J. Gen. Appl. Microbiol., 48, 125–133 (2002) Full Paper Isolation and characterization of a Gram-positive polyphosphate-accumulating bacterium Shin Onda and Susumu Takii* Department of Biological Sciences, Graduate School of Science, Tokyo Metropolitan University, Hachioji, Tokyo 192–0397, Japan (Received October 31, 2001; Accepted March 9, 2002) A Gram-positive polyphosphate-accumulating bacterium was isolated from phosphate-removal activated sludge using pyruvate-supplemented agar plates. The isolate was oval or coccobacilli 0.7؋0.5–1.0 mm) that occurred singly, in pairs or irregular clumps. Polyphosphate granules–0.4) in the cells were observed by toluidine blue staining. The pure culture of the isolate rapidly took up phosphate (9.2 mg-P/g-dry weight) in the 3-h aerobic incubation without organic substrates, after anaerobic incubation with organic substrates containing casamino acids. When acetate was the sole carbon source in the anaerobic incubation, the isolate did not remove phosphate. These physiological features of the isolate were similar to those of Microlunatus phosphovorus. However, unlike M. phosphovorus the P-removal ability of the isolate was relatively low and was not accelerated by repeating the anaerobic/aerobic incubation cycles. Phylogenetic analysis and comparison of several characteristics showed that the isolate was identified as Tetrasphaera elongata which was recently proposed as a new polyphosphate-accumulating species isolated from activated sludge. As the isolate contained menaquinone (MK)-8(H4) as the predominant iso- prenoid ubiquinone, it may be significantly responsible for phosphate removal, because MK- 8(H4) has reportedly been found in fairly high proportions in many phosphate-removing activated sludges. Key Words——activated sludge; menaquinone-8(H4); phosphate-removal; polyphosphate; Tetra- sphaera elongata Introduction ganic pollutants from wastewater. Since phosphate in wastewater causes eutrophication, phosphate-removal Activated sludge systems are widely used for treat- (P-removal) is important to prevent water pollution. In ment of municipal sewage and industrial wastewater. the EBPR system, microbes called polyphosphate-ac- An enhanced biological phosphorus removal (EBPR) cumulating organisms (PAOs) are considered to be re- system is one of the activated sludge systems modi- sponsible for phosphate removal. Under anaerobic fied by attaching an anaerobic tank in front of the aera- conditions, PAOs take up carbon substrates (e.g., ac- tion tank to efficiently remove phosphate along with or- etate, propionate) and store them as a form of polyhy- droxyalkanoate (PHA) while releasing phosphate, uti- lizing polyphosphate (poly-P) in their cells as an en- * Address reprint requests to: Dr. Susumu Takii, Department of Biological Sciences, Graduate School of Science, Tokyo Met- ergy source. Under subsequent aerobic conditions, uti- ropolitan University, 1–1 Minami-ohsawa, Hachioji, Tokyo 192– lizing the carbon storage in their cells, PAOs take up 0397, Japan. more phosphate than they released under anaerobic E-mail: [email protected] conditions and accumulate it in the form of poly-P 126 ONDA and TAKII Vol. 48 (Comeau et al., 1986; van Groenestijn et al., 1987; Ko- and Morikawa, 1999). rtstee et al., 2000; Mino et al., 1998). Acinetobacter spp. were first investigated as PAOs Materials and Methods possibly responsible for P-removal (Kortstee et al., 2000; Mino et al., 1998). Microlunatus phosphovorus Reactor and activated sludge sample. An acti- (Nakamura et al., 1995; Ubukata and Takii, 1997) and vated sludge sample was collected from the outlet of Lampropedia sp. (Stante et al., 1997) were also iso- the aeration tank of a laboratory-scale anaerobic/aero- lated as PAOs which showed a similar ability to metab- bic continuous flow reactor. The reactor consisted of olize phosphate to P-removal activated sludges. How- an anaerobic tank (5 L working volume) and an aero- ever, some of their characteristics are different from bic tank (2 L working volume). The artificial sewage the P-removal activated sludges, and culture-indepen- fed was composed of (mg/L): CH3COOH (100), dent methods showed that they did not dominate in P- C2H5COOH (75), peptone (140), yeast extract (60), removal activated sludges (Bond et al, 1999; Crocetti H3PO4 (15.5 as P), KHCO3 (65), NaHCO3 (100), et al., 2000; Hesselmann et al., 1999; Hiraishi et al., MgSO4 ·7H2O (100), CaCl2 ·5H2O (25) and NaCl (100). 1998; Kortstee et al., 2000; Mino et al., 1998). The influent was loaded at 30 L/day, and the return Culture-independent methods such as molecular sludge was 120 L/day (Maejima and Matsuo, 1997). At methods and quinone profiling have been used to in- sampling time, mixed liquor suspended solid (MLSS) vestigate bacterial communities in P-removal activated and mixed liquor volatile suspended solid (MLVSS) sludges. Recently, Hesselmann et al. (1999) and Cro- were 5,160 mg/L and 3,790 mg/L, respectively. The pH cetti et al. (2000) reported that a bacterium closely re- was 6.8 in the anaerobic tank, 7.6–8.2 in the aerobic lated to Rhodocyclus, named Candidatus Accu- tank. The concentration of PO4-P in the effluent was mulibacter phosphatis (Hesselmann et al., 1999), was 0.2 mg/L and the total phosphorus content of the responsible for P-removal by cloning of 16S rDNA and sludge was 8.4% (Onda et al., 2002). fluorescent in situ hybridization (FISH). In our previous Pure isolation and media. The 10-ml activated report (Onda et al., 2002) this bacterium was also sludge sample was diluted with 40 ml of pyrophos- shown to be the dominant PAO in a laboratory-scale phate solution (sodium tripolyphosphate 5 mg/L, NaCl anaerobic/aerobic continuous flow reactor by using de- 8.5 g/L) and homogenized (11,500 rpm) with a waring naturing gradient gel electrophoresis of PCR-amplified blender (Ace Homogenizer: Nihon Seiki Co., Ltd., 16S rDNA, quinone profiling, and FISH. However, it Tokyo, Japan) for 5 min. One milliliter of sludge sus- has never been successfully isolated, so far as we pension was diluted with 19 ml pyrophosphate solution know. Culture-independent methods also suggested and homogenized again for 10 min. The sludge sus- that actinobacteria having menaquinone (MK)-8(H4) pension was serially diluted with the pyrophosphate was another important member of the EBPR microbial solution and spread on agar plates. For isolation and community (Hiraishi et al., 1998). Recently, poly-P ac- growth, R2A (Reasoner and Geldreich, 1985), 1/2PYG cumulating Tetrasphaera spp. were isolated and de- supplemented with pyruvate (Sugitate and Morikawa, scribed from P-removal activated sludges (Hanada et 1999), and artificial sewage supplemented with pyru- al., 2002; Maszenan et al., 2000). However, these vate (ASP) media were used. The compositions of the strains have not been reported to show characteristics media were as follows: R2A medium (g/L); yeast ex- similar to the phosphate metabolism in relation to P-re- tract (0.5), peptone (0.5), casamino acids (Difco, De- moval activated sludges. troit, MI, USA) (0.5), glucose (0.5), soluble starch (0.5), To better understand EBPR systems, isolation of sodium pyruvate (0.3), K2HPO4 (0.3) and MgSO4 · dominant PAOs is important. In this experiment we 7H2O (0.05); 1/2PYG with pyruvate medium (g/L); pep- tried to isolate and characterize new PAOs from the tone (1), yeast extract (0.5), glucose (0.25) and laboratory-scale reactor described above (Onda et al., sodium pyruvate (0.15); ASP medium (mg/300 ml); 2002). For isolation, pyruvate-supplemented agar sodium acetate (50), sodium propionate (50), yeast ex- plates were used in this study, because it was reported tract (50), casamino acids (Difco) (100), sodium pyru- that the addition of sodium pyruvate to agar media vate (45), activated sludge extract (60 ml), MgSO4 · greatly increased the numbers of colonies on agar 7H2O (15), CaCl2 ·2H2O (5), KH2PO4 (50) and NaCl plates from some environmental samples (Sugitate (100). The media for the agar plates were supple- 2002 A polyphosphate-accumulating bacterium 127 mented with 1.5% agar. All media were adjusted to the Gram-staining and the toluidine blue staining. PHA pH 7.5. After the inoculation, the agar plates were in- granules were stained with Nile blue according to the cubated at 28°C for about three weeks. Colonies ap- protocol of Ostle and Holt (1982) and were examined pearing on the agar plates were picked up and purified under an epifluorescent microscope (BH-2) with filter by repeated streaking. set BH2-DMU. Cultivation of isolates and P-uptake experiments. Physiological and biochemical characteristics. Ni- Isolates were aerobically grown at 20°C or 30°C in trate reduction, gelatin hydrolysis, indole production, shaking flasks (500 ml) plugged with cotton plugs. The catalase and oxidase were examined according to culture media were the same in composition as the Smihert and Krieg (1981) using R medium of Martin et isolation media. To obtain sufficient cells for P-uptake al. (1997) as the basal medium. The GϩC content of experiments, organic substrates (casamino acids 1.2 DNA was determined by the HPLC method (Suzuki et g/L, acetate 1.2 g/L and yeast extract 0.4 g/L) were al., 1987). Menaquinones were analyzed by reverse added to the APS medium. phase HPLC (Yamada and Kuraishi, 1982). The ability of the isolates to remove phosphate was Total cell counting. Total cell numbers of the acti- tested under similar conditions to those in EBPR sys- vated sludge sample and liquid cultures were deter- tems as follows. Organic substrates (mixture of mined by acridine orange staining. The sample was di- casamino acids 1.2 g/L, acetate 1.2 g/L and yeast ex- luted with the pyrophosphate solution and dispersed tract 0.4 g/L) were first added to their cultures at the by sonication (60 s, 5 W/ml, Sonicator 5202; Ohtake stationary phase, and the gas phase was replaced Works, Tokyo, Japan).