Articles & News Stories

Activity and Community Composition of Denitrifying Bacteria in Poly(3-hydroxybutyrate-co-3-hydroxyvalerate)-Using Solid-phase Denitrification Processes

Description
Two laboratory-scale solid-phase denitrification (SPD) reactors, designated reactors A and B, for nitrogen removal were constructed by acclimating activated sludge with pellets and flakes of poly(3-hydoxybutyrate-co-3-hydroxyvalerate) (PHBV) as the
Published
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  http://wwwsoc.nii.ac.jp/jsme2/  Microbes Environ.Vol. 22, No. 1, 20–31, 2007 Activity and Community Composition of Denitrifying Bacteria in Poly(3-hydroxybutyrate- co -3-hydroxyvalerate)-Using Solid-phase Denitrification Processes S HAMS  T ABREZ  K  HAN 1 , Y OKO  H ORIBA 1 , N AOTO  T AKAHASHI 1  and A KIRA  H IRAISHI 1 * 1  Department of Ecological Engineering, Toyohashi University of Technology, Toyohashi 441–8580, Japan (Received September 4, 2006—Accepted November 13, 2006)Two laboratory-scale solid-phase denitrification (SPD) reactors, designated reactors A and B, for nitrogenremoval were constructed by acclimating activated sludge with pellets and flakes of poly(3-hydoxybutyrate- co -3-hydroxyvalerate) (PHBV) as the sole added substrate under denitrifying conditions, respectively. The averagedenitrification rate in both reactors was 60 mg NO 3 − -N g − 1  (dry wt) h − 1  under steady-state conditions, whereaswashed sludge taken from the reactors showed an average denitrification rate of 20 mg NO 3 -N g − 1  (dry wt) h − 1 with fresh PHBV as the sole substrate. The difference in the denitrification rate between the two might be due tothe bioavailability of intermediate metabolites as the substrate for denitrification, because acetate and 3-hydroxy-butyrate were detected in the reactors. Most of the predominant denitrifiers isolated quantitatively by the plate-counting method using non-selective agar medium were unable to degrade PHBV and were identified as mem-bers of genera of the class  Betaproteobacteria  by studying 16S rRNA gene sequence information. nirS   and nosZ  gene clone library-based analyses of the microbial community from SPD reactor A showed that most of the nirS  and nosZ   clones proved to be derived from members of the family Comamonadaceae  and other phylogeneticgroups of the  Betaproteobacteria . These results suggest that the efficiency of denitrification in the PHBV-SPD process is affected by the availability of intermediate metabolites as possible reducing-power sources as well asof the solid substrate, and that particular species of the  Betaproteobacteria  play the primary role in denitrificationin this process. Key words:  Solid-phase denitrification, poly(3-hydroxybutyrate), nirS   gene, nosZ   gene, microbial community Biological denitrification is a series of distinct bioener-getic reactions in which nitrate is reduced to dinitrogen gasthrough nitrite, nitric oxide and nitrous oxide 46) . This bio-chemical process is important not only as a key step in thenitrogen cycle in nature but also as a means of nutrientremoval in engineered wastewater treatment processes. Bio-logical denitrification processes using biodegradable solid polymers, termed the solid-phase denitrification (SPD) pro-cess, have received attention as promising methods for removing nitrogen from water and wastewater  17) . The SPD process has some advantages over conventional nitrogenremoval systems using soluble substrates, e.g., a constantsupply of reducing power and ease in handling and opera-tion. Promising solid substrates for denitrification are poly-hydroxyalkanoates (PHAs) and some other biodegradablealiphatic polyesters. To date, SPD processes using poly(3-hydroxybutyrate) (PHB) 31)  and its copolymer, poly(3-hydroxybutyrate- co -hydroxyvalarate) (PHBV) 6,26,30) , havebeen most intensively studied. The application of poly( ε -caprolactone) (PCL) to SPD processes has also beenreported 4–6,22) .Functional and structural analyses of the microbial com-munities involved in SPD processes are necessary to pro-vide a basis for their practical application to nitrogenremoval. In this connection, several strains of PHB- andPHBV-degrading denitrifying bacteria have been isolatedand characterized from activated sludge and PHA-utilizing *Corresponding author. E-mail address: hiraishi@eco.tut.ac.jp; Tel.: + 81–532–44–6913; Fax: + 81–532–44–4929.   Denitrifying Microbial Communities with PHBV  21 SPD processes 3,24–26,30,38) . A representative of PHB-degrad-ing denitrifiers is  Diaphorobacter nitroreducens 25) . Our pre-vious study on a 16S rRNA gene clone library constructedfrom a PHBV-utilizing SPD reactor has shown that mem-bers of the family Comamonadaceae , a major phylogeneticgroup of the  Betaproteobacteria , predominated in this process 26) . However, the question of what kinds of microor-ganisms are actually involved in denitrification in PHBV-SPD processes has remained unanswered.Although 16S rRNA genes are powerful molecular mark-ers for studying the entire prokaryotic community structurein an environment, they can provide no direct informationabout the physiologic and functional nature of the commu-nity. In the present study, therefore, nirS   and nosZ   genes,encoding cytochrome cd  1  nitrite reductase and nitrous oxidereductase, respectively 45) , were used as molecular markersto assess the community composition of denitrifying bacte-ria in the PHBV-SPD process. These denitrifying-enzymegenes as well as the Cu-containing nitrite reductase gene, nirK  , have been used for the molecular characterization of denitrifying microbial communities in different environ-ments including wastewater treatment systems 2,13,32,44,45) . Inaddition, the predominant culturable denitrifying bacteriafrom PHBV-SPD reactors were isolated, phylogeneticallyidentified, and studied in terms of their substrate specificity.Relationships between the activity and community structureof denitrifiers in the PHBV-SPD process are discussed. Materials and Methods  Biodegradable plastics PHBV    pellets   and    powder    containing   5%    poly(3-hydroxy-valerate) (PHV) were obtained from Japan Monsanto Co.,Tokyo, Japan. PHBV sheets (5% PHV) were kindly pro-vided by the Mikawa Textile Research Institute, Aichi Pre-fectural Government, Gamagori, Japan. The PHBV sheetswere cut into small flakes (5 mm×5 mm). All PHBV pelletsand flakes were washed with ethanol and then pure water  prior to use. Construction of PHBV-acclimated reactors Activated sludge samples were collected from the mainaerobic treatment tank of a sewage treatment plant in Toyo-hashi, Japan, and used as the seed for constructing denitrifi-cation reactors. Two glass flasks (2 L capacity) containing1,000 ml each of culture medium was inoculated with thesludge and semi-anaerobically incubated at 25°C for morethan 2 months in the presence of nitrate added. One of thereactors was acclimated with PHBV pellets (reactor A) andthe other, with PHBV flakes (reactor B). At the beginning,mineral base RM2 (pH 7.0) 18) , 20 g of PHBV and 20 mMKNO 3  were added to the reactors. During the first 4 weeks,half of the supernatant in the reactor was exchanged withfresh mineral medium containing 40 mM KNO 3  every 3–4days of operation. During the subsequent period, the reac-tors were loaded daily with the same medium by exchang-ing half of the supernatant and also supplemented with 5 geach of PHBV every week. The concentration of microbialsludge was adjusted to ca. 700 mg dry wt L − 1  every week.During the acclimation, the reactors were gently stirred at70 rpm with a magnetic stirrer; under these conditions, thedissolved oxygen (DO) tension in the core of the reactorswas less than 0.5 mg L − 1 , indicating that the reactors werecontinuously operated under semi-anaerobic conditions. Measurement of denitrifying activity The rate of denitrification in the reactors was measuredby monitoring the change in the concentration of nitrate andnitrite in each batch cycle, for which ion chromatographywas used as described previously 24,26) . Also, the denitrifica-tion rate was measured separately using washed sludge andfresh PHBV powders as the sole substrate. For this, sludgesamples from the reactors at the end of each batch cyclewere collected by centrifugation, washed twice with 50 mM phosphate buffer (pH 7.0), and concentrated in a small vol-ume of this buffer. An aliquot of the concentrated sludgesuspension was introduced into rubber-plugged test tubes(30-ml capacity) containing 20 mM KNO 3 , 0.5 g of PHBV powder, and 25 ml of RM2 mineral medium (pH 7.0) fromwhich (NH 4 ) 2 SO 4  was eliminated. The tubes were spargedwith argon to establish anoxic conditions and incubated on areciprocal shaker at 25°C for 24–48 h. Then, the concentra-tions of nitrate, nitrite, nitrous oxide, and N 2  gas were moni-tored by ion chromatography and gas chromatography asdescribed previously 24,26) . Preliminary experiments showedthat the amounts of nitrite and nitrous oxide produced asintermediates both in the reactors and in the test tubes werenegligible in almost all cases. Therefore, the nitrate removalrate was considered as the denitrification rate in this study. Measurement of organic acids Samples were taken from the reactors at appropriateintervals and centrifuged at 12,600×  g   for 10 min. Theresultant supernatant samples were filtered through a mem-brane filter (pore size, 0.2 µ m) and stored at − 20°C untilanalyzed. Lower fatty acids and 3-hydroxybutyrate (3HB)were measured by ion-pair HPLC as described previously 33) .  K  HAN   et al. 22 Stoichiometric analyses The overall mixture of microbial sludge and PHBV pel-lets remaining in the reactors were collected at the end of operations. The sludge and PHBV were separated manuallyfrom each other using tweezers, washed twice with purewater, and subjected to a dry weight analysis as described previously 24)  to estimate the growth yield coefficient Y  x/s .The substrate consumed/oxidant consumed (S/O) ratio wascalculated based on the amounts of PHBV and nitrate con-sumed. The theoretical relationship between Y  x/s  and the S/Oratio was based on information from Hiraishi and Khan 17) (see also chemical reaction formula [ 1 ] and [ 2 ] given below).  Direct cell counting  Ten milliliters of sludge suspension from the reactors atthe end of batch cycles was sonicated on ice for 90 sec with2-sec intermittent bursts (20 kHz; output power, 50 W) anddiluted with filter-sterilized phosphate-buffered saline (PBS).Aliquots (10–50 µ l) of these diluted samples were taken andused for direct cell counting. Total bacterial counts weremeasured by epifluorescence microscopy with ethidiumbromide (EtBr) or SYBR Green II (Molecular Probes, Inc.,Eugene, USA) staining as described previously 15,43) . Detec-tion of viable cells using a LIVE/DEAD BacLight Bacte-rial Viability kit (Molecular Probes) was performed accord-ing to the manufacturer’s instructions and as described previously 43) . Stained specimens were observed under anOlympus BX-50 epifluorescence microscope equipped witha Flovel FD-120 M digital CCD camera (Flovel Co., Tokyo,Japan). The number of stained cells was counted using theimage analysis program WINROOF (Flovel): 10–15 fields per sample and a total of 1,000–2,000 cells per sample.  Enumeration, isolation and characterization of denitrifiers Sludge and mixed liquor samples taken from the reactor were diluted with PBS as described above, and appropriatedilutions were used for the enumeration of viable bacteria.Plate counts of aerobic heterotrophic bacteria were obtainedby the smear-plating method using 1/10-diluted PBY 12)  agar medium or 1/10-diluted tryptic soy agar medium. Denitrify-ing bacteria were enumerated by the pour-plating methodusing agar media containing mineral base RM2, 20 mMKNO 3 , vitamin solution PV1 21)  (1 ml L − 1 ) and 1.8% agar asthe basal medium. The agar media were supplemented witheither 0.2% tryptic soy broth, 10 mM acetate, 10 mM 3HBor 2% PHBV powder (w/v) as the sole carbon and energysource: designated TSN, ACN, 3HBN and PHBN agar media, respectively. For plating, 1 ml of each of the dilutedsamples noted above was plated and incubated anaerobi-cally at 25°C using the AnaeroPak system (Mitsubishi GasChemicals, Niigata, Japan). Inoculated plates were incu-bated for 2 to 4 weeks before the final reading of CFUs.Few if any detectable colonies appeared on the 4 agar mediawithout nitrate added, thereby confirming that coloniesrecovered on the test media with nitrate were denitrifiers.For counting CFUs of PHBV-degrading denitrifiers, colo-nies exhibiting a cleared halo formation were taken as being positive. Colonies recovered on the TSN plates were ran-domly selected and purified by repeatedly streaking on thesame agar medium under anaerobic conditions. Since allisolates thus obtained were aerobic chemoorganotrophicdenitrifying bacteria, they were maintained aerobically onPBY agar slants. 16S rRNA gene amplification and sequencing  Crude cell lysates as the DNA source were prepared for PCR use as described previously 19) . 16S rRNA gene frag-ments that corresponded to positions 8 to 1543 in  Escheri-chia coli  16S rRNA 8)  were PCR-amplified from the celllysate with a set of bacterial universal primers, fD1 andrP1 42) , as described previously 19) . PCR products were puri-fied by the polyethylene glycol precipitation method 16) ,sequenced with a SequiTherm Long Read cycle sequencingkit (Epicentre Technologies, Madison, USA), and analyzedwith an Amersham-Pharmacia ALF express  DNA sequencer. Quinone profiling  Quinones from sludge samples and the isolates wereextracted with a chloroform-methanol mixture and partially purified by column chromatography. Quinone componentswere separated and identified by reverse-phase HPLC and photodiode array and mass spectrometric detection withexternal ubiquinone (Q- n ) and menaquinone (MK- n ) stan-dards. Detailed information on the analytical procedures hasbeen given previously 14,15) .  DNA extraction and purification Bulk DNA was extracted from sludge samples as des-cribed previously 26) . The crude DNA extracted was further  purified by a standard method including deproteinizationwith chloroform-isoamylalcohol and RNase treatment 29) .The DNA solution thus obtained was diluted in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) as needed andused for PCR experiments.  Amplification and sequencing of denitrifying enzyme genes Fragments of nitrite reductase genes, nirK   and nirS  , from   Denitrifying Microbial Communities with PHBV  23 the bulk DNA purified from sludge samples were PCR-amplified with the oligonucleotide primer pairs nirK1F(GG[A/C]ATGGT[G/T]CC[C/G]TGGCA)-nirK5R (GCC-TCGATCAG[A/G]TT[A/G]TGG) 7)  and nirS2F’ (5'-GAAT-A[C/T]CACCC[C/G]GAGCCG-3')-nirS6R’ (5'-A[C/G][A/G]CGTTGAACTT[A/G]CCGGT-3'), respectively. The pri-mer pair nirS2F’-nirS6R’ was a modification of the nirS2F-nisS6R pair described by Braker et al. 7) . The expected sizesof PCR-amplified nirK   and nirS   fragments were 0.51 and0.80 kb, respectively. For PCR amplification of nosZ   frag-ments, we first used previously described primers 36) , but thisattempt gave unsatisfactory results. In this study, therefore,we designed a new pair of PCR primers, nosZ2f (5'-GT-GCCGAAGAACCC[G/C]CACGG-3' and nosZ3r (5'-T[C/G]GCCGGAGATGTCGATCA-3'), based on nosZ   genesequences retrieved from databases (see Fig. 4). The se-quences of nosZ2f and nosZ3r correspond to positions 1006to 1025 and 1466 to 1484 in the nosZ   gene of  Ralstonia eu-tropha  strain H16 (database accession number AY305378),and the expected size of PCR-amplified nosZ   fragments was0.48 kb. All thermocycling reactions were performed by thetouchdown PCR method 10)  using an r  Taq  DNA polymerasekit (Takara, Otsu, Japan), one of the primer sets and a TakaraThermal Cycler. The first half of the PCR procedure includ-ed an initial preheating step of 2 min at 94°C and 20 cyclesof a touch-down procedure consisting of denaturation for 1min at 94°C, annealing for 1 min at a temperature decreas-ing from 60 to 51°C with 1°C decremental steps of 2 cycleseach and extension for 1 min at 72°C. Following this, addi-tional 20 cycles of the thermocycling reaction with anneal-ing at 50°C was performed. The final step was followed byextension at 72°C for 5 min. PCR products were separatedby agarose gel electrophoresis, cut from the gel, and then purified using a GENECLEAN Spin kit (Bio 101, Vista,USA). Purified PCR fragments were subcloned using a pT-Blue Perfectly Blunt cloning kit (Novagen, Madison, USA).Transformation of  Escherichia coli  competent cells wascarried out according to a standard manual of molecular cloning 35) . Plasmid DNA was isolated and purified by usingthe Wizard Plus Minipreps DNA Purification System(Promega Inc., Madison, USA) according to the manufac-turer’s instructions. Nucleotide sequences of nirS   and nosZ  gene clones were determined as described for 16S rRNAgene sequences. For comparison, nosZ   genes from the pre-viously described denitrifying bacteria  Brachymonas deni-trificans  strain AS152 20)  and  Diaphorobacter nitroreducens strain NA10B T   24,25)  were PCR-amplified, subcloned andsequenced as described above. In addition,  Aquitalea  sp.strain PGP-1 and Comamonas  sp. strain TSL-h, both of which were isolated from SPD reactor A in this study, weresubjected to the nosZ   gene analysis. This study is the first todetermine the nosZ   gene sequences of these 4 denitrifyingspecies. Sequence comparisons and phylogenetic analyses All sequence data were compiled with the GENETYX-MAC program (Software Development Co., Osaka, Japan).16S rRNA gene sequence data were compared with thoseretrieved from Ribosomal Database Project II 9) . Nucleotidesequences and translated amino acid sequences of nirS   and nosZ   clones were analyzed using the BLAST homologysearch system 1,37) . The multiple alignment of sequences andcalculation of the nucleotide substitution rate with Kimura’stwo parameter model 27)  were performed using the CLUSTALW program 40) . Distance matrix trees were constructed by theneighbor-joining method 34) , and the topology of the treeswas evaluated by bootstrapping with 1,000 resamplings 11) .  Nucleotide sequence accession numbers The nucleotide sequences determined in this study havebeen deposited with the DDBJ under the accession numbersAB277845 to AB277852 for 16S rRNA genes, AB278071to AB278090 for nirS   genes, and AB278091 to AB278115for nosZ   genes. Results  Denitrifying activity of PHBV-using reactors PHBV-containing reactors A and B were operated at 3- to4-day batch intervals during 30 days of start-up and with a24-h batch cycle thereafter. Both the reactors exhibited sim-ilar denitrifying activity during the overall period of opera-tion. As an example, changes in denitrifying activity inreactor A are shown in Fig. 1. Nitrate remained in detect-able amounts at the end of each batch cycle during the first2 weeks of operation but was not detected thereafter (Fig.1a). Also, little nitrite was detected after 2 weeks of opera-tion. During the first 40 days of operation, the denitrifi-cation rate exhibited by the reactor in each batch cycleincreased markedly. After that, the denitrification rateexhibited by the reactor became constant at around 60 mg NO 3 − -N g − 1  (dry wt) h − 1  (Fig. 1b). To estimate the actualdenitrification rate with PHBV more accurately, sludgesamples were taken from the reactors at the end of eachbatch cycle, washed and separately tested for denitrificationwith fresh PHBV as the sole substrate. The denitrificationrate with the fresh PHBV also increased linearly during thefirst 30 days of operation and reached around 20 mg NO 3 − -  K  HAN   et al. 24  N g − 1  (dry wt) h − 1  under steady-state conditions (Fig. 1b).HPLC experiments revealed that significant amounts of acetate and 3HB were constantly produced in both the reac-tors under steady-state conditions. The concentrations of acetate and 3HB detected in the supernatant at the end of operation were 2.9–4.8 and 1.9–2.4 mmol L − 1 , respectively.Thus, the marked difference in the denitrification ratebetween the reactor itself and the washed sludge with freshPHBV was possibly due to the bioavailability of intermedi-ate metabolites, e.g., organic acids, as a usable substrate for denitrification. In fact, the washed sludge taken from thereactors at the end of operation exhibited 2.0–2.5 foldhigher denitrification rates with acetate than with PHBV(data not shown).Stoichiometric data on PHBV, nitrate, and microbial bio-mass in reactors A and B during the overall period of opera-tion are shown in Table 1. Based on the amounts of PHBVconsumed and excess biomass produced, the growth yieldcoefficient for the reactors was estimated to be 0.43–44.These growth yields are slightly lower than those recorded previously for a PHB-using denitrifying reactor  31)  and a pure culture of  Diaphorobacter nitroreducens  with PHB asthe substrate 24) . The denitrification reactions with PHB asthe substrate and with or without NH 4 +  as the nitrogensource are given by the following formula 17) :5[C 4 H 6 O 2 ] + 18NO 3 − → 9N 2 + 18HCO 3 − + 2CO 2 + 6H 2 O( 1 )10[C 4 H 6 O 2 ] + 14NO 3 − + 6NH 4 + → 7N 2 + 10CO 2 + 6[C 5 H 7 O 2  N] + 12H 2 O + 18OH − ( 2 )Based on the biochemical reactions ( 1 ) and ( 2 ), it can be predicted that a Y  x/s  value of 0.43–44 corresponds to a S/Oratio of 0.58–0.59 17) . However, the experimental values of the S/O ratio obtained (0.624–0.640) were considerablyhigher than the predicted values. This suggests that PHBVas the substrate was consumed not only for denitrificationbut also for other biochemical processes, e.g., aerobic respi-ration. Quinone profiling of microbial communities Sludge samples were collected from reactors A and Bunder steady-state conditions (on days 48, 56 and 66) for quinone profiling. In both the reactors, ubiquinone Q-8accounted for 83–85% of the total quinone content. Theremaining fractions were composed mainly of Q-9 (3–5%)and Q-10 (7–9%). Menaquinones constituted only minor  proportions of the total content. In light of the availableinformation on microbial quinone systems 14) , it was evidentthat Q-8-containing proteobacterial species, especially thoseof the  Betaproteobacteria , constituted the overwhelmingmajority of the microbial communities in the reactors. The Fig.1.Changes in nitrate removal activity of SPD reactor A duringthe acclimation with PHBV under denitrifying conditions. Figure1a shows changes in the concentrations of NO 3 − -N and NO 2 − -N inthe reactor at each batch cycle: diamonds, concentration of NO 3 − - N at the start of each batch cycle; open squares, concentration of  NO 3 − -N at a 24 h-interval of sampling; open triangles, concentra-tion of NO 2 − -N at a 24 h-interval of sampling. Figure 1b showschanges in the nitrate removal rate as estimated the reactor (dia-monds) and the washed sludge with fresh PHBV (squares).Table1.Stoicheometric estimation of microbial biomass, PHBVand nitrate in the reactorParameterMeasured value forReactor AReactor BTotal PHBV added (g)45.045.0PHBV consumed (g)20.120.6Total NO 3 − -N added and removed (g)7.287.28Excess biomass produced (g dry wt)8.79.1 Y  x/s  (g g − 1 )0.430.44S/O ratio (g g − 1 )0.6240.640
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks