Bacterial, archaeal, and eukaryotic diversity across distinct microhabitats in an acid mine drainage

18 Pages·2017·5.15 MB·English

Bacterial, archaeal, and eukaryotic diversity across
distinct microhabitats in an acid mine drainage
Victoria Mesa, José Luis. R. Gallego, Ricardo González-Gil, Béatrice Lauga,
Jesus Sánchez, Célia Méndez-García, Ana I. Peláez
To cite this version:
Victoria Mesa, José Luis. R. Gallego, Ricardo González-Gil, Béatrice Lauga, Jesus Sánchez, et al..
Bacterial, archaeal, and eukaryotic diversity across distinct microhabitats in an acid mine drainage.
Frontiers in Microbiology, 2017, 8 (SEP), ￿10.3389/fmicb.2017.01756￿. ￿hal-01631843￿
HAL Id: hal-01631843
https://hal.archives-ouvertes.fr/hal-01631843
Submitted on 14 Jan 2018
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est
archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
fmicb-08-01756 September9,2017 Time:16:9 #1
ORIGINALRESEARCH
published:12September2017
doi:10.3389/fmicb.2017.01756
Bacterial, Archaeal, and Eukaryotic
Diversity across Distinct
Microhabitats in an Acid Mine
Drainage
VictoriaMesa1,2*,JoseL.R.Gallego3,RicardoGonzález-Gil4,BéatriceLauga5,
JesúsSánchez1,CeliaMéndez-García6†andAnaI.Peláez1†
1DepartmentofFunctionalBiology–IUBA,UniversityofOviedo,Oviedo,Spain,2VedasResearchandInnovation,VedasCII,
Medellín,Colombia,3DepartmentofMiningExploitationandProspecting–IUBA,UniversityofOviedo,Mieres,Spain,
4DepartmentofBiologyofOrganismsandSystems–UniversityofOviedo,Oviedo,Spain,5EquipeEnvironnementet
Microbiologie,CNRS/UniversitédePauetdesPaysdel’Adour,InstitutdesSciencesAnalytiquesetdePhysico-chimiepour
l’EnvironnementetlesMatériaux,UMR5254,Pau,France,6CarlR.WoeseInstituteforGenomicBiology,Urbana,IL,
UnitedStates
Editedby:
AxelSchippers,
FederalInstituteforGeosciences Acid mine drainages are characterized by their low pH and the presence of dissolved
andNaturalResources,Germany toxic metallic species. Microorganisms survive in different microhabitats within the
Reviewedby: ecosystem, namely water, sediments, and biofilms. In this report, we surveyed the
RichardAllenWhiteIII(RickWhite),
microbial diversity within all domains of life in the different microhabitats at Los
IdahoStateUniversity,UnitedStates
OlivierPringault, Rueldos abandoned mercury underground mine (NW Spain), and predicted bacterial
InstituteofResearch
function based on community composition. Sediment samples contained higher
forDevelopment,France
MarkDopson, proportions of soil bacteria (AD3, Acidobacteria), as well as Crenarchaeota and
LinnaeusUniversity,Sweden Methanomassiliicoccaceae archaea. Oxic and hypoxic biofilm samples were enriched
*Correspondence: in bacterial iron oxidizers from the genus Leptospirillum, order Acidithiobacillales, class
VictoriaMesa
Betaproteobacteria,andarchaeafromtheclassThermoplasmata.Watersampleswere
[email protected]
†Theseauthorshavecontributed enriched in Cyanobacteria and Thermoplasmata archaea at a 3–98% of the sunlight
equallytothiswork. influence, whilst Betaproteobacteria, Thermoplasmata archaea, and Micrarchaea
dominated in acid water collected in total darkness. Stalactites hanging from the Fe-
Specialtysection:
rich mine ceiling were dominated by the neutrophilic iron oxidizer Gallionella and other
Thisarticlewassubmittedto
ExtremeMicrobiology, lineages that were absent in the rest of the microhabitats (e.g., Chlorobi, Chloroflexi).
asectionofthejournal
Eukaryotes were detected in biofilms and open-air water samples, and belonged
FrontiersinMicrobiology
mainly to clades SAR (Alveolata and Stramenopiles), and Opisthokonta (Fungi). Oxic
Received:19May2017
Accepted:29August2017 and hypoxic biofilms displayed higher proportions of ciliates (Gonostomum, Oxytricha),
Published:12September2017 whereas water samples were enriched in fungi (Paramicrosporidium and unknown
Citation: microbial Helotiales). Predicted function through bacterial community composition
Mesa V,Gallego JLR,
suggested adaptive evolutive convergence of function in heterogeneous communities.
González-Gil R,Lauga B,Sánchez J,
Méndez-García CandPeláez AI Our study showcases a broad description of the microbial diversity across different
(2017)Bacterial,Archaeal,and
microhabitats in the same environment and expands the knowledge on the diversity
EukaryoticDiversityacrossDistinct
MicrohabitatsinanAcidMine ofmicrobialeukaryotesinAMDhabitats.
Drainage.Front.Microbiol.8:1756.
doi:10.3389/fmicb.2017.01756 Keywords:acidminedrainage,Bacteria,Archaea,Eukarya,sediment,ore,biofilm,stalactite
FrontiersinMicrobiology|www.frontiersin.org 1 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #2
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
INTRODUCTION involvement in the biogeochemical cycles in AMDs is still
limited(Amaral-Zettler,2012;Aguilera,2013;Volantetal.,2016).
Acid mine drainages (AMD) form when sulfide minerals (e.g., Thediversityofeukaryoticmicroorganismsinhabitingopenair
pyriteandFeS )areexposedtooxygenandwaterduringmetal AMD systems includes microscopic algae, which are primary
2
ore mining (Nordstrom and Alpers, 1999). Pyrite dissolution is producers, protozoans (ciliates, flagellates, rotifers, amoebae),
enhancedbytheactivityofautochthonousironoxidizingbacteria contributing to primary or secondary production, and fungi,
and archaea, which contribute to the formation of acidic and which act as decomposers and contribute to carbon recycling
metal-rich solutions that drain from mine wastes and mining (Méndez-Garcíaetal.,2015).Fungiandprotistsconferstructure
activities,generatingAMDs(SilvermanandEhrlich,1964;Baker to the biofilms and impact the community composition by
and Banfield, 2003). Low pH increases the solubility of certain grazing on resident bacteria and archaea (Baker et al., 2004).
metallicspeciespresentinsecondaryminerals,therebyincreasing In soils, protozoa can affect the structure of the bacterial
themetalloadofthedrainage(Lariosetal.,2012).Communities communities (Rosenberg et al., 2009), or might impact their
ofautotrophic/heterotrophicbacteriaandarchaeathriveinthese dispersal(Brocketal.,2011).
conditions,andultimatelycontrolthecyclingofbiogeochemical Acidminedrainagesconstitutethemainsourceofpollutionof
elementsFe,S,C,N,andHinAMDs(BakerandBanfield,2003). freshsurfacewatersonEarth.Thesemetal-richmineoutflowsare
Themainenvironmentalvariablesinfluencingdistributionof highlytoxicand,whenmixedwithgroundwater,surfacewater,or
microbial species in acidic mine outflows are pH, temperature, soil,theybecomeresponsibleforthecontaminationofdrinking
concentrations of dissolved metals, total organic carbon, and water, disruption of growth and reproduction of aquatic plants
dissolved oxygen (Méndez-García et al., 2015). AMD systems and animals, or corrosion of infrastructures. Most microbes
constitute approachable models for microbial ecology analysis thriving in these ecosystems obtain their energy through the
due to their relatively low microbial species richness and the oxidation of reduced metallic species and have potential for
existence of a tight coupling of biological and geochemical mineralbioleaching.Asthegenerationofacidicleachatesoccurs
processes. during this process, a possible solution for the remediation of
Los Rueldos Hg mine (Asturias, NW Spain) constitutes a AMDs consists in preventing the oxic conditions that allow
recently explored example of an AMD formation (Méndez- theactivityofthemicrobesinvolved.Nevertheless,theselective
Garcíaetal.,2014).Theemplacementappearsasacaveopened pressures operating in these extreme environments (low pH,
in a cinnabar (HgS)-rich mountain slope and was associated to toxicconcentrationofmetals)haveequippedthemwithdiverse
the process of Hg recovery until its abandonment, more than adaptivemechanismsthatmaketheirbiologyathrillingsubject
40 years ago. Los Rueldos displays a drainage of pH 2, with ofstudy(JohnsonandHallberg,2003).
levels of arsenic and aluminum above the limits allowed by the The current report describes the diversity of the
Spanish legislation for direct discharge (Méndez-García et al., microorganisms pertaining to the Bacteria, Archaea, and
2014). Its more distinctive characteristic, as compared to other Eukaryadomainsoflifeacrossdistinctmicrohabitats(acidwater,
AMD ecosystems, is the presence of massive streamer biofilms mineral fractions including ore and sediments, and different
developing in shallower regions of a rather static drainage, biofilms morphologies) in Los Rueldos AMD using low as well
whereasthickmicrobialbiofilmsthriveunderhypoxicconditions ashigh-throughputtaxonomicprofilingofphylogeneticmarkers
in stagnant ponds across its course. While the microbial (16S/18SrRNAgenes).Wefurtherpredictedmicrobialfunction
diversity and function of biofilms thriving at oxic and suboxic bymetagenomicreconstructionthrough16SrRNAgenessurvey.
conditionshavebeeninvestigated(Méndez-Garcíaetal.,2014), With our scrutiny, we attempted to expand our knowledge
the microbial diversity, including microbial eukaryotes, across on the microbial eukaryotes and bacteria/archaea present in
the identifiable microhabitats within the drainage remained unexplored microhabitats within Los Rueldos AMD. Thus, this
unexplored. extremeecosystemrepresentsanexcellentnaturallaboratoryin
ThemicrobialecologyatdiscretemicrohabitatswithinAMDs which complex questions about evolution and functionality of
acrosstheglobehasbeenextensivelyreviewedrecently(Méndez- microbialcommunitiescouldbeinvestigated.
García et al., 2015; Chen et al., 2016). Nevertheless, there are
only few studies on the comparative microbial diversity across
MATERIALS AND METHODS
different microhabitats within the same AMD (Amaral-Zettler,
2012;Falagánetal.,2014;Jonesetal.,2015).Thus,ourknowledge
Samples Description and Measurement
onthemicrobialcompositionatdifferentmicrohabitatsinAMD
systemshasbeeninferredfromdiscretestudiesondifferentAMD of Environmental Variables
microhabitats. For example, bacteria inhabiting thick biofilms The geological features and geochemistry of Los Rueldos mine
differ from bacteria inhabiting thin biofilms and acid waters in inAsturiashavebeenextensivelydescribedelsewhere(Méndez-
distinct ecosystems (Bond et al., 2000a; Hallberg et al., 2006; García et al., 2014). AMD samples were aseptically collected in
Haoetal.,2010),andarchaeadivergetaxonomicallyinseparate triplicatesinMarch2013from10samplingpointsinsterile50mL
microhabitats from different AMDs (Baker and Banfield, 2003; tubes and kept at 4◦C until being processed in the laboratory
Bruneeletal.,2008;Haoetal.,2010;Méndez-Garcíaetal.,2014). (within 2 h of collection). Samples included water (WOUT,
Eukaryotes inhabiting AMD systems have been provided WEN,andWIN),mineralfractions(sediment,S,andoresamples
comparatively little attention, and knowledge about their collected from stalactites, ST), and microbial biofilms (BS, BF,
FrontiersinMicrobiology|www.frontiersin.org 2 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #3
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
TABLE1|Listofsamplesanalyzedinthisstudy. 340i, WTW, Germany). Luminance (Lx) was determined using
Lunasix3s(Gossen).
Typeofsample Sample Description
Morphological information about the eukaryotes present in
Water WOUT Acidwaterfrompondoutsideofthegallery theWOUTsamplewasobtainedbyphasecontrastmicroscopy,
WEN Acidwaterfrompondattheentranceofthe using a Nikon Eclipse E-200 microscope coupled to a Nikon
gallery Cool-Pix990digitalcamera.
WIN Acidwaterfilteredthrougha0.22µmporesize
membranefilter.Analysisontheretained
fraction. Denaturing Gradient Gel Electrophoresis
Sediment S AMDbedsediment (DGGE)
Stalactite ST Surfaceofstalactitesgeneratedatthecave
Sequence-specificseparationbyDGGE(Muyzeretal.,1993)was
ceiling
performed in B1A, B1B, and B2 samples collected during the
Biofilm BF Subaerialbiofilmattheinterfacerock/acid
period 2007–2012 using 300–500 ng of pooled PCR products
water
obtained by the amplification of the V3–V5 variable regions of
BS Biofilmattheinterfacesediment/waterata
depthof15cm the 16S rRNA gene with primer pairs 341fGC/907R (Muyzer,
B1A/B1B Stratifiedstreamerwithuppermost(A)and 1999; Muyzer et al., 1993) for Bacteria and 344fGC/915R
lowermost(B)strata (Stahl and Amann, 1991 and Raskin et al., 1995) for Archaea.
B2 Submergedmicrobialmatatdrainagedepth An annealing temperature of 55◦C provided the best results
∼50cm for all group amplifications. DGGE ran in a DCodeTM
UniversalMutationDetectionSystem(Bio-Rad)inTris-acetate-
ethylenediaminetetraacetic acid (TAE) buffer using 6% (w/v)
B1A, B1B, and B2; see Table 1 for a description of samples) poly-acrylamide1mmthickgels(40%Acrylamide/BisSolution
(Figure1andTable1). 37.5:1,Bio-Rad).Solutionsof0,30,and70%ofdenaturingagents
Conductivity (mS cm−1), temperature (◦C), redox potential [100% denaturing solution with 7 M urea and 40% formamide
(mV), and pH were measured at three discrete points in each (v/v)] were used to build the gradient. The electrophoresis
samplinglocationusingaportablemulti-parameterprobe(Multi coursedat70Vfor16hat60◦C.Apre-runningstepfor5minat
FIGURE1|MapsoftheLosRueldosemplacementinAsturias(NWSpain)(A).DepictionofthesamplinglocationsalongtheAMD(B).
FrontiersinMicrobiology|www.frontiersin.org 3 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #4
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
110Vwasincludedtoavoiddispersionofthesample.Thegelwas and singletons were excluded from the analysis. Taxonomic
stainedwithSYBRGoldNucleicAcidGelStain(Invitrogen,Life annotation was performed through comparison against the
TechnologiesTM, United States) and scanned with a Kodak Gel Greengenes database (release May, 2013) for 16S rRNA gene
Logic200ImagingSystem(Kodak,UnitedStates).Bandprofiles amplicon data, and with the SILVA non-redundant database,
were analyzed using Phoretix 1D Pro software (TotalLab Ltd., release 128 (September, 2016) for the 18S rRNA gene amplicon
United Kingdom) and clustered by UPGMA (Unweighted Pair data.
GroupMethodwithArithmeticmean)analysis.
Statistical Analyses of High Throughput
Large-Scale Parallel Pyrosequencing of
Sequencing Results
16S and 18S rRNA Genes
QIIME data were imported into the R version 3.2.4
MicrobialcommunitycompositionwithinthedomainsBacteria,
(RDevelopmentCoreTeam,2016). Analyses and subsequent
Archaea, and Eukarya was addressed by high-throughput
visualizations were performed using functions from ggplot2
454 pyrosequencing. DNA extraction was carried out from
version 1.0.0 (Wickham, 2009). Alpha diversity inferences
three replicate samples collected at each sampling point.
included determination of observed number of OTUs per
The three replicates were then composted together before
sample, species richness estimation (Chao1), and calculation
statistical analysis. Bacterial V1–V3 region of the 16S rRNA
of diversity indices (Shannon and Simpson) using the package
genewasamplifiedusingtheprimerset27F(5(cid:48)-AGAGTTTGA
{phyloseq}, version 1.7.12 (McMurdie and Holmes, 2013).
TCCTGGCTCAG-3(cid:48)) and 338R (5(cid:48)-TGCTGCCTCCCGT
Rarefaction curves were generated using the {vegan} package
AGGAGT-3(cid:48)) (Hongoh et al., 2003; Fierer et al., 2008);
(Oksanen et al., 2016). Beta diversity was evaluated through
Archaeal V3–V5 region of the 16S rRNA gene was amplified
principal coordinate analysis (PCoA) of weighted UniFrac
using Arch349F (5(cid:48)-CCCTACGGGGTGCASCAG-3(cid:48)) and
distancesobtainedthroughQIIME(LozuponeandKnight,2005)
Arch806R (5(cid:48)-GGACTACVSGGGTATCTAAT-3(cid:48)) (Hugoni
tovisualizesimilaritiesinthemicrobialcomposition.Canonical
et al., 2013), and Eukaryotic V9 region of the 18S rRNA
correspondence analysis (CCA) was performed to explore the
using 1380F (5(cid:48)-CCCTGCCHTTTGTACACAC-3(cid:48)) and
environmental factors that had the most significant influence
1510R (5(cid:48)-CCTTCYGCAGGTTCACCTAC-3(cid:48)) (Amaral-
on the microbial community structure. We used the adonis()
Zettler et al., 2009). Primers were bound to 10-nucleotide
function from the vegan library, which fits linear models to
multiplexing tags and to the 454 FLX sequencing adaptors A
distance matrices and uses a permutation test with Pseudo
(5(cid:48)-CCATCTCATCCCTGCGTGTCTCCGACTCAG-3(cid:48)) and
F-ratios.
B (5(cid:48)-CCTATCCCCTGTGTGCCTTGGCAGTCTCAG-3(cid:48)) in
the forward and reverse regions, respectively. PCR reactions
wereperformedusingAmpliTaqGold360MasterMix(Applied Prediction of Functional Content from
Biosystems, United States) following the instructions provided Marker Gene Survey and Statistical
bythemanufacturer.Productsobtainedfromseparatereactions Analysis of PICRUSt Results
werepooledandpurifiedusingGelBandPurificationColumns
Bacterial function was inferred using the 16S rRNA gene
(GE Healthcare, United Kingdom), and DNA concentration
highthroughputsequencingdata.QIIMEbiologicalobservation
was determined using Qubit dsDNA assay kit (Invitrogen).
matrices were imported into PICRUSt (Langille et al., 2013),
Sequencing was carried out on a 454 Life Sciences Genome
andmetagenomesinferredusingtheproposedpipeline.PICRUSt
Sequencer FLX (Roche 454 Life Sciences, Branford, CT,
uses marker gene data to query a reference database for the
United States) at Macrogen Inc. (Seoul, South Korea). The raw
closestreferencegenomeavailable.Genomic-driveninferenceof
reads have been deposited into the NCBI short-reads archive
functionisthenusedtopredictgenefamilies,whicharetherefore
database(BioProjectAccession:PRJNA391850).
combinedtoestimatethecompositemetagenome(Langilleetal.,
2013).Briefly,aPICRUSt-compatibleOTUtablewasconstructed
Bioinformatic Analyses of High
inQIIME.Normalizationby16SrRNAcopynumberperOTU
Throughput Sequencing Data
was performed with the normalize_by_copy_number.py script,
16S/18S rRNA gene amplicon sequences were analyzed using followed by metagenome functional prediction for each sample
QIIME version 1.9 (Caporaso et al., 2010a). Reads were quality (predict_metagenomes.py and categorize_by_function.py). The
checked (minimum length was set to 200 bp, ambiguous bases accuracy for the predicted metagenome was tested through the
andmismatchesinprimersequenceswerenotallowed,minimum NearestSequencedTaxonIndex(NSTI),reflectingthepresence
qualityscorewas25)andassignedtosamplesbasedonnucleotide of reference genomes that are closely related to the samples in
barcodes.Quality-controlledsequencesweredenoizedusingthe analysis.
QIIME denoiser. Chimeras were predicted with ChimeraSlayer PICRUStresultswereimportedintoR(RDevelopmentCore
and removed prior to downstream analyses. Chimera-free Team, 2016). Probability distribution of data was determined
sequences were clustered into operational taxonomic units using the function descdist from the {fitdistrplus} package
(OTUs) defined by a 97% sequence similarity cutoff using and uniformity of variance was tested with the Levene test
UCLUST (Edgar, 2010). Representative sequences from each (leveneTest{car}).Differencesbetweenexperimentalgroupswere
OTU were aligned using PyNAST (Caporaso et al., 2010b) evaluatedusingtheKruskal–Wallistestforcontinuousvariables
FrontiersinMicrobiology|www.frontiersin.org 4 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #5
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
(kruskal.test{stats}).Ap-valueoflessthan0.05wasconsideredto
indicatestatisticalsignificance.
SSU rRNA Gene (18S) Clone Libraries
Construction, Sequencing, and Gene
Phylogeny Reconstruction
The diversity of eukaryotes was analyzed in parallel through
full length 18S rRNA gene clone library construction. DNA
from water, mineral fractions, and biofilms samples was
extracted using the PowerSoil DNA Isolation Kit (MoBio,
United States) according to manufacturer’s recommendations.
The 18S rRNA gene was amplified employing the primer
set EK-42F (5(cid:48)-CTCAARGAYTAAGCCATGCA-3(cid:48))/EK-1498R
(5(cid:48)-CACCTACGGAAACCTTGTTA-3(cid:48)) (Marande et al., 2009).
PCR amplicons were generated using AmpliTaq Gold 360
(Applied Biosystems, Life TechnologiesTM, United States)
following the instructions provided by the manufacturer. The
amplifications were performed in a MJ Mini Thermal Cycler
(Bio-Rad, United States) with an annealing temperature of
55◦C for 35 cycles. The PCR reactions obtained from separate
reactions were pooled and purified using the Illustra GFX
purificationkit(GEHealthcare,UnitedKingdom).PCRproducts
were cloned into pGEM-T Easy Vector System I (Promega,
United States) following the manufacturer’s instructions. The
ligation mixtures were then transformed into Escherichia coli
competent JM109 cells (Promega Corporation). Transformant
FIGURE2|Spatialvariationofgeochemicalvariables[conductivity,Cond.;
colonieswerescreenedviaPCRamplificationoftheinsertswith redoxpotential,Eh;pH;T,◦C;andoutsidelight(%)]oftheAMDsamples
flanking vector primers (M13F 5(cid:48)-GTTTTCCCAGTCACGAC- acrosscollectionsites.Symbols(dotsortriangles)andverticalerrorbars
3(cid:48); M13R 5(cid:48)-GAAACAGCTATGACCATG-3(cid:48)), and amplicons representthemeanofthethreemeasurementsateachsamplinglocationand
were sequenced according to the protocol of the BigDye theircorrespondingstandarderrors,respectively.
Terminator v3.1 sequencing kit (Applied Biosystems) and
subjected to capillary electrophoresis in an ABI PRISM 3130xl
GeneticAnalyzer(AppliedBiosystems). (14.25◦C), a lower redox potential (53 mV), and a higher
Sanger electropherograms were corrected and checked for conductivity (8.89 mS cm−1). The samples WOUT and WEN
chimera presence using the USEARCH (Edgar, 2010) using as werelocatedundertheinfluenceofthesunlight,receiving98and
reference the SILVA database (release 123_1) (Pruesse et al., 3% of total irradiance, respectively. The rest of samples (WIN,
2007). 18S rRNA gene sequences were automatically aligned S, BS, BF, B1A, B1B, and B2) were collected in total darkness
using the SINA aligner against the SILVA SSURef_123_1 (Figure2).
reference alignment (Pruesse et al., 2012). The resulting
Microbial Diversity of Bacteria, Archaea,
alignment was manually examined to correct erroneously
situatedbaseswithARB(Ludwig,2004).Amaximumlikelihood and Eukaryotes in the Different
phylogeny was constructed using general time reversible Microhabitats of Los Rueldos AMD
distances to define the nucleotide substitution model with
Bacterial 16S rRNA genes were successfully amplified from all
RAxML(Stamatakis,2014).
samples, whereas archaeal 16S rRNA amplicons were obtained
only from the samples WOUT, WEN, WIN, S, BS, and ST.
Our research team had previously explored the bacterial and
RESULTS
archaeal diversity in B1A, B1B, and B2 biofilms present in the
drainage(Méndez-Garcíaetal.,2014).Thesedatawereusedfor
Environmental Conditions across
comparative purposes in the present work due to the existence
Collection Sites
ofverystablepopulationsacrosstime,asdemonstratedthrough
The AMD samples were characterized by a low pH (2 ± 0.95, denaturinggradientgelelectrophoresis(DGGE)analysisinB1AB
mean ± SD), and a mean (±SD) temperature, redox potential, and B2 throughout a 5-year period (Supplementary Figure S1).
and conductivity of 12.49 ± 4.98◦C, 232 ± 114 mV, and Eukaryotic 18S rRNA genes were successfully amplified from
4.38 ± 2.23 mS cm−1, respectively. An exception to this samplesWOUT,WEN,andthestreamerB1A.
homogeneitywasthestalactite(ST)sample,inwhichthedripping The raw sequence data of the samples consisted of 456,928
water showed a higher pH (6.5 ± 0.94), higher temperature (155,513 bacterial, 140,769 archaeal, and 160,646 eukaryotic
FrontiersinMicrobiology|www.frontiersin.org 5 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #6
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
FIGURE3|(A)Rarefactioncurvesata3%ofdissimilaritycut-offamongsequences.(B)Alphadiversitymetrics(observedspecies,Chao1,Shannon,andSimpson
indices)forBacteria,Archaea,andEukaryainLosRueldosAMDsamples.
reads), which were ultimately clustered into 844 bacterial, 535 been needed for total recovery of the archaeal and eukaryotic
archaeal, and 176 eukaryotic OTUs. Rarefaction curves showed diversity (Figure 3A). The stalactite sample, ST displayed the
that the bacterial diversity was recovered satisfactorily in all highest bacterial diversity among all samples (metrics observed
samples,whereasamoreextensivesequencingeffortwouldhave OTUs, Chao1, and Shannon and Simpson indices were 410,
FrontiersinMicrobiology|www.frontiersin.org 6 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #7
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
700, 4.8, and 0.97, respectively). The samples showing a higher Generaandraretaxa
archaealdiversitywereST(observedOTUs,andChao1,Shannon, Dominantbacterial generaincludedAcidimicrobium(7%WIN,
andSimpsonindiceswere125,173,1.94,and0.71,respectively) 3.15% WOUT, 3.35% BF, 12.15% BS, 3.85% WEN, and 5.4%
and WIN (Shannon and Simpson indices were 1.97 and 0.78, S); Leptospirillum (35.9% BF, 23.55% BS, 14.8% S, 6.45% WIN,
respectively). The highest diversity of eukaryotes was estimated 6.45%WEN,1.35%WOUT,and0.2%ST);Acidiphilium(0.35%
forthewaterinthepondoutsideofthegallery(WOUT),which WEN, 0.2% BF, 0.1% S, and 0.1% WOUT); Acidobacterium
receivedthemajorinfluenceofthesunlight(observedOTUs,and (6.8% BS, 4% S, 3.2% WEN, 2.25% BF, and 2% ST); Thiomonas
indices Chao1, Shannon and Simpson were 105, 119, 1.78, and (6.55% ST); Gallionella (13% ST); Acidithiobacillus (1.8% BF,
0.78,respectively)(Figure3B). 1.2% WIN, 1.1% S, and 0.2% WEN), and an undetermined
genus belonging to the Ca. Saccharibacteria (18% WIN, 9.8%
DiversityofBacteria WOUT, 6.3% BS, 2.2% S, 1.7% WEN, 0.5% ST, and 0.3%
Bacterialphylaandclasses BF). The sediment (S) was characterized by the presence of
Proteobacteriawasthemostwidelydetectedphylum(accounting RCP1-48 Gammaproteobacteria, Xanthomonadaceae bacteria,
for 51.9% of the total reads), followed by Cyanobacteria Leptospirillum, and AD3 bacteria; the stalactite (ST) was
(12.4%, present only in WOUT and WEN), Nitrospirae dominated by Gallionella, uncharacterized Actinomycetales
(10%), Actinobacteria (7.2%), TM7 (hereinafter referred to as (Actinobacteria), and Xanthomonadaceae; Leptospirillum,
CandidatusSaccharibacteria,Albertsenetal.,2013)(4.2%),AD3 unknown Betaproteobacteria, Xanthomonadaceae, and
(2.7%),Acidobacteria(2.6%),andFirmicutes(0.64%). RCP1-48 bacteria predominated in the subaerial biofilm
Bacterial taxa with relative abundances higher than 1% BF; Leptospirillum, RCP1-48, and AD3 bacteria dominated in
affiliated within the phyla Actinobacteria, Cyanobacteria, BS, and unknown genera affiliating within the Acidobacteria,
Nitrospirae, Proteobacteria, and Ca. Saccharibacteria in water Acidimicrobiales, and Ca. Saccharibacteria were present in
samples (WOUT, WEN, WIN); within the Acidobacteria, proportions above 6%; WEN and WOUT were dominated
Actinobacteria, AD3, Firmicutes, Nitrospirae, Proteobacteria, by Cyanobacteria. Uncharacterized Ca. Saccharibacteria and
and Ca. Saccharibacteria in sediment (S); within Acidobacteria, RCP1-48bacteriawerethethirdmostabundantgroupsdetected
Actinobacteria,AD3,Firmicutes,Nitrospirae,andProteobacteria in WOUT and WEN; respectively. Filtered water (WIN) was
in the subaerial biofilm (BF); and within Acidobacteria, characterizedbythepresenceofBetaproteobacteria,unclassified
Actinobacteria, AD3, Bacteroidetes, Chlorobi, Chloroflexi, Ca. Saccharibacteria, and Leptospirillum (Supplementary
Cyanobacteria, Firmicutes, Nitrospirae, Proteobacteria, and TableS1A).
WPS-2inthesurfaceofthestalactite(ST)(Bacteria,Figure4A). Rare taxa were defined as representing 0.1–1% relative
Afractionofthesequencescouldnotbeassignedtoanytaxaat abundance based on 16S rRNA hypervariable region sequence
the phylum level and remained unclassified (1.3–5.6% in water counts (Hugoni et al., 2013). Phyla with abundances < 1%
samples, 9.8% in sediment sample, 12.4% in BS sample, 1.5% included Armatimonadetes (0.05%), Bacteroidetes (0.15%),
in the subaerial biofilm sample, and 11.6% in the stalactite). Chlorobi (0.4%), Chloroflexi (0.1%), Gemmatimonadetes
Proteobacteria dominated in all samples (61.1% WIN, 53.4% (0.04%),NC10(0.01%),andOD1(0.02%)inthestalactite.Inthe
BF, 53.3% ST, 45.2% S, 39.6% WEN, and 24.5% BS), except sediment, minority taxa included Planctomycetes (0.01%) and
in the acid water sample collected outside of the gallery WSP2(0.2%).Elusimicrobia(0.03%)weredetectedinST,S,and
(WOUT), where Cyanobacteria were predominant (Bacteria, WENandTM6(0.01%)wereidentifiedinWIN(Supplementary
Figure4A). TableS1A).
Main bacterial classes detected in the outside water sample
(WOUT) were unclassified Cyanobacteria (82.7%) and Ca. DiversityofArchaea
Saccharibacteria from an undetermined class (8.7%) (Bacteria, Archaealphylaandclasses
Figure 4B); classes detected in the water sample from the Archaea affiliating within the Euryarchaeota were predominant
pond at the entrance of the gallery (WEN) were unclassified in all samples (99.25% WEN, 99.2% BS, 99.1% WOUT,
Cyanobacteria (37.9%), Gammaproteobacteria (29.8%), and 87.6% S, 83.6% ST, and 79.6% WIN). Crenarchaeota were
Alphaproteobacteria (9.2%); the water inside of the gallery detected only in stalactites (15.4% ST), sediment (3.4% S),
(WIN) contained higher proportions of Alpha-, Beta- and and in the subaerial biofilm (0.4% BS). Ca. Parvarchaeota
Deltaproteobacteria (51.7, 50.5, and 6.3%, respectively). were detected in WIN (3.8%) (Archaea, Figure 4A). Class
Acidobacteria were predominant in the sediment and Thermoplasmata was predominant in WEN (93.7%), WOUT
biofilm at the interface sediment/water, as compared to (93.1%), and BS (99.1%). The acidic water (WIN) displayed
the other samples (7.2% in BS and 4.3% in S). BS was a higher proportion of Ca. Micrarchaea (3.75%). A variable
enriched in Nitrospira (22.3%), whereas S displayed a higher proportion of sequences remained unclassified (0.61–17.08% in
percentage of Gammaproteobacteria (36.1%). The subaerial water samples, 8.75% in sediment sample, 0.38% in BS sample,
biofilm contained distinctively more Nitrospira (33.1%), and1%inthestalactite).Thaumarchaeota(13.25%),unclassified
and the stalactite was enriched in Actinobacteria (13.6%), Crenarchaeota(0.3%),andCa.Parvarchaeota(0.1%)weremore
Ignavibacteria (3.4%), Bacteroidia (1.3%), and Alpha-, Beta-, abundantinthestalactite(ST).Thesediment(S)wasenrichedin
and Gammaproteobacteria (41.2, 34.2, and 10.3%, respectively) MBGA(3.35%)andunclassifiedEuryarchaeota(0.7%)(Archaea,
(Bacteria,Figure4B). Figure4B).
FrontiersinMicrobiology|www.frontiersin.org 7 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #8
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
FIGURE4|TaxonomyprofilesdisplayingthemicrobialdiversityintheAMDsamples,revealedbyhighthroughputsequencingofthe16S/18SrRNAgenes.(A)Bar
plotsdisplayingtherelativeabundancesatthephylumlevelwithinBacteria,Archaea,andEukarya.(B)Heatmapsshowingthemainclassesdetectedwithineach
phyla.
Archaealgenera fungal groups Ascomycota (30.4% WOUT), Chytridiomycota
Archaeal genera Ferroplasma (2.6% FW and 9.15% ST) and (1.6% WOUT), Basidiomycota (0.3% WEN), LKM11 (78.9%
Thermogymnomonas (93% POUT, 72% S, 74% FW, 69% ST, WEN,0.6%WOUT,4.9%B1A),Nucletmycea(4.5%WEN,0.1%
84% DZ, and 93% PIN) predominated in Los Rueldos samples WOUT, and 0.7% B1A), and LKM15 (0.1% WOUT and 0.2%
(SupplementaryTableS1B). B1A)(Eukarya,Figure4B).
Taxa with abundances < 1% included representatives of
DiversityofEukaryotes the supergroups Amoebozoa (1.6%) and Excavata (0.7%).
Eukaryotes across different microhabitats in Los Rueldos Unclassifiedsequencesaccountedfora12.7–14.2%inallsamples
AMD primarily belonged to the supergroup SAR (SupplementaryTableS1C).
(Stramenopiles + Alveolata + Rhizaria) and Opisthokonta Micrographs showing the morphology of the different
taxa.SARweredetectedinallsamples,beingdominantinB1A eukaryotic species present in open-air water samples collected
and WOUT (78.9 and 52.1%, respectively). The Stramenopiles at Los Rueldos (filamentous algae, cyanobacteria, diatoms, and
(1.7% WEN and 50.7% WOUT) were Diatomea of the class protozoans)arepresentedinSupplementaryFigureS2.
Bacillariophyceae (1.7% in WEN and 35% in WOUT) and Using OTU count data, a PCoA of weighted UniFrac
Chrysophyceae (15.7% WOUT). Alveolata included Ciliophora distanceswasconductedtoexplorethecompositionalsimilarities
of the class Spirotrichea (78.7% B1A, 0.5% WOUT, and 0.1% among the bacterial, archaeal, and eukaryotic communities
WEN). Detected Rhizaria belonged to the Cercozoa (0.9% across all distinct microhabitats (Figure 5). PCoA on the
WOUTand0.1%WEN)(Eukarya,Figure4A). phylogenetic distances among samples revealed the main taxa
Opisthokontawerepresentinallsamples(83.7%WEN,32.9% contributingtodifferencesintheirmicrobialdiversity.Sediment
WOUT, and 6% B1A), and were mainly represented by the and biofilm at the interface sediment/water samples (S, BS)
FrontiersinMicrobiology|www.frontiersin.org 8 September2017|Volume8|Article1756
fmicb-08-01756 September9,2017 Time:16:9 #9
Mesaetal. MicrobialDiversityacrossAMDMicrohabitats
FIGURE5|Principalcoordinatesanalysis(PCoA)basedontheweightedUniFracdistancemetricforBacteria,Archaea,andEukarya.PCoAonthephylogenetic
distancesamongsamplesrevealingthemaintaxacontributingtodifferencesintheirmicrobialdiversity(A).PCoAtovisualizesimilaritiesinthemicrobial
composition(B).
werecharacterizedbytheGammaproteobacteriaorderRCP1-48 was associated to the acid water accumulating right at the
and Methanomassiliicoccaceae archaea. The Nitrospira genus entrance (WEN) and outside the gallery (WOUT). Ciliate
Leptospirillum was distinctively contributing to community protozoa (Spirotrichea) mainly contributed to differences in
structure differences in the subaerial biofilm (BF). Ferroplasma the microbial eukaryotic community of the streamer B1.
archaea were differentially detected in the acid mine water Bacillariophyceae (diatoms) were found in the acid water
inside of the gallery, whereas the genus Thermogymnomonas accumulating outside the gallery (WOUT), and unclassified
FrontiersinMicrobiology|www.frontiersin.org 9 September2017|Volume8|Article1756