,againsupportingourphylogeneticanal-
than that of Hole 1200D) is thought to be due to the ad- ysis (Figure 6). Distinctly separate from the other Crenar-
ditional presence of the indicator Euryarchaeota phylotypes chaeota phylotypes discovered, recent findings suggest the
(Figure6),whichislikelydependentinpartonitsgeographic MBGBs may have a syntrophic relationship with these Eu-
proximitytoanactivespringlocatedatthesummitofSouth ryarchaeotawithrespecttoAMO.Theirrelativelyhighabun-
ChamorroSeamount.Upwellingofsedimentporefluidsslows danceinassociationwithbenthiccoldseepsindicatesthatthey
with distance from thespring; Hole 1200D, located approxi- maybeworkinginconjunctionwithAMOEuryarchaeota,act-
mately80mfromthespring,hadaporefluidupwellingspeed ingasanewlydiscoveredtypeofsulfatereducer(Biddleetal.
of0.2cmperyearrelativetothemudmatrix. 2006; Knittel et al. 2005; Sørensen and Teske 2006; Tourova
IncontrastHole1200E,locatedonlyafewmetersfromthe etal.2002).
spring,hadaporefluidupwellingspeedestimatedat3cm/yr ClusteranalysisoftheT-RFLPdatafromacrosstheMar-
(Fryer and Salisbury 2006). More recently, maximum up- iana forearc system revealed the detection of additional ar-
wellingspeedsof10and36cm/yrhavebeenmodeledforBlue chaeal communities from sites on Blue Moon and Big Blue
Moon and Big Blue, respectively (Hulme et al. 2010), which Seamounts similar that of South Chamorro Seamount ODP
haveshownsimilararchaealcommunitystructures(Figure7). Hole 1200E (Figure 7). These unique archaeal fingerprints
Theupwardflowofporefluidsfromdepthcoupledwithactive resulted from an abundance of indicator Euryarchaeota phy-
serpentinization processes transports slab-derived nutrients lotypes,whichmaybeassociatedwithmethane-oxidizingand
thatappearstofuelarchaealchemolithotrophicmetabolisms sulfate-reducingchemolithotrophicmetabolisms.Thecombi-
andfasterflowappearstocorrelatewiththeaddedcomplexity nation of our phylogenetic analysis of the dominant popu-
ofthearchaealcommunity. lations within the archaeal communities at ODP Site 1200
Between the two sampling locations at South Chamorro inconjunctionwiththeT-RFLPclusteranalysisfrommulti-
Seamount(ODPHoles1200Dand1200E)33OTUswerede- plelocationsindicatesthatthedistributionofthesearchaeal
tectedamongthe122archaealclonesexaminedwithatotalof populationswithinnearsurfacehabitatsismostlikelydueto
11OTUsdetectedasdominantcommunitymembers(Tables1 environmentalparameters(e.g.,flowrates,pH,availabilityof
and 2). Sequence comparisons of representative clones indi- methane,sulfate,etc.).SamplesfromBlueMoonandBigBlue
catedthatallDSAOTUs(1through5)andasingleESAOTU Seamounts that had corresponding archaeal communities to
(1)groupedwithintheMarineGroupI(MGI)Crenarchaeota South Chamorro Seamount, ODP Hole 1200E are hypoth-
(Figure 3). T-RFLP data from Hole 1200D allowed for dis- esized to have occurred at similar regions of high flow and
criminationamongalltheseMGIpopulations(Figure5).The ensuing pore water geochemistry (e.g., Hulme et al., 2010;
efficacyofthisapproachwasconfirmedbythecomparisonof Mottl et al. 2003). Serpentinite seamount environments are
T-RFLPdatafromHole1200E,whereonlyasingleMGIphy- dynamicsystems,anditislikelythatthesummitlocationsof
lotype was detected (Figure 6), supporting our phylogenetic high flow conduitsare notonlypatchy butchange over time
analysis(Figure3). (Fryeretal.2006).
TheseMGIphylotypesappeartorepresentthemostabun- Thisstudyallowedforthediscoveryanddetectionofnovel
dant archaeal populations within the microbial community Euryarchaeota as well as Marine Benthic Group B Crenar-
atSouthChamorroSeamount.Thisisconsistentwithprevi- cheaotaphylotypes,whichweredeterminedaspossibleeffica-
ousreportsthattheMGICrenarchaeotadominatetheglobal ciousarchaealindicatorpopulationsinvolvedwithAMOand
oceanbiosphere(Karneretal.2001;Takaietal.2004b).The sulfate reduction fueled by deep subsurface serpentinization
serpentinite muds investigated at South Chamorro revealed reactions.Theseindicatorpopulationsareapparentlyliving(in
geochemical gradients caused by the rising fluids that origi- relativelyhighnumbers)ontheedgeofthehabitablezonefor
nate at depth and mix with seawater near the surface (Fryer lifeinthisnear-surfaceserpentinitemudhabitat.Inthefuture,
andSalisbury2006).ThepersistenceoftheMGIphylotypes weplantouseadditionalmolecularmicrobialtechniquessuch
amongalltheMarianaforearcsamplesislikelyanindication asfluorescentinsituhybridization(FISH)toaidinthelocal-
ofthisnearsurfacemixingphenomenonwithinallofthemud izationofmetabolicallyactivemethane-oxidizingandsulfate-
volcanoes. reducingarchaealpopulations(e.g.,Biddleetal.2006;Knittel
Phylogenetic analysis of representative clones from ODP et al. 2005; Schippers et al. 2005; Schrenk et al. 2004) and
Hole 1200E grouped ESA OTUs 3 and 6 within the Ma- quantitative-PCR(q-PCR)inanefforttodeterminetheloca-
rineBenthicGroupB(MBGB)Crenarchaeota(Figure3).Ad- tionandidentityofbothphylogeneticgroupsandtheirfunc-
ditional comparisons revealed euryarchaeotal classifications tionalgeneswithinthesearchaealcommunities(e.g.,Dhillon
with ESA OTUs 2 and 5 grouping within the Methanobac- etal.2005;Inagakietal.2004;Lloydetal.2006;Takaietal.
teria and ESA OTU 4 grouping within the Methanosarci- 2004b). Further localization of these extremophilic archaeal
nales (Figure 4). This discovery is intriguing because of all populationsalongwithabetterunderstandingoftheircorre-
the Methanosarcinales isolates known, none have been able spondingmetabolicpotentialwillaidindeterminingifthese
to grow above pH 10 (Boone et al. 2001). The MBGBs, populations are actuallyendemic tohigh flow systems along
MethanobacteriaandMethanosarcinalesphylotypeswereonly theMarianaforearcoriftheyhavetheabilitytoexistinother
detectedintheHole1200Eclonelibrary(Tables1and2). habitats.