JOURNAL OF BIOSCIENCE AND BIOENGINEERING Vol. 99, No. 2, 150164. 2005 DOI: 10.1263/jbb.99.

150

2005, The Society for Biotechnology, Japan

Microbial Community Analysis of Mesophilic Anaerobic Protein Degradation Process Using Bovine Serum Albumin (BSA)-Fed Continuous Cultivation
YUEQIN TANG,1 TORU SHIGEMATSU,1* SHIGERU MORIMURA,1 AND KENJI KIDA1,2
Department of Materials and Life Science, Graduate School of Science and Technology, Kumamoto University, 2-39-1 Kurokami, Kumamoto City, Kumamoto 860-8555, Japan1 and Department of Applied Chemistry and Biochemistry, Faculty of Engineering, Kumamoto University, 2-39-1 Kurokami, Kumamoto City, Kumamoto 860-8555, Japan2
Received 15 September 2004/Accepted 16 November 2004

Two mesophilic anaerobic chemostats, one without added Ni2+ and Co2+ (chemostat 1) and the other with added Ni2+ and Co2+ (chemostat 2), were supplied with synthetic wastewater containing bovine serum albumin (BSA) as the sole carbon and energy source in order to study the capacity of protein degradation, microbial community structure and the effects of the addition of trace metals. Volatile fatty acids and ammonia were the main products of chemostat 1, while methane, CO2 and ammonia were the main products of chemostat 2, and critical dilution rates of 0.15 d1 and 0.08 d1 were obtained, respectively. Fluorescence in situ hybridization (FISH) with archaeal and bacterial domain-specific probes showed that archaeal cells were very limited in chemostat 1 while large populations of several types of archaeal cells were present in chemostat 2. Phylogenetic analyses based on 16S rRNA gene clonal sequences, DGGE, and quantitative real-time polymerase chain reaction (PCR) showed that, within the domain Archaea, methanogens affiliated with the genera Methanosaeta and Methanoculleus were predominant in chemostat 2. Within the domain Bacteria, rRNA genes obtained from chemostat 1 were affiliated with the three phyla; Firmicutes (43%), Bacteroidetes (50%) and Proteobacteria (7%). A total of 56% of rRNA genes obtained from chemostat 2 was affiliated with the three phyla, Firmicutes (32%), Bacteroidetes (11%) and Proteobacteria (13%) while 44% of rRNA genes remained unclassified. Phylogenetically distinct clones were obtained in these two chemostats, suggesting that different protein degradation pathways were dominant in the two chemostats: coupled degradation of amino acids via the Stickland reaction in chemostat 1 and uncoupled degradation of amino acids via syntrophic association of amino acid degraders and hydrogenotrophic methanogens in chemostat 2.
[Key words: bovine serum albumin (BSA), anaerobic degradation, methane fermentation, phylogenetic analysis]

In addition to carbohydrates and lipids, proteins are regarded as the main organic substrates in anaerobic digestion of sludge and wastewaters. The protein composition of suspended solid (SS) of domestic sewage is reportedly more than 40% (1). Industries that process food such as whey, cheese, casein and fish also typically produce wastewater containing significant amounts of protein. However, proteins have been demonstrated to degrade more slowly than carbohydrates under anaerobic conditions (2). It has also been documented that the anaerobic degradation of protein-rich wastes is often incomplete (35). These findings imply the importance of protein degradation processes in anaerobic ecosystems. Under methanogenic conditions, proteins are hydrolyzed to peptides and amino acids, which are subsequently fer* Corresponding author. e-mail: shige@kumamoto-u.ac.jp phone: +81-(0)96-342-3668 fax: +81-(0)96-342-3679
150

mented to volatile fatty acids and finally converted to methane and CO2 by an association of microorganisms. Species from genera such as Clostridium (58), Peptostreptococcus (911), Acidaminobacter (12), Aminomonas (13), Thermanaerovibrio (14), Sporanaerobacter (15), Sedimentibacter (16) and Aminobacterium (17, 18) have been shown to degrade amino acids anaerobically. Most anaerobic amino acid degrading bacteria are affiliated with the phylum Firmicutes (the low G + C gram-positive bacteria). In anoxic environments, the oxidation of certain amino acids is endergonic under standard conditions and only proceeds if the reducing equivalents produced are removed by a biological or a chemical electron scavenger. This is possible by interspecies hydrogen transfer in the presence of hydrogen utilizing microorganisms, and when chemical compounds, such as amino acids (Stickland reaction) (5), act as the final electron acceptor. It has been considered that amino acids are preferentially fermented via the Stickland reaction during anaero-

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

151

bic treatment of protein-rich wastewater or waste because mixed amino acids are readily supplied by protein hydrolysis. Despite this knowledge of protein and amino acid degradation that was primarily obtained by investigating pure culture assays, the contribution of these microorganisms to protein degradation in anaerobic ecosystems remains unknown. Because only a fraction of the total microorganism population in an ecosystem can be cultured (19), the actual diversity of protein and/or amino acid degraders remains unclear. Culture-independent methods using molecular biological techniques have successfully been applied to analyze microbial communities in several ecosystems, including anaerobic digesters. Although there are several reports about the anaerobic treatment of protein-rich wastewaters concerning the optimization of treatment processes or morphological observation (3, 4, 2023), there have not been any reports concerning the analysis of microbial community structure among anaerobic protein digesters to date. In this study, using anaerobic digested sludge as the source of microorganisms, two chemostats supplied with synthetic wastewater containing bovine serum albumin (BSA) as the sole carbon and energy source were constructed as treatment models for protein-rich wastewater. The capacity for protein degradation, microbial community structure and the effects of the addition of trace metals were then investigated.

FIG. 1. Operation schedules for chemostats 1 (a) and 2 (b). Arrows refer the times FISH, and DNA extraction for 16S rRNA gene analyses were conducted.

MATERIALS AND METHODS

Synthetic wastewater Synthetic wastewater with BSA as the sole carbon and energy source was prepared without Ni2+ and Co2+ as follows (g/l): BSA (albumin, bovine, general grade; Nacalai Tesque, Kyoto), 17; KH2PO4, 0.3; KHCO3, 4.0; NH4Cl, 1.0; NaCl, 0.6; MgCl2 6H2O, 0.82; CaCl2 2H2O, 0.08; cysteinHCl H2O, 0.1; 10 ml of the trace element solution of DSMZ medium 318 [Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH: Catalogue of Strains 2001. http://www.dsmz.de/dsmzhome. htm] without NiCl2 6H2O and CoCl2 6H2O; and 10 ml of the vitamin solution of DSMZ medium 318 without B12. For the synthetic wastewater with Ni2+ and Co2+, NiCl2 6H2O and CoCl2 6H2O were added to give Ni2+ and Co2+ concentrations of 0.10 and 0.12 mg/l, respectively. The pH of the synthetic wastewater was adjusted to 7.0 with HCl. The total organic carbon (TOC) concentration of the synthetic wastewater was 8000 mg/l. Operation of BSA-fed chemostats Two anaerobic chemostats were operated using two completely stirred tank reactors (CSTRs), each with a working volume of 1.7 l and were mixed thoroughly using a magnetic stirrer. The temperature during the continuous cultivation was maintained at 37C by immersing the reactors in a thermostated water-bath, and pH was maintained automatically at 7.0 by feeding 1 N HCl solution through a pump which was controlled by a pH controller. A 1.2-l portion of mesophilic digested sludge from the Kumamoto-Hokubu sewage treatment plant (Kumamoto City) was washed twice with a synthetic wastewater without BSA, Ni2+ and Co2+ under anaerobic conditions, was diluted to 1.7 l, and was then placed in each CSTR. Synthetic wastewaters containing BSA as the sole carbon source, with or without Ni2+ and Co2+ for chemostat 2 and chemostat 1, respectively, were fed into the CSTR. Biogas was collected in a gasholder. Figure 1 summarizes the time schedule for dilution rate changes of chemostat 1 and chemostat 2.

Fluorescence in situ hybridization (FISH) A 0.5-ml portion of culture broth was centrifuged at 16,000g for 5 min at 4C and the resulting pellet was used as the sample for FISH. FISH was carried out according to the method of Amann (24). EUB338 (S-D-Bact-0338-a A-18) (25), which was 5-end labeled with Cy5 (Amersham Bioscience, Piscataway, NJ, USA), and ARC915 (S-D-Arch-0915-a-A-20) (26), which was 5-end labeled with Cy3 (Amersham Bioscience), were used for the detection of bacterial and archaeal cell, respectively. A confocal laser scanning microscope FV300 (Olympus, Tokyo) was used for microscopic observation. DNA extraction, PCR amplification and cloning DNA from the microbial community was extracted as previously reported (27). Amplification of 16S rRNA gene from the extracted community DNA was performed by PCR using AmpliTaq Gold (Applied Biosystems, Foster City, CA, USA) according to the manufacturers instructions (100 ng template DNA, 1 PCR buffer, 2.5 units AmpliTaq Gold polymerase, 250 mM dNTPs and 40 pmol of each primer in a 100-ml reaction volume). The primer sets Ar28F (5-TGGTTGATCCTGCCAGAGG-3) and 1490R (5-GGTTACC TTGTTACGACTT-3), and Eu27F (5-AGAGTTTGATCCTGGC TCAG-3) and 1490R, were used to amplify 16S rRNA gene from the total-community DNA and targeted Archaea and Bacteria, respectively. The reactions were performed using a Gene Amp PCR System 2400 (Applied Biosystems) with the following cycle conditions: preincubation at 95C for 9 min; and 30 cycles of 95C for 1 min, 50C for 1 min and 72C for 2 min. The amplified 16S rRNA gene fragments were cloned into a pT7Blue T-vector plasmid (Novagen, Madison, WI, USA) using the DNA Ligation Kit ver. 2 (Takara, Kyoto). Libraries of clones were constructed using Eschericha coli DH5a-competent cells (Takara) according to the manufacturers instructions. Sequencing and phylogenetic analysis Cloned 16S rRNA genes were prepared from randomly selected recombinants and were used as templates for sequencing with a CEQ8000 genetic analysis system (Beckman Coulter, Fullerton, CA, USA) and a CEQ DTCS-Quick Start Kit (Beckman Coulter) in accordance with the manufacturers instructions. About 1400 bases were sequenced for all clones selected. All sequences were checked manually for chimeric artifacts using the CHIMERA_CHECK program ver. 2.7 of the Ribosomal Database Project II (RDP-II) (28).

152

TANG ET AL.

J. BIOSCI. BIOENG.,

Searches for similar sequences were performed using the BLASTN program (29). Multiple alignments were generated with the Clustal X program ver. 1.8 (30) and phylogenetic trees were constructed using the MEGA program ver. 2.1 (31) based on evolutionary distances that were calculated using the Neighbor-Joining method (32) and Kimuras two-parameter model (33). Bootstrap resampling analysis (34) for 500 replicates was performed in order to estimate degrees of confidence in tree topologies. Identical sequences were recognized by analysis of phylogenetic trees and manual comparisons, in which sequences that were more than 99.8% similar were defined as identical, and such sequences were used for further phylogenetic analysis as an operational taxonomic unit (OTU). Four libraries of clones were constructed with total-community DNA that had been extracted from culture broth from the chemostats. An archaeal 16S rRNA gene library (BSA1A library) and a bacterial 16S rRNA gene library (BSA1B library) were prepared from the community DNA from chemostat 1 at the dilution rate of 0.15 d1. Twenty-three clones from BSA1A library and 44 clones from BSA1B library were sequenced. And, an archaeral 16S rRNA gene library (BSA2A library) and a bacterial 16S rRNA gene library (BSA2B library) were prepared from chemostat 2 at the dilution rate of 0.08 d1. Thirty-nine clones from BSA2A library and 54 clones from BSA2B library were sequenced. Denaturing gradient gel electrophoresis (DGGE) The V3 regions of 16S rRNA genes from the community DNA extracted from the chemostats were amplified by PCR and the amplified products were used for DGGE analysis. PCR was performed using AmpliTaq Gold (Applied Biosystems) in a total volume of 100 ml in 0.2-ml reaction tubes, and the GeneAmp PCR system 2400 (Applied Biosystems). The primer sets and GC clamp used were the same as those described by vres et al. (35). For amplification of bacterial rRNA gene, PRBA338F and PRUN518R were used as primers. The thermal profile was as follows: incubation at 95C for 9 min; and then 30 cycles of incubation at 95C for 1 min, 55C for 1 min and 72C for 2 min. Archaeal rRNA gene was obtained by nested PCR with a thermal profile similar to that used for bacterial rRNA gene, with the following exceptions: there were 35 cycles of amplification and the annealing temperature was set at 53.5C. For nested PCR, PRA46F and PREA1100R primers were used to produce fragments of rRNA gene, and these fragments were then used as a template for the second PCR with primers PARCH340F and PARCH519R. DGGE was performed with a DCode Universal Mutation Detection System (Bio-Rad Laboratories, Hercules, CA, USA). The amplified fragments of rRNA gene were loaded onto a 10% (w/v) polyacrylamide gel in 0.5 TAE (pH 8.0). Gel prepared with a denaturing gradient ranging from 30% to 60% was used for analysis of archaeal rRNA gene while gel prepared with a denaturing gradient ranging from 20% to 60% was used for analysis of bacterial rRNA gene. Electrophoresis was performed at 60C, first for 20 min at 25 V and then at 130 V for 8 h and 12 h for analysis of archaeal and bacterial rRNA gene, respectively. Representative clones from 16S rRNA gene libraries were analyzed simultaneously to determine the correlation between the bands detected in DGGE and the clones obtained in the rRNA gene libraries. Real-time quantitative PCR For detection and quantification of methanogens by real-time quantitative PCR, the TaqMan fluorogenic PCR system (Applied Biosystems) was used and the method was the same as reported previously (27). The primer/probe sets for the genera Methanosaeta, Methanosarcina, Methanoculleus and Methanospirillum were MS1b/SAE835R/SAE761TAQ, MB1b/SAR835R/SAR761TAQ, AR934F/MG1200b/MCU1023TAQ (27) and AR934F/MG1200b/MSP1025TAQ (5-GAATGATAGTC GGGATGAAGACTCTA-3, 10001025, E. coli position), respectively. The cloned 16S rRNA gene fragments, EA02 (AB092914), DA16 (AB092915), AA01 (AB092916), BA03 (AB092917), were

used as standards for the quantification of the genera Methanosaeta, Methanosarcina, Methanoculleus and Methanospirillum, respectively. Other analysis methods All of the following parameters of the culture broth in the reactor, except for total solids (TS), total volatile solids (TVS), suspended solids (SS) and volatile suspended solids (VSS), were analyzed in supernatants obtained after centrifugation at 10,000 rpm for 10 min. TS, TVS, SS and VSS were analyzed in accordance with standard methods (36). Soluble total organic carbon (TOC) was analyzed using a TOC auto analyzer (TOC-V; Shimadzu, Kyoto) according to the testing methods for industrial wastewater, JISK0102-1986 (36). Volatile fatty acids (VFAs) were analyzed as described previously (37). NH4+ and NO3 were measured using the HACH methods (HACH, Loveland, CO, USA) (38). Methane content of the biogas was measured using a gas chromatograph equipped with a thermal conductivity detector (TCD) (GC323; GL Sciences, Tokyo) and a packed column (Porapack N; GL Sciences). The protein concentration of culture broth was measured using the LowryFolin method (39). Nucleotide sequence accession numbers Archaeal and bacterial 16S rRNA gene sequences obtained in this study were deposited with the DDBJ and are available under the accession nos. AB175336 to AB175393.

RESULTS Continuous anaerobic degradation of BSA using anaerobic chemostats Continuous anaerobic degradation of BSA was carried out for about one year using synthetic wastewater containing BSA as the sole carbon and energy source at 37C. Synthetic wastewater without Ni2+ and Co2+ was fed into chemostat 1, while that with Ni2+ and Co2+ was fed into chemostat 2. The dilution rate (D) of chemostat 1 was varied in steps from 0.01 to 0.2, as shown on the schedule in Fig. 1. The concentrations of TOC, VFA, ammonia and VSS were stable at a dilution rate between 0.05 and 0.15 d1, indicating that chemostat 1 was at steadystate at the dilution rates (Fig. 2a). The critical dilution rate (Dc), where the system was able to maintain stable conditions, was 0.15 d1. The VSS concentration at a dilution rate of 0.15 d1 was maintained at about 1.6 g/l for about 270 d. The measured outputs of carbon and nitrogen accounted for about 100% of the inputs (Table 1 and 2), and the system was thought to be in the stable condition. High concentrations of VFAs (800012,000 mg/l) and ammonia (26003000 mg/l) were detected at this dilution rate (Tables 1 and 2). Under steady-state conditions at a dilution rate of 0.15 d1, the protein degradation ratio was approximately 75% and the main products of degradation were VFAs and ammonia (Fig. 2a). Negligible amounts of gas were produced. Acetic acid, propionic acid, butyric acid and iso-valeric acid accounted for most of the VFA produced (Table 1). The result suggested that acetogenesis and methanogenesis were completely inhibited by the lack of Ni2+ and Co2+. The dilution rate of chemostat 2 (with Ni2+ and Co2+) was varied in steps from 0.01 to 0.2, as shown on the schedule in Fig. 1. The concentrations of TOC, VFA, ammonia and VSS were stable at a dilution rate between 0.05 and 0.08 d1, indicating that chemostat 2 was at steady-state at the dilution rates (Fig. 2b). Based on the fermentation results, the Dc at which BSA could be converted to methane was only 0.08 d1 (Fig. 2B). The VSS concentration at a dilution rate of 0.08 d1

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

153

FIG. 2. Effect of dilution rate on TOC, VFA, NH4+, and VSS concentrations and gas production rate in chemostat 1 (a) and chemostat 2 (b). TABLE 1. The concentration of BSA and volatile fatty acids in influent and effluent, and the gas production in each chemostat Protein (g l1) Total VFA (g l1) Acetate (g l1) Propionate Butyrate (g l1) (g l1) N.D. 2.02 Isovalerate (g l1) N.D. 2.37 Gas production (ml d1) Inorganic carbon (g l1) Carbon in SS (g-C d1) 0.00 0.21 Total C per day (g-c d1) 2.435 2.600

Chemostat 1 (0.15 d1) 0.00 N.D. N.D. Influent 17.00 (2.30 g-C d1) 11.10 4.69 1.99 Effluent 4.25 (0.57 g-C d1) (1.48 g-C d1) Chemostat 2 (0.08 d1) 0.00 N.D. N.D. Influent 17.00 (1.22 g-C d1) 1.07 0.31 0.76 Effluent 1.70 (0.12 g-C d1) (0.07 g-C d1) N.D., Not detected. SS, Suspended solid. The calculated amounts of carbon are shown in parentheses.

0.5 (0.135 g-C d1) 120 1.05 (0.06 g-C d1) (0.28 g-C d1) 0.50 (0.072 g-C d1) 1780 0.78 (0.95 g-C d1) (0.11 g-C d1)

N.D. N.D.

N.D. N.D.

0.00 0.07

1.292 1.320

TABLE 2. The balance of nitrogen element of the BSA-fed chemostats Protein-N NH4+-N NO3-N (g-N d1) (g-N d1) (g-N d1) Chemostat 1 (0.15 d1) Influent 0.735 Effluent 0.183 Chemostat 2 (0.08 d1) Influent 0.392 Effluent 0.039 Solid compo- Total N nent-N (g-N d1) (g-N d1) 0.000 0.045 0.000 0.023 0.803 0.884 0.429 0.408

0.061 0.620 0.033 0.335

0.007 0.036 0.004 0.011

was about 1.7 g/l. The output and input of carbon and nitrogen were shown in Tables 1 and 2, and the outputs accounted almost 100% of inputs. Under steady-state conditions at this dilution rate, the protein degradation ratio was over 90% (Tables 1 and 2). The main products of the BSA degradation were biogas (about 1050 ml/l d) and ammonia (about 3000 mg/l), thus suggesting that acetogenesis and methanogenesis had occurred, and approximately 100% of the TOC was detected in the effluent and the biogas evolved. The methane concentration in the biogas was approximately 50%. Fluorescence in situ hybridization (FISH) The two chemostats constructed here showed different properties as described above. This difference would be caused by the

154

TANG ET AL.

J. BIOSCI. BIOENG.,

FIG. 3. Phase-contrast photomicrographs (a, c) and results of FISH (b, d) for microorganisms among the mesophilic anaerobic BSA digesters. Panels a and b show cells grown in chemostat 1. Panels c and d show cells grown in chemostat 2. The florescent probes used were a combination of ARC915 (Archaea, green) and EUB338 (Bacteria, red). The bar represents 10 mm.

difference of microbial communities of both chemostats. The microbial community analyses on the chemostats were therefore conducted. To determine the relative properties of bacterial and archaeal cells, Cy5-labeled EUB338 and Cy3labeled ARC915 were used simultaneously in the culture broth at dilution rates of 0.15 d1 for chemostat 1 and 0.08 d1 for chemostat 2. As shown in Fig. 3a and 3b, for chemostat 1, the number of cells hybridized with the ARC915 probe was small and almost all cells hybridized with the EUB338 probe, thus indicating that bacterial cells were absolutely dominant. Most of the cells hybridized with the EUB338 probe were rod shaped. In contrast, as shown in Fig. 3c and 3d, for chemostat 2, a significant number of rod-shaped, filamentous and coccoid cells were hybridized with the ARC915 probe, although the majority of cells hybridized with the EUB338 probe. Filamentous cells were dominant among the cells hybridized with the ARC915 probe. General phylogenetic analysis All the clones sequenced in two archaeral libraries (BSA1A and BSA2A) were affiliated with the phylum Euryarchaeota. Among the clones analyzed, 10 different sequences (OTUs) were found in BSA1A library and 9 OTUs in the BSA2A library (Table 3). Among the clones sequenced in the bacterial libraries, 18 OTUs were found in BSA1B library and 21 OTUs in the BSA2B library, and their classification were shown in Table 3.

Domain Archaea In the BSA1A library, 10 OTUs (23 clones) were obtained. All of them were assigned to the order Methanomicrobiales in the phylum Euryarchaeota (Fig. 4). Seven OTUs of 14 clones were closely related to Methanocelleus bourgensis with 9899% sequence similarities (BSA1A-01, one clone; BSA1A-02, one clone; BSA1A03, eight clones; BSA1A-04, one clone; BSA1A-05, one clone; BSA1A-06, one clone; BSA1A-07, one clone). Two OTUs of the eight clones were closely related to MethanoTABLE 3. Distribution of 16S rRNA gene clones detected in the culture broth of the BSA-fed chemostats Chemostat 1 Chemostat 2 (without Ni2+ and Co2+)a (with Ni2+ and Co2+)b Number of Number of Number of Number of OTUs clones OTUs clones 23 19 22 3 44 9 10 3 4 4 21 39 17 6 7 24 54

Taxon (phylum)

Archaea Euryarchaeota 10 Bacteria Firmicutes 11 Bacteroidetes 5 Proteobacteria 2 Unclassified Total (Bacteria) 18 a Dilution rate (d1) = 0.15 b Dilution rate (d1) = 0.08

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

155

FIG. 4. Phylogenetic tree showing the genetic relationships among the clones affiliated with the order Methanosarcinales and the order Methanomicrobials. The tree was constructed by the Neighbor-Joining method (34) using partial sequences of 16S rRNA gene. BSA1A and BSA2A refer to clones derived from chemostat 1 and chemostat 2, respectively. Numbers of clones with identical sequences are shown in parentheses. The bar represents two substitutions per 100 nucleotide positions. Bootstrap probabilities (36) are indicated at the branch nodes. The DDBJ/EMBL/GenBank accession numbers for reference strains and clones obtained in this study are shown in parentheses in this and other figures. The tree was rooted using Methanobacterium bryantii as the outgroup.

culleus chikugoensis with 98% sequence similarities (BSA1A-08, one clone; BSA1A-09, seven clones). In addition, one clone (BSA1A-10) was closely related to Methanospirillum hungatei with a 96% sequence similarity. In the BSA2A library, nine OTUs (39 clone) were obtained. Three OTUs (28 clones) were assigned to the order Methanosarcinales and six OTUs (11 clones) were assigned to the order Methanomicrobiales in the phylum Euryarchaeaota (Fig. 4). The most dominant sequence was closely related to Methanoseata concilii with a 100% sequence similarity (BSA2A-08, 25 clones). Five OTUs (10 clones) were closely related to Methanoculleus bourgensis with 9899% sequence similarities. One clone (BSA2A-06) was closely related to Methanospirillum hungatei with a 97% sequence similarity.

Domain Bacteria Phylum Firmicutes (Low G + C gram-positive bacteria) In the BSA1B library, 11 OTUs (19 clones, 43% of total clones in the library) were affiliated with the phylum Firmicutes. The OTUs were divided into five phylogenetic groups. Two groups, six OTUs (13 clone) and one OTU (one clone), were assigned to the Clostridium cluster XII (40) and Desulfotomaculum cluster I (41), respectively (Fig. 5). Two groups, one OTU (one clone) and two OTUs (two clones), were closely related to Clostridium cluster XIII and X (40), respectively (Fig. 5). Another group, one OTU (two clones), was assigned to a cluster containing clone PeH17 (42). Four OTUs (BSA1B-08 through 11, 11 clones) in Clostridium cluster XII were closely related to Clostridium sp. strain PO (the nucleotide sequence accession no. AJ002593), isolated from an anaerobic digester, which uses gelatin as a sole

156

TANG ET AL.

J. BIOSCI. BIOENG.,

FIG. 5. Phylogenetic tree showing the genetic relationships among the clones affiliated with the phylum Firmicutes (low G+C gram-positive bacteria). The tree was constructed by the Neighbor-Joining method (34) using partial sequences of 16S rRNA gene. The bar represents two substitutions per 100 nucleotide positions. Numbers in parentheses and numbers at the branch nodes are the same as those in Fig. 4. The tree was rooted using Arthrobacter globiformis as the outgroup.

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

157

source of carbon and energy (Javis, G.N., personal communication), with 98100% sequence similarities. OTUs BSA1B-06 and 07 in this cluster were closely related to Sporanaerobacter acetigenes with a 94% sequence similarity. OTU BSA1B-05 (one clone) was closely related to Sedimentibacter hydroxybenzoicus with a 96% sequence similarity. OTU BSA1B-01 (one clone) in the cluster related to Clostridium cluster X was closely related to Aminobacterium colombiense, a mesophilic asaccharolytic syntrophic anaerobe that degrades amino acid (17), with a 98% sequence similarity. Another OTU named BSA1B-02 in this cluster was closely related to Aminomonas paucivorans, a mesophilic asaccharolytic amino-acid-degrading bacterium (13), with a 92% sequence similarity. In the BSA2B library, 10 OTUs (17 clones, 32% of total clones in the library) were affiliated with the phylum Firmicutes. The phylogenetic positions of the OTUs in the BSA2B library were different from those of OTUs from the BSA1B library. The OTUs of the BSA2B library were divided into three phylogenetic groups. Five OTUs (nine clones) were closely related to Clostridium cluster X (40). One OTU (one clone) and 4 OTUs (seven clones) were

assigned to Clostridium cluster III (40) and a cluster containing clone ML635J-14 (43), respectively (Fig. 5). All 5 OTUs (BSA2B-01 through 05, nine clones) in the cluster related to Clostridium cluster X were closely related to Aminobacterium mobile, an asaccharolytic mesophilic syntrophic amino-acid degrading bacterium (18), with approximately 95% sequence similarity. Four OTUs (BSA2B-06 to 09) were not related to any pure cultured microorganisms, but were related to the uncultured clone ML635J-14 obtained from Mono Lake (44). Phylum Bacteroidetes Five OTUs (22 clones, 50% of total clones in library) of the BSA1B library were affiliated with the phylum Bacteroidetes. Of these, three OTUs (20 clones) were assigned to the order Bacteroidales (Fig. 6). OTUs BSA1B-12 and 13 were closely related to the uncultured clone vadin BC27, obtained from an anaerobic reactor treating wine distillation waste (45), with approximately 95% sequence similarity. The other two OTUs (BSA1B-15 and 16, two clones) were assigned to the order Sphingobacteriales. In contrast to the BSA1B library, only three OTUs (six clones, 11% of total clones in library) of the BSA2B library

FIG. 6. Phylogenetic tree showing the genetic relationships among the clones affiliated with the phylum Bacteroidetes. The tree was constructed by the Neighbor-Joining method (34) using partial sequences of 16S rRNA gene. The bar represents five substitutions per 100 nucleotide positions. Numbers in parentheses and numbers at the branch nodes are the same as those in Fig. 4. The tree was rooted using Spirochaeta africana as the outgroup.

158

TANG ET AL.

J. BIOSCI. BIOENG.,

FIG. 7. Phylogenetic tree showing the genetic relationships among the clones affiliated with the phylum Proteobacteria. The tree was constructed by the Neighbor-Joining method (34) using partial sequences of 16S rRNA gene. The bar represents two substitutions per 100 nucleotide positions. Numbers in parentheses and numbers at the branch nodes are the same as those in Fig. 4.

were affiliated with the phylum Bacteroidetes. OTUs BSA2B11 (one clone) and 12 (4 clones) were assigned to the order Bacteroidales and were closely related to the uncultured clone WCHB1-29 obtained from a hydrocarbon- and chlorinated-solvent-contaminated aquifer (46) (Fig. 6). Another OTU (BSA2B-13, one clone) formed a cluster with species of the order Sphingobacteriales. Phylum Proteobacteria Within the Proteobacteria, two OTUs (BSA1B-17 and 18, three clones, 7% of total clones in BSA1B) in the BSA1B library and 4 OTUs (BSA2B-14 through 17, seven clones, 13% of total clones in BSA2B) in the BSA2B library were detected (Fig. 7). OTU BSA1B-17 (one clone) was closely related to a sulfate-reducing bacteria, Desulfovibrio desulfuricans, which belongs to the class Deltaproteobacteria (47), with a 99% sequence similarity. OTU BSA1B-18 (2 clones) was closely related to Comamonas testosteroni, which belongs to the class Betaproteobacteria (47) with a 97% sequence similarity. Four OTUs (BSA2B-14 through 17, seven clones) in the BSA2B-library were closely related to Arcobacter skirrowi, which belongs to the class Epsilonproteobacteria (47), with approximately 98% sequence similarity. Unclassified clones Four OTUs (24 clones) in the BSA2B library were not affiliated with any phyla determined to date. The phylogenetic position of these in the domain Bacteria was related with the phyla Bacteroidetes and

Spirochaetes (Fig. 8), but was distinct from both. OTUs BSA2B-18 (one clone) and BSA2B-19 (one clone) showed a significant relationship with the phylum Bacteroidetes with a bootstrap value of 100. The other two OTUs, OTU BSA2B-21 (20 clone) and BSA2B-20 (two clones), were apparently distinct from the phylum. These OTUs were closely related to the uncultured clone R5p16 (nucleotide sequence accession no. AF482444) obtained from granular sludge concerning fatty acid oxidization, with a 99% and a 93% sequence similarity, respectively. Analysis of community structure by DGGE To evaluate the coverage of the microbial communities in the chemostats by our 16S rRNA gene library analysis, the community DNAs from the culture broths of chemostat 1 at a dilution rate of 0.15 d1 and chemostat 2 at a dilution rate of 0.08 d1 were analyzed by DGGE. Fragments of approximately 200 bp in the V3 region of archaeal and bacterial 16S rRNA gene were amplified from the community DNAs and clones obtained by the rRNA gene library analysis, separately, and used to DGGE. The position of bands derived from the community DNAs and rRNA gene clones were compared to consider the correlation. At least five and six bands were detected after DGGE using the amplified archaeal rRNA gene derived from the community DNA of chemostat 1 and chemostat 2, respectively (lanes 1 and 2 in Fig. 9). Among these bands, A2 and

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

159

FIG. 8. Phylogenetic tree showing the genetic relationships among the clones unclassified so far. The tree was constructed by the NeighborJoining method (34) using partial sequences of 16S rRNA gene. The bar represents five substitutions per 100 nucleotide positions. Numbers in parentheses and numbers at the branch nodes are the same as those in Fig. 4. The tree was rooted using Methanobacterium bryantii as the outgroup.

A3 were common to both chemostats. Band A2, which was the dominant band in the two systems, was correlated with OTUs BSA1A-01 to 07 and OTUs BSA2A-01 to 05, which were closely related to Methanoculleus bourgensis. Band A3 was correlated with OTU BSA1A-10 and BSA2A-06, which were closely related to Methanospirillum hungatei. Band A1, which was also predominant in chemostat 1, was correlated with OTUs BSA1A-08 and 09, which were closely related to Methanoculleus chikugoensis. Band A4, which was the most predominant band in chemostat 2, was correlated with OTUs BSA2A-07 through 09, which were closely related to Methanosaeta concilii. These observations supported the results of archaeal rRNA gene clone library analysis, which showed that the genus Methanoculleus was predominant in chemostat 1 and the genera Methanosaeta and Methanoculleus were predominant in chemostat 2. All the OTUs obtained in the analysis of archaeal libraries were detected as bands in DGGE. Two bands from chemostat 1 and three bands from chemostat 2 were not correlated to the rRNA gene clones, indicating the limited coverage of the rRNA gene clone libraries. However, the libraries could cover most of the archaeal communities, because all the bands with no correlation with the rRNA gene clones were not predominant.

At least 21 and 19 bands were detected after DGGE using the amplified bacterial rRNA genes derived from culture broths of chemostat 1 and chemostat 2, respectively (lanes 3 and 4 in Fig. 9). The band patterns for the two chemostats were quite different, indicating different community structures. For chemostat 1, two intense bands, B2 and B6, were detected. Band B2 was correlated with the predominant OTUs BSA1B-12 to 14 (20 clones), which were closely related to the clone vadin BC27 (45). Band B6 was correlated with the predominant OTUs BSA1B-08 to 11 (11 clones), which were closely related to Clostridium sp. strain PO. Band B7 was correlated with the OTUs BSA1B-04 (two clones) and 05 (one clone). Bands B1, B3, B4, B5, B8 and B9 were correlated with OTUs BSA1B-16 (one clone), 18 (two clones), 06 (one clone), 07 (one clone), 15 (one clone) and 02 (one clone), respectively. For chemostat 2, four intense bands, B10, B11, B12 and B18, were detected. Bands B10 and B11 were correlated with OTUs BSA2B-20 (two clones) and 21 (20 clones), which were closely related to the clone R5p16. The two bands were detected from a single clonal sequence, due to electrophoresis conditions. Band B12 was correlated with OTU BSA2B-14 to 17 (seven clones), which were closely related to Arcobacter skirrowi, and OTU BSA2B-18 (one clone) and 19 (one clone). A sin-

160

TANG ET AL.

J. BIOSCI. BIOENG.,

FIG. 9. Results of denaturing gradient gel electrophoresis (DGGE) for the fragments of 16S rRNA gene from Archaea (lanes 1 and 2) and Bacteria (lanes 3 and 4) in BSA-fed chemostats. Lanes 1 and 3 show results for chemostat 1. Lanes 2 and 4 show the results for chemostat 2. Bands are numbered as A1 to B19. The fragments of representative clones with approximate migratory positions in the DGGE profile are schematically illustrated beside the result of DNA samples and the ratio of clone in its library is shown in parenthesis.

gle band of the same position was detected from these OTUs, although the amplified sequence of OTUs BSA2B14 to 17 was different from BSA2B-18 and 19. Band B18 was correlated with OTU BSA2B-01 (one clone), 03 to 05 (seven clones), which were affiliated with Aminobacterium mobilis. Band B15 was correlated with OTU BSA2B-10 (one clone) and 12 (four clones). Band B19 was correlated with OTU BSA2B-06 to 07 (five clones). Bands B13, B14, B16 and B17 were correlated with OTU BSA2B-11 (one clone), 13 (one clone), 09 (one clone) and 02 (one clone), respectively. These observations supported the results of bacterial rRNA gene clone library analysis and almost all OTUs obtained in the bacterial libraries were detected as bands in DGGE. The trace level-bands, 12 bands from chemostat 1 and nine bands from chemostat 2, were not correlated to any clones in the rRNA gene libraries. These results indicated that the bacterial rRNA gene clone libraries could cover most of the bacterial communities in the chemostats.

Real-time quantitative PCR Real-time quantitative PCR experiments were performed in order to quantify the four methanogens using DNA extracted from the culture broths of chemostat 1 at a dilution rate of 0.15 d1 and chemostat 2 at a dilution rate of 0.08 d1. For chemostat 1, only the 16S rRNA gene of the genus Methanoculleus was detected with a copy number of approximately 1.5 106 copies/50 ng DNA (Table 4). For chemostat 2, rRNA genes of the genera Methanosaeta and Methanoculleus were detected, both at a copy number of approximately 2.5 106 copies/50 ng DNA. A relatively small amount of rRNA gene of the genus Methanosarcina was also detected in chemostat 2, although clones affiliated with Methanosarcina were not obtained in 16S rRNA gene clone libraries. In addition, 16S rRNA gene of the genus Methanospirillum were not detected in the two chemostats, although clones affiliated with Methanospirillum were obtained in 16S rRNA gene clone analysis. These data supported the results of 16S rRNA gene library and DGGE analysis, which confirmed

TABLE 4. Quantification of 16S rRNA gene of methanogens in the culture broth of the BSA-fed chemostats Primer/probe set Target organism Chemostat 1a MS1b/SAE835R/SAE761TAQ Methanosaeta spp. N.D. MB1b/SAR835R/SAR761TAQ Methanosarcina spp. N.D. AR934F/MG1200b/MCU1023TAQ Methanoculleus spp. 1.51 106 AR934F/MG1200b/MSP1025TAQ Methanospirillum spp. N.D. N.D., Not detected. The values represent the averages of experiments conducted in duplicates. Unit, 16S rRNA gene copies/50 ng-DNA. a Dilution rate (d1)= 0.15 b Dilution rate (d1)= 0.08 Chemostat 2b 2.36 106 1.39 105 2.77 106 N.D.

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

161

that, among the Archaea, the genus Methanoculleus and the genera Methanosaeta and Methanoculleus were predominant in chemostat 1 and chemostat 2, respectively. Our results of real-time quantitative PCR suggested that a significant population of the genus Methanoculleus (approximately 29% of total methanogens in chemostat 2) existed in chemostat 1, which showed almost no methane production ability. This discrepancy might be caused by the difference of population in the two chemostats. From 30 ml of culture broth of chemostats 1 and 2, 39.6 mg and 91.5 mg of total DNA were obtained, respectively. This means that microbial concentration in chemostat 2 was 23 times higher than that in chemostat 1. Because same amount of the extracted DNA was used for the real-time quantitative PCR experiment, the estimated value for chemostat 1 would be larger than those for chemostat 2 at a same volume of culture broth. DISCUSSION Two mesophilic anaerobic BSA-fed chemostats were constructed: one without the trace metals Ni2+ and Co2+ (chemostat 1) and the other with Ni2+ and Co2+ (chemostat 2). For chemostat 1, the major products of BSA degradation were volatile fatty acids and ammonia. Acetogenesis and methanogenesis did not occur in this chemostat. In constrast, acetogenesis and methanogenesis after BSA degradation proceeded in chemostat 2. The population of cells belonging to the domain Archaea was very small in chemostat 1, as demonstrated by our FISH experiments, while a significant number of archaeal cells were detected in chemostat 2. The archaeal cells in our chemostats were classified into three genera of methanogens, Methanoculleus, Methanospirillum and Methanosaeta, by analysis of 16S rRNA gene clonal sequences. Methanogens reportedly play an important role in acetogenesis from fatty acids by hydrogen removal (48) as well as in methanogenesis from H2 and CO2 and/or acetate. The inhibition of acetogenesis and methanogenesis in chemostat 1 could, therefore, account for the small populations of methanogens in the chemostat. Methanogens require Ni2+ and Co2+ for their growth and methanogenic activity (49). The lack of Ni2+ and Co2+ in chemostat 1 may have inhibited methanogen growth, thus resulting in small populations. The lack of a methanogen population in the chemostat would, therefore, cause the inhibition of methanogens-dependent reactions, such as acetogenesis and methanogenesis. In chemostat 1, two genera of hydrogenotrophic methanogens, Methanoculleus and Methanospirillum, were detected by the analysis of 16S rRNA gene clonal sequences. These genera could possibly act as hydrogen scavengers during amino acid degradation and/or acetogenesis. However, their metabolic functions among the total microbial flora of the chemostat could essentially be considered negligible because of their poor populations. In contrast, large populations of methanogenic archaea were detected in chemostat 2 and were classified into three genera; Methanoculleus, Methanospirillum and Methanosaeta. Among the three genera, Methanosaeta, an aceticlastic methanogen, and Methanoculleus, a hydrogenotrophic methanogen, were

found to be dominant. The genus Methanosaeta would play a role in mineralization of acetate produced by amino-acid degradation and subsequent acetogenesis. Hydrogen produced during amino-acid degradation and acetogenesis would be converted to methane by the genus Methanoculleus in this chemostat. In both chemostats, cells belonging to the phylum Firmicutes were predominant among the Bacteria. Nineteen (43% in BSA1B library) and 17 (31% in BSA2B library) rRNA gene clonal sequences were obtained from chemostat 1 and chemostat 2, respectively. Because most of the strains related to protein and/or amino acid degradation are affiliated with this phylum, the Firmicutes would play a central role in protein and/or amino acid degradation in our chemostats. Interestingly, our phylogenetic analysis of the rRNA gene clonal sequences affiliated with the Firmicutes showed that the major phylogenetic positions in the Firmicutes differed between clones from chemostat 1 and chemostat 2. Most of the clones affiliated with Firmicutes from chemostat 1 (6 OTUs, 13 clones) fell into the Clostridium cluster XII (42). The genus Sporanaerobacter was found to degrade certain amino acids via the Stickland reaction (15). Strains in the genus Sedimentibacter, which was phylogenetically related to this cluster, also reportedly carry out the Stickland reaction (16). In contrast, no clones from chemostat 2 were closely related to the Clostridium cluster XII. A significant number of clones affiliated with Firmicutes from chemostat 2 (five OTUs, nine clones) were assigned to the cluster containing the genus Aminobacterium. Aminobacterium mobile and Aminobacterium colombiense have been demonstrated to degrade certain amino acids in syntrophic association with hydrogenotrophic methanogens but not to perform the Stickland reaction (17, 18). The other four OTUs (seven clones) were found to form a cluster with uncultured clones ML635J-14 and ML635J-38, which were obtained from Mono Lake (44). The metabolic functions of the bacteria in these four OTUs are not understood, because these OTUs formed a cluster distinct from rRNA gene from any of the cultured strains in this phylum. These analyses on microbial communities suggest that the major amino acid degradation pathway in chemostat 1 is different from that in chemostat 2. The major pathway in chemostat 1 would be the Stickland reaction, which would be carried out by an amino acid degrader belonging to the Clostridium cluster XII, while bacteria closely related to the genus Aminobacterium, which is associated with hydrogenotrophic methanogens, would play a central role in amino acid degradation in chemostat 2. Twenty-two (50% in BSA1B library) and 6 (11% in BSA2B library) clones affiliated with the phylum Bacteroidetes were obtained from chemostat 1 and chemostat 2, respectively. No obvious phylogenetic differences between the two chemostats were observed. Because most of these clones formed a cluster distinct from the rRNA gene of cultured strains, their metabolic functions could not be elucidated. Some strains in this phylum, including Bacteroides fragilis, have been shown to have the ability to degrade proteins and amino acids (50, 51), although data on amino acid degradation by strains belonging to this phylum are currently very limited. These clones in our chemostats would

162

TANG ET AL.

J. BIOSCI. BIOENG.,

FIG. 10. Possible scheme of protein (BSA) degradation pathways and microorganisms detected in chemostats 1 and 2 as suggested by the results of this study.

play some role in protein and/or amino acids degradation. Twenty-two rRNA gene clones (41% in BSA2B library) found only in chemostat 2 formed a cluster with an uncultured clone R5p16 and could not be assigned to any phyla. Because the uncultured clone R5p16 was obtained from methanogenic granular sludge concerning fatty acid degradation, these clones in chemostat 2 would be likely to contribute to acetogenesis. The possible roles of microorganisms detected in our chemostats in protein degradation are summarized in Fig. 10. We demonstrated by microbial community analyses that the major amino acid degraders were different in the two chemostats. Surprisingly, no rRNA gene clones closely related to known amino acid degraders performing the Stickland reaction were obtained from chemostat 2, into which the trace metals Ni2+ and Co2+ were added and methane was the final product. In contrast, a significant number of clones related to known bacteria capable of syntrophic amino acid degradation with hydrogenotrophic methanogens were obtained from this chemostat. These findings suggest that syntrophic degradation of amino acids is favorable in chemostat 2 when compared with the Stickland reaction. It is considered that the Stickland reaction would dominate in environments rich in amino acids but that the reducing equivalents might be preferentially channeled to methanogenesis in environments with low supplies of amino acids and high methanogenic activity (52). In our chemostats, addition of the trace metals Ni2+ and Co2+ enabled methanogens to have higher populations in chemostat 2 when compared with chemostat 1. The higher methanogenic activity might lead to the dominance of syntrophic amino acid degraders in chemostat 2, thus resulting in syntrophic degradation of amino acids being the major pathway. Some bacteria that are able to degrade amino acids syntrophically, are able to perform the Stickland reaction. One of these bacteria, Caloramator proteoclasticus, was reported to use glycine as an electron acceptor and degrade amino

acids via the Stickland reaction when cultured under pure culture conditions. However, a large portion of the reducing equivalents were channeled to methanognesis in the case of co-culture with a hydrogenotrophic methanogen, Methanothermobacter thermautotrophicus strain Z245 (52). Moreover, the proteolitic activity of C. proteoclasticus increased in co-culture with M. thermautotrophicus (53). The syntrophic amino acid degradation would overcome the Stickland reaction in protein degradation in ecosystems where a significant number of hydrogenotrophic methanogens are present, although information on such comparative ecological aspects is very limited at present. In spite of the advantages for syntrophic degradation of amino acids, the growth rates of syntrophic amino acid degraders were significantly lower, with an apparent specific growth rate of 0.08 d1, as calculated from the Dc value of chemostat 2. In contrast, chemostat 1, in which bacteria performing the Stickland reaction dominated, exhibited a higher Dc value of 0.15 d1. It is believed to be difficult for amino acid degraders to grow rapidly in syntrophic association with hydrogenotrophic methanogens in a CSTR. In order to induce a higher rate of treatment for protein-rich wastewater and waste, a two-step methane fermentation process, including VFA production from proteins mainly by the Stickland reaction as a first step, and acetogenesis and methanogenesis as a second step, might be favorable. This is supported by several studies that found a two-stage anaerobic process more efficient for the treatment of proteinrich wastewater (20, 43). In conclusion, we cultured protein degrading anaerobic microbial communities using BSA-fed chemostat cultivation. Microbial community analysis suggested that the major pathway of amino acid degradation was via the syntrophic association of amino acid degraders and hydrogenotrophic methanogens in the presence of the trace metals Ni2+ and Co2+ and methane was the final product, and via the Stickland reaction in the absence of these metals and volatile

VOL. 99, 2005

MICROBIAL COMMUNITIES OF BSA-FED CHEMOSTAT CULTIVATIONS

163

fatty acid was the final product. Our results suggest the ecological importance and advantages of syntrophic degradation of amino acids in methanogenic ecosystems. ACKNOWLEDGMENTS
This work was conducted as Research and development of a highly efficient bioprocess for production of valuable chemicals from protein-containing biomass and organic waste as part of the International Co-operation Program for Global Environment Promotion which is handled by the Research Institute of Innovative Technology.

16.

17.

18.

REFERENCES
1. Ikbal, Morimura, S., Shigematsu, T., and Kida, K.: Bacterial populations and degradation of sludge components in a modified two-series digestion process including sludge liquefaction. Jpn. J. Water Treat. Biol., 38, 193201 (2002). 2. Gujer, W. and Zehnder, A. J. B.: Conversion processes in anaerobic digestion. Water Sci. Technol., 15, 127167 (1983). 3. Breure, A. M., Mooijman, K. A., and van Andel, J. G.: Protein degradation in anaerobic digestion: influence of volatile fatty acids and carbohydrates on hydrolysis and acidogenic fermentation of gelatin. Appl. Microbiol. Biotechnol., 24, 426431 (1986). 4. Fang, H. H. P., Chui, H. K., Li, Y. Y., and Chen, T.: Performance and granule characteristics of UASB process treating wastewater with hydrolyzed proteins. Water Sci. Technol., 30, 5563 (1994). 5. McInerney, M. L.: Anaerobic hydrolysis and fermentation of fats and proteins, p. 373409. In Zehnder, A. J. B. (ed.), Biology of anaerobic organisms. John Wiley, New York (1988). 6. Barker, H. A.: Amino acid degradation by anaerobic bacteria. Ann. Rev. Biochem., 50, 2340 (1981). 7. Elsden, S. R. and Hilton, M. G.: Amino acid utilization patterns in Clsotridial taxonomy. Arch. Microbiol., 123, 137 141 (1979). 8. Mead, G. C.: The amino acid fermenting Clostridia. J. Gen. Microbiol., 67, 4756 (1971). 9. Cato, E. P., Johnson, J. L., Hash, D. E., and Holdeman, L. V.: Synonymy of Peptococcus glycinophilus with Peptostreptococcus micros and electrophoretic differentiation of Peptostreptococcus micros from Peptecoccus magnus. Inst. J. Syst. Bacteriol., 33, 207210 (1983). 10. Drre, P., Spahr, R., and Andreesen, J. R.: Glycine fermentation via glycine reductase in Peptococcus glycinophilus and Peptococcus magnus. Arch. Microbiol., 134, 127135 (1983). 11. Whiteley, H. R.: Fermentation of amino acids by Micrococcus aerogenes. J. Bacteriol., 74, 324330 (1957). 12. Stams, A. J. M. and Hansen, T. A.: Fermentation of glutamate and other compounds by Acidaminobacter hydrogenoformas gen. nov., an obligate anaerobe isolated from black mud. Studies with pure cultures and mixed cultures with sulfate reducing and methanogenic bacteria. Arch. Microbiol., 137, 329337 (1984). 13. Baena, S., Fardeau, M.-L., Ollivier, B., Labat, M., Thomas, P., Garcia, J.-L., and Patel, B. K. C.: Aminomonas paucivorans gen. nov., sp. nov., a mesophilic, anaerobic, aminoacid-utilizing bacterium. Int. J. Syst. Bacteriol., 49, 975982 (1999). 14. Nanninga, H. J., Drent, W. J., and Gottschal, J. C.: Fermentation of glutamate by Selenomonas acidaminophila sp. nov. Arch. Microbiol., 147, 152157 (1987). 15. Hernandez-Eugenio, G., Fardeau, M.-L., Cayol, J.-L., Patel, B. K. C., Thomas, P., Macarie, H., Garcia, J.-L., and Ollivier, B.: Sporanaerobacter acetigenes gen. nov., sp. nov., 19.

20. 21. 22.

23. 24.

25.

26.

27.

28.

29.

30.

31.

a novel acetogenic, facultatively sulfur-reducing bacterium. Int. J. Syst. Evol. Microbiol., 52, 12171223 (2002). Breitenstein, A., Wiegel, J., Haertig, C., Weiss, N., Andreesen, J. R., and Lechner, U.: Reclassification of Clostridium hydroxybenzoicum as Sedimentibacter hydroxybenzoicus gen. nov., comb. nov., and description of Sedimentibacter saalensis sp. nov. Int. J. Syst. Evol. Microbiol., 52, 801807 (2002). Baena, S., Fardeau, M.-L., Labat, M., Ollivier, B., Thomas, P., Garcia, J.-L., and Patel, B. K. C.: Aminobacterium colombiense gen. nov. sp. nov., and amino acid-degrading anaerobe isolated from anaerobic sludge. Anaerobe, 4, 241 250 (1998). Baena, S., Fardeau, M.-L., Labat, M., Ollivier, B., Garcia, J.-L., and Patel, B. K. C.: Aminobacterium mobile sp. nov., a new anaerobic amino acid-degrading bacterium. Int. J. Syst. Evol. Microbiol., 50, 259264 (2000). Amann, R. I., Ludwig, W., and Schleifer, K. H.: Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev., 59, 143169 (1995). Fang, H. H. P. and Yu, H. Q.: Mesophilic acidification of gelatinaceous wastewater. J. Biotechnol., 93, 99108 (2002). Ramsay, I. R. and Pullammanappallil, P. C.: Protein degradation during anaerobic wastewater treatment: derivation of stoichiometry. Biodegradation, 12, 247257 (2001). Tommaso, G., Varesche, M. B., Zaiat, M., Vazoller, R. F., and Foresti, E.: Morphological observation and microbial population dynamics in anaerobic polyurethane foam biofilm degrading gelatin. Brazil. J. Chem. Eng., 19, 287292 (2002). Yu, H. Q. and Fang, H. H. P.: Acidogenesis of geletin-rich wastewater in an upflow anaerobic reactor: influence of pH and temperature. Water Res., 37, 5566 (2003). Amann, R. I.: In situ identification of micro-organisms by whole-cell hybridization with rRNA-targeted nucleic acid probes, section 3.3.6, p. 115. In Akkermans, A. D. L., Elsas, J. D., and de Bruijn, F. J. (ed.), Molecular microbial ecology manual. Kluwer Academic Publishers, The Netherlands (1995). Amann, R. I., Binder, B. J., Olson, R. J., Chisholm, S. W., Devereux, R., and Stahl, D. A.: Combination of 16S rRNAtargeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl. Environ. Microbiol., 56, 19191925 (1990). Stahl, D. A. and Amann, R. I.: Development and application of nucleic acid probes, p. 205248. In Stackebrandt, E. and Goodfellow, M. (ed.), Nucleic acid techniques in bacterial systematics. John Wiley & Sons, New York (1991). Shigematsu, T., Tang, Y. Q., Kawaguchi, H., Ninomiya, K., Kijima, J., Kobayashi, T., Morimura, S., and Kida, K.: Effect of dilution rate on structure of a mesophilic acetatedegrading methanogenic community during continuous cultivation. J. Biosci. Bioeng., 96, 547558 (2003). Maidack, B. L., Cole, J. R., Lilburn, T. G., Parker, C. T., Jr., Saxman, P. R., Farris, R. J., Garrity, G. M., Olsen, G. J., Schmidt, T. M., and Tiedje, J. M.: The RDP-II (Ribosomal Database Project). Nucleic Acids Res., 29, 173174 (2001). Altschul, S. F., Madden, T. L., Schffer, A. A., Zhang, Z., Miller, W., and Lipman, D. J.: Gapped BLAST and PSIBLAST: a new generation of protein database search programs. Nucleic Acids Res., 25, 33893402 (1997). Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., and Higgins, D. G.: The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res., 24, 48764882 (1997). Kumar, S., Tamura, K., Jakobsen, I. B., and Nei, M.: MEGA2: molecular evolutionary genetics analysis software. Bioinformatics, 17, 12441245 (2001).

164

TANG ET AL.

J. BIOSCI. BIOENG.,

32. Saitou, N. and Nei, M.: The Neighbor-Joining method: a new method for constructing phylogenetic trees. Mol. Biol. Evol., 4, 406425 (1987). 33. Kimura, M.: A simple method for estimating evolutionary rates of base substitution through comparative studies of nucleotide sequences. J. Mol. Evol., 16, 111120 (1980). 34. Felsenstein, J.: Confidence limits of phylogenesis: an approach using the bootstrap. Evolution, 39, 783791 (1985). 35. vres, L., Forney, L., Daae, F. L., and Torsvik, V.: Distribution of bacterioplankton in meromictic Lake Slenvannet, as determined by denaturing gradient gel electrophoresis of PCR-amplified gene fragments coding for 16S rRNA. Appl. Environ. Microbiol., 63, 33673373 (1997). 36. Namiki, H.: Testing methods for industrial wastewater, JIS K 0102-1986. Japanese Standards Assosiation, Tokyo (1986). 37. Kida, K., Morimura, S., and Sonoda, Y.: Accumulation of propionic acid during anaerobic treatment of distillery wastewater from barely-shochu making. J. Ferment. Bioeng., 75, 213216 (1993). 38. Yeoh, B. G., Morimura, S., Ikbal, Shigematsu, T., Razreena, A. R. P., and Kida, K.: Simultaneous removal of TOC compounds and NO3 in a combined system of chemical and biological processes for the treatment of wastewater from rubber threat manufacturing. Jpn. J. Water Treat. Biol., 38, 5767 (2002). 39. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.: Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193, 265275 (1951). 40. Collins, M. D., Lawson, P. A., Willems, A., Cordoba, J. J., Fernandez-Garayzabal, J., Garcia, P., Cai, J., Hippe, H., and Farrow, J. A. E.: The phylogeny of the genus Clostridium: proposal of five new genera and eleven new species combinations. Int. J. Syst. Bacteriol., 44, 812826 (1994). 41. Stackebrandt, E., Sproer, C., Rainey, F. A., Burghardt, J., Puker, O., and Hoppe, H.: Phylogenetic analysis of the genus Desulfotomaculum: evidence for the misclassification of Desulfotomaculum guttoideum and description of Desulfotomaculum orientis as Desulfosporosinus orientis gen. nov., comb. nov. Int. J. Syst. Bacteriol., 47, 11341139 (1997). 42. Egert, M., Wagner, B., Lemke, T., Brune, A., and Friedrich, M. W.: Microbial community structure in midgut and hindgut of the humus-feeding larva of Pachnoda ephippiata (Coleoptera: Scarabaeidae). Appl. Environ. Microbiol., 69, 66596668

(2003). 43. Hulshoff Pol, L. W. and Lettinga, G.: New technologies for anaerobic wastewater treatment. Water Sci. Technol., 18, 41 53 (1986). 44. Humayoun, S. B., Bano, N., and Hollibaugh, J. T.: Depth distribution of microbial diversity in Mono Lake, a meromictic soda lake in Califonia. Appl. Environ. Microbiol., 69, 10301042 (2003). 45. Godon, J.-J., Zumstein, E., Dabert, P., Habouzit, F., and Moletta, R.: Molecular microbial diversity of an anaerobic digestor as determined by small-subunit rDNA sequence analysis. Appl. Environ. Microbiol., 63, 28022813 (1997). 46. Dojka, M. A., Hugenholtz, P., Haack, S. K., and Pace, N. R.: Microbial diversity in a hydrocarbon- and chlorinatedsolvent-contaminated aquifer undergoing intrinsic bioremediation. Appl. Environ. Microbiol., 64, 38693877 (1998). 47. Garrity, G. M. and Holt, J. G.: The road map to the manual. p. 119166. In Boone, D. R., and Castenholz, R. W. (ed.) Bergeys manural of systematic bacteriology, 2nd ed., vol. 1. Springer, New York (2001). 48. Schink, B.: Energetics of syntrophic cooperation in methanogenic degradation. Microbiol. Mol. Biol. Rev., 61, 262280 (1997). 49. Kida, K., Shigematsu, T., Kijima, J., Numaguchi, M., Mochinaga, Y., Abe, N., and Morimura, S.: Influence of Ni2+ and Co2+ on methanogenic activity and the amounts of coenzymes involved in methanogenesis. J. Biosci. Bioeng., 91, 590595 (1999). 50. Gibson, S. A. and Macfarlane, G. T.: Characterization of proteases formed by Bacteroides fragilis. J. Gen. Microbiol., 134, 22312240 (1988). 51. Gibson, S. A. and Macfarlane, G. T.: Studies on the proteolytic activity of Bacteroides fragilis. J. Gen. Microbiol., 134, 1927 (1988). 52. Schink, B. and Stams, A. J. M.: Syntrophism among prokaryotes. In Dworkin, M. (ed.), The Prokaryotes: an evolving electronic resource for the microbiological community, 3rd ed., release 3.8, http://link.springer-ny.com/link/service/books/ 10125/. Springer-Verlag, New York (January 31, 2002). 53. Tarlera, S. and Stams, A. J. M.: Degradation of proteins and amino acids by Caloramator proteoclasticus in pure culture and in coculture with Methanobacterium thermoformicicum Z245. Appl. Microbiol. Biotechnol., 53, 133138 (1999).