60632 (099)

Biosci. Biotechnol. Biochem., 71, 60632-1–9, 2007

Reducing Sludge Production and the Domination of by Reducing the Oxygen Supply in the Wastewater Treatment Procedure of a Food-Processing Factory

Tamiko SADAIE,1 Aya SADAIE,1 Masao TAKADA,2 Keiichi HAMANO,3 Junichi OHNISHI,4 y Niji OHTA,4 Kouji MATSUMOTO,4 and Yoshito SADAIE4;

1Clarus Environment Co., Ltd., Nishi Sinjuku 6-12-7, Shinjuku, Tokyo 160-0023, Japan 2Takada Co., Ltd., Miyoshi-cho 901-9, Iruma, Saitama 354-0041, Japan 3Taitec Co., Ltd., Koshigaya 2693-1, Saitama 343-0822, Japan 4Department of Molecular Biology, Faculty of Science, Saitama University, Saitama 338-8570, Japan

Received November 8, 2006; Accepted December 19, 2006; Online Publication, March 7, 2007 [doi:10.1271/bbb.60632]

Sludge production was reduced remarkably by re- limited oxygen supply requires sufficient degradation of ducingAdvance the dissolved oxygen supply to less than View 1 mg/‘ nutrients to maintain cellular activity.1,2) in the conventional wastewater treatment procedure In this communication, we report an improvement in of a food-processing factory that produced 180 m3 of the conventional large-scale activated sludge process wastewater of biochemical oxygen demand (BOD) of by lowering of the oxygen supply to less than 1 mg/‘, about 1,000 mg/‘ daily. DNA was extracted from the which resulted in a reduction in sludge production. Since sludge and subjected to PCR amplification. The PCR 80% of the total microorganisms detected by DAPI product was cloned into a plasmid and sequenced. (40-6-Diamidino-2-phenylindole) staining in activated Estimation of the resident bacterial distribution by sludge were hybridizable with bacterial DNA probes3) 16S rDNA sequences before and after improvement of and most are not culturable,4,5) we performed the system suggested a remarkable gradual change in a phylogenetic orProofs community structure analysis6) by the major bacterial population from Anaerolinaeceae comparing sequences of cloned 16S ribosomal RNA (15.6%) to Comamonadaceae (52.3%), members of genes (rDNA) of a bacterial population in the improved denitrifying bacteria of . Although we sludge. With a low oxygen supply, the wastewater did not directly confirm the ability of denitrification of treatment system showed limited production of sludge. the resulting sludge, a change in the major final electron Community structure analysis suggested that nitrate- acceptors from oxygen to nitrate might explain the reducing Comamonadaceae became the dominant bac- reduction in sludge production in a conventional terial species. We suggest that sludge containing a large activated sludge process when the oxygen supply was fraction of the members of Comamonadaceae is capable limitted. of effecting a biological nutrient removal process with a reduced increase in cell mass when the oxygen supply is Key words: activated sludge; DO reduction; reduced limited. sludge production; community structure analysis Materials and Methods

The conventional wastewater treatment process re- Activated sludge process. The improved wastewater quires a higher dissolved oxygen concentration of more treatment process of a food-processing factory consisted than 1 mg/‘ in the reactor to activate sludge. This of a first adjusting tank (capacity 200 t), a second reactor process inevitably produces excess sludge. A reduction tank (capacity 400 t), a third setting tank (capacity 70 t), in sludge production should lead to a reduction in the and a fourth sludge-digesting tank (capacity 200 t) for amount of energy required for sludge treatment. Re- the treatment of sludge with lower dissolved oxygen for ducing the oxygen supply to less than 1 mg/‘ in the one week. Sedimented sludge in the first tank was also wastewater treatment system often results in a reduction digested in a fifth tank (capacity 50 t) and returned to the in sludge production, possibly because energy produc- fourth tank. A fraction of the supernatant and suspended tion by bacterial respiration and metabolism with a solids of the fourth sludge-digesting tank was returned

y To whom correspondence should be addressed. Tel: +81-48-858-3399; Fax: +81-48-858-3384; E-mail: [email protected] 60632-2 T. SADAIE et al. to the first adjusting tank every week. After November MEGA3 version 3.1,11) run on a PC. Unambiguously 25, 2004, the operation of the second reactor tank was aligned parts are further analyzed to make a neigh- performed with a lower oxygen supply (DO < 1). bor-joining tree, which was visualized with the Tree- Before improvement, the second reactor tank was kept View program (http://taxonomy.zoology.gla.ac.uk/rod/ running with higher oxygen supply (DO > 1), and the treeview.html).12) The 34 reference sequences shown fourth tank was used as a reactor. Before improvement, in Fig. 2 are C. gracilis (AB109889), P. lanceolata the sedimented sludge of the first and third tanks was (AB021390), R. antarcticus (AY609198), R. tataoui- condensed in the fifth tank, followed by a drying nensis (AF144383), V. paradoxus (DQ241396), H. pal- process. leronii (AF019073), X. azovorans (AF285414), A. kon- jaci (AF137507), A. metamorphum (AY780904), Colony formation. Colonies of the sludge sample were A. psychrophilum (AF078755), D. acidovorans (AB23- formed after serial dilutions with distilled water on 1159), Diaphorobacter NA5 (DQ294626), C. terrigena Luria-Bertani broth7) agar after incubation at room (AB021418), C. denitrificans (AF233880), B. denitrifi- temperature (20 C) for 3 d. cans (D14320), A. hongkongensis (DQ489306), L. dis- cophora (Z18533), I. dechloratans (X72724), R. benzo- DNA extraction. DNA was extracted8) from the lyticum (AJ888903), B. brasilensis (AJ238360), P. api- sludge. It was obtained by precipitating the sample sta (AY268172), L. thiooxidans (AJ289885), S. albido- from the reactor at 3,000 RPM for 15 min. The sludge flavus (AY965999), O. formigenes (U49758), S. ster- sample was resuspended in TE buffer containing phenol coricanis (AJ566849), A. defragrans (AB195161), and glass beads (BZ-04, diameter 0.35–0.5 mm, AS N. halophila (AF272413), F. limneticum (Y17060), ONE, Osaka, Japan), and vortexed vigorously. DNA was T. chlorobenzoica (AF229887), A. anaerobius (Y147- extracted twice with phenol-chroloform-isoamylalchol 01), M. pratensis (AY298905), H. denitrificans (AY82- afterAdvance RNase1 treatment. View0184), N. flava (AJ239301), and E. coli (J01695).

PCR and TA cloning. 16S rDNA was PCR amplified Nucleotide sequence accession numbers. The 16S with template sludge DNA, AmpliTaq DNA polymerase rDNA sequences of the 318 uncultured clones deter- (Applied Biosystems, Branchburg, NJ) and a primer mined in this study have been deposited under DDBJ pair, 50AGAGTTTGATCCTGGCTCAG30 (E. coli 16S accession nos AB286331 to AB286648. rDNA positions 8–27)/50AAGGAGGTGATCCAGCC- GCA30 (E. coli 16S rDNA positions 1541–1530). Prim- Results and Discussion ers were custom made (Espec Oligo Service, Tsukuba, Japan). Amplified DNA was cloned with plasmid pCRII- Improvement ofProofs activated sludge process TOPO (Invitrogen, Carlsbad, CA) and introduced into A conventional activated sludge and wastewater Escherichia coli strain DH5-T1 by the procedures of treatment process was performed in a food-manufactur- the supplier. Apr transformant colonies were randomly ing factory which daily produced 180 m3 wastewater of selected and regrown on LB agar medium (1cm 1cm) biochemical oxygen demand (BOD) of about 1,000 containing antibiotics. Cells of the regrown colony were mg/‘. Dissolved oxygen (DO) of more than 1 mg/‘ was collected with toothpicks, and recombinant plasmids kept in the reactor tank before improvement. We were purified with a purification kit (ABI PRISMÔ reduced DO to less than 1 mg/‘ in the reactor tank Miniprep Kit, Applied Biosystems, Foster City, CA). and aerated the sedimented sludge in the fourth tank Cloned 16S rDNA was PCR amplified with plasmid with a lower oxygen supply, in which DO was also kept DNA and a primer pair, as described above. to less than 1 mg/‘ for one week. A fraction of super- natant and suspended solids of the fourth tank was DNA sequencing. DNA sequencing was performed returned to the first tank of the system to stabilize the with an ABI PRISM 3100 Genetic Analyzer after system. Although the DO value fluctuated, it was kept to sequencing reaction of the cloned segment by the Dye less than 1.0 (from 0.8 to 0.2) after improvement. terminator method as described by the supplier. Se- Before improvement, 10 tons of wet sedimented quencing was performed from both directions using the sludge was extracted every day, although careful same primer pair. Some of the sequencing reactions management of the whole system may have reduced were also performed by Hitachi HiTec Science Systems the production of sludge. No sludge was extracted for (Hitachi-Naka, Ibaraki, Japan). more than 8 months after improvement. We measured pH, suspended solids (SSs), mixed Bioinformatics. DNA sequences were subjected to the liquor suspended solids (MLSSs), chemical oxygen de- Classifier and Sequence Match programs of Ribosomal mand (COD), BOD, N-hexane-extractable fraction (N- Data Base Project II (http://rdp.come.msu.edu/).9) Co- hexane), nitrate, and sulfate in the system before im- mamonadaceae and sequences were provement and 2 months after improvement (Table 1). aligned using ClustalX program version 1.83 (http:// Discharged water from the setting tank showed exten- bips.u-strasbg.fr/fr/Documentation/ClustalX/)10) and sively lowered values of COD and BOD as compared to Activated Sludge and Community Structure Analysis 60632-3 Table 1. Parameters of Activated Sludge Process with Lower Oxygen Content

Adjusting Reactor Setting Sludge digesting tank tank tank tank pH 5.4(5.6) 7.2(7.6) 6.2(6.1) SS 680(217) 13(30.6) MLSS 11,000(10,100) 19,000(12,600) COD 680(303) 6,900(4,850) 18(35.1) 8,400(5,430) BOD 460(862) 2,000(1,260) 11(7.6) 1,300(1,670) N-hexan 260(98.6) <5(<5) Nitrate <0:2(0.8) <0:1(<0:2) <0:2(82.6) Sulfate <1(15.8) 25(24) 61(75.6)

Data are from activated sludge system on January 20, 2005. The numbers in parentheses are those of the system with higher oxygen content (November 25, 2004). The numbers, except for pH, indicate concentrations in mg/l. those of incoming wastewater (adjusting tank). The uncultured clones were subjected to the Classifier and suspended solid (SS) value of discharged water was Sequence Match programs of the Ribosome Data Base about 10.6 mg/‘ after improvement, indicating only a (http://rdp.cme.msu.edu/).9) The results are shown in slight release of sedimented sludge. It was notable that Tables 2, 3, 4, and 5. We analyzed a total of 318 clones the quantity of nitrate fell remarkably in the fourth originating from the samples from the reactor tank before sludge-digesting tank after the reduction in oxygen improvement (#01), and one (#03), two (#07), and three supply. It was thought that sludge consumed nitrate for (#15) months after improvement, as well as from the respiration as the oxygen supply was limited. The sulfate samples from the raw incoming wastewater (#13), the contentAdvance fell also. Oxidation-reduction potential View (ORP) in first adjusting tank (#14), and the fourth sludge-digesting the second reactor tank was kept at about þ200 mV. tank (#16) 3 months after improvement. Sequencing revealed that they were all bacterial 16S rRNA genes. Colony forming ability of bacteria in sludge The average number of nucleotides of the sequences Samples of the improved system were taken from obtained was 1,437, including 2.6 uncertain (N) nucleo- each tank 2 months after improvement. Samples from tides. We found only two putative chimeras, which are raw wastewater (sample 1) and the adjusting tank denoted in the footnotes to Tables 2 and 4. The pre- (sample 2) remained turbid even after 2 h. On the other dominant members in the reactor tank were those of hand, samples from the reactor tank (sample 3) and the Proteobacteria (about 45 to 80%), especially the mem- fourth sludge-digesting tank (sample 4) showed thick bers of BetaproteobacteriaProofs(about 30 to 60%). The sedimenting sludge and clear supernatant. Sludge of members of Bacteroidetes, especially those of Sphingo- sample 4 separated, probably due to gas production by bacteriales, constituted a significant fraction (about the sludge. Colony-forming units per ml on Luria- 10%) in the reactor tank, while members of Firmicutes Bertani broth agar of samples 1, 2, 3, and 4 from raw were few in number. It is notable that the sample from the wastewater, the adjusting tank, the reaction tank, and the reactor tank with a higher oxygen supply (November 25, fourth sludge-digesting tank were 10 106, 250 106, 2004, #01) contained a significant fraction of members of 1:6 106, and 1:7 106 respectively. The number of Chloroflexi (15.6%), while the sample from the reactor colony forming units increased once in sample 2, and tank 3 months after improvement with a lower oxygen declined greatly in samples 3 and 4, while the quantity supply (February 3, 2005, #15) contained few Chloro- per ml of sedimenting sludge and extracted DNA was 10 flexi. On the other hand, the latter sample contained large times greater in samples 3 and 4 than in samples 1 and 2. numbers of members of (52.3%) of The reason for the lower number of colony-forming Proteobacteria, while the former sample contained units in the reactor tank (sample 3) in spite of an smaller numbers of Burkholderiales (17.8%) (Table 2). increased quantity of sedimentation and extracted DNA The members of Sphingobacteriales in the reactor tanks might be the formation of biofilms containing aggre- did not change in numbers upon reduction of the oxygen gated cells in the reactor tank, or the formation of viable supply. but unculturable conditions (unable to form colony) in The change in the fraction of the estimated major the sludge under the lower oxygen supply. Another bacterial population in the reactor tanks was gradual, as explanation might be simply a large number of bacteria shown in Table 2, 1 (December 22, 2004, #03) and 2 unable to form colonies under aerobic conditions. (January 20, 2005, #07) months after improvement, Chloroflexi and Burkholderiales were 15.6% and 17.8%, Community structure of activated sludge and 2.1% and 38.3%, respectively. Only members of Sludge samples were taken for bacterial 16S rDNA Anaerolinea were found in Chloroflexi. Detailed analysis analysis to estimate the community structure of activated of Burkholderiales revealed that the major family was sludge, although we did not confirm community structure unclassified Comamonadaceae (Table 3 and Fig. 1). by direct observation. Sequences of 16S rDNA of the In the improved activated sludge process (February 3, 60632-4 T. SADAIE et al. Table 2. Classification of Bacterial 16S rDNA Clones Retrieved from the Sludge of the Reactor Tank of the Wastewater Treatment Procedure of the Food-Processing Factory

Reactor Tank 01 03 07 15 Munich Bacteria 45(100.0) 45(100.0) 47(100.0) 44(100.0) 62(100.0) OP10 1( 2.2) 0( <2:2)0(<2:1)0(<2:3)0(<1:6) OP10 1( 2.2) 0( <2:2)0(<2:1)0(<2:3)0(<1:6) Acidobacteria 1( 2.2) 0( <2:2)0(<2:1)0(<2:3)0(<1:6) Acidobacteria 1( 2.2) 0( <2:2)0(<2:1)0(<2:3)0(<1:6) Verrucomicrobia 1( 2.2) 1( 2.2) 0( <2:1)0(<2:3) 1( 1.6) Verrucomicrobia 1( 2.2) 1( 2.2) 0( <2:1)0(<2:3) 1( 1.6) Planctomycetes 2( 4.4) 5( 11.1) 2( 4.3) 0( <2:3)0(<1:6) Planctomycetes 2( 4.4) 5( 11.1) 2( 4.3) 0( <2:3)0(<1:6) Chloroflexi 7( 15.6) 7( 15.6) 1( 2.1) 1( 2.3) 1( 1.6) Anaerolineae 7( 15.6) 7( 15.6) 1( 2.1) 1( 2.3) 1( 1.6) Bacteroidetes 6( 13.3) 3( 6.7) 8( 17.0) 4( 9.1) 0( <1:6) Bacteroidetes 0( <2:2)0(<2:2) 1( 2.1) 0( <2:3)0(<1:6) Bacteroidales 0( <2:2)0(<2:2) 1( 2.1) 0( <2:3)0(<1:6) Sphingobacteria 6( 13.3) 3( 6.7) 7( 14.9) 4( 9.1) 0( <1:6) Sphingobacteriales 6( 13.3) 3( 6.7) 7( 14.9) 4( 9.1) 0( <1:6) Firmicutes 1( 2.2) 0( <2:2) 1( 2.1) 0( <2:3) 6( 9.7) Bacilli 1( 2.2) 0( <2:2) 1( 2.1) 0( <2:3) 3( 4.8) Bacillales 0( <2:2)0(<2:2) 1( 2.1) 0( <2:3)0(<1:6) Lactobacillales 1( 2.2) 0( <2:2)0(<2:1)0(<2:3) 3( 4.8) Mollicutes 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 1( 1.6) AdvanceIncertae sedis 8 0( View<2:2)0(<2:2)0(<2:1)0(<2:3) 1( 1.6) Clostridia 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 2( 3.2) Clostridiales 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 2( 3.2) Proteobacteria 20( 44.4) 21( 46.7) 33( 70.2) 34( 77.3) 54( 87.1) Alphaproteobacteria 4( 8.9) 3( 6.7) 9( 19.1) 3( 6.8) 2( 3.2) Rhizobiales 1( 2.2) 0( <2:2)1( 2.1) 0( <2:3)0(<1:6) Caulobacterales 0( <2:2)0(<2:2) 1( 2.1) 0( <2:3)0(<1:6) Rhodospirillales 0( <2:2) 1( 2.2) 2( 4.3) 1( 2.3) 0( <1:6) Sphingomonadales 2( 4.4) 1( 2.2) 4( 8.5) 0( <2:3) 2( 3.2) Rhodobacterales 0( <2:2)0(<2:2) 1( 2.1) 1( 2.3) 0( <1:6) Rickettsiales 0( <2:2)0(<2:2)0(<2:1) 1( 2.3) 0( <1:6) Unclassified Alphaproteobacteria 1( 2.2) 1( 2.2) 0( <2:1)0(<2:3)0(<1:6) Deltaproteobacteria 3( 6.7) 1( 2.2) 1( 2.1)Proofs 2( 4.5) 0( <1:6) Myxococcales 1( 2.2) 0( <2:2)0(<2:1) 2( 4.5) 0( <1:6) Bdellovibrionales 1( 2.2) 0( <2:2)0(<2:1)0(<2:3)0(<1:6) Unclassified Deltaproteobacteria 1( 2.2) 1( 2.2) 1( 2.1) 0( <2:3)0(<1:6) Betaproteobacteria 13( 28.9) 17( 37.8) 22( 46.8) 28( 63.6) 33( 53.2) Burkholderiales 8( 17.8) 8( 17.8) 18( 38.3) 23( 52.3) 28( 45.2) Rhodocyclales 2( 4.4) 8( 17.8) 3( 6.4) 3( 6.8) 1( 1.6) Neisseriaceae 2( 4.4) 0( <2:2)0(<2:1)0(<2:3)0(<1:6) Unclassified Betaproteobacteria 1( 2.2) 1( 2.2) 1( 2.1) 2( 4.5) 4( 6.5) Gammaproteobacteria 0( <2:2)0(<2:2) 1( 2.1) 1( 2.3) 10( 16.1) Psudomonadales 0( <2:2)0(<2:2) 1( 2.1) 0( <2:3) 7( 11.3) Xanthomonadales 0( <2:2)0(<2:2)0(<2:1) 1( 2.3) 0( <1:6) Aeromonadales 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 1( 1.6) Unclassified Gammaproteobacteria 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 2( 3.2) Epsilonproteobacteria 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 9( 14.5) Campylobacterales 0( <2:2)0(<2:2)0(<2:1)0(<2:3) 9( 14.5) Unclassified Bacteria 6( 13.3) 8( 17.7) 2( 4.3) 5( 11.4) 0( <1:6)

The numbers are the numbers of clones from the sludge of the reactor tank on Nov. 25, 2004 (01), Dec. 22, 2004 (03), Jan. 20, 2005 (07), and Feb. 3, 2005 (15). , Clone 0719 is perhaps a putative chimera. , Munich clones are described in Table 2 of Ref. 3. The numbers in parenthesis indicate % of total clones of each tank.

2005), the bacterial population in the raw wastewater dales were detectable in the incoming wastewater and and the adjusting tank was estimated to contain bacteria the first adjusting tank (#13 and #14), members of of different phyla (Table 4, #13 and #14). There were Sphingobacteriales appeared in the reactor tank and the members of Bacteroidetes, Clostridiales, and Gammap- fourth sludge-digesting tank (#15 and #16). Detailed roteobacteria, as well as Betaproteobacteria. The first classification of Burkholderiales of Betaproteobacteria three classes disappeared in the reactor tank and the showed that Brachymonas was dominant in the raw sludge-digesting tank and members of Betaproteobac- wastewater and the adjusting tank, while unclassified teria became dominant. While members of Bacteroi- Comamonadaceae became dominant in the third reactor Activated Sludge and Community Structure Analysis 60632-5 Table 3. Classification of Burkholderiales 16S rDNA Clones Retrieved from Sludge from the Reactor Tank of the Food-Processing Factory

Reactor Tank 01 03 07 15 Burkholderiales 8( 17.8) 8( 17.8) 18( 38.3) 23( 52.3) Comamonadaceae 4( 8.9) 6( 13.3) 17( 36.2) 23( 52.3) Brachymonas 0( <2:2)0(<2:2) 1( 2.1) 0( <2:3) Unclassified Comamonadaceae 4( 8.9) 6( 13.3) 16( 34.0) 23( 52.3) Incertae sedis 5 1( 2.2) 2( 4.4) 0( <2:1)0(<2:3) Rubrivivax 1( 2.2) 0( <2:2)0(<2:1)0(<2:3) Unclassified Incertae sedis 5 1( 2.2) 2( 4.4) 0( <2:1)0(<2:3) Unclassified Burkholderiales 2( 2.2) 0( <2:2) 1( 2.1) 0( <2:3)

The numbers are the numbers of clones from the sludge of the reactor tank on Nov. 25, 2004 (01), Dec. 22, 2004 (03), Jan. 20, 2005 (07), and Feb. 3, 2005 (15). The numbers in parenthesis indicate % of total clones of each tank.

1535 1542 1537 1502 1533 1509 1534 1521 1544 Advance View1522 1516 1512 1540 1501 1550 1528 1525 1532 1519 1513 Proofs 1541 1526 AB109889 AF137507 1523 AF019073 T60 Z93993 ATCC 11996 sp. OS-3 AB076853 OS1L-4 AB076862 sp. NSP4 AB076849 TSL-1-1 AB076876 TSL-1-5 AB076878 OS1L-19 AB076871 TSL-2-2 AB076880 OS1L-18 AB076870 OS1L-20 AB076872 TSL-2-3 AB076881 TSL-2-9 AB076882 TSL-1-7 AB076879 TSL-1-3 AB076877 D14320 X60646

Fig. 1. Phylogenic Tree of the Members of Comamonadaceae from Activated Sludge of a Food-Processing Company. The clones of unclassified Comamonadaceae listed in Table 6 except for T96 and T99 were subjected to the ClastalX program. The bar indicates 2% dissimilarity. Boot strap values greater than 50% were reported at the nodes. 60632-6 T. SADAIE et al. Table 4. Classification of Bacterial 16S rDNA Clones Retrieved from the Activated Sludge System of the Food-Processing Factory

Tank 13 14 15 16 Bacteria 47(100.0) 48(100.0) 44(100.0) 42(100.0) Actinobacteria 0( <2:1)0(<2:1)0(<2:3) 1( 2.4) Actinobacteria 0( <2:1)0(<2:1)0(<2.3) 1( 2.4) Unclassified Actinobacteria 0( <2:1)0(<2:1)0(<2:3) 1( 2.4) Chloroflexi 0( <2:1)0(<2:1)1(<2:3) 1( 2.4) Anaerolineae 0( <2:1)0(<2:1)1(<2:3) 1( 2.4) Bacteroidetes 2( 4.3) 4( 8.3) 4( 9.1) 3( 7.1) Bacteroidetes 2( 4.3) 4( 8.3) 0( <2:3)0(<2:4) Bacteroidales 2( 4.3) 4( 8.3) 0( <2:3)0(<2:4) Sphingobacteria 0( <2:1)0(<2:1) 4( 9.1) 3( 7.1) Sphingobacteriales 0( <2:1)0(<2:1) 4( 9.1) 3( 7.1) Firmicutes 6( 12.8) 10( 20.8) 0( <2:3) 1( 2.4) Bacilli 1( 2.1) 4( 8.3) 0( <2:3) 1( 2.4) Bacillales 0( <2:1)0(<2:1)0(<2:3) 1( 2.4) Lactobacillales 1( 2.1) 4( 8.3) 0( <2:3)0(<2:4) Clostridiales 5( 10.6) 6( 12.5) 0( <2:3)0(<2:4) Clostridiales 5( 10.6) 6( 12.5) 0( <2:3)0(<2:4) Proteobacteria 37( 78.7) 33( 68.8) 34( 77.3) 27( 64.3) Alphaproteobacteria 1( 2.1) 1( 2.1) 3( 6.8) 5( 11.9) Rhodospirillales 0( <2:1)0(<2:1) 1( 2.3) 2( 4.8) Sphingomonadales 0( <2:1) 1( 2.1) 0( <2:3) 1( 2.4) Rhodobacterales 1( 2.1) 0( <2:1) 1( 2.3) 0( <2:4) AdvanceRickettsiales View0( <2:1)0(<2:1) 1( 2.3) 0( <2:4) Unclassified Alphaproteobacteria 0( <2:1)0(<2:1)0(<2:3) 2( 4.8) Deltaproteobacteria 0( <2:1)0(<2:1) 2( 4.5) 3( 7.1) Myxococcales 0( <2:1)0(<2:1) 2( 4.5) 1( 2.4) Bdellovibrionales 0( <2:1)0(<2:1)0(<2:3) 2( 4.8) Betaproteobacteria 18( 38.3) 14( 29.2) 28( 63.6) 19( 45.2) Burkholderiales 14( 29.8) 10( 20.8) 23( 52.3) 14( 33.3) Rhodocyclales 0( <2:1)0(<2:1) 5( 11.4) 4( 9.5) Neisseriaceae 4( 8.5) 4( 8.3) 0( <2:3)0(<2:4) Unclassified Betaproteobacteria 0( <2:1)0(<2:1) 2( 4.5) 1( 2.4) Gammaproteobacteria 17( 36.2) 17( 35.4) 1( 2.3) 0( <2:4) Enterobacteriales 0( <2:1) 2( 4.2) 0( <2:3)0(<2:4) Aeromonadales 8( 17.0) 6( 12.5)Proofs 0( <2:3)0(<2:4) Pseudomonadales 9( 19.1) 8( 16.7) 0( <2:3)0(<2:4) Chromatiales 0( <2:1) 1( 2.1) 0( <2:3)0(<2:4) Xanthomonadales 0( <2:1)0(<2:1) 1( 2.3) 0( <2:4) Epsilonproteobacteria 0( <2:1) 1( 2.1) 0( <2:3)0(<2:4) Campylobacteriales 0( <2:1) 1( 2.1) 0( <2:3)0(<2:4) Unclassified Bacteria 2( 4.3) 1( 2.1) 5( 11.4) 9( 21.4)

The numbers are the numbers of clones from the sludge of the incoming wastewater (13), first adjusting tank (14), reactor tank (15), and the sludge digesting tank (16). The samples were taken on Feb. 3, 2005. , One clone (1332) is perhaps a putative chimera. The numbers in parenthesis indicate % of total clones of each tank.

Table 5. Classification of Burkholderiales 16S rDNA Clones Retrieved from Sludge from the Wastewater Treatment System of the Food- Processing Factory

Tank 13 14 15 16 Burkholderiales 14( 29.8) 10( 20.8) 23( 52.3) 14( 33.3) Comamonadaceae 14( 29.8) 10( 20.8) 23( 52.3) 14( 33.3) Brachymonas 14( 29.8) 10( 20.8) 0( <2:3)0(<2:4) Unclassified Comamonadaceae 0( <2:1)0(<2:1) 23( 52.3) 14( 33.3)

The numbers are the numbers of clones from the sludge of the incoming wastewater (13), first adjusting tank (14), reactor tank (15), and sludge digesting tank (16). The samples were taken on Feb. 3, 2005. The numbers in parenthesis indicate % of total clones of each tank.

tank and in the fourth sludge-digesting tank (Table 5). dominated under such conditions. Raw wastewater ran through a leading pipe, that was not The significance of the disappearance of Chloroflexi aerated well and thus provided conditions with lower members and of the appearance of unclassified Coma- dissolved oxygen. Members of Brachymonas may have monadaceae members was well supported by chi square Activated Sludge and Community Structure Analysis 60632-7 Table 6. Classification of the Members of Comamonadaceae Clones

Comamonadaceae (81) Clone name Diaphorobacter (4) KSP3,KSP4,OS1L-2,OS1L-9 Comamonas (10) T30,T54,NSP5,NSP7,NSP8,OS1L-7 OS1L-10,OS1L-11,OS1L-16,TSL-23 Hydrogenophaga (10) T19,T25,T35,T41,T47,T70,T71 T83,T90,T98 Rhodoferax (3) T3,T14,T67 Acidovorax (14) T20,T22,T36,T49,T59,T87, KSP1,KSP2,OS-6,OS-9,OS-19 OS1L-1,OS1L-5,OS1L-6 Curvibacter (1) T73 Unclassified Comamonadaceae (39) 1501,1502,1509,1512,1513,1516 1519,1521,1522,1523,1525,1526 1528,1532,1533,1534,1535,1537 1540,1541,1542,1544,1550,T60 T96,T99,NSP4,OS3,OS1L-18,,TSL-1-1 TSL-1-3,TSL-1-5,TSL-1-7,TSL-2-2 TSL-2-3,TSL-2-9,OS1L-4,OS1L-19 OS1L-20

All 44 clones from the #15 sludge sample, 67 clones from the Munich sludge3) and 33 selected Comamonadaceae clones from laboratory sludge14) were subjected to classifier program by RDP (http://rdp.come.msu.edu/). Eighty-one members of Comamonadaceae are shown. analysisAdvance (< 0:1%). The significance of the presence View of bacteria) are abundant in activated sludge.13) The Brachymonas members in the first two tanks only was members of Chloroflexi in our sludge were from the also supported. genus Anaerolinea, classified with the members of subgroup 1.5,12,13) Sixteen clones classified as members Comparison of community structures of different of Anaerolinea from the five reactor tank samples in sludges Table 2 were closely affiliated with a clone from an Phylogenetic analysis of activated sludge of a large aerated lagoon (Kmlps6.20, AF289914), except for municipal wastewater treatment plant revealed that one from the 01 sample, which was affiliated with a member of Betaproteobacteria of Proteobacteria con- mesophilic UASB sludge clone (MUG7, AB011299).14) stituted a large fraction of the sludge population.3) The clones classifiedProofs as Betaproteobacteria, described We subjected the 16S rDNA sequences described in in Tables 2 and 4, were subjected to the ClustalX the study just mentioned to the classifier program of program with 34 reference sequences and 13 represen- Ribosomal Data Base Project II (http://rdp.come.msu. tative sludge sequences.3,15) The results are shown in a edu/)9) and compared them with those of this study. It is phylogenetic tree in Fig. 2. They were categorized into 7 surprising that our system and the system reported share OTUs (operational taxonomic units) (> 99% nucleotide a large fraction of Burkholderiales (Table 2). We fur- identity), four pairs of highly similar sequences (> 99% ther classified these and found that they belonged to identity), and 10 single sequences. The members (1343, members of Comamonadaceae (Table 6). Members of 1338, and OTU2) of Brachymonas described in Table 5 Hydrogenophaga and Acidovorax constituted a large were closely affiliated with Brachymonas denitrificans. fraction of Comamonadaceae in the plant in that study, The most bacteria and clones closely related to the while the dominant members were unclassified Coma- dominant members (OTU1 and OTU6) of unclassified monadaceae in the sludge in this study. Comamonadaceae were Culvibacter gracilis, Pseudo- The dominant Burkholderiales in the activated sludge monas lanceolata, and clones of Cluster V (T65 and described in the municipal plant in the other study might T15) of Munich municipal wastewater treatment plant. imply that the system was not aerated enough since the sludge was from the first basin of a phosphate removing Evaluation of the improved system two-stage system, which usually comprises aerobic and A reduction in sludge production is favored because anaerobic systems. This is also corroborated by the fact extraction and treatment of evolved sludge in the that Burkholderiales dominated when the oxygen supply conventional activated sludge process requires high was limited in the plant in this study (Table 2). energy costs. Because our improved process did not Only one Chloroflexi clone (1.6%) was detected in the include the addition of exogenous sludge, the usual system in the other study, while our system showed a facilities of the conventional activated sludge process large fraction of members of Chloroflexi before im- can easily be improved without extensive modification provement. Chloroflexi fell in numbers and Burkholder- of the system. iales dominated when the oxygen supply was limited in Incoming wastewater contains significant amounts of our system. Filamentous Chloroflexi (green non-sulfur Brachymonas of Comamonadaseae, probably because of 60632-8 T. SADAIE et al.

1633 OTU1 OTU6 Munich Cluster V

Munich Cluster I Munich Cluster III

Burkholderiales

1523 KSP2 Comamonadaceae

Munich Cluster II OS1L-5 NOS3

OS1L-2 sp. NA5 Munich Cluster IV

OS1L-11

TSL-23 NSP5 NSP4

1343 1338 AdvanceOTU2 View 0121 unclassified Burkholderiales 0103

Burkholderiaceae

Oxalobacteriaceae 0101 Alcaligenaceae 0732 1507 OTU7 NitrosomonadalesProofs 1503 1613 Rhodocyclales

OTU5 0106 OTU3 Methylophilales Hydrogenophilales 0119 Neisseriales 1303 OTU4

Fig. 2. Phylogenic Tree of the Members of Betaproteobacteria from Activated Sludge of a Food-Processing Company and a Municipal Wastewater Treatment Plant. The reference bacteria and clones of Betaproteobacteria listed in Tables 2 and 4 were subjected to the MEGA3 program. The OTUs (operational taxonomic units) OTU1, OTU2, OTU3, OTU4, OTU5, OTU6, and OTU7 contained 59 (01 4 þ 03 4 þ 07 16 þ 15 21 þ 16 14), 23 (07 1 þ 13 12 þ 14 10), 12 (01 1 þ 03 8 þ 07 2 þ 15 1), 6 (13 2 þ 14 4), 5 (07 1 þ 15 2 þ 16 2), 3(03 2 þ 15 1), and 3 (01 1 þ 03 1 þ 15 1) clones respectively. Seven consensus sequences were extracted from each OTU and used for affiliation. Clones 0350, 0305, 0125, and 0107 were quite similar to clones 0103, 0121, 0106, and 0119 respectively and are not shown. Members of Munich clusters I (T25, Z93971), II (T33, Z93960), III (T67, Z93955), IV (T60, Z93993), and V (T65, Z93959) are described elsewhere.3) Clones TSL-2-3 (AB076885), NSP4 (AB076849), NSP5 (AB076850), OS1L-11 (AB076868), KSP2 (AB076843), OS1L-5 (AB076863), NOS3 (AB076845), and OS1L-2 (AB076861) are described elsewhere.15) The bar indicates 2% dissimilarity. Boot strap vaules greater than 50% were reported at the nodes. anaerobic conditions inside leading pipes, which might filamentous Chloroflexi (green non-sulfur bacteria). allow the growth of nitrate reducing bacteria. Enough Reducing the dissolved oxygen concentration in the aeration in the reactor reduces these bacteria and reactor reduced the growth of members of Chloroflexi enhances the growth of bacterial species such as and allowed the growth of members of Comamonada- Activated Sludge and Community Structure Analysis 60632-9 seae, which did not contain Brachymonas. The dominant 4) Amann, R. I., Ludwig, W., and Schleifer, K. H., members were unclassified Comamonadaceae, presum- Phylogenetic identification and in situ detection of ably belonging to nitrate reducing bacterial members, individual microbial cells without cultivation. Microbiol. which were closely affiliated with nitrate reducing Rev., 59, 143–169 (1995). Comamonadaceae (Table 6, Fig. 1) isolated in the 5) Hugenholtz, P., Goebel, B. M., and Pace, N. R., Impact of culture-independent studies on the emerging phylo- activated sludge acclimated with degradable substrate 15) genetic view of bacterial diversity. J. Bacteriol., 180, poly(3-hydroxybutyrate-co-3-hydroxyvalerate) under 4765–4774 (1998). a low oxygen supply (DO ¼ 0:1). It is thought that 6) Woese, C. R., Bacterial evolution. Microbiol. Rev., 51, sludge containing a large fraction of members of 221–271 (1987). Comamonadaceae is capable of performing a biological 7) Sambrook, J., Fritsch, E. F., and Maniatis, T., ‘‘Molecu- nutrient removal process with a reduced increase in cell lar Cloning, a Laboratory Manual’’ 2nd ed., Cold Spring mass when the oxygen supply is limited. Harbor Laboratory Press, Cold Spring Harbor (1989). The activated sludge with a lower oxygen supply de- 8) Ohshima, H., Matsuoka, S., Asai, K., and Sadaie, Y., scribed in this report was derived from a food-process- Molecular organization of intrinsic restriction and ing factory. The incoming wastewater was kitchen modification genes BsuMofBacillus subtilis Marburg. wastewater, which was rich in nutrients for bacteria. J. Bacteriol., 184, 381–389 (2002). 9) Cole, J. R., Chai, B., Farris, R. J., Wang, Q., Kulam, S. On the other hand, general wastewater of lower BOD A., McGarrell, D. M., Garrity, G. M., and Tiedje, J. M., is poor in nutrients. The bacterial population of such The ribosomal database project (RDP-II): sequences and wastewater may be quite different from that of wastewa- tools for high-throughput rRNA analysis. Nucleic Acids ter from the food-processing factory, described in this Res., 33, D294–D296 (2005). study. It is desirable to compare the sludge community 10) Thompson, J. D., Gibson, T. J., Plewniak, F., structures of different wastewater treatment plants. Jeanmougin, F., and Higgins, D. G., The ClustalX AdvanceMore rapid structure analysis of sludge is needed. View We windows interface: flexible strategies for multiple are performing experiments to compare sludge com- sequence alignment aided by quality analysis tools. munity structures by analyzing genome profiling pat- Nucleic Acids Res., 24, 4876–4882 (1997). terns16) derived from sludge DNA samples. 11) Kumar, S., Tamura, K., and Nei, M., MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief. Bioinform., 5, 150–163 Acknowledgment (2004). 12) Page, R. D. M., TREEVIEW: an application to display We are grateful to R. H. Doi for critical reading of the phylogenetic trees on personal computers. Comp. Appl. manuscript. Biosci., 12, 357–358 (1996). This study was performed as a part of the Rational 13) Bjornsson, L.,Proofs Hugenholtz, P., Tyson, G. W., and Evolutionary Design of Advanced Biomolecules Blackall, L. L., Filamentous Chloroflexi (green non- (REDS) Project of the Saitama Prefecture Collaboration sulfur bacteria) are abundant in wastewater treatment of Regional Entities for the Advancement of Techno- processes with biological nutrient removal. Microbiol., logical Excellence program, supported by Japan Science 148, 2309–2318 (2002). and Technology Agency (JST). 14) Sekiguchi, Y., Takahashi, H., Kamagata, Y., Ohashi, A., and Harada, H., In situ detection, isolation, and physio- logical properties of a thin filamentous microorganism References abundant in methanogenic granular sludges: a novel isolate affiliated with a clone cluster, the green non- 1) Maier, R. M., Pepper, I. L., and Gerba, C. P., ‘‘Environ- sulfur bacteria, subdivision I. Appl. Environ. Microbiol., mental Microbiology,’’ Academic Press, San Diego, 67, 5740–5749 (2001). California (2000). 15) Khan, S. T., Horiba, Y., Yamamoto, M., and Hiraishi, 2) Tran, Q. H., and Unden, G., Changes in the proton A., Members of the family comamonadaceae as potential and the cellular energetics of Escherichia coli primary poly(3-hydroxybutyrate-co-3-hydroxyvalerate)- during growth by aerobic and anaerobic respiration or by degrading denitrifiers in activated sludge as revealed by fermentation. Eur. J. Biochem., 251, 538–543 (1998). polyphasic approach. Appl. Environ. Microbiol., 68, 3) Snaidr, J., Amann, R., Huber, I., Ludwig, W., and 3206–3214 (2002). Schleifer, K. H., Phylogenetic analysis and in situ 16) Naimuddin, M., and Nishigaki, K., Genome analysis identification of bacteria in activated sludge. Appl. technologies: towards species identification by genotype. Environ. Microbiol., 63, 2884–2896 (1997). Brief. Funct. Genomic. Proteomic., 1, 356–371 (2003).