Recently, biological mechanisms, in a process known as enhanced biological phosphorus removal (EBPR), have tended to be used for phosphorus (P) removal from wastewater, rather than chemical precipitation. The mechanism of EBPR occurs under anaerobic conditions, and involves phosphate accumulating organisms (PAOs) taking up volatile fatty acids (VFAs) and converting them into poly-hydroxyl-alkanoates (PHAs) by releasing P; whereas, under aerobic conditions they incorporate more P than is released using PHAs as an energy source[1, 2]. However, our understanding of the properties of the bacteria involved in P removal is quite sparse. It has been reported that the bacterial community structure of phosphate removing sludge was different from those of non-phosphate removing sludges[3], and an analysis of the bacterial community in phosphate removing sludge revealed that PAOs, such as Rhodocyclus sp.[4, 5], were dominant in the efficient P-removal phase of EBPR system. During the performance of inefficient EBPR, the occurrence of bacteria with a distinctive morphology, named G-bacteria by Cech and Hartman[6], and later glycogen accumulating nonpolyphosphate organisms (GAOs), based on their glycogen accumulating ability, have also been frequently observed [1]. Most recently, these bacteria occurred in clumps and packages of tetrads in EBPR systems from several countries, and have been named tetrade forming organisms (TFOs), based on their similar cell morphologies, as confirmed by light microscopy. Some of these TFOs were identified using molecular techniques, such as fluorescence in situ hybridization (FISH), as Amaricoccus sp.[7], and 16S rRNA sequencing evidence suggests that TFOs belong to various eubacterial groups[8]. Conversely, some Gram-positive bacteria also exhibited microscopically similar cell morphologies to TFOs, known as inhibitors of P removal, but they could accumulate polyphosphate[9], similarly to other PAOs. This implies that the cell form arranged in clumps and packages of tetrads may not be a unique characteristic of specific bacteria, but maybe just a temporary phenomenon appearing under unknow conditions during the process of EBPR system.

Therefore, it is important to check the status of TFOs if the cells that form into clumps and packages of tetrads remain unchanged, even when their ecological conditions, such as the presence /absence of oxygen, are varied

The objective of this study was to characterize the TFOs that frequently occur during the deteriorated EBPR process of a wastewater treatment plant, including their isolation

Bacteria arranged in packages of tetrads (TFOs) were obtained from a dairy facility, Wisconsin Dairy State Cheese (WDSC) wastewater treatment plant (WWTP), located in Wisconsin, USA, where those bacteria have often been observed, not only in mixed liquor, but also in effluent. Wastewater discharged from the cheese production facility is initially equalized in a 30,000 gallon tank, and then fed into a 450,000 gallon sequencing batch reactor (SBR) on a daily basis. This wastewater has a chemical oxygen demand (COD) in the range 1,000-5,000 mg/L, and the concentration of NO3-N is commonly below 10 mg/L, but sometimes jumped up to 40 mg/L. The total phosphorus (TP) of the influent varied from 20 to 80 mg/L. Overall, the wastewater has a high nutrient concentration, but this fluctuates widely[2]

TFOs floating in effluent were sampled, filtered using 3.0 μm polycarbonate filters (Millipore, USA) to separate the TFOs from the other bacteria and subsequently treated for scanning electron microscopy (SEM; Hitachi S-4300; Hitachi Ltd., Tokyo, Japan) and observed. After the microscopy, the elemental composition of intact TFOs, with extracellular polymeric substance (EPS) and aerobically cultured TFOs (short-rod), were analyzed using a backscatter detector for compositional contrast (electron dispersive X-ray analyzer [EDS]; Jeol model JSM-5800LV, Japan)

After separating the TFOs obtained from effluent using filtration (3.0 μm), the TFOs and filtrate were checked for purity under a phase contrast microscope, and the TFOs then aerobically cultured using the medium[10] in a shaking incubator, and anaerobically cultured in a candle jar with a Gas Pak (BBL, England) at 30℃ as long as the cultured TFOs indicated a sufficient cell density for further DNA extraction

While the DNA of the aerobically cultured TFOs that had converted into unicellular short-rods (cultured TFOs) was extracted by the previously described protocol [8], the DNA from intact TFOs surrounded by large amounts of EPS required pretreatment prior to the extraction of DNA: TFOs were first ultra sonicated for 3-5 hours to free the EPS layer, then cell lysis, followed by phenol-chloroform extraction, and finally ethanol precipitation. This DNA preparation was used as a DNA template in a polymerase chain reaction (PCR) performed with 1×PCR buffer (Gibco BRL, Gaithersburg, MD, USA), containing 200 μM each of the deoxynucleoside triphosphates, 1.5 mM MgCl2, and 0.1 μM of each primer (Operon Technologies, Inc., Alameda, CA, USA), and 2.5 U of Taq polymerase (Pharmacia Biotech Inc., Piscataway, NJ, USA) in a final volume of 100 μL. For amplification of the 16S rDNA for denaturing gradient gel electrophoresis (DGGE), the forward primer; 968F (5'-AACGCGAAGAACCTTAC- 3') with a GC clamp (5'-CGCCCGGGG CGCGCCCCGGGCGGGG CGGGGGCACGGGGGG-3'), and the reverse primer; 1392R (5'- ACGGGCGGTGTGTAC-3'), were used. The PCR conditions consisted of preheating at 94°C for 5 minutes, with a subsequent 30 cycles of denaturation (1 minute at 94°C), annealing (1 minute at 40°C) and extension (1.5 minutes at 72°C), with a final extension at 72°C for 10 minutes. Amplified DNA was verified by electrophoresis of 2 μL of the PCR product on a 0.5% agarose gel prior to the DGGE analysis.

Using the PCR products, DGGE was performed to Confirm if the intact TFO_s, as well as the aerobically cultured and subsequently converted into unicellular short-rod from TFOs, were the same organism. The DGGE conditions were optimized to the PCR products and those product were separated using a Dcode Universal Mutation Detection System (Bio-Rad Laboratories, Hercules, CA, USA) at 180 V and 60°C for 4 hours. Samples were loaded onto a 6% polyacrylamide gel (acrylamide; N, N'-methylene- bisacrylamide at a ratio of 37.5:1 [Bio-Rad]) in 1 ×TAE buffer (20 mM Tris acetate [pH 7.4], 10 mM sodium acetate and 0.5 mM disodium EDTA). The denaturing gradient in the gel was generated by mixing two stock solutions of 6% polyacrylamide, containing 40/60% denaturing agent: the 100% denaturing agent was 7 M urea and 40% (w/v) formamide. After electrophoresis, the gel was stained with silver stain (1g AgNO₃/L) and photographed with a charge coupled device camera.

The eluted 16S rDNA band from the DGGE gel was amplified by PCR; the primers for the DGGE were as described above, and purified for further cloning and sequencing. Cleaned PCR fragments were ligated into PCR 4-TOPO plasmid vectors, as prescribed by the manufacturer (TOPO Cloning Kit for Sequencing; Invitrogen, Carlsbad, CA, USA). Ligated plasmids were transformed into Top 10 one shot competent cells, also as prescribed by the manufacturer. Using S.N.A.P. mini-prep kits (Invitrogen), the plasmid DNA was removed from competent cells by centrifugation for 5 min at 20,000 rpm

The 3 clones obtained from both intact TFOs and aerobically cultured TFOs were partially sequenced (-410 bp) and compared to available 16S rRNA sequences in GenBank using the NCBI Blast program. Using the neighbor-joining method[11] a phylogenetic tree was constructed

To identify the intact TFOs floating in the effluent, as well as those aerobically cultured (unicellular short-rods), FISH (fluorescent in situ hybridization) was performed using gene probes for α-, β- and γ-subclasses of proteobacteria[12]. In the case of the intact TFOs, due to their thick EPS the gene probe could not penetrate into the cell, so intact TFOs with EPS were first ultrasonicated for 3-5 hours to free the EPS layer, and pretreated with lysozyme[7] prior to performing the FISH

As clearly shown in the phase-contrast and SEM micrographs in Fig. 1, the TFOs obtained from the effluent of a SBR treatment plant, with high nutrient strength, were surrounded by a thick layer of EPS. Under phase-contrast microscopy, the TFOs appeared as 4er cells, but in fact, 4er clustered cells were again arranged in packages of tetrads to form 8er or 32er clustered cells (Fig. 1). The major function of the EPS layer is unknown. However, as previously reported[13], that EPS layer functions at least as a permeability barrier, allowing the passage of low molecular substances, while excluding large molecules. As performing FISH for their identification was very difficult, even with pretreatment to allow penetration of the gene probes into the cells (data not shown), while aerobically cultured TFOs (unicellular short-rods) without the EPS layer could be identified as γ-subclass Proteobacteria using FISH. The results of the elemental analyses by EDS supported that the EPS layer on the TFOs may have inhibited the function of the uptake of phosphate (Table 1). The C:O ratio of the intact TFOs was around 2:1 (C-K; 60%, O-K; 35.30%), such as with the lipid-like compound, poly-β-hydroxybutyrate (PHB[C₄H₆O₂]n), known as a carbon source for biological P removal, while the ratio of the cultured TFOs (unicellular short-rods) reached around 3:1 (C-K; 66.98%, O-K; 22.63%). Also, the elemental composition of the cultured TFOs (unicellular short-rods) differed from those of intact TFOs. Interestingly, when TFOs obtained from the effluent were stained for PHB and Neisser, the EPS layer was positive stained for both PHB and Neisser, according to Jenkins and coworkers[14]. Further detailed study for the EPS layer is needed. Moreover, TFOs frequently occurred in the effluent with lower phosphate content (1.28%) than the cultured TFOs (unicellular short-rods) without the EPS layer (2.91%).

[Fig. 1.] Phase-contrast micrograph of tetrade forming organisms (TFOs), with a thick EPS layer (left, ×1,000), and SEM micrographs of TFOs not arranged in 4er clustered cell form, but 8er or 32er clustered cell forms (right, ×10,000). EPS: extracellular polymeric substance, SEM: scanning electron microscopy.

[Table 1.] Comparative analyses of the elemental composition of intact TFOs and aerobically cultured unicellular short rods using EDS

Comparative analyses of the elemental composition of intact TFOs and aerobically cultured unicellular short rods using EDS

Furthermore, the TFOs that occurred in WDSD were able to be cultured aerobically, but not anaerobically. This suggests that the TFOs in WDSD may be obligate aerobic bacteria, and the cell morphology of those TFOs in clumps and packages of tetrads is not unique, but just a temporary phenomenon appearing under anaerobic conditions during the deteriorated P removal process of an EBPR system. The thick EPS layer produced might then play an important role for their survival, finally resulting in distinctive TFO cell forms

As seen in the DGGE profile (Fig. 2), the 16S rDNA extracted from the intact TFOs floating in the effluent of the WDSD, and that of the aerobically cultured TFOs (unicellular short rods) demonstrated the same banding patterns. This implies that both of the bacteria could be identical because DGGE enables DNA fragments of the same length, but with different base-pair sequences, to be separated[15]. As a result of performing the FISH, both the intact and aerobic cultured TFOs (unicellular short-rods) belonged to γ-subclass Proteobacteria, which was also supported by the DGGE-result. In order to evaluate the identity of both bacteria, along with related sequences retrieved from GenBank, the partial 16S rDNA sequences (-410 bp) from the intact TFOs clones in the effluent of the full-scale WDSD WWTPs (W) were compared with the aerobically cultured unicellular TFOs clones (M), with the phylogenetic tree constructed (Fig. 3). The similarities between the reference strain, Acinetobacter iwoffii and the W-7 clone from intact TFOs, and Acinetobacter sp. and the W- 5 clone were 100 and 98% respectively. Historically, the genus Acinetobacter, known as strict aerobes[16], has also been thought to be a primary P remover prior to the introduction of molecular techniques for the analysis of bacterial community structures in sludges and several ecosystems[17]. This suggests that an alternation to the aerobic-anaerobic regime, especially in the SBR system, where bacteria must suffer from the changes in dissolved oxygen (DO), might cause the occurrence of TFOs. It is well known that the WWTP’s environment in a bed, even within a floc, is not homogenous. Therefore, the strict aerobe, Acinetobacter sp., could survive by producing an EPS layer in the anaerobic phase. Furthermore, even though Acinetobacter sp. is known as an efficient P-remover[16], the thick layer of EPS produced for survival may inhibit, not only the uptake of phosphate, but reduce the general metabolic activity[18], as revealed by the difficulty in penetrating with the gene probes during FISH. Moreover, as reported in other studies on the effect of temperature on TFO_s, the growth ability of Acinetobacter sp. over a wide range of temperatures[15] allows them to out-compete the PAOs, especially at high temperature[19], leading to their rapid proliferation due to the higher metabolic activities[20]. In the case of WDSD’s WWTP, TFOs were found floating in the effluent in large amounts during the warmer months (personal communication), probably due to the lower density of the thick EPS layer surrounding the Acinetobacter sp. than that of the water medium. The buoyant TFO_s with thick EPS layers resulted in extremely high content of suspended solids in effluent[2] and the poor P-removal efficiency of the EBPR process.

[Fig. 2.] DGGE profile of intact tetrade forming organisms (TFOs)floating in the effluent (W; floating TFOs during the deterioratedEBPR process in the effluent of the full scaled WDSC WWTP, USA)and those aerobically cultured (unicellular short-rod) in a reactor(R; the sampled TFOs from the same site and then cultured in thepresence of oxygen). DGGE: denaturing gradient gel electrophoresis,EBPR: enhanced biological phosphorus removal, WDSC:Wisconsin Dairy State Cheese, WWTP: wastewater treatmentplant.

[Fig. 3.] Phylogenetic tree for the partial 16S rDNA sequences (-410 bp) from intact tetrade forming organisms (TFOs) clones retrieved from the effluent of a full-scale WDSD WWTPs (W) and from aerobically cultured TFOs (M), along with the related sequences retrieved from GenBank. The units at the bottom of the tree indicate the number of substitution events. The dotted lines on a dendrogram indicate negative branch lengths. WDSC: Wisconsin Dairy State Cheese, WWTP: wastewater treatment plant.

This study revealed the identity of TFOs floating in a large amount of effluent at a dairy WWTP (SBR) during the deteriorated EBPR process. TFOs were separately collected from the effluent of a cheese production facility in Wisconsin, USA, using 3.0 μm polycarbonate filters. The results are summarized as follows:

When the TFOs were cultured aerobically, they were converted into unicellular short-rods, but the TFOs did not grow anaerobically. The cultured unicellular short-rods (TFOs) belonged to γ-subclass Proteobacteria

In order to identify them both of intact TFOs and the unicellular short-rods) the DNA from each was extracted, analyzed using DGGE-profile and sequenced for comparison to the 16S rDNA sequences in GenBank using the NCBI Blast program. Both bacteria showed the same banding pattern in their DGGE-profiles, and based on their sequencing data, were identified as Acinetobacter sp.

Taken together, this study suggests that the TFOs frequently occurring in the EBPR process are not always specific bacteria, such as Amaricoccus [7], having a unique cell form, but the distinctive cell morphology could be just a temporary phenomenon that appears under unfavorable environmental conditions during the inefficient P-removal process of EBPR system. Further ecological studies will be required to evaluate the operating conditions that facilitate the dominant occurrence of TFOs, as well as the detailed characteristics of their EPS layer against P-removal.