Jorrnxer or BroscmNcr eNo BlorucnrrrnrNcVol. 88, No. 1,92-97. 1999

Physiological and Chemotaxonomical Studies on Microflora within aComposter Operated at High Temperature

MANNIX SALVADOR PEDRO,I$ NOBUHIRO ROBERT HAYASHI,I TAKASHI MUKAI,2 MASAHARU ISHII,IAKIRA YOKOTA,3 eNp YASUO IGARASHII*

Department of Biotechnology, Graduate School of Agricultural and Life Sciences, The University of Tokyo, l-Ll Yayoi,Bunkyo-ku, Tokyo 113-8657,t Appliances and Pipeline R&D Department, Commercial Appliances R&D Center,

Tokyo Gas Co. Ltd., 3-13-l Minami-Senju, Arakawa-ku, Tokyo I 16-0003,2 and Institute of Molecular andCellular Biosciences, The University of Tokyo, 1-1-l Yayoi, Bunkyo-ku, Tokyo 113-0032,3 Japan

Received 25 December l998/Accepted 2 April 1999

Physiological and chemotaxonomical studies were carried out on microorganisms isolated from a composteroperated at around 50 to 60oC. Only gram-positive rods were isolated and the microbial counts on agar plateswere relatively low (103 to 10s cfu,/g-sample). The chemotaxonomic properties of the ten isolates revealed thatthey were typical of the genus Bacillus.Identification by the Biolog system concurred with the chemotaxonomy;all the strains were determined to be Bacillus. 165 rDNA analysis closely identified five of the isolates as B.lichenifurmis, one as B. subtilis, and two as B. thermoamylovorans, while the remaining two strains seemed tobelong to a new species. Ilowever, in several cases 165 rDNA analysis and the Biolog system gave differentresults at the species level. The stability of microflora within the composter was also investigated using theBiolog system. Analysis of colonies obtained during three sampling periods-December 1997, April 1998, andJune 1998-showed that the microflora in the composter remained fairly stable.

[Key words: Bacillus, microflora, chemotaxonomy, composter, Biolog system, 165 rDNA analysis,phylogenetic treel

With rapidly increasing amounts of waste, especiallyin urban areas, placing a large burden on the global en-vironment, measures to reduce the quantity of waste areclearly needed.

Among the various kinds of urban waste, we havefocused on organic material (house hold waste/garbage).Organic waste can be incinerated, used in land recla-mation, or degraded by microbial treatment. However,secondary problems are associated with the first two ofthese procedures; CO2 can be given off during incinera-tion, while the inherent problems associated with recla-mation are well known, particularly those related to thetime needed for site restoration and stabilization. Fur-thermore, neither of these procedures is economical. Inthis context, microbial treatment is preferable becauseof its sustainability, i.e., the low burden it puts on theenvironment and the recyclability of organic and/or energyresources achieved through the process. We set the follow-ing aims for our research: (i) establishment of a micro-biological method to degrade urban organic waste effi-ciently and (ii) enhancement of the microbial degradationof such waste through the introduction of microfloraable to degrade organic material rapidly and efficiently.

To obtain suitable microflora, we investigated microor-ganisms in a well-controlled composter that had beenoperated over a long period of time at a relatively con-stant temperature (50 to 60'C) with daily supplements oforganic waste. The research reported here was initiatedto clarify the dominant microorganisms and assess theirstability within the composter. Current methodologies-chemotaxonomy, the Biolog system, and 165 rDNA se-

I Corresponding author.$ Present address: National Institute of Molecular Biology and

Biotechnology, University of the Philippines at Los Banos College,Laguna 4031, Philippines.

quencing analysis-were used to identify and character-ize the isolated strains.

MATERIALS AND METHODS

Collection of samples The composter selected forthe study, which was manufactured by Tokyo Gas Co.,Tokyo (model 50 CG), has a maximum treatment capae-ity of 40 kg garbage/d. It consists of three vessels whosecontents are stirred by means of stirring bars. Garbage isput into the flrst vessel daily. When the contents of thisvessel reach a certain point, the overflow goes into thesecond vessel. In a similar way, the overflow from thisgoes into the third vessel. Finally, the overflow from thethird vessel goes to a small vessel from which the finalproduct can be collected. A detailed description of thecomposter can be found elsewhere (1). In our study, theinside temperature was maintained at around 50 to 60'Cby circulating hot water through a pipe around the wallof the composter. Samples from the composter werethus comparable in texture to dry soil. The composterwas operated with a daily supplementation of wastefrom Monday to Friday at around 2 o'clock. This wasteconsisted primarily of vegetables and egg shells (about10 to 20 kgld) from the cafeteria where the composterwas installed. Throughout the experiments, samples ofabout 100 g were taken from the first vessel just beforethe supplementation of waste on Wednesday, Thursday,or Friday.

Isolation of strains An aliquot of each sample wasserially diluted, and 100 pl of the 10 to 103 dilutionswere plated onto tryptic soy agar (TSA; Difco, Detroit,MI, USA), 30g/1, malt extract agar (malt extract, l0g/l;glucose, l0 g/l; peptone, I g/l; agar, 15 e/l IpH 6.01), orlactic acid bacteria agar (yeast extract, l0 g//; glucose, 10g//; sodium acetate,3gl/; sodium chloride, 5g/l; agar,

92

Vor. 88, 1999

15 g/l lpH 6.51). Antibiotics were omitted from all mediato allow all possible culturable microorganisms to growunder laboratory conditions. After 2 to 4 d incubation at37oC or 50'C, single colonies were acquired. These werepurifled by three successive subcultures. Pure cultures werestored at -85"C wilh 5% dimethyl sulfoxide.

Morphological and physiological characteristics Cellmorphology was observed by phase-contrast micro-scopy after growth on TSA. Spore formation was inducedby using Schaeffer's medium (2). Spores were observedunder a light microscope. Motility was determined bythe hanging drop technique. Catalase activity was deter-mined by the formation of bubbles in 3fr hydrogen per-oxide solution. Oxidase activity was analyzed by oxidationof l% tetramethyl-p-phenylene diamine on a filter paper.

Analysis of isoprenoid quinones Isoprenoid qui-nones were extracted from 200 mg of freeze-dried cellsby shaking with chloroform : methanol (2: l, by vol.) for2h at room temperature. After evaporation, acetone-soluble substances were flltrated and quinone was puri-fied by thin-layer chromatography (TLC) (Silica Gel G,20x20cm; Merck, Darmstadt, Germany). The solventsystem for development was hexane: ethyl ether (85: 15,by vol.). After puriflcation, the flnal yellowish compoundwas analyzed by high-performance liquid chromatography(HPLC) using a Waters HPLC model 510 apparatus(Waters, Tokyo) equipped with a Wakosil-Il 5Cl8 HGcolumn (Wako, Osaka) (3-5).

Analysis of cellular fatty acids Methylated cellularfatty acids were obtained from 50 mg of freeze-driedcells by acid methanolysis for 3 h at 100'C using a l0%HCl-methanol solution. The methylated fatty acids wereextracted by hexane. The hexane solution was analyzed bygas-liquid chromatography (GLC). GLC was performedat 150'C using a Shimadzu GC-8A apparatus (Shimadzu,Kyoto) equipped with a Supelco-SP 2330 50 m capillarycolumn (3-6).

Analysis of cell-wall diamino acids Cell wall wasprepared from l.5g of wet cells as follows. First, a cellsuspension was sonicated at 180W for 30min. The totalhomogenate was treated with SDS (final, 3%) at 100"Cfor 40 min. Cell wall was recovered by a series of cen-trifugations (18,000xg, 15 min) and freeze-dried.

The hydrolysate was prepared by suspending 2.5 mg ofthe freeze-dried cell wall in 0.2 ml of 4 N HCI at 100'Covernight. Treatment with triethylamine and phenyliso-thiocyanate followed to obtain the diamino acid deriva-tive as phenylthiocarbamoyl according to the supplier'sinstructions (7). The final reaction mixture was analyzedby HPLC using a Shimadzu class VP apparatus equippedwith a Wakopak WS-PTC column (Wako) to analyzediamino acids within the cell wall. A part of the hy-drolysate was also analyzed by cellulose TLC (CelluloseF; 20x20cm, Merck) to determine isomers of diamino-pimelic acid (DAP) (8). The solvent system used for devel-opment was methanol : pyridine : 6N HCI : water (64 : 8 :4 :4, by vol.).

Analysis of G*C content DNA was isolated bythe method of Saito and Miura (9). The Gf C contentof the DNA was determined by HPLC following themethod of Mesbah et al. (10) after treatment with Plnuclease and alkaline phosphatase. HPLC was perform-ed using a Shimadzu model LC-6AD apparatus equippedwith a Cosmosil 5C18-AR column (4.6 x 150 mm; NacalaiTesque, Kyoto).

Identification of strains by Biolog system A culture

MICROFLORA IN COMPOSTER 93

grown for 24h on Biolog Universal Growth Mediumsupplemented with lfi glucose (v,/v) was prepared asdescribed in the Biolog user's manual (Biolog Inc.,Hayward, CA, USA). Cell suspensions were also made ac-cording to the manual. In an effort to obtain a uniformcell density, suspensions were adjusted with 0.85% NaClto give a perc€nt transmittance of 37 to 39 when mea-sured by a Biolog turbidimeter. Under this condition, thecell concentration was approximately 4.8 x 108 cells,/ml.The microtiter plates were inoculated according to theBiolog manual. After incubation for 24h at 37"C, read-ings were obtained by using an automated Biolog micro-plate reader.

165 ribosomal DNA sequence determination and phylo-genetic analysis Wet cells (l-3 mg) were suspendedin 200 pl of sterile distilled water and sonicated twice at200W for 5min. Forty microliters of the cell lysate wasresuspended in 50 pl of 40 mM Tris-HCl buffer (pH 8.0)containing l% Tween 20, 0.2mM EDTA, and 02%Nonidet P-40 (BDH Laboratory Supplies, Poole, UK).Final lysis was achieved by Proteinase K treatment for20min at 60"C. After centrifugation at 10,000x9 for5 min, the supernatant was subjected to PCR amplifica-tion. PCR was performed for 3 min at 94"C at first andconsisted of 30 cycles under the following conditions:1 min at 94"C, I min at 50oC, and 2min at 72"C. Afterthe cycles, the final elongation step was performed forl0min at 4'C. The primers used for the PCR reactionswere as follows: forward primer, 8F (5'-AGAGTTTGATCCTGGCTCAG-3'); reverse primer, 15 12R (5'-ACGGCTACCTTGTTACGACT-3'). The foregoing numbering isfrom the Escherichiq coli system (l 1). The PCR productgenerated with TaKaRa Ex Taq polymerase (Takara Shuzo,Kyoto) was cloned using an Original TA Cloning Kit (Invi-trogen, CA, USA).

The sequencing reaction was performed with a BigDyervI Terminator Cycle Sequencing Kit (Perkin ElmerJapan, Applied Biosystems Division, Chiba). Sequencingwas carried out with an ABI Model 377 DNA Sequencer(Perkin Elmer Japan). The 165 rDNA sequences ob-tained for each strain were multiply aligned usingCLUSTAL W ver. 1.72 (12) with a selection of 46 refer-ence sequences obtained from the DNA Data Bank ofJapan (DDBJ) or from the Ribosomal Database Project(RDP, Department of Microbiology, University of Illi-nois at Urbana-Champaign, USA; Release Number 5.0,INIay 17, 1995). The reference strains used and accessionnumbers of the reference sequences are listed in Table 1.

A phylogenetic tree was constructed from the evolu-tionary distance data by applyrng the algorithm of theneighbor-joining method (13) to the Koo" values (14). Toevaluate the robustness of branches in the inferred tree,the bootstrap resampling method of Felsenstein (15) with1000 replicates was used.

Nucleotide sequence accession numbers The nucleo-tide sequences reported in this paper have been submit-ted to the DDBJ, GenBank, and EMBL databases underthe following accession numbers: A8020192 (strainTGS137), AB020193 (TGS237), A8020t94 (TGS337),4B020195 (TGS437), A8020196 (TGS537), A8020197(TGS650), AB020198 (TGS750), 4B020199 (TGS850),A8020200 (TGS950), and AB020201 (TGSI050).

RESULTS AND DISCUSSION

Plate observation and isolation of strains When a

94 PEDRO ET AL. J. Broscr. BroENG.,

TABLE l. Reference strains irsed and their accession numbers

Reference strain Accession number Reference strain Accession number

Bacillw atrophaeus NCIMB 12899Bacillus pallidus DSM 3670 (T)Bacillus caldovelox DSM 411Bacillus lentimorbus ATCC 14707 (T)Bacillus popilliae ATCC 14706 (T)Bacillus thermoalkalophilzs DSM 6866Bacillus thermodenitrificans DSM 465Paenibacillus lautus NCIMB 12780Bacillus caldolyticus DSM 405Bacillus denitrificans DSM 466Bacillus methanolicus NCIMB 13114Bacillus sporothermodurans DSM 10599 (T)Bacillus thermocatenulalas DSM 730 (T)Bacillus thermoglucosidasius ATCC 43742 G)Bacillus caldotenax DSM 406Bacillus flavothermus DSM 2647Bacillus oleronius DSM 9356 (T)Bacillus thermoamylovorans CIP CNCMI-1378Bacillus thermocloacae DSM 5250 (T)Bacillus thermoleoyorons DSM 5366Bacillus kaustophilus NCIMB 8547A licyclobacillus acidoterrestris DSM 3922 (T)Bacillw amyloliquefociens ATCC 23350 (T)

x60607226930226925x60622x60633226931226928x60621226924226927s42879u49079226926x60641x62180226932x8u92L27478226939226923x60618x60602x6060s

from RDPx64887x60514x70648x60609x55060x77790x60631x68416x60629x60638x60637D78316D78320x60632x60612x60620x94195xs7309x60646u20384x60614Lt40t1

Lactobacillus casel subsp. casei ATCC 334Lactococcus lactis strain IL 1403Streptomyces coelicolor A3Q)Staphylococcw aureus NCDO 949Bacillus a4otoformans ATCC 29788 (T)Bacillus cereus NCDO 177 1

Bacillus badius ATCC 14574 (T)Bacillus pasteuril NCIMB 8841 (T)Bacillus licheniformis DSM 13 (T)Bacillus megaterium DSM 32 (T)Bacillus simplex DSM 1321 (T)Bacillus pumilus NCDO 1766Bacillus smithi, JCM 9076 (T)Paenibacillus validus DSN/ 3037Paenibacillus polymyxa NCDO 1774Brevibacillus brevis NCIMB 9372Brevibacillus latesporus NCDO 1763Aneurinibacillus migulanus DSM 2895Bacillus stearothermophilus strain 10Bacillus subtilis NCDO 1769B aci I I us infernus str un TH-22Bacillus coagulans NCDO 1761Bacillus sphaericus strain 2362

(T), Type strain.

sample was analyzed using three different plates (TSA,malt extract agar, or lactic acid bacteria agar), no appar-ent qualitative difference could be discerned among thecolonies, either at 37 or 50"C. In addition, neither yeastsnor molds could be observed on plates. Therefore, TSAwas mainly used for further experiments.

The plate counts of samples on TSA showed smallvariations among the three sampling periods (Table 2)although the viable counts themselves were low. Theselow counts are likely to reflect, to some extent, a limiteddiversity of microorganisms inside the composter. Be-cause the makeup of the garbage did not vary much, itis highly probable that the microorganism diversity wasrelatively low. Alternatively, there is a possibility thatunculturable strains were involved in the degradation inaddition to the colony-forming strains. Higher countswere noted at 50'C than at 37"C (Table 2), which can beattributed to thermophilic nature of the microorganisms.

Five strains were finally purified at 37"C (TGS137,TGS237, TGS337, TGS437, and TGS537) and anotherfive strains at 50"C (TGS650, TGS750, TGS850,TGS950, and TGS1050), although optimum growth tem-perature for each strain has not yet been demonstrated.

Phenotypic and chemotaxonomic characteristicsThe chemotaxonomic characteristics of the isolates aresummarized in Table 3. Although strains TGS237,TGS537, and TGS750 showed a negative gram reaction,their respiratory quinone system (MK-7) strongly sug-gests that they actually belong to the gram-positivegroup. Spores were observed from all isolates. HPLC

TABLE 2. Microbial counts

Cultivationtemperature

('c)Counts (cfu/g-sample)

December 1997a April 1998a June 1998u

analysis revealed that they all had a menaquinone systemwith seven isoprene units (MK-7). The major cellular fattyacid component was of the iso- or anteiso-branchedtype. An examination of amino acids from the cell wallby HPLC revealed that the isolates contained diamino-pimelic acid (DAP) and that its configuration was meso asdetermined by TLC. The G*C molar ratio ranged from37.2 to 54.8%. In combination, the above results readilyidentify the isolates as belonging to the genus Bacillus(6, 16).

Biolog analysis Carbon utilization patterns havebeen found increasingly useful in environmental andecological microbiology over the past few years (17-19).Hence, the isolates were also identified using the Biologsystem, which employs a redox reaction by tetrazoliumviolet to test the ability of a strain to utilize a carbonsource. Tetrazolium dye can function as an artificial elec-tron acceptor when NADH is generated during metab-olism of the test substrate (20). As a result, the dyeis reduced to a colored, insoluble formazan that can bemeasured qualitatively and quantitatively (19). Severalcautions have been noted in previous studies on the useof this system: (i) The inoculum concentration can affectthe length of the lag period for the development of colorin the microtiter plate wells (20, 2l); (ii) factors like theoxygen concentration (18) can affect the results; and (iii)large amounts of nutrients on the plate when slow-grow-ing cells are introduced may cause or accelerate celldeath (17, 22). Using this system, all the isolates wereidentifled as belonging to the genus Bacillus at the spe-cies differentiation level (Table 4).

165 rDNA analysis Nearly complete 165 rDNAsequences (approximately 1500 nucleotides) of the tenisolates were determined. The primary structures werealigned and compared with those of 46 reference strainsobtained from DDBJ and RDP. Sequence similaritieswere calculated using data for 1240 nucleotides. Aphylogenetic tree based on the K,," values is shown inFig. 1. The tree qlearly demonstrates that all the isolatesbelong to the genus Bacillus and that they can be grouped

al50

2.0 x 103

9.0 x 104

2.8 x 103

8.0 x 104

1.9 x 103

7.3x 7U

TSA was used.a Sampling period.

Vor. 88, 1999 MICROFLORA IN COMPOSTER 95

TABLE 3. Morphological and chemotaxonomic characteristics of isolates

lsolateIsolation

temperature("c)

Motility Spore ..g:ffi,catalase oxidase G+9 Quinone Fatty acid compositiontest lest moloo

Cell walldiamino acid

TGSI37TGS237TGS337TGS437TGS537TGS65OTGS75OTGS85OTGS95OTGSlO5O

JIal3737375050505050

++

++++++++++

++++

+++++

37.2 MK-747.4 MK-742.4 MK-746.8 MK-754.8 MK-737.2 MK-746.3 MK-745.1 MK-745.7 MK-737.2 MK-7

i15, nl6, i17,il5, nl6, i17, atl7il5, n16, il7, ail7il5, n16, il7, ailTil5, nl6, il7, ail7il5, nl6, il7, ai17il5, n16, il7, ai77il5, n16, il7, at17il5, n16, il7, a117

il5. n16. il7, ail1

meso-DAPMCSO-DAPMCSO-DAPmeso-DAPmeso-DAPmeso-DAPmeso-DAPmeso-DAPmeso-DAPmeso-DAP

n, Normal; i, iso-branched, ai, anteiso-branched.

into two subgroups-a B. subtilis subgroup (TGS237,TGS437, TGS750, TGS850, TGS950, and TGSI050) anda B. methqnollctts subgroup (TGS137, TGS337, TGS537,and TGS650) (RDP data). From Fig. 1, it is obviousthat different types of strains were obtained at 37 and50"C. The only possible exception is the similarity betweenstrains TGS537 and TGS650, which according to the 165rDNA data was 98.9% (Table 5). However, the GtCmol)( analysis (Table 3) showed marked difference (a17.6% difference), which suggests that they are in factdiflerent strains.

Analysis of the 165 rDNA sequences closely identifiedTGS237 as ,8. subtilis, TGS437, TGS750, TGS850,TGS950 and TGS1050 as B. licheniJormis, and TGS537and TGS650 as B. thermoamylovorons. However, as not-ed above, it is thought that the latter two isolates shouldbe differentiated at the species level. Further work isneeded to clarify their taxonomic position(s).

Isolates TGSI37 and TGS337 seemed to form a differ-ent branch arising from the B. methqnolicus and B.infernus cluster. This supposition is based on the follow-ing facts: (i) TGS137 and TGS337 form a monophyleticcluster with a bootstrap confidence level of 99% and (ii)the level of confidence with respect to the neighboringcluster of B. methanolicus and B. infernus was only78%. When the G+C mol% is taken into account, thereis a difference of more than 5% between the value forTGS137 (37.2%) and that for TGS337 (42.4%). If thereis difference in G*Cmol% of more than 5%, strainsshould not be placed in the same species (12). This stipu-lation indicates that TGSI37 and TGS337 should beplaced under different species. Further work is now inprogress to clarify the taxonomical positions of TGS137and TGS337.

TABLE 4. Identification of isolates by Biolog system

Isolate Closest identiflcation % Similarity

TGS137TGS237TGS337TGS437TGS537TGS65OTGS75OTGS85OTGS95OTGSIO5O

Bacillus coagulansBacillus megateriumBacillus brevisBacillus lichenifurmisBacillus brevisBacillus coagulansBacillus pasteuriiBacillus pasteuriiBacillus pasteuriiBacillus sp.

60.8074.3071.1172.3078. l07 5.1078.0055.3051.0030.40

Based on the results of chemotaxonomy, the Biologsystem, and 165 rDNA analysis, the isolates were clearlyshown to belong to the genus Bucillus. However, differ-ences were noted at the species level between the Biologand 165 rDNA analyses. Two examples are discussedbelow. (i) The Biolog system closely identified isolatesTGS750, TGS850, and TGS950 as B. pasteurii, while165 rDNA analysis closely identified these isolates as B.lichenifurmis. Chemotaxonomically, B. pasteurii shouldcontain LL-DAP as a diamino acid in the cell wall (19).However, TGS750, TGS850, and TGS950 were found tocontain meso-DAP (Table 3), which is in disagreementwith the result assumed from the Biolog analysis. Takingall the above factors into consideration, the three strainsshould be identifled as B. lichenifurmis. (ii) The Biologsystem closely identified TGS137 as B. coqgulans andTGS337 as B. brevis. However, as discussed above, the165 rDNA analysis indicates that they form one clusterand seem to belong to a new species.

These differences in identiflcation at the species levelcan be explained by the fact that the database in the Bio-

TABLE 5. Similarity values of 165 rDNA sequences among TGS strains

% Similarity

TGSI37 TGS237 TGS337 TGS437 TGS537 TGS650 TGS750 TGS85O TGS95O

TGSI37TGS237TGS337TGS437TGS537TGS65OTGS75OTGS85OTGS95OTGSIO5O

92.798.295.394.293.993.292.893.594.2

93.295.592.091.796.095.796.597.3

93.993.593.s92.892.393.293.8

93.893.59'.7.2

96.89'.7.4

98.1

98.992.992.593.293.8

92.992.393.193.5

97.398.198.5

98. I98.4

Total number of nucleotides: 1240.

99.3

96 PEDRO ET AL. J. Broscr. BrorNc.,

TABLE 6. Percentages of Bacillus strains determined by Biolog system

Percentage of isolates per sampling period Percentage in relation tototal number of isolatesDecember 1997 April 1998 June 1998

B. brevisB. megateriumB. coagulansB. pasteuriiB. licheniformisB. amyloliquefociensB. badiusB. subtilisBacillus sp.Not identified

32.210.721.410.73.6

NDNDND7.1

14.3

26.518.48.2

14.34.12.02.O4.18.2

12.2

26.233.37.1

11.92.4

NDND4.82.4

11.9

27.72r.910.912.63.40.80.83.45.9

12.6

Total number of isolates tt9424928

ND, Not detected.

Kruc 0.02 Bacillus lentimorbwBacillus atrophaeus

Bacillus pumilus- Bacillus popilliae- B a c i I lu s anry lo liqu ef a c i e nsBacillus subtilis

@. megaterium)Paenibacillus lautus

O @acillus sp.)lichenifurmis

;5950 ' (8. pasteuril

TGS850 (8. pasteuril'G5750 (B. pasreurii)

(8. licheniformis)

-

Snphylococcus aureusBacillus cereus

Bacillus simpluBacillus megaterim

-Bacillre

uotoformansB ac illus s porothemodurans

Bacillus oleronireAlicyclobacillus acidoterres tr is

Bacillus coagulareBacillus badius

Bacillus smithiiBacillus methanoLicus

infunus

log system contains only already known species for com-parison. Optimization of the culture conditions (18) andappending a more extensive database to the system (23)will improve identification rates. In sum, while weshould rely on 165 rDNA analysis for the final identifica-tion step, the Biolog system can be used to know rough-ly the microflora within a composter.

Stability of microflora The Biolog system was usedto assess the stability of microflora in the composter.Samples were taken during three periods; the colonieswere obtained at random at 50oC. Table 6 shows the per-centages of the different strains obtained in each sam-pling period. B. brevis, B. megaterium, B. coqgulqns, B.pqsteurii and B. licheniformis were consistently obtainedfrom the December 1997, April 1998, and June 1998

samples. In contrast, B. amyloliquefociens and B. badiuswere found only in the samples obtained in April 1998,while B. subtilis was found in the April 1998 and June1998 samples. These findings indicate that Bacillus isinvolved in the degradation of waste.

In each sampling period, around l0% of the isolatescould not be identified using the Biolog system (Table 6)although the unidentifled microorganisms may have beenBqcillus species. Overall, the results indicate that themicroflora inside the composter were relatively stable,from which it can be deduced that they were specializedfor the degradation of the garbage. However, becausethe makeup of the garbage was very similar throughoutthe study period, it is possible that new microflorawould be needed if the constituents of the garbagechanged. Introduction of diflerent materials into the gar-bage and analysis of any new resulting microflora will benecessary in future experiments.

General discussion There is a possibility that isolat-ed cells may exist inside the composter in the form ofspores because (i) the sample from the c*omposter had atexture comparable to that of dry soil, and (ii) the platecounts were themselves low. However, even if this is thecase, the fact that the plate counts at 50"C were higherthan those at 37"C indicates that spores can germinateas soon as waste is disposed of. Also, in the originalreport on the system, a thermophilic aerobe was foundto be mainly involved in the degradation of organicmaterials (l). Therefore, the isolated strains, especiallythose obtained at 50oC, could be involved in the degra-dation of waste. In order to clarify this, work to extractDNA from a sample and subject it to detailed analysis isnow in progress.

Future studies will focus on the possible role of each

1

67.

5337 (8. brevis)137 (8. coagulare\

Bacillus th e r mo amy lovorans-TGS650 (8. coagulare)TGS537 (p. brevis)B a c il lw th er mo a I kn lop h i lus

Bacillrc palliduBacillus flavothermw

Bacillus caldatenmBacillusde nitrificarc

B a c il lu s tho m o denitr ifr c an sBacillus kaustophilusB ac il lus th ermo c atenulatus

B a ci I lu s ther mo I eov or a nsBacillrc caldoveloxBacillus caldolyticus

B acillus ste arother omophilusBacillw sphaericus

-Bocillus

pasteurii

B acillus thu mo g lucosidasius

Bacillus thermocbacaePaenibacillus validus

Paenibacillus polymymB revibacillw later o spor w

Brevibacillw brevisA n eur inib a c i I lrc mi g u I anu s

Laclococc$ lactisLactobacillus caseiStreptotrry ce s co elico lor

FIG. l Phylogenetic tree based on 165 rDNA sequences. Thephylogenetic tree was constructed by the neighbor-joining methodwith 46 strains obtained from DDBJ and RDP and with Lactococcuslactis, Lactobacillus casei, and Streptomyces coelicolor as outgroups.The total number of nucleotides for comparison was 1240. The num-bers on the branches are confidence limits (expressed as percentages)by bootstrap estimation. Confidence limits are indicated if they aregreater than 500/1000. The isolates obtained in this study (TGSstrains) are indicated in bold. Names in parentheses shows the re-sults of identiflcation by the Biolog system.

Vor. 88, 1999

isolate in the degradation of waste. Also, ln situ experi-ments are expected to provide many insights into theprocess of degradation.

ACKNOWLEDGMENTS

MSP was a researcher in the 25th International PostgraduateUniversity Course in Microbiology supported by the JapaneseGovernment, UNESCO, and the International Cell Research Or-ganization (ICRO). We thank Mr. Tatsuya Ishida of the Instituteof Molecular and Cellular Biosciences, The University of Tokyo,and Mr. Hirobumi Imai, The University of Tokyo, for their techni-cal helps. This research was supported by the Research for theFuture Program from Japan Society for the Promotion of Science(JSPS).

REFERENCES

l. Anan, F., Fujii, K., Fukuda, M., Shindou, Y., Kitabatake, S.,and Ishida, M.: Shingata baio-shiki namagomi-shoriki nokaihatsu. Kankyou Ken-kyu (Research on Environment), 95,13-19 (1994). (in Japanese)

2. Schaeffer, P. and Ionesco, H.: Contribution a l'etude genetiquede la sporogenese bacterienne. Comptes Rendus, 251, 3725-3127 (1960).

3. Kuroshima, K., Sakane, T.,Tzkata, R., and Yokota, A.: Bacil-lus ehimensis sp. nov. and Bacillus chitinolyticus sp. nov., newchitinolytic members of the genus Bacillus. Int. J. Syst. Bac-teriol., 46, 76-80 (1996).

4. Nakagawa, Y., Sakane, T., and Yokota, A.: Emendation ofthe genus Planococcus and transfer of Flavobacteriumokeanokoites Zobell and Upham 1944 to the genus Planococ-cus as Planococcus okeanokoites comb. nov. Int. J. Syst. Bac-teriol., 46, 866-870 (1996).

5. Yokota, A., Takeuchi, M., and Weiss, N.: Proposal of twonew species in the genus Microbscterium: Microbacteriumdextranolyticum sp. nov. and Microbacterium aurum sp. nov.Int. J. Syst. Bacteriol., 43,549-554 (1993).

6. Goodfellow, M., and O'Donnell, A.: Chemical methods inprokaryotic systematics, p. 13-139. John Wiley & Sons,Chichester, UK (1994).

7. Wako Pure Chemical Inilustries, Ltd.: Technical note on thesystem of PTC-amino acid analysis. Wako Pure Chemical In-dustries, Osaka (1989).

8. Ilasegawa, T., Takizawa, M., and Tanida, S.: A rapid analysisfor the chemical grouping of aerobic actinomycetes. J. Gen.Appl. Microbiol., 29, 319-322 (1983).

9. Saito, H. and Miura, K,: Preparation of transforming deox-yribonucleic acid by phenol treatment. Biochim. Biophys.

MICROFLORA IN COMPOSTER 97

Acta, 72, 619-629 (1963).10. Mesbah, M., Premachandran, IJ., and Whitman, W. B.: Pre-

cise measurement of the G*C content of deoxyribunucleicacid by high-performance liquid chromatography. Int. J. Syst.Bacteriol., 39, 159-167 (1989).

11. Brosius, J., Palmer, M.L., Kennedy, J.P., and Noller, H.F.:Complete nucleotide sequence of a 165 ribosomal RNA genefrom Escherichis coli. Proc. NatI. Acad. Sci. USA, 75, 4807-4805 (1978).

12. Thompson, J. D., Higgins, D. G., and Gibson, T. J.:CLUSTAL W: improving the sensitivity of progressive multiplesequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. NucI. AcidsRes., 22, 4673-4680 (1994).

13. Saituo, N. and Nei, M.: The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol. Biol. Evol.,4,406425 (1987).

14. Kimura, M.: A simple method for estimating evolutionary ratesof base substitutions through comparative studies of nucleotidesequences. J. Mol. Evol., 16, lll-120 (1980).

15. Felsenstein, J.: Confidence limits on phylogenies: an approachusing the bootstrap. Evolution., 39,783-791 (1985).

16. Sneath, P. II. A.: Endospore-forming gram positive rods andcocci, p. 1104-1139. In Mair, N. S., Sharpe, M. E., and Holt,J. G. (ed.), Bergey's manual of systematic bacteriology, vol. 2.Williams and Wilkins, Baltimore (1986).

17. Kanopka, A., Oliver, L., and Turco, R. F., Jr.: The use of car-bon substrate utilization patterns in environmental and ecological microbiology. Microb Ecol., 35, 103-115 (1998).

18. Baath, E., Diaz-Ravina, M., Frostegard, A., and Campbell,C. D.: Effect of metal-rich sludge ammendments on the soilmicrobial community. Appl. Environ. Microbiol., 64, 238-245(1ee8).

19. Mauchline, W. S. and Keevil, C. \Y.: Development of the Bio-log substrate utilization system for identification of Legionellaspp. Appl. Environ Microbiol., 57,3345-3349 (1991).

20. ffaack, S. K., Garchow, II., Klug, M. J., and Forney, L. J.:Analysis of factors affecting the accuracy, reproducibility, andinterpretation of microbial community carbon source utiliza-tion patterns. Appl. Environ. Microbiol., 61,1458-1468 (1995).

21. Garland, J. L. and Mills, A. L.: Classification and characteriza-tion of heterotrophic microbial communities on the basis ofpatterns of community-level sole-carbon-source utilization.Appl. Environ. Microbiol., 57, 2351-23 59 (1991).

22. Posgate, J, R.: Viability measurements and the survival ofmicrobes under minimum stress. Adv. Microb. Physiol., 1, 1-23(1967).

23. Jones, M. N., Turnbull, P. C., and Manchee, R. J.: Evaluationof the Biolog system for the identification of Bacillus anthracis.Lett. Appl. Microbiol., 20,209-2ll (1993).