Diversity and dynamics of Archaea in an activated sludge wastewater treatment plant

18 Pages·2012·2.07 MB·English

Fredrikssonetal.BMCMicrobiology2012,12:140
http://www.biomedcentral.com/1471-2180/12/140
RESEARCH ARTICLE Open Access
Diversity and dynamics of Archaea in an activated
sludge wastewater treatment plant
Nils Johan Fredriksson1*, Malte Hermansson2 and Britt-Marie Wilén1
Abstract
Background: The activated sludge process is one of the most widely used methods for treatment of wastewater
and the microbial community composition in thesludgeis important for the process operation. While the bacterial
communities have been characterizedin various activated sludge systemslittleis known about archaeal
communities in activated sludge. The diversity and dynamics of the Archaea communityin a full-scale activated
sludge wastewater treatment plant were investigated by fluorescence in situ hybridization,terminal restriction
fragment length polymorphismanalysis and cloning and sequencing of 16S rRNA genes.
Results: The Archaea communitywas dominated byMethanosaeta-like species. During a 15 month period major
changes in thecommunitycomposition were only observed twice despite seasonal variations inenvironmental
and operatingconditions. Water temperature appeared to be the process parameter that affected thecommunity
composition the most. Several terminal restriction fragments also showed strong correlations with sludge properties
and effluent water properties.The Archaea were estimated to make up 1.6% of total cell numbers inthe activated
sludge and were presentboth as single cells and colonies of varying sizes.
Conclusions: The results presentedhere show that Archaea can constitute a constant and integral part ofthe
activated sludge and that it can therefore be useful to include Archaea infuture studies ofmicrobialcommunities
inactivated sludge.
Background (FISH), e.g. [10]. Methanogens [11,12] and putative
The activated sludge process is one of the most widely ammonia-oxidizing Archaea (AOA) [13-15] have been
used methods for treatment of wastewater. An import- detected by amplification of 16S rRNA and archaeal
ant factor for the operation of activated sludge waste- ammonia monooxygenase subunit A genes. Although
water treatment plants (WWTPs) is the solid-liquid present, Archaea seem to be of minor importance for
separation in which flocculation and settling are import- both nitrogen and carbon removal [11,16]. However, it is
ant processes [1]. Poor flocculation and settling of the still possible that the Archaea have other functions or
activated sludge lead to poor effluent quality and can affect the properties of the activated sludge. Addition
cause environmental problems in the receiving waters. of methanogens to the sludge in intermittently aerated
Thesludge characteristics dependonthemicrobial com- bioreactors increased the rates of specific oxygen uptake,
munity composition [2-4], the microbial activity [5] and denitrification and nitrification suggesting a symbiotic
the properties of the extra-cellular polymeric substances relationship with Bacteria [17]. The composition of the
in the flocs [6,7]. The bacterial community has been methanogenic community in anaerobic sludge has been
characterized in a number of activated sludge systems shown to be crucial for the structure and integrity of
[8,9]butverylittleisknownaboutarchaealcommunities granules [18-20] and if methanogens are present in acti-
in sludge. The presence of Archaea in activated sludge vatedsludgetheymaycontribute totheflocstructure.
has been shown by fluorescence in situ hybridization This study had three aims. The first was to describe
the Archaea community in the activated sludge of a full-
*Correspondence:[email protected] scale WWTP by cloning and sequencing of 16S rRNA
1DepartmentofCivilandEnvironmentalEngineering,WaterEnvironment
genes. Although there are many studies where activated
Technology,ChalmersUniversityofTechnology,Gothenburg,Sweden
Fulllistofauthorinformationisavailableattheendofthearticle sludge samples have been screened for the presence of
©2012Fredrikssonetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsofthe
CreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,
distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
Fredrikssonetal.BMCMicrobiology2012,12:140 Page2of18
http://www.biomedcentral.com/1471-2180/12/140
AOA (e.g. [13-15]), to our knowledge there are only two 16S rRNA gene sequences of lengths between 756 and
published studies on the diversity of Archaea in acti- 862 bases. Based on DNA similarity the sequences were
vated sludge from a full-scale WWTP [11,12]. One of assigned to operational taxonomic units (OTUs). The
the studies investigated two small WWTPs [11] and the sequences were assigned to OTUs corresponding to 25
other a seawater-processing WWTP [12]. The Rya species of 10 genera, 7 families/classes and 6 different
WWTP is a large WWTP treating municipal and indus- phyla. The Archaea community richness was estimated
trial wastewater, thus different from the WWTPs in to be at least 43 species of 19 different genera. Thus, the
those two studies. Since little is known about Archaea clone library covered at most 58% of the species and
in WWTPs and, importantly, sequence coverage for 53% of the genera present in the activated sludge. Accu-
Archaea from WWTPs is still modest, the 16S rRNA mulation curves (Figure 1) also illustrate that the clone
sequences we obtained here would indicate if published library does notfully coverthe Archaeacommunity.
FISH probes were relevant. If so, the second aim was to
quantify the Archaea by confocal microscopy and FISH Evennessandfunctionalorganization
and to determine their localization in the flocs. The Figure 2 shows a Pareto-Lorenz evenness curve of the
third aim was to follow the dynamics of the Archaea Archaea community based on the relative abundances of
community for a longer period of time using terminal the 25 OTUs obtained by applying a 98.7% sequence
restriction fragment length polymorphism (T-RFLP) similarity threshold. The functional organization (Fo)
analysis. For the third aim, the samples that were used index, the combined relative abundance of 20% of the
were collected for previous studies of the dynamics of OTUs, is 56%, meaning that more than half of the
the floc composition and flocculation and settling prop- observed sequences belong to only five of the observed
erties of the activated sludge at the Rya WWTP [21,22]. OTUs. A high Fo index is an indication of a specialized
Analyzing the Archaea community in the same samples community since it means that a big part of the popula-
could give an indication of possible causes or effects of tion belongs to a small number of OTUs and performs
Archaeacommunity changes. a small number of ecological functions. In a completely
even community all OTUswould have the same number
Results of individuals and it would be possible for a large
ObservedandestimatedrichnessoftheArchaea number of different functions to be equally abundant. In
communityintheactivatedsludge the clone library, the five most abundant OTUs, which
A 16S rRNA gene clone library was constructed from a include 56% of the sequences, all belong to Methano-
sample of activated sludge collected at the aeration tank saeta andpresumably areall methanogens. Furthermore,
of the Rya WWTP at a time of normal operating condi- the composition of the clone library indicates that the
tions. There were no atypical process parameter values community includes a small number of ecological func-
or extreme events prior to sample collection. However, tions since 13 of 25 OTUs, including 77% of the
the F/M-ratio was higher at the time of the clone library sequences,wereidentifiedasMethanosaeta(Figure 3).
sample collection (May 2007) compared with the times
when samples were collected for FISH (December 2007) ComparisonwithavailablesequencesinGenBankandSILVA
and T-RFLP analyses (May 2003 - August 2004) SearchesinGenBankusingBLAST[25]andintheSILVA
(Table 1). Cloning and sequencing generated 82 archaeal rRNA database [26] found sequences with a sequence
Table1ComparisonofWWTPparametersatthedifferentsamplecollectiontimesa
Parameterb,c,d May03-Aug04e May07f Dec07g Comment
Tempb 15±3 15±1 11±1**
SRTb 3±1 3±0 2±0
F/Mb 0.008±0.002 0.014±0.004** 0.008±0.002 MaxvalueinMay2007
CODb 1058±240 999±194 1068±97±
NO23-Nb 48±8 46±9 42±22 MinandmaxvaluesinDec07
SSVIc 80±15 54 79
EffluentNSSc 23±17 26 31
aThethreeperiodswerecomparedusingtheKruskal-Wallistest.Astatisticallysignificantdifference,p(same)<0.05,ismarkedwithasterisks(**).bAveragevalues
(±standarddeviation)fromallsampledatesandthesixdaysprecedingthesampledates.cDataonlyfromsampledates,notincludingthesixprecedingdays.
dTheparametersarewatertemperature(Temp,°C),solidsretentiontime(SRT,days),foodtomassratio(F/M,g/kg*s),CODgoingintotheactivatedsludgetanks
(COD,g/s),nitrite/nitratelevelsgoingintotheactivatedsludgetanks(NO23-N,g/s),standardizedsludgevolumeindex(SSVI,ml/g)andeffluentnon-settleable
solids(EffluentNSS,mg/l).eSamplescollectedduringthisperiodwereusedforT-RFLPanalysis.fAsamplecollectedduringthisperiodwasusedforT-RFLPand
clonelibraryanalysis.gAsamplecollectedduringthisperiodwasusedforFISHanalysis.
Fredrikssonetal.BMCMicrobiology2012,12:140 Page3of18
http://www.biomedcentral.com/1471-2180/12/140
30 100%
98.7%
95%
s 25
U 90% 80%
T
O n
80% oe
ed 20 rtinc
rv poda 60%
e on
obs 15 e prabu 56%
r of ativTU 40%
be 1100 ulO
m mf
u uo
N C
5 20%
0
0%
0 50 100
0% 20% 40% 60% 80%100%
Number of collected clones
Cumulative proportion of OTUs
Figure1Accumulationcurvesofarchaeal16SrRNAgene
Figure2Pareto-Lorenzevennesscurve.82archaeal16SrRNA
sequences.82archaeal16SrRNAgenesequenceswereassigned
genesequencesweredividedin25OTUsbasedonasequence
toOTUsbasedonsimilaritythresholdsrepresentingthedivisionin
similaritythresholdof98.7%andtheOTUswererankedfromhigh
phylum(80%),family/class(90%),genus(95%)andspecies(98.7%)
tolow,basedontheirabundance.ThePareto-Lorenzevenness
levels[23,24].Theaccumulationcurvesshowtherateofdiscoveryof
curveistheplotofthecumulativeproportionofOTUabundances
newOTUsatthedifferentsequencesimilaritythresholds.
(y-axis)againstthecumulativeproportionofOTUs(x-axis).TheFo
index,i.e.thecombinedrelativeabundanceof20%oftheOTUs,
isshown.ThedottedstraightlineisthePareto-Lorenzcurveofa
similarityof98.7%orhigherfor22of25OTUs,including
communitywithperfectevenness.
78 of the 82 sequences (Table 2). With 100% coverage,
4 sequences could only be matched with sequence sim-
ilarities lower than 98.7% and may therefore represent terminal restriction fragment (TRF) profile of the first
new species belonging to the genera Methanosaeta sample in the time series, May 16, 2003, and all follow-
(OTU10 and OTU16) or the Thermoplasmatales, Cluster ing samples was on average only 5.8±6.3% and
B (OTU20). The most similar sequences in the databases 5.4±7.1% for the AluI and RsaI analysis, respectively
were from various types of soil environments, water (Figure 6). On two occasions, in October 2003 and in
environments and anaerobic bioreactors in North Amer- January2004,theBray-Curtisdistancepeaked,indicating
ica,EuropeandAsia.For30ofthe82sequences,thebest a deviation from the community composition at the be-
matchcamefromananaerobicbioreactor. ginning of the time series. The difference between the
TRF profiles of May 16, 2003 and May 22, 2007, four
Phylogenetictreeanalysis years later, was 10% and 0% for the AluI and RsaI anal-
The phylogenetic affiliation of the obtained 16S rRNA ysis, respectively. However, the sensitivity of theT-RFLP
gene sequences was determined by phylogenetic tree analysis is limited and TRFsof low abundance cannotbe
analysis. A phylogenetic tree for Euryarchaea inferred by detected. Thus, a Bray-Curtis distance of 0% between
maximum likelihood analysis is shown in Figure 4. A theTRF profiles of two samples does not indicate identi-
phylogenetic tree for Crenarchaea and Thaumarchaea cal community composition since there may be differ-
inferred by maximum likelihood analysis is shown in ences in the composition of Archaea of low abundance.
Figure 5. The majority of the sequences were determined To get a rough estimate of the sensitivity of theT-RFLP
tobeofgenusMethanosaeta(Figure3).Severalsequences method,a comparison wasmade between the theoretical
alsoaffiliatedwithdivisionsofunculturedArchaea. TRF profile of the clone library and the observed TRF
profile from the same sample. The comparison showed
DynamicsoftheArchaeacommunity that only TRFs with a relative abundance higher than
The community composition, as assessed by T-RFLP, 20% in the clone library were observed in the TRF
changed little throughout the monitored period.Thedif- profile (Table 3). Thus, the T-RFLP analysis shows the
ference, measured as Bray-Curtis distance, between the dynamics of the major groups of Archaea in the
Fredrikssonetal.BMCMicrobiology2012,12:140 Page4of18
http://www.biomedcentral.com/1471-2180/12/140
Figure3Communitycomposition.The8216SrRNAgenesequenceswereclassifiedaccordingtothephylogenetictreeanalysis.Thenumber
ofsequenceswithineachgroupisgiven.
activated sludge. The relative abundances of the or Thaumarchaeota sequences in the databases have
observed TRFs in all TRF profiles in the time series are otherpredicted TRFlengths.
showninFigures 7and8.
To identify the TRFs the observed TRF lengths were
compared with the predicted TRF lengths of sequences Correlationanalysis
in the clone library. The predicted TRFs from the Several TRFs showed a significant correlation with
sequences in the clone library were between 4 and 6 process parameters (Table 5). The parameters that cor-
bases longer than the observed TRFs (Table 3). Such a related with most TRFs were water temperature and ni-
discrepancy between observed and predicted TRF sizes trogen concentration. There were also significant
is commonly observed [32,33]. Not all observed TRFs in correlations between several TRFs and the sludge and
the time series could be matched with the predicted effluent water properties (Table 6). The parameter efflu-
TRFs of the clone library sequences. To explore the ent non-settleable solids (NSS) and the concentration of
possibility that the TRFs in the TRF profiles from the carbohydrates in extracted extracellular polymeric sub-
samples from 2003 and 2004 come from sequences stances (EPS) correlated with most TRFs. No TRF
other than those found inthe clone libraryacomparison showed a significant correlation with the sludge volume
was also made with a database of 5802 archaeal 16S orshearsensitivity.
rRNA gene sequences matching the primers used in this
study. The database was checked for sequences that
would result in any of the observed combinations of QuantificationandlocalizationofArchaeaintheactivated
TRFs generated by AluI and RsaI. The result of the sludgeflocs
analysis was a number of possible identities for each The 16S rRNA gene clone library indicated that pub-
observed combination of AluI and RsaI TRFs (Table 4). lished FISH probes would cover the Archaea at Rya
Although the database comparison may result in false WWTP. Archaea could be observed in the activated
identities of the TRFs, it is valuable because it gives an sludge flocs, both centrally located and close to the
indication about the range of species that could give the edges of the flocs. FISH analyses showed that the aver-
observed TRF combinations. By comparison with the age relative abundance of Archaea in the activated
clone library sequences the dominating TRFs (AluI 176, sludge of the aeration tank was 1.6% (Figure 9). In the
AluI 184, RsaI 74 and RsaI 238) were determined to anaerobic digester and in the water recycled into
represent Methanosaeta-like species. Comparisons with the activatedsludgetanks(reject water) there weremore
the predicted TRFs of 5802 Archaea sequences in the Archaea than Bacteria (Figure 9). In most images of
database (Table 4) showed that it is possible that the activated sludge flocs the percentage of Archaea was
dominating TRFs are from other species of the Eur- lower than 2% (Figure 10). Occasionally there were
yarchaeota than Methanosaeta. However, it is unlikely larger colonies of Archaea (Figure 11, panel A) but in
that the origin of theseTRFs is species of the Crenarch- most images Archaea were either present as individual
aeota or Thaumarchaeota, since all the Crenarchaeota cells orsmall colonies(Figure11,panelB).
Fredrikssonetal.BMCMicrobiology2012,12:140 Page5of18
http://www.biomedcentral.com/1471-2180/12/140
Table2Databasecomparisons
Databasematcha
OTU Matchingclones Acc.no. Identityb Taxonomy Sourceenvironment
OTU1 1 CU917405 99.8 Methanosaeta Digester
6 CU917423 99.6-100 Methanosaeta Digester
6 CU917466 99.8-100 Methanosaeta Digester
2 JF280185 100.0 Methanosaeta Highmountainhotspring
OTU2 2 AJ831102 99.9 Archaea Landfilldrainagelayer
4 CP002565 100.0 Methanosaetaconcilii StrainGP6
1 CU916678 100.0 Methanosaeta Digester
3 CU917245 99.9-100 Methanosaeta Digester
2 FR832406 99.9-100 Methanosaetaconcilii Digester
OTU3 6 CP002565 99.9-100 Methanosaetaconcilii StrainGP6
1 CU915936 100.0 Methanosaeta Digester
1 CU916215 99.9 Methanosaeta Digester
OTU4 3 AF050611 99.6-99.9 Methanosaeta Contaminatedaquifer
3 EU155906 99.3 Archaea Richminerotrophicfen
OTU5 2 AJ831108 99.9 Archaea Landfilldrainagelayer
3 CP002565 99.6-100 Methanosaetaconcilii StrainGP6
OTU6 4 EU155906 99.0-99.2 Archaea Richminerotrophicfen
OTU7 4 GU591511 98.8-99.1 Archaea Microbialfuelcell
OTU8 4 GU591511 98.6-99.1 Archaea Microbialfuelcell
OTU9 3 EU155906 98.7-99.2 Archaea Richminerotrophicfen
1 AY667272 98.7 Archaea TCE-dechlorinatinggroundwater
OTU10 1 EU155954 93.5 Archaea Richminerotrophicfen
1 FN691755 93.0 Archaea LakeLlebreta
OTU11 1 CU917466 99.9 Methanosaeta Digester
1 CU916809 99.8 Methanosaeta Digester
OTU12 2 AJ576227 99.5-99.9 Archaea Landfillleachate
OTU13 1 HM244086 99.0 Archaea Lakesediment
1 AF050611 100.0 Methanosaeta Contaminatedaquifer
OTU14 1 HQ592619 99.5 Archaea Activatedsludge
OTU15 1 FR749947 98.9 Methanocorpusculumsinense StrainDSM4274T
OTU16 1 AY693812 97.6 Euryarchaea Anaerobicsludge
OTU17 1 FR832415 99.8 Methanosaetaconcilii Digester
OTU18 1 CU917031 100.0 Archaea Digester
OTU19 1 AJ576235 99.8 Archaea Landfillleachate
OTU20 1 AF050619 98.4 Euryarchaeota Contaminatedaquifer
OTU21 1 AB353220 99.2 Euryarchaeota Thermophilicdigestedsludge
OTU22 1 HQ316970 100.0 Crenarchaeota Wastewatertreatmentplant,oilrefinery
OTU23 1 FR832415 98.8 Methanosaetaconcilii Digester
OTU24 1 EU399655 99.2 Archaea Phenol-degradingsludge
OTU25 1 CU917014 99.9 Archaea Digester
aBestmatchingentryinGenBankortheSILVArRNAdatabasewith100%coverage.bIdentityin%.
Fredrikssonetal.BMCMicrobiology2012,12:140 Page6of18
http://www.biomedcentral.com/1471-2180/12/140
Figure4Phylogenetictreeofarchaeal16SrRNAgenes.Consensustreeconstructedfrom100maximumlikelihoodtrees.Thebranchlengths
andthescalebarareproportionaltonucleotidedifferences.Bootstrapvaluesoutofatotalof100aregivenatthenodes.Thesequenceof
Aquifexpyrophiluswasusedasoutgroup.TheOTUnumbersoftheRyaWWTPsequencesaregivenwiththetotalnumberofsequenceswithin
thatOTUinparentheses.TheclusternamesareinaccordancewithKemnitz[27],Grosskopf[28]andChouari[29].
Discussion group of Archaea, then the T-RFLP data show that the
In this study the abundance, localization, composition main part of the Archaea community was the same in
and dynamics of Archaea in the activated sludge of a 2003, 2004 and in May 2007 (Figures 7 and 8) and that
full-scale WWTP were assessed using FISH analysis, 16S we can use the clone library data to identify the TRFs
rRNAgene clonelibrary analysis andT-RFLPtimeseries in the T-RFLP time series. We further assume that the
analysis. These three analyses were all done on samples ArchaeacommunitystayedmainlythesameinDecember
collected at different times. However, for most process 2007, which make it possible to use the clone library data
parameters there were no significant differences between tochooseappropriateprobesfortheFISHanalysis.
these times (Table 1). The WWTP was also operated the The clone library sequences indicated that already
same way at all times, except for four months, May 24 published FISH probes were relevant for an estimation
to September 24, 2004, when the primary settlers were of the relative abundance of major Archaea groups. The
bypassed. The samples were therefore considered relative abundance of the Archaea has been estimated in
comparable. other investigations to be low, based on activity mea-
TheT-RFLP time series analysis showed that the most surements [11], and up to 8% of Bacteria [10] or 10% of
abundant TRFs were the same throughout 2003 and total cell numbers [16]. In this study Archaea was esti-
2004 as well as in May 2007 (Figures 7 and 8). If we mated, by FISH, to make up 1.6% of total cell numbers
assume that the same TRF always represent the same in the activated sludge, a relatively low abundance.
Fredrikssonetal.BMCMicrobiology2012,12:140 Page7of18
http://www.biomedcentral.com/1471-2180/12/140
U62817_SCA1170_soil Cluster 1.1b
93 AY278106_54D9_soil
AY278073_ROB17_soil
U62811_SCA1145_soil
AF058724_KBSCul13_soil
50
AY487103_TREC89-34_rhizospere
AB353209_TDS-J-r-C01_sludge
HM639792_C01_Rya, OTU22 (1)
96
DQ782340_ARC_10SAF3-A_clean_room
FJ784306_TX1H04_soil
DQ782354_ARC_1SAF1-45_NASA_clean room
92
AB050221_SAGMA-P_gold_mine_water
EU281334_N.gargensis
99 M88076_SBAR12_marine Cluster 1.1a
AF083071_Cenarchaeum_symbiosum
X96695_FFSB10_soil Cluster 1.1c
AJ131314_VAL96_freshwater Cluster 1.3
U81774_vadinDC69_sludge
Thaumarchaeota U59986_pGrfC26_sediment
89
AJ227956_ABS13_soil
AM503243_LLKoral29_rhizosphere
DQ399814_PMA2_sludge
AY835793_2LOC5_sludge
74 HM639819_G13_Rya, OTU18 (1)
CU917031_QEEE1CB011_mesophilic_anaerobic_digester
AJ227948_ARR38_rhizosphere
AY278093_GRU5_freshwater
AB661712_Desulfurococcus_amylolyticus_strain_JCM_91 88
NR_040922_Pyrobaculum_calidifontis_JCM_11548_strain_VA1
100
97
NR_029127_Sulfolobus_solfataricus_strain_IFO_15331
X14835_T.pendens Crenarchaeota
NC_013741_Archaeoglobus profundus_DSM_5631
NC_003901_Methanosarcina_mazei_Go1
98 CP002565_M.concilii_GP-6
AB071701_M.thermophila_PT
87 96
55
Euryarchaeota AY970347_M.harundinacea_6Ac
60 AY260435_M.parvum_DSM3828
NR_040881_Halosimplex_carlsbadense
49
M38637_Thermoplasma_acidophilum
HQ591420_Methanobacterium_formicicum_strain_MG-134
Bacteria
M83548_A.pyrophilus
0.07
Figure5Phylogenetictreeofarchaeal16SrRNAgenes.Consensustreeconstructedfrom100maximumlikelihoodtrees.Thebranchlengths
andthescalebarareproportionaltonucleotidedifferences.Bootstrapvaluesoutofatotalof100aregivenatthenodes.Thesequenceof
Aquifexpyrophiluswasusedasoutgroup.TheOTUnumbersoftheRyaWWTPsequencesaregivenwiththetotalnumberofsequences
withinthatOTUinparentheses.TheclusternamesareinaccordancewithDeLong[30]andJurgens[31].
Fredrikssonetal.BMCMicrobiology2012,12:140 Page8of18
http://www.biomedcentral.com/1471-2180/12/140
40%
AluI RsaI
35%
o
e t 30%
c
distanay-03 2205%%
s M
Curti16- 15%
-y 1100%%
a
Br 5%
0%
Figure6Communitystability.Thestabilityofthearchaealcommunityisillustratedbyplottingthedifference,measuredasBray-Curtisdistance,
betweentheTRFprofileofthefirstsampleinthetimeseries,May16,2003,andallfollowingsamples.TheBray-Curtisdistanceiscalculatedby
comparingtherelativeabundancesoftheTRFsintheTRFprofilesandislowforsimilarprofiles.ResultsfromboththeAluIandtheRsaIanalysis
areshown.
However, the importance ofa microbial group cannotbe The composition of the Archaea community was
deduced by abundance alone. Putative AOA were 1-10% investigated by analysis of 82 16S rRNA sequences. The
of total cell numbers in activated sludge, but despite this community richness was estimated to be 43 species of
abundancetheydidnotcontributesignificantlytonitrifi- 19 genera. As expected, the clone library does not fully
cation [16], whereas foaming organisms have great cover the Archaea community (Figure 1). However, the
impact on floc structure and sludge properties even 25 species of 10 genera that were observed are assumed
when present in numbers around 1% [35,36]. Another to represent the most abundant groups. The Archaea
example is ammonium oxidizers, which at an abundance community was functionally specialized (Figure 2) with
of 3-5% (of total bacteria), could perform the first step 77% of the sequences in the clone library being
in a successful 80% reduction of nitrogen in an activated Methanosaeta-like organisms (Figure 3). Specialized
sludge system [37]. Thus, despite their relatively low communities dominated by methanogens, although of
abundance, a possible contribution of Archaea to sludge other genera than Methanosaeta, have also been
propertiescannotberuledout. observedinactivatedsludgefromotherWWTPs[11,12].
Table3ObservedandpredictedTRFlengthsa
ObservedTRFlengthsT-RFLP2007-05-22 PredictedTRFlengths,clonelibrary2007-05-22
AluI %b RsaI %b AluI RsaI %c Identity
84 80 1% Thaumarchaeota/Cluster1.3
126 608 1% Thermoplasmatales/ClusterB
166 609 2% Thermoplasmatales/ClusterB
180 80 4% Methanosaeta
183 100% 74 86% 188 80 48% Methanosaeta
238 14% 188 242 21% Methanosaeta
188 NDd 2% Methanosaeta
189 264 1% ARCI
359 262 1% Thaumarchaeota/Cluster1.1b
400 609 1% Thermoplasmatales/RCIII
515 264 1% Methanocorpusculum
551 609 10% Thermoplasmatales/RCIII
ND 80 1% Methanosaeta
ND 613 1% Thermoplasmatales/ClusterC
ND ND 4% Methanosaeta
aObservedandpredictedTRFlengthsfromT-RFLPandclonelibraryanalysisofasamplefrom2007-05-22.bRelativeabundancebasedontotalfluorescence.
cRelativeabundancebasedonfrequencyinclonelibrary.dNDindicatescaseswhereaTRFcouldnotbepredictedorwherethepredictedTRFwasoutsidethe
detectionrange.
Fredrikssonetal.BMCMicrobiology2012,12:140 Page9of18
http://www.biomedcentral.com/1471-2180/12/140
Figure7RelativeabundancesofAluITRFs.RelativeabundancesofTRFsinnormalizedTRFprofilesgeneratedbydigestionwithAluI.Together
withtheRsaITRFs,theAluITRFswerecomparedwiththepredictedTRFsoftheclonelibrarysequences(identitiesinbold)andthesequences
fromtheRDPdatabase(identitiesinitalics)(Table4).
The T-RFLP time series analysis showed that the partsof activated sludge flocsanoxic microenvironments
Archaea community was practically the same in most can exist [38] which may allow growth of anaerobic
samples (Figures 7 and 8), despite variations in environ- organisms. In the activated sludge at Rya WWTP,
mental conditions such as organic loading rate and methanogens were observed both deep within the flocs
temperature [22]. Only in a few samples more than two and close to the surface (Figure 11). Although exposed
TRFs were observed. However, as shown in Table 3, tooxygen,themethanogensatthesurfacearenotneces-
the sensitivity of the T-RFLP analysis was low, so it is sarily inactive since methanogens have been shown
possible that there were changes in the composition of to be able to maintain viability [11] and activity [39] in
the less abundant groups of Archaea. A comparison be- thepresenceofoxygen.
tween the observed TRF lengths and the predicted TRF To avoid washout from the activated sludge, micro-
lengths of the clone library sequences identified the two organisms need to be active and have a doubling time
main TRFs as coming from Methanosaeta sequences, shorter than the sludge retention time. Pure cultures
given the assumption that all TRFs represent the same of Methanosaeta concilii have a temperature optimum
groupsofArchaeainallsampleswheretheyareobserved, at 35-40°C [40] and a doubling time of 4-7 days at
as discussed above. An alternative way of identifying the 37°C [41]. The low water temperature at Rya WWTP,
TRFs would be to compare the observed AluI and RsaI 10-20°C, does not necessarily prevent activity since
TRF combinations with the predicted TRF combinations Methanosaeta-like species have been shown to grow at
from Archaea sequences in the RDP database. A com- 9-14°C in bioreactors [42,43] and dominate methano-
parison with 5802 Archaea 16S rRNA sequences showed genic cultures from rice field soil at 15°C [44]. The
that sequences of Methanosaeta or other Euryarchaea solids retention time (SRT) at Rya WWTP is typically
would give the observed AluI and RsaI TRF combi- calculated as 5-7 days and could also allow for growth
nations, but no Crenarchaeota or Thaumarchaeota of Methanosaeta-like organisms. In this study, Methano-
sequences. In the following discussion we therefore saeta-like TRFs dominated throughout 15 months, and
assumethatthetwomainTRFscomefrommethanogens. correlation analysis showed that some of the Methano-
Methanogens are anaerobic and the oxygen concentra- saeta-like TRFs increased in abundance with increasing
tion in activated sludge is high. However, in the deeper temperature and increasing SRTs (Table 5), i.e.
Fredrikssonetal.BMCMicrobiology2012,12:140 Page10of18
http://www.biomedcentral.com/1471-2180/12/140
Figure8RelativeabundancesofRsaITRFs.RelativeabundancesofTRFsinnormalizedTRFprofilesgeneratedbydigestionwithRsaI.Together
withtheAluITRFs,theRsaITRFswerecomparedwiththepredictedTRFsoftheclonelibrarysequences(identitiesinbold)andthesequences
fromtheRDPdatabase(identitiesinitalics)(Table4).
theoretically more favorable conditions. Both the con- RC-IIIArchaeaexists,but a studyofamethanogenicen-
stant dominance in theTRF profiles and the response to richment culture suggests that RC-III Archaea are meso-
changed environmental conditions would be explained philic anaerobes growing heterotrophically on peptides
by an active population of Methanosaeta-like organisms. with a doubling time of approximately three days [27].
The detection of methanogens by FISH analysis also RC-III has been detected in soil [27], anaerobic biore-
showed the presence of rRNA, which is expressed actors [46] and groundwater [47]. This study shows that
in active cells. However, high rRNA levels may be main- RC-IIIArchaeacanalso bepresent inactivated sludge.
tained despiteinactivity. In conclusion, theactivity of the Thermoplasmatales-related sequences of Cluster B
Methanosaeta-likeorganismsisanopenquestion. and C were also found in the clone library. There are
If the Methanosaeta-like species do not grow fast currentlynoculturedrepresentativesorproposedpheno-
enough to avoid washout, their constant presence types for these groups. Cluster B and C sequences have
requires that they are constantly added to the sludge. been retrieved from environments with methanogenic
Possiblesourcesaretheinfluent wastewater andrecycled communities and complete or partially anoxic zones,
water from an anaerobic bioreactor. By FISH analysis, such as water [48], landfill leachates [49], sediments [50],
Archaeawasconfirmedtobepresentinhighnumbersin bioreactors [51] and the digestive tract of animals [52].
both the anaerobic bioreactor and in the reject water Thisstudyaddsactivatedsludgetothatlist.
(Figure 9). Thus the bioreactor might seed the activated One sequence, clone G15, belongs to a yet unde-
sludge with Archaea. This is supported by the fact that a scribed lineage of Archaea: ARC I [29]. The ARC I
majority of the detected 16S rRNA sequences cluster lineage is well-represented in anaerobic bioreactors and
with sequences from anaerobic sludge (Figure 4). Fur- in reactors with a high abundance of ARC I, the abun-
thermore, no sequences matched typical methanogens in dance of species related to M. concilii is low and vice
human fecal matter, such as Methanobacter smithii and versa [53], which could indicate a competition for acet-
Methanosphaera stadtmanae [45], indicating that fecal ate between these two lineages. The clone library in this
matter from the influent water was not an important study followed the same pattern with low abundance of
inputtothemethanogensintheactivatedsludge. ARC I and high abundance of M. concilii. The same pat-
The second largest group in the clone library was tern was also seen in the TRF profiles since the only
Thermoplasmatales-related sequences affiliated with time that the TRFs corresponding to sequence G15 was
Rice Cluster III (RC-III). No cultured representative of observed (January 28, 2004) the relative abundance of