Micromanipulation and further identification of FISH-labelled microcolonies of a dominant...

Environmental Microbiology (2004)

6

(5) 470ndash479 doi101111j1462-2920200400580x

copy 2004 Blackwell Publishing Ltd

Blackwell Science LtdOxford UKEMIEnvironmental Microbiology 1462-2912Society for Applied Microbiology and Blackwell Publishing Ltd 20046

5470479

Original Article

Identification of dominant denitrifier in activated sludgeT R Thomsen J L

Nielsen N B Ramsing and P H Nielsen

Received 10 September 2003 accepted 11 December 2003 Forcorrespondence E-mail phnbioaucdk Tel (+45) 96358503 Fax(+45) 96350558

Micromanipulation and further identification of FISH-labelled microcolonies of a dominant denitrifying bacterium in activated sludge

TR Thomsen

1

J L Nielsen

1

N B Ramsing

2

and P H Nielsen

1

1

Department of Life Sciences Aalborg University Sohngaardsholmsvej 57 DK-9000 Aalborg Denmark

2

Department of Microbial Ecology Institute of Biological Sciences Aarhus University DK-8000 Aarhus C Denmark

Summary

The activated sludge process relies on the formationof strong microbial flocs The knowledge about dom-inant floc-forming bacteria is at present very limitedespecially from a phylogenetic perspective In thisstudy numerous microcolonies in the activatedsludge flocs were found to be targeted by a

Betapro-teobacteria-

group-specific oligonucleotide probeusing fluorescence

in situ

hybridization (FISH) Someof these were micromanipulated and further identifiedby reverse transcriptase polymerase chain reaction(RT-PCR) and sequencing to belong to the

Aquaspir-illum

genus in the

Neisseriaceae

family A specificoligonucleotide probe Aqs997 was designed to tar-get the identified bacteria A survey in nine differentwastewater treatment plants with nutrient removal(WWTP) showed a high abundance of bacteria hybrid-izing to the oligonucleotide probe developedMicroautoradiography (MAR) combined with FISH onactivated sludge incubated with radiolabelled sub-strate showed uptake of substrate with oxygennitrate and nitrite as electron acceptor demonstratinga denitrifying potential of the bacteria investigatedThe

Aquaspirillum

-related bacteria seemed to beabundant denitrifiers in WWTPs with nitrogen removaland they were particularly numerous in plants mainlyreceiving domestic wastewater where they consti-tuted up to 30 of all bacteria

Introduction

In activated sludge wastewater treatment plants (WWTP)

formation of strong microbial flocs is of key importancefor an efficient separation of activated sludge and treatedwastewater (Eikelboom and van Buijsen 1983 Bidault

et al

1997) Therefore dominant floc-forming bacteriaare important for the floc properties and thus the qualityof the solid-liquid separation in wastewater treatmentplants

The microbial population in activated sludge has oftenbeen described from a functional point of view as aerobicheterotrophs nitrifiers denitrifiers sulphate reducers ironreducers and phosphate-accumulating organisms (PAOs)eg (Nielsen and Nielsen 2002a) However a substantialamount of these bacteria are still phylogenetically unde-scribed especially the aerobic heterotrophs and the den-itrifiers from WWTPs with nitrogen removal These twofunctional groups of bacteria represent 70ndash80 of all bac-teria in activated sludge (Nielsen and Nielsen 2002a) andthey can thus be anticipated to be of key importance tothe floc formation Most bacteria in activated sludge plantsbelong to the

Betaproteobacteria

eg (Juretschko

et al

2002 Wagner

et al

2002 Nielsen

et al

2004 Schmid

et al

2003) and recently it was shown that microcoloniesbelonging to this group form very strong microcoloniesthat can be very resistant to shear stress in contrast toother phylogenetic groups (Nielsen

et al

2004) Amongknown genera in this group present in activated sludgeare the

Nitrosomonas

(Wagner

et al

2002)

Zoogloea

(Lajoie

et al

2000)

Azoarcus

(Wagner and Loy 2002)and bacteria belonging to the

Rhodocyclus

genus eg

Candidatus

lsquoAccumulibacter phosphatisrsquo (Hesselmann

et al

1999 Crocetti

et al

2000) However there are stillmany unknown bacterial species hidden in this group andvery little is understood about their function

In several WWTPs with nitrogen removal we haveobserved unidentified large microcolony-forming bacteriabelonging to the

Betaproteobacteria

They appeared to beabundant in some of the WWTPs so the aim of this studywas to identify these bacteria capable of forming largemicrocolonies Instead of performing an extensive clonelibrary we used micromanipulation (Skerman 1968Blackall 1991 Hornsby and Horan 1994 Bradford

et al

1996) and the full cycle rRNA approach (Amann

et al

1995 Snaidr

et al

2002) to identify the bacteria andconfirm their abundance by fluorescence

in situ

hybridiza-

Identification of dominant denitrifier in activated sludge

471


Environmental Microbiology

6

470ndash479

tion (FISH) The great advantage of this technique is thata highly enriched sample can be obtained for further phy-logenetic investigations Here we present an additionalimprovement of the method where the micromanipulationwas performed on probe-labelled samples that enabledenrichment and identification of specific bacteria ofinterest which could not be otherwise distinguished bymorphological means The method was tested on somefilamentous bacteria and subsequently improved for iden-tifying microcolony-forming bacteria A new specific oligo-nucleotide probe was developed confirming the presenceof an abundant denitrifying and microcolony-formingspecies closely related to bacteria in the

Aquaspirillum

genus in the

Neisseriaceae

family

Results

Micromanipulation and phylogenetic analysis

The new combination that uses micromanipulation onsamples hybridized with specific oligonucleotide probeswas initially tested on filamentous bacteria from an indus-trial WWTP as filamentous bacteria are easier to micro-manipulate than microcolonies One

Thiothrix

filamentwas micromanipulated on the basis of morphologicalcharacteristics (Eikelboom and van Buijsen 1983) andfour filamentous bacteria all

Meganema perideroedes

were micromanipulated after hybridization to a species-specific oligonucleotide probe They all gave a suitablePCR product for cloning Two PCR products were clonedone of each type The

Thiothrix

was identified as

Thiothrixramosa

(980 similarity) whereas the other filament wasconfirmed to be


(1000 simi-larity) These results demonstrated the applicability of themethod to micromanipulate one single filamentous bac-terium and directly identify it further without previouscultivation

In activated sludge from Aalborg East WWTP

Betapro-teobacteria

constituted a high fraction (40ndash50) of allbacteria present and most were present in large micro-colonies (Nielsen

et al

2004) By applying various morespecific gene probes for ammonia oxidizing bacteria

Azoarcus Zoogloea Thauera

and

Candidatus

lsquoAccumuli-bacter phosphatisrsquo belonging to

Rhodocyclus

in a prelim-inary screening it became clear that there were someunidentified microcolonies belonging to the

Betaproteo-bacteria

Therefore these microcolonies were microman-ipulated for further phylogenetic analysis

Suspensions of FISH-labelled activated sludge werecompared with traditional FISH and no differences insignal intensities were observed Subsequently the FISHin suspension was used for micromanipulation of micro-colonies using a glass capillary needle A homogenizationof the sludge sample before hybridization and microman-

ipulation was important to make a rough separation of theflocs Micromanipulation was performed on microcoloniesfrom sludge hybridizing to the BONE23A probe It wasgenerally difficult to manipulate single colonies withoutcontaminants from single cells in the sample It was how-ever possible to get a highly enriched sample A total of20 microcolonies were micromanipulated Polymerasechain reaction products resulting from RT-PCR on micro-manipulated microcolonies were obtained for 14 of whichthree (samples A B C) were cloned Two to three typesof DGGE band representing different sequences wereachieved from each PCR product indicating that severalspecies were obtained by means of the micromanipula-tion Plasmids representing the different DGGE bandsfrom sample A B and C were then sequenced

Phylogenetic trees based on aligned nucleic data wereestimated by using distance matrix parsimony and max-imum likelihood The methods applied resulted in congru-ent tree topologies and in Fig 1 a maximum likelihoodtree is shown Sequences obtained from sample A affili-ated to the genus

Aquaspirillum

in the family

Neisseri-aceae

(sequence AquaspiA was 993 similar to

Pseudomonas lanceolata

)

Escherichia coli

and asequence rather distantly to clone OPB90 from a Yellow-stone hot spring Sample B sequences affiliated to thegenus

Aquaspirillum

(sequence AquaspiB was 990similar to


) and one sequencegrouped with

Peptostreptococcus

-like organisms SampleC also consisted of the genus

Aquaspirillum

-related bac-teria (AquaspiC was 991 similar to

Pseudomonas lan-ceolata

) and a sequence very similar to

Caulobacter

Thus even though the micromanipulation was performedon bacteria hybridizing to an oligonucleotide probe for asubgroup of the

Betaproteobacteria

sequences fromother groups were also obtained However in all threesamples very similar Aquaspi-sequences were present(993ndash995 identical) and they were closely related to


and

Aquaspirillum delicatum

inthe genus

Aquaspirillum

according to the nomenclatureused on the new RDP preview release (Cole

et al

2003)Only one other sequence (AY050604 uncultured bacte-rium clone GOUTB19) from a recent study on microbialdiversity in an

in situ

reactor system treating monochlo-robenzene-contaminated groundwater (Alfreider

et al

2002) was closely related to the sequences obtained inthis study (989ndash991 similarity to AquaspiA-C)

Probe design and test

The sequences closely related to the genus

Aquaspirillum

in the family

Neisseriaceae

were used to design an oligo-nucleotide probe (Aqs997) which matches these verysimilar sequences perfect in theory (AquaspiA AquaspiBand AquaspiC)


ATCC14667

472

T R Thomsen J L Nielsen N B Ramsing and P H Nielsen



6

470ndash479


AB021390

Comamonas test-osteroni

AF50188

Comamonas

sp AF501881 and a fewuncultured bacteria S--Aquaspi-997-a-A-18 (Aqs997)5

cent

-CTC TGG TAA CTT CCG TAC-3

cent

Escherichia coli

num-bering is used (Brosius

et al

1981)Six reported sequences affiliated in the

Comamona-daceae

and

Neisseriaceae

family (

Pseudomonas spinosa

AB021387

Hydrogenophaga

sp AF019037

Aquaspiril-lum gracile

AF078753

Acidovorax konjaci

AF078760

Comamonas

sp AF078773 and

Comamonas baltica

Z46256) have one central mismatch to Aqs997 makingthe gene probe less specific than intended Two compet-itors were designed to exclude the binding of probeAqs997 to cells with one mismatch comp1Aqs997 5

cent

-CTC TGG CAA CTT CCG TAC-3

cent

and comp2Aqs997 5

cent

-CTC TGG TCA CTT CCG TAC-3

cent

All other knownsequences have a weighted mismatch above 15 Com-pared with probe Cte (cited in Loy

et al

2003) whichtarget above 100 sequences of

Comamonas

spp

Aci-dovorax

spp

Hydrogenophaga

spp and

Aquaspirillum

spp probe Aqs997 has a much more narrow target Notall targets of Aqs997 are targeted by probe Cte eg Cte

has one mismatch to


whereasAqs997 matches perfectly Aquaspi-sequences obtainedin this study have theoretically perfect match with both theAqs997 and the Cte oligonucleotide probes Howevermore positive cells were observed with the Aqs997 probethan the Cte probe when hybridized to sludge from Aal-borg East WWTP The Aquaspi-sequences obtained bymicromanipulation of microcolonies hybridizing to theBONE23A oligonucleotide probe have also theoreticalperfect match with this probe which is also the case for


On the contrary

Aquaspirillumdelicatum

has one mismatch to the BONE23A probeProbe-binding profiles of the pure culture (


ATCC14667) and Aalborg East sludge samplesshowed that probe Aqs997 hybridized well up to 35formamide (results not shown) When the competitorprobes were used in combination with Aqs997 probe nodifferences were observed indicating that very few of thebacteria with one mismatch to probe Aqs997 were presentin the sludges investigated Applying the new probe inactivated sludge from Aalborg East WWTP revealed thatindeed 20ndash30 of the total number of bacteria in the

Fig 1

A 16S rRNA tree showing the phylogenetic affiliation of the micromanipulated microcolonies (AquaspiA AquaspiB and AquaspiC from samples A B and C) The tree was calculated using the maximum-likelihood method using default settings in

ARB

Branch points supported (bootstrap values

gt

75) by all inference methods used are indicated by solid circles and open circles indicate those supported by most methods The scale bar corresponds to 10 estimated sequence divergence


473



6

470ndash479

sludge sample were targeted by the Aqs997 probe par-ticularly microcolonies (Fig 2AndashF) Most Aqs997-positivebacteria showed a quite distinct morphology being rela-tively large coccoid cells with a diameter of 1ndash15

m

m(Fig 2A) The size of the microcolonies varied a lot Mostwere 10ndash15

m

m in diameter but colonies of up to 40

m

mwere observed Some of the microcolonies were not asdense as observed with other colonies indicating pres-ence of varying amounts of exopolymers Interestingly inapproximately 50 of the Aqs997-positive microcoloniesother bacteria were observed in between the cells formingmixed co-culture in the microcolonies Some of the otherbacteria were positive with the probe for the

Betaproteo-bacteria

In addition the specific probe also targeted afilamentous bacterium and some single cells with aslightly different morphology than cells within themicrocolonies

An extra oligonucleotide probe targeting the Aquaspi-sequences at position 460 was designed but did nothybridize to the sludge samples and isolates tested Thisposition is well known to be a rather inaccessible positionfor probe binding on the bacteria studied by Behrens

et al

(2003)

Oligonucleotide probes were also designed whenpossible to target the sequences obtained from themicromanipulation which were not related to the

Betaproteobacteria

However the probes hybridized tovery few microcolonies and single cells in the activatedsludge from Aalborg East WWTP so no further work onthese was performed It was not possible to design aspecific oligonucleotide probe for

Caulobacter

which isone of the bacteria suggested to contribute to floc forma-tion (Winkler and Cox 1980)

Escherichia coli ATCC25922was used as a non-target organism for the probes (havingfour mismatches to the Aquaspi-sequences) and positivehybridization signals were never observed

FISH survey in different WWTPs

Probe Aqs997 hybridized to microcolonies in all activatedsludges tested (Table 1) The plants all had nitrification-denitrification and in particular in the six WWTPs treatingdomestic wastewater a high number of Aqs997-positivebacteria were found ranging between 10 and 30 of allbacteria Less but still many were observed in the indus-trial WWTPs The majority of the bacteria hybridizing tothe new gene probe were growing in microcolonies withthe same characteristic morphology as that observed inAalborg East WWTP Some of the microcolonies consist-ing of Aqs997-positive cells were mixed with other bacte-ria so far unidentified especially in plants receivingdomestic wastewater Some of these bacteria belongedto the Betaproteobacteria Microcolonies consisting ofbacteria hybridizing with the probe for Thauera were

observed in all treatment plants investigated whereasZoogloea was mostly found in the industrial WWTPs

Autoradiography

Microautoradiography was applied to study the physiologyof bacteria targeted by the Aqs997 probe in sludge fromAalborg East WWTP A mixture of amino acids was addedto the sludge under conditions with oxygen nitrate ornitrite as electron acceptor and under anaerobic condi-tions without any added electron acceptor Substrateuptake by Aqs997-positive microcolonies and single cellswere evident with oxygen nitrate and nitrite as electronacceptor but not under anaerobic conditions An exampleis shown in Fig 2GndashJ where the silver grains on top ofmany Aqs997-positive cells indicate substrate uptakeunder denitrifying conditions and thus illustrate a denitri-fying potential of this bacterium Labelled acetate etha-nol glucose lactate and propionate was also tested butno substrate uptake was found by Aqs997-positive cellsunder any conditions tested (See Table 2) However otherbacteria in the flocs were found to take up the testedsubstrates thus serving as positive controls

Discussion

Evaluation of method

The combination of micromanipulation of FISH-labelledfilaments or microcolonies in activated sludge sampleswith RT-PCR for further identification was successful Agreat advantage of using FISH-labelled samples is thatthe identity is known to a certain extent eg to belongingto the Beta 1 group of Betaproteobacteria which can betested after sequencing so the number of oligonucleotide

Table 1 Screening of bacteria in nine different wastewater treatmentplants with the new oligonucleotide probe Aqs997

PlantTypewastewater Process designa Aqs997 ()

1 Aalborg Eastb Domestic C N DN CP BP 20ndash302 Aalborg West Domestic C N DN CP BP 20ndash303 Egaring Domestic C N DN BP 20ndash304 Horsens Domestic C N DN CP 20ndash305 Aalbaeligk Domestic C N DN CP 10ndash206 Aabybro Domestic C N DN CP 10ndash207 Daka Loesning Industrial C N DN 5ndash108 Daka Randers Industrial C N DN 5ndash109 Daka Ringsted Industrial C N DN 5ndash10

a C = carbon removal N = nitrification DN = denitrificationBP = biological phosphorus removal CP = chemical phosphorusremovalb Aalborg East WWTP where the micromanipulation and furtheridentification of BONE23A-positive microcolonies in this study wasperformed

474 T R Thomsen J L Nielsen N B Ramsing and P H Nielsen

copy 2004 Blackwell Publishing Ltd Environmental Microbiology 6 470ndash479

Identification of dominant denitrifier in activated sludge 475


probes to be designed and tested can be minimized Theapplied method was more successful on filamentousbacteria than on microcolonies as the filaments weregenerally easier to separate from the floc material It wasnot possible to obtain microcolonies consisting of only onesingle species but only highly enriched samples from themicromanipulated microcolonies In some cases this wasdue to the presence of several species within the micro-colony in others probably contamination during themicromanipulation procedure This might vary accordingto sludge type and bacteria species and according to ourexperience some sludges and microcolonies are moresuitable than others for micromanipulation This simplemethod could probably be improved but as it wasgenerally impossible to micromanipulate single puremicrocolonies colonies from several micromanipulationscan be pooled followed by RT-PCR to minimize the furtherwork

Several other methods of sorting bacteria exist egmicromanipulation using a Bactotip (Froumlhlich and Koumlnig1999) and cell fishing based on biotinylated polyribonu-cleotide probes (Stoffels et al 1999) In complex sam-ples only flow cytometry has been used successfully tosort bacteria (Wallner et al 1997) The advantages ofmicromanipulation compared with flow cytometry are thatthe sorting can be based on morphology in case no rele-vant oligonucleotide probe exists the sensitivity is higherit is possible to isolate microcolonies despite the 3D-struc-ture of the cell aggregate and that the method is rathersimple and the costs are low

Identity and significance of microcolony-forming bacteria

The identified bacterium represented by the sequences

AquaspiA-C was closely related to Pseudomonas lan-ceolata and Aquaspirillum delicatum in the family Neisser-iaceae and a clone GOUTB19 described by Alfreider et al(2002) The clone GOUTB19 was present in one out of87 sequences obtained in a study of a reactor treatingpolluted ground water The nomenclature we have usedis in accordance with the new RDP preview release (Coleet al 2003) and Bergeyrsquos manual (Garrity et al 2002)The only difference is that in the new RDP preview releasePseudomonas lanceolata is placed within the Betaproteo-bacteria in the genus Aquaspirillum whereas in Bergeyrsquosmanual it affiliates with the Gammaproteobacteria In theolder version of RDP Aquaspirillum delicatum andPseudomonas lanceolata are both placed in the Acidovo-rax group within the Betaproteobacteria

The experiments using MAR showed that mostAquaspirillum-related bacteria were active under both aer-obic and anoxic conditions with nitrate and nitrite as elec-tron acceptor suggesting that they are denitrifiers with acapability to grow under such conditions This is in agree-ment with other recent studies showing that other bacteriarelated to the family Neisseriaceae and Comamona-daceae are denitrifiers Denitrifiers have been isolatedfrom activated sludge several times (eg Gumaelius et al2001 Khan and Hiraisi 2001 Khan et al 2002 Hiraishiand Khan 2003) A few of these strains are relativelyclosely related to the Aquaspirillum-related sequencesobtained in our study (eg 948 similarity with aAquaspirillum psycrophilum-related strain NOS8 fromKhan et al 2002)

The closest isolated relatives to the Aquaspi-sequences Aquaspirillum delicatum and Pseudomonaslanceolata are heterotrophs grow under aerobic condi-tions but are unable to perform full denitrification (canreduce nitrate to nitrite) (Leifson 1962 Krieg 1984 Pot

Table 2 Uptake of radiolabelled substrates by Aquaspirillum-related bacteria (Aqs) or other floc-forming bacteria (Floc) with oxygen nitrate andnitrite as electron acceptor and under anaerobic conditions in Aalborg East WWTP - no uptake + uptake observed

Oxygen Nitrate Nitrite Anaerobic

Aqs Floc Aqs Floc Aqs Floc Aqs Floc

Acetate ndash + ndash + ndash ndash ndash +Propionate ndash + ndash + ndash ndash ndash +Lactate ndash + ndash + ndash ndash ndash +Amino acid mixture + + + + + + ndash +Ethanol ndash + ndash + ndash ndash ndash +Glucose ndash + ndash + ndash ndash ndash +

Fig 2 FISH micrographs showing two microscopic fields (AndashC and DndashF) of activated sludge from Aalborg East WWTP revealing the clustering and abundance of Aqs997-positive cells Panels GndashJ show the MAR-FISH micrographs of a microcolony consisting of both MAR-positive and MAR-negative Aquaspirillum-related cells (I) incubated under anaerobic conditions with nitrate as electron acceptor and with a mixture of 3H-labelled amino acids Panels A D and H show the sludge hybridized with Cy3-labelled Aqs997 probe and panels B E and G show the hybridization with FLUOS-labelled EUBmix Panel C and F are superimposed images of A+B and C+D respectively whereas panel J is the overlay of panels GndashI The scale bars correspond to 5 mm



et al 1991) Clone GOUTB19 and GOUTB4 from Alfre-ider et al (2002) were found in the monochlorobenzene-contaminated groundwater flowing out from a reactor Inthe Neisseriaceae family denitrifying isolates are knowneg strain NOS8 NSP4 and KSP2 from Khan et al(2002) which were isolated under denitrifying conditionsusing poly(3-hydroxybutyrate-Co-3-hydroxyvalerate) assubstrate However most relatives to the Aquaspi-sequences are still undescribed species and clones (SeeFig 1)

So far the dominant denitrifiers in nutrient removalplants treating municipal wastewater are unknown Inindustrial wastewater treatment plants bacteria belongingto the genera Azoarcus Thauera and Zoogloea are abun-dant and believed to be the dominant denitrifiers(Juretschko et al 2002 Wagner and Loy 2002) Ourstudy strongly suggests that bacteria belonging to thegenus Aquaspirillum were dominant denitrifiers in themunicipal treatment plants investigated They werepresent in lower numbers in the WWTPs treating industrialwastewater where Thauera and Zoogloea were oftenfound in large numbers and were thus likely to be thedominating denitrifyers It will be interesting to reveal whythis difference in the dominant population exists It mightbe related to the incoming wastewater composition wheremunicipal wastewater is relatively similar among the treat-ment plants at least in Denmark whereas industrialwastewater can be very different from plant to plant Thecapability to consume amino acids could indicate that theyare important in the degradation of proteins in the waste-water However there was also a difference in the phos-phorus removing operation of the plants so although theAquaspirillumndashrelated bacteria did not exhibit typicalphosphorus accumulating capabilities with uptake oforganic substrate under anaerobic conditions the growthconditions in these treatment plants may affect the domi-nance Future studies must be carried out to enlightenthis

It was interesting to observe that some microcoloniesconsisting of Aqs997-positive cells were mixed with otherso far unidentified bacteria As these types of coloniesare as abundant as the pure microcolonies and presentin all plants investigated this could indicate that some sortof interaction takes place

The new oligonucleotide probe developed Aqs997seems rather specific and targets particularly micro-colony-forming bacteria consisting of characteristic coc-coid cells However a few filamentous bacteria and a fewsingle cells were also targeted by probe Aqs997 If thedifferent morphotypes represented by a filamentous mor-photype and the characteristic coccid cells primaryfound as microcolonies but also as single cells corre-spond to different species it might be necessary todevelop 23S rRNA probes to reveal this difference How-

ever the MAR experiments indicated that it is the samespecies as the substrate uptake patterns were very simi-lar for the various morphotypes Other bacteria areknown also to grow as filaments and single cells depend-ing on the growth conditions (eg Foot et al 1992Gumaelius et al 2001)

The Aquaspirillum-related bacteria were the mostabundant Betaproteobacteria in Aalborg East WWTP andas bacteria belonging to this group have previously exhib-ited very strong floc-forming properties (Nielsen et al2004) it is likely that it was in fact the Aquaspirillum-related bacteria that were responsible for these proper-ties It is known that Zoogloea and Thauera produceextreme water containing exopolymers that can give set-tling and dewatering problems (Lajoie et al 2000) in away that is rarely seen in municipal treatment plants withN-removal In any case as dominant bacteria in thesenutrient removal plants they will strongly influence the flocproperties and this will be investigated further in futurestudies

Experimental procedures

Activated sludge sampling

The experiments were carried out with activated sludge fromAalborg East and Grindsted WWTP Denmark Aalborg EastWWTP has carbon removal nitrification denitrificationchemical and biological phosphorus removal and a mean cellresidence time (sludge age) of 20ndash30 days Grindsted WWTPis described by Nielsen et al (2000) The activated sludgeused for fluorescence in situ hybridization and micromanipu-lation was fixed according to Amann (1995) and sludge usedfor microautoradiography was collected and treated asdescribed by Nielsen and Nielsen (2002b)

FISH

Fluorescence in situ hybridization (FISH) was performedaccording to Amann (1995) The following oligonucleotideprobes were used in the preliminary screening of microcolonyforming bacteria related to the Betaproteobacteria in acti-vated sludge from Aalborg East WWTP NEU and Nso190(ammonia-oxidizing bacteria) Azo644 (targeting most Azoar-cus) PAO651 (specific for Candidatus lsquoAccumulibacter phos-phatisrsquo) ZRA (Zoogloea ramigera) BONE23A (Beta 1 groupof Betaproteobacteria) BTWO23A (Beta 2 group of Betapro-teobacteria) and Bet42a (Betaproteobacteria) Further infor-mation about the probes can be found in probeBase (Loyet al 2003) The probe Thau646 (Thauera spp) is describedby Lajoie et al (2000)

Suspension of FISH-labelled activated sludge was con-ducted using a modified version of the protocol published byFuchs et al (1998) Ethanol fixed sludge from Aalborg EastWWTP was gently homogenized using a glass tissue grinder(Thomas Scientificreg) Forty microlitres of the suspendedsludge was pelleted by centrifugation for 5 min at 1924 g andresuspended in 40 ml hybridization buffer with the used oligo-



nucleotide probes (final concentration 50 ng probe ml-1) Afterhybridization for two hours at 46infinC the sludge was pelletedfor 5 min at 1924 g and washed with preheated washingbuffer for 15 min at 48infinC The following probes were used insuspension of FISH EUB338 EUB338-II and EUB338-III (allBacteria) (Daims et al 1999) BONE23A (Beta 1 group ofBetaproteobacteria) and BTWO23A (Beta 2 group of Betap-roteobacteria) as competitor for BONE23A (Loy et al 2003)and Meg983 (Meganema perideroedes) (Thomsen et al2004)

Micromanipulation

Glass slides were covered with 1 agarose and 5 ml sus-pension of FISH-labelled activated sludge was placed on thesolidified agarose A Skerman micromanipulator (Skerman1968) was mounted on a long distance 33 yen objective on anAxioskop 2 plus microscope (Carl Zeiss) and the desiredFISH-labelled microcolony was micromanipulated using aglass capillary needle and directly transferred to a PCR tubecontaining 5 ml DEPC-treated water (deionized high qualitymolecular biology grade water) The setup was tested onfilamentous bacteria abundant in an industrial WWTP inGrindsted Denmark

Reverse transcriptase PCR (RT-PCR)

Superscript One-Step RT-PCR with Platinum Taq (InvitrogenLife Technologies) was applied for cDNA synthesis and sub-sequent PCR according to the manufacturerrsquos protocols usinga gene-specific primer PCR was performed using the prim-ers 27F and 1390R (Lane 1991) Only micromanipulationson ethanol fixed samples were successfully used for RT-PCRwhereas paraformaldehyde was found to inhibit the reversetranscriptase

Cloning

Fresh PCR amplificates were purified using a QIAquick PCRPurification Kit (Qiagen GmbH) The amplificates wereligated into a pCRreg4-TOPOreg vector and transformed intoONE SHOT Escherichia coli cells following the manufac-turerrsquos directions (Invitrogen) Plasmids of 20 randomlyselected clones were then recovered using a QIAprep spinminiprep kit (Qiagen GmbH)

Screening of clones by denaturing gradient gel electrophoresis (DGGE)

Polymerase chain reaction products were obtained from theselected plasmids using the primers 341F-GC (Muyzer et al1993) and 907r (Lane 1991) Denaturing gradient gel elec-trophoresis was performed using the D-Gene System (Bio-Rad Herlev Denmark) according to the manufacturerrsquosguidelines using polyacrylamide gels with a denaturing gra-dient of 30ndash70 Gels were run for 15 h at 100 V in 1yen TAEbuffer at a temperature of 60infinC and were stained in100 ng ml-1 SYBR Gold (Molecular Probes) Plasmids show-ing different band patterns were chosen for sequencing

Sequencing

Polymerase chain reaction amplification was performed ondiluted plasmids using M13 primers Sequence reactionswere performed using 05 ml primers (10 pmol ml-1) 5 mlDYEnamic ET dye terminator sequencing kit (AmershamBiosciences) and 2ndash7 ml purified PCR product in each set-upAfter amplification the reactions were loaded on aMegaBACE DNA sequencer (Amersham Biosciences)

Phylogeny and oligonucleotide probe design

Near complete (1363 bases) 16S rDNA sequences werecompiled and aligned using the automatic nucleic acid alignerin the ARB software package (httpwwwarb-homede) andalignments were refined manually Checks for chimericsequences were conducted using the CHECK_CHIMERA

program of Ribosomal Database Project (httprdpcmemsuedu) and the program BELLEROPHON developedby Hugenholtz and Huber (2003) Only unambiguouslyaligned sequences were used for calculation of trees bydistance matrix parsimony and maximum likelihoodapproaches using default settings in the ARB software Fur-ther phylogenetic analysis was performed using PAUP ver-sion 50b (written by David Swofford) as described inHugenholtz et al (2001) A distance matrix was calculatedusing default settings in PAUP version 50b

Oligonucleotide probes were designed using the probedesignmatch tools of the ARB software package Oligonucle-otides were 5cent-labelled with 5(6)-carboxyfluorescein-N-hydroxy-succinimide ester (FLUOS) or with sulphoindocya-nine dyes (Cy5 and Cy3) (Thermo Hybaid Ulm Germany)Probe-binding profiles were performed using hybridizationbuffers containing 0 to 70 formamide at 5ndash10 incre-ments on fixed activated sludge samples from Aalborg EastWWTP Aquaspirillum delicatum ATCC14667 (having perfectmatch to oligonucleotide probe Aqs997) and Escherichia coliATCC25922 (chosen as non-target organism with four mis-matches to the target site) Quantification of signal intensitywas performed using image analysis The most stringentconditions which gave a bright signal with the presumedtarget cells and the pure culture were chosen as optimalhybridization conditions for the oligonucleotide probes andused in the subsequent survey

The FISH survey was conducted on different activatedsludge samples fixed according to the procedures describedfor Gram-negative and Gram-positive bacteria by Amann(1995) The samples were immobilised on gelatine-coatedslides dehydrated in series of ethanol and subsequentlyhybridized

Autoradiography

Microautoradiography (MAR) experiments were performedaccording to the procedure described by Nielsen and Nielsen(2002b) The incubations were carried out using 2 ml dilutedactivated sludge (1 g SS l-1) A [3H]-labelled amino acid mix-ture (16 amino acids ICN Biomedicals) was used The finalconcentration of each amino acid was adjusted to 01 mM byadding unlabelled amino acids The amount of radioactivity



of each amino acid varied from 68 to 288 mmCi2 ml [3H]-labelled acetate ethanol glucose lactate and [14C]-propi-onate (Amersham Biosciences) were adjusted to a final con-centration of 1 mM and 10 mmCi2 ml The potential use ofnitrate and nitrite as electron acceptor was tested using apreincubation step with unlabelled substrate for 2 h beforethe labelled substrate was added (Andreasen and Nielsen2000) A preincubation step was also included under anaer-obic conditions An incubation time with labelled substratesof 3 h was used under aerobic anoxic conditions with nitrateand nitrite and under anaerobic conditions Nitrate and nitriteconcentrations were 2 mM and 05 mM in anoxic incubationsrespectively As a control for chemography sludge was pas-teurized at 70infinC for 10 min just before the incubation Afterincubation the activated sludge was fixed and prepared asdescribed under the FISH procedure before applying thephotographic emulsion (Lee et al 1999)

Nucleotide accession numbers

The GenBank accession numbers for the 16S rRNAsequences reported in this paper are AY322151- AY322153

Acknowledgements

The study is a part of the framework programme lsquoActivity andDiversity of Complex Microbial Systemsrsquo funded by the Dan-ish Technical Research Council We are grateful to Jane Ildaland Marianne Stevenson for their valuable technical assis-tance Caterina Levantesi is acknowledged for teaching TRTmicromanipulation Ulrike Purkhold Alexander Loy andMichael Wagner are acknowledged for helping TRT withARB

References

Alfreider A Vogt C and Babel W (2002) Microbialdiversity in an in situ reactor system treating mono-chlorobenzene contaminated groundwater as revealed by16S ribosomal DNA analysis Syst Appl Microbiol 25 232ndash240

Amann RI (1995) In situ identification of microorganisms bywhole cell hybridization with rRNA-targeted nucleic acidprobes In Molecular Microbial Ecology Manual Akker-mans ADL van Elsas JD and de Bruijn FJ (eds)London Kluwer Academic Publications pp MMEMndash3361ndashMMEM-33615

Amann RI Ludwig W and Schleifer KH (1995) Phylo-genetic identification and in situ detection of individualmicrobial cells without cultivation Microbiol Rev 59 143ndash169

Andreasen K and Nielsen PH (2000) Growth of Microthrixparvicella in nutrient removal activated sludge plants Stud-ies of in situ physiology Wat Res 34 1559ndash1569

Behrens S Ruumlhland C Inacio J Huber H Fonseca AgraveSpencer-Martins I et al (2003) In situ accessibility ofsmall-subunit rRNA of members of the domains Bacteria

Archaea and Eucarya to Cy3-labeled oligonucleotideprobes Appl Environ Microbiol 69 1748ndash1758

Bidault A Clauss F Helaine D and Balavoine C (1997)Floc agglomeration and structuration by a specific talc min-eral composition Wat Sci Technol 36 57ndash68

Blackall LL (1991) Use of the skerman micromanipulatorfor isolating actinomyceetes in the wastewater field Acti-nomycetes 2 8ndash12

Bradford D Hugenholtz P Seviour EM CunninghamMA Stratton H Seviour RJ and Blackall LL (1996)16S rRNA analysis of isolates obtained from Gram-nega-tive filamentous bacteria micromanipulated from activatedsludge Syst Appl Microbiol 19 334ndash343

Brosius J Dull TJ Sleeter DD and Noller HF (1981)Gene organization and primary structure of a ribosomalRNA operon from Escherichia coli J Mol Biol 148 107ndash127

Cole JR Chai B Marsh TL Farris RJ Wang QKulam SA et al (2003) The Ribosomal Database Project(RDP-II) previewing a new autoaligner that allows regularupdates and the new prokaryotic taxonomy Nucleic AcidsRes 31 442ndash443

Crocetti GR Hugenholtz P Bond PL Schuler A KellerJ Jenkins D and Blackall LL (2000) Identification ofpolyphospate-accumulating organisms and design of 16SrRNA-directed probes for their detection and quantificationAppl Environ Microbiol 66 1175ndash1182

Daims H Bruumlhl A Amann R Schleifer K-H andWagner M (1999) The domain-specific probe EUB338 isinsufficient for the detection of all Bacteria Developmentand evaluation of a more comprehensive probe set SystAppl Microbiol 22 434ndash444

Eikelboom DH and van Buijsen HJJ (1983) Microscopicsludge investigation manual Deft the Netherlands TNOResearch Institute for Environmental Hygiene MGndashTNO A94a

Foot RJ Kocianova E and Forster CF (1992) Variablemorphology of Microthrix parvicella in activated sludge sys-tems Wat Res 26 875ndash880

Froumlhlich J and Koumlnig H (1999) Rapid isolation of singlemicrobial cells from mixed natural and laboratory popula-tions with the aid of a micromanipulator Syst Appl Micro-biol 22 249ndash257

Fuchs BM Wallner G Beisker W Schwippl I LudwigW and Amann R (1998) Flow cytometric analysis of thein situ accessibility of Escherichia coli 16S rRNA for fluo-rescently labeled oligonucleotide probes Appl EnvironMicrobiol 64 4973ndash4982

Garrity GM Johnson KL Bell JA and Searles DB(2002) Taxonomic Outline of the Procaryotes BergeyrsquosManual of Systematic Bacteriology New York Springer-Verlag

Gumaelius L Magnusson G Pettersson B andDalhammer G (2001) Comamonas denitrificans sp novan efficient denitrifying bacterium isolated from activatedsludge Int J Syst Evol Microbiol 51 999ndash1006

Hesselmann RPX Werlen C Hahn D van der MeerJR and Zehnder AJB (1999) Enrichment phylogeneticanalysis and detection of a bacterium that performsenhanched biological phosphate removal in activatedsludge Syst Appl Microbiol 22 454ndash465

Hiraishi A and Khan ST (2003) Application of polyhy-



droxyalkanoates for denitrification in water and wastewatertreatment Appl Microbiol Biotechnol 61 103ndash109

Hornsby LA and Horan NJ (1994) Isolation of filamen-tous bacteria from activated sludge using micromanipula-tion Wat Sci Technol 28 2033ndash2034

Hugenholtz P and Huber T (2003) Chimeric 16S rDNAsequences of diverse origin are accumulating in the publicdatabases Int J Syst Evol Microbiol 53 289ndash293

Hugenholtz P Tyson GW Webb RI Wagner AM andBlackall LL (2001) Investigation of candidate divisionTM7 a recently recognized major lineage of the domainBacteria with no known pure-culture representatives ApplEnviron Microbiol 67 411ndash419

Juretschko S Loy A Lehner A and Wagner M (2002)The microbial community composition of a nitrifying-dentri-fying activated sludge from an industrial sewage treatmentplant analyzed by the full-cycle rRNA approach Syst ApplMicrobiol 25 84ndash99

Khan ST and Hiraisi A (2001) Isolation and characteriza-tion of a new poly (3-hydroxybutyrate) -degrading denitri-fying bacterium from activated sludge FEMS Microbiol Let205 253ndash257

Khan ST Horiba Y Yamamoto M and Hiraishi A (2002)Members of the family Comamonadaceae as primary poly(3-hydroxybutyrate-co-3-hydroxyvalerate) -degrading deni-trifiers in activated sludge as revealed by a polyphasicapproach Appl Environ Microbiol 68 3206ndash3214

Krieg NR (1984) Aerobicmicroaeropilic motile helicalvibroid gram-negative bacteria In Bergeyrsquos Manual of Sys-tematic Bacteriology Krieg NR and Holt JG (eds) Bal-timore The Williams amp Wilkins Co pp 71ndash84

Lajoie CA Layton AC Gregory IR Sayler GS TaylorDE and Meyers AJ (2000) Zoogleal clusters andsludge dewatering potential in an industrial activated-sludge wastewater treatment plant Wat Environ Res 7256ndash64

Lane DJ (1991) 1623S rRNA sequencing In NucleicAcid Techniques in Bacterial Systematics Stackebrandt Eand Goodfellow M (eds) Chichester UK Wiley pp113ndash175

Lee N Nielsen PH Andreasen KH Juretschko SNielsen JL Schleifer K-H and Wagner M (1999)Combination of fluorescent in situ hybridization andmicroautoradiography- a new tool for structure functionanalyses in microbial ecology Appl Environ Microbiol 651289ndash1297

Leifson E (1962) The bacterial flora of distilled and storedwater III New species of the genera Corynebacteria Fla-vobacterium Spirillum and Pseudomonas Int Bull Bacte-riol Nom Tax 12 161ndash170

Loy A Horn M and Wagner M (2003) probeBase ndash anonline resource for rRNA-targeted oligonucleotide probesNucleic Acids Res 31 514ndash516

Muyzer G De Waal EC and Uitterlinden AG (1993)

Profiling of complex microbial populations by denaturinggradient gel electrophoresis analysis of polymerase chainreaction-amplified genes coding for 16S rRNA Appl Envi-ron Microbiol 59 695ndash700

Nielsen JL and Nielsen PH (2002a) Quantification offunctional groups in activated sludge by microautoradiog-raphy Wat Sci Technol 46 389ndash395

Nielsen JL and Nielsen PH (2002b) Enumeration of ace-tate-consuming bacteria by microautoradiography underoxygen- and nitrate respiring conditions in activatedsludge Wat Res 36 421ndash428

Nielsen PH Muro MA and Nielsen JL (2000) Studieson the in situ physiology of Thiothrix spp present in acti-vated sludge Environ Microbiol 2 389ndash398

Nielsen PH Thomsen TR and Nielsen JL (2004) Bac-terial composition of activated sludge ndash importance for flocand sludge properties Wat Sci Technol in press

Pot B Gillis M and de Ley J (1991) The genus Aquaspir-illum In The Prokaryotes A Handbook of the Biology ofBacteria Ecophysiology Isolation Identification Applica-tion Balows A Truumlper HG Dworkin M Harder W andSchleifer KH (eds) New York Springer-Verlag pp2569ndash2582

Schmid M Thill A Purkhold U Walcher M Bottero JYGinestet P et al (2003) Characterization of activatedsludge flocs by confocal laser scanning microscopy andimage analysis Wat Res 37 2043ndash5052

Skerman VBD (1968) A new type of micromanipulator andmicroforge J Gen Microbiol 54 287ndash297

Snaidr J Beimfort C Levantesi C Rossetti S WaardeJVD Geurkink B et al (2002) Phylogenetic analysisand in situ identification of lsquoNostocoida limicolarsquo- like fila-mentous bacteria in activated sludge from industrial waste-water treatment plants Wat Sci Technol 46 99ndash104

Stoffels M Ludwig W and Schleifer KH (1999) rRNAprobe-based cell fishing of bacteria Environ Microbiol 1259ndash271

Thomsen TR Blackall LL Aquino de Muro M NielsenJL and Nielsen PH (2004) Meganema perideroedesgen nov sp nov a new filamentous Alphaproteobacteriafrom activated sludge Internat J Syst Evol Microbiol (inpress)

Wagner M and Loy A (2002) Bacterial community compo-sition and function in sewage treatment systems Cur OpinBiotechnol 13 218ndash227

Wagner M Loy A Nogueira R Purkhold U Lee N andDaims H (2002) Microbial community composition andfunction in wastewater treatment plants Antonie Van Leeu-wenhoek 81 665ndash680

Wallner G Fuchs B Spring S Beisker W and AmannRI (1997) Flow sorting of microorganisms for molecularanalysis Appl Environ Microbiol 63 4223ndash4231

Winkler MA and Cox EJ (1980) Stalked bacteria in acti-vated sludge Eur J Appl Microbiol Biotechnol 9 235ndash242


471



6

470ndash479

tion (FISH) The great advantage of this technique is thata highly enriched sample can be obtained for further phy-logenetic investigations Here we present an additionalimprovement of the method where the micromanipulationwas performed on probe-labelled samples that enabledenrichment and identification of specific bacteria ofinterest which could not be otherwise distinguished bymorphological means The method was tested on somefilamentous bacteria and subsequently improved for iden-tifying microcolony-forming bacteria A new specific oligo-nucleotide probe was developed confirming the presenceof an abundant denitrifying and microcolony-formingspecies closely related to bacteria in the

Aquaspirillum

genus in the

Neisseriaceae

family

Results

Micromanipulation and phylogenetic analysis

The new combination that uses micromanipulation onsamples hybridized with specific oligonucleotide probeswas initially tested on filamentous bacteria from an indus-trial WWTP as filamentous bacteria are easier to micro-manipulate than microcolonies One

Thiothrix

filamentwas micromanipulated on the basis of morphologicalcharacteristics (Eikelboom and van Buijsen 1983) andfour filamentous bacteria all


were micromanipulated after hybridization to a species-specific oligonucleotide probe They all gave a suitablePCR product for cloning Two PCR products were clonedone of each type The

Thiothrix

was identified as

Thiothrixramosa

(980 similarity) whereas the other filament wasconfirmed to be


(1000 simi-larity) These results demonstrated the applicability of themethod to micromanipulate one single filamentous bac-terium and directly identify it further without previouscultivation

In activated sludge from Aalborg East WWTP

Betapro-teobacteria

constituted a high fraction (40ndash50) of allbacteria present and most were present in large micro-colonies (Nielsen

et al

2004) By applying various morespecific gene probes for ammonia oxidizing bacteria

Azoarcus Zoogloea Thauera

and

Candidatus

lsquoAccumuli-bacter phosphatisrsquo belonging to

Rhodocyclus

in a prelim-inary screening it became clear that there were someunidentified microcolonies belonging to the

Betaproteo-bacteria

Therefore these microcolonies were microman-ipulated for further phylogenetic analysis

Suspensions of FISH-labelled activated sludge werecompared with traditional FISH and no differences insignal intensities were observed Subsequently the FISHin suspension was used for micromanipulation of micro-colonies using a glass capillary needle A homogenizationof the sludge sample before hybridization and microman-

ipulation was important to make a rough separation of theflocs Micromanipulation was performed on microcoloniesfrom sludge hybridizing to the BONE23A probe It wasgenerally difficult to manipulate single colonies withoutcontaminants from single cells in the sample It was how-ever possible to get a highly enriched sample A total of20 microcolonies were micromanipulated Polymerasechain reaction products resulting from RT-PCR on micro-manipulated microcolonies were obtained for 14 of whichthree (samples A B C) were cloned Two to three typesof DGGE band representing different sequences wereachieved from each PCR product indicating that severalspecies were obtained by means of the micromanipula-tion Plasmids representing the different DGGE bandsfrom sample A B and C were then sequenced

Phylogenetic trees based on aligned nucleic data wereestimated by using distance matrix parsimony and max-imum likelihood The methods applied resulted in congru-ent tree topologies and in Fig 1 a maximum likelihoodtree is shown Sequences obtained from sample A affili-ated to the genus

Aquaspirillum

in the family

Neisseri-aceae

(sequence AquaspiA was 993 similar to


)

Escherichia coli

and asequence rather distantly to clone OPB90 from a Yellow-stone hot spring Sample B sequences affiliated to thegenus

Aquaspirillum

(sequence AquaspiB was 990similar to


) and one sequencegrouped with

Peptostreptococcus

-like organisms SampleC also consisted of the genus

Aquaspirillum

-related bac-teria (AquaspiC was 991 similar to

Pseudomonas lan-ceolata

) and a sequence very similar to

Caulobacter

Thus even though the micromanipulation was performedon bacteria hybridizing to an oligonucleotide probe for asubgroup of the

Betaproteobacteria

sequences fromother groups were also obtained However in all threesamples very similar Aquaspi-sequences were present(993ndash995 identical) and they were closely related to


and


inthe genus

Aquaspirillum

according to the nomenclatureused on the new RDP preview release (Cole

et al

2003)Only one other sequence (AY050604 uncultured bacte-rium clone GOUTB19) from a recent study on microbialdiversity in an

in situ

reactor system treating monochlo-robenzene-contaminated groundwater (Alfreider

et al

2002) was closely related to the sequences obtained inthis study (989ndash991 similarity to AquaspiA-C)

Probe design and test

The sequences closely related to the genus

Aquaspirillum

in the family

Neisseriaceae

were used to design an oligo-nucleotide probe (Aqs997) which matches these verysimilar sequences perfect in theory (AquaspiA AquaspiBand AquaspiC)


ATCC14667

472




6

470ndash479


AB021390


AF50188

Comamonas


cent


cent

Escherichia coli


et al


Comamona-daceae

and

Neisseriaceae

family (

Pseudomonas spinosa

AB021387

Hydrogenophaga

sp AF019037


AF078753

Acidovorax konjaci

AF078760

Comamonas

sp AF078773 and

Comamonas baltica


cent


cent

and comp2Aqs997 5

cent


cent


et al


Comamonas

spp

Aci-dovorax

spp

Hydrogenophaga

spp and

Aquaspirillum


has one mismatch to




On the contrary





Fig 1


ARB


gt



473



6

470ndash479


m


m


m


Betaproteo-bacteria



et al

(2003)


Betaproteobacteria


Caulobacter






Autoradiography


Discussion


































FISH






Micromanipulation




Cloning




Sequencing







Autoradiography







Acknowledgements


References





















































472




6

470ndash479


AB021390


AF50188

Comamonas


cent


cent

Escherichia coli


et al


Comamona-daceae

and

Neisseriaceae

family (

Pseudomonas spinosa

AB021387

Hydrogenophaga

sp AF019037


AF078753

Acidovorax konjaci

AF078760

Comamonas

sp AF078773 and

Comamonas baltica


cent


cent

and comp2Aqs997 5

cent


cent


et al


Comamonas

spp

Aci-dovorax

spp

Hydrogenophaga

spp and

Aquaspirillum


has one mismatch to




On the contrary





Fig 1


ARB


gt



473



6

470ndash479


m


m


m


Betaproteo-bacteria



et al

(2003)


Betaproteobacteria


Caulobacter






Autoradiography


Discussion


































FISH






Micromanipulation




Cloning




Sequencing







Autoradiography







Acknowledgements


References






















































473



6

470ndash479


m


m


m


Betaproteo-bacteria



et al

(2003)


Betaproteobacteria


Caulobacter






Autoradiography


Discussion


































FISH






Micromanipulation




Cloning




Sequencing







Autoradiography







Acknowledgements


References
















































































FISH






Micromanipulation




Cloning




Sequencing







Autoradiography







Acknowledgements


References
























































Micromanipulation




Cloning




Sequencing







Autoradiography







Acknowledgements


References


























































Acknowledgements


References





















































Micromanipulation and further identification of FISH-labelled microcolonies of a dominant...

Documents

Transcript of Micromanipulation and further identification of FISH-labelled microcolonies of a dominant...