Abstract

Microthrix parvicella is a filamentous bacterium that frequently causes severe bulking events in wastewater treatment plants (WWTPs) worldwide. In this study, sludge properties and dynamics of filamentous bacteria in a Beijing WWTP seasonally suffering from M. parvicella bulking were continuously monitored over a duration of 15 months, and the correlations between M. parvicella and operating parameters were evaluated. The predominance of M. parvicella was observed at low temperatures (14–18.8 °C) with the relative abundance of around 30% (estimated by both qPCR and FISH analysis). Using micromanipulation technology, 545 filaments of M. parvicella were micromanipulated from bulking sludge (SVI > 180 mL g−1) on six different media. After 3-month purification and enrichment, six strains, phylogenetically closely related to Candidatus Microthrix parvicella, were successfully acquired on R2A medium (20 °C) in pure cultures. Considering the limitation and extremely slow growth rate of M. parvicella filaments, newly isolated strains represent valuable sources for further investigations on the physiology and behavior of this filamentous bacterium, with the focus on the establishment of bulking control strategy.

INTRODUCTION

Activated sludge process has been widely used for wastewater treatment, but one of the main challenges for the stable operation in wastewater treatment plants (WWTPs) is the excessive growth of filamentous bacteria known as filamentous bulking (Eikelboom & Van Buijsen 1983; Jenkins et al. 1993; Jenkins et al. 2004). As the most common filamentous bacteria in bulking events, CandidatusMicrothrix parvicella is detected based on its distinct morphological features of an unsheathed, non-branched, and Gram-positive filament (Blackall et al. 1995). Several strains have been successively isolated and phylogenetically identified as a novel member of Actinomycetes (Slijkhuis 1983a; Blackall et al. 1996) since the first isolation of M. parvicella from a WWTP in the Netherlands (Van Veen 1973). Most analyzed isolates of M. parvicella, originating from different sources, were affiliated to the same species, suggesting a taxonomic uniformity within this morphotype (Rossetti et al. 1997; Levantesi et al. 2006). To date, however, only few strains are still viable under axenic conditions. The extremely limited availability of actively growing strains makes it difficult to elucidate the key mechanisms allowing M. parvicella to surge in WWTPs and cause severe bulking events (Rossetti et al. 2005).

With the extensive construction of municipal WWTPs in China, filamentous sludge bulking is drawing much concern because of its severe impact on effluent water quality. In the past several years, large efforts have been made for the monitoring of bulking events in full-scale WWTPs and M. parvicella was frequently found as dominant filament in the aerobic zone of activated sludge-based biological WWTPs (Wang et al. 2014a, 2014b; Xie et al. 2007). Using high-throughput pyrosequencing and molecular quantification-based approaches, Wang et al. (2016) found that M. parvicella was the predominant filamentous bacterium in bulking sludge and its seasonal dynamic change was related to environmental and operational factors, such as temperature and dissolved oxygen (DO) concentrations. Xie et al. (2007) also observed an intensive growth of M. parvicella during the cold winter and spring in the Tangshan WWTP with a triple oxidation ditch. Despite the extensive occurrences of M. parvicella bulking in Chinese large WWTPs, no successful isolation of M. parvicella has been achieved in this country, which primarily limits the further researches on its physiochemical properties and metabolisms. Moreover, the taxonomic diversity of the filamentous bacteria with similar morphology (e.g. Nocardioform bacteria) raises two questions: whether the M. parvicella strains in Chinese WWTP are in the same genus with those from other countries and the differences among them; and whether geographic isolation or barriers exist within the group.

This study reports on the dynamics of M. parvicella in a large WWTP and the correlations between filamentous abundance and key parameters. Based on the full-scale observations and previous investigations (Fan et al. 2017), six CandidatusMicrothrix parvicella strains were firstly isolated from a Chinese WWTP, and their taxonomic and basic biological characteristics were described here. These results provide a valuable source for further study on the behavior of this filamentous bacterium with a focus on bulking control.

MATERIALS AND METHODS

Seed sludge

M. parvicella-dominating activated sludge samples were collected from the end of aerobic zone in Xiaohongmen WWTP in Beijing, China. This plant was designed for 1,930,000 population equivalents with an anaerobic-anoxic-oxic (A2/O) process (Figure S1). The system operated well during the experimental period and main parameters were summarized in Table S1. (Figure S1 and Table S1 are available with the online version of this paper.)

Chemical analysis and microscopic examination

Samples for chemical analysis were filtered through 0.45 μm PES membranes (Millipore, USA) in advance. Ammonium nitrogen (NH4+-N), nitrate nitrogen (NO3N), nitrite nitrogen (NO2N), orthophosphate phosphorus (PO43–P), sludge volume index (SVI), and mixed liquor sludge solids (MLSS) were weekly measured according to the standard methods (APHA 1998). Temperature, pH and DO were daily monitored using thermometer and Orion Star A portable DO/pH instrument (Thermo Scientific, USA).

Fresh mixed liquor samples used for microscopic filaments quantification were weekly collected in duplicates from the end of aerobic zone. Filamentous bacteria and sludge properties were monitored regularly under microscopy according to Jenkins et al. (2004). The abundance of filamentous bacteria was roughly determined using filament index (FI) on a scale of 0–6 (Jenkins et al. 2004). Generally, sludge bulking was considered to be severe at FI ≥ 5.

DNA extraction and filamentous quantification

Approximately 2 mL of sludge samples were centrifuged at 8,000 g (Sigma Centrifuge, Germany) and cell pellets were stored at −20 °C. Genomic DNA was extracted following the manufacturer's instructions of Fast DNA Spin Kit for Soil (MP Biomedicals, USA). The extracted products were visualized on 1% agarose gel.

Extracted DNA was amplified in a polymerase chain reaction (PCR) Cycler (Light Cycler 96, Roche, USA) using the primer sets listed in Table S2 (available online). Detailed process of M. parvicella quantification was performed according to a previous description (Fan et al. 2018). Briefly, amplification was carried out in 25 μL reactions by the program of 95 °C 10 min, 95 °C 15 s, 62 °C 30 s, 60 °C 30 s, and 39 cycles. Average amplification efficiency of the specific qPCR primer set was 98% with an r2 of 0.997 for the standard curve. The analysis of samples and two negative controls (one without primers and the other without template DNA) were performed in duplicate assays on an ABI7300 apparatus (ABI, USA). The 16S rRNA gene copies of total bacteria were quantified using the primer set S-D-Bact-0509-S-17/S-DBact-0784-A-22 according to the procedures described by Rupf et al. (1999).

Fluorescence in situ hybridization (FISH)

For FISH analysis, 2 mL samples were fixed in 9 mL 4% (w/v) paraformaldehyde for 2 h at 4 °C and stored in PBS/ethanol (1:1, v/v) at –20 °C until hybridization (Nielsen et al. 2009a). The fixed cells were sonicated for 60 s with an ultrasonic processor (Vibra-Cell, Sonics) at a pulse of 5 s and an output power of 4 W to break the floc structure for obtaining a better view of the filaments during microscopic observation. The oligonucleotide probes EUBmix (equimolar concentrations of EUB338, EUB338II, and EUB338III) labeled with fluorescein isothiocyanate (FITC) and MPAmix (equimolar concentrations of MPA60, MPA223, and MPA645) labeled with the sulfoindocyanine dye Cy3 (Table S3, available online), were used for FISH analysis (Amann et al. 1990; Erhart et al. 1997). After hybridization, the slides were covered with ProLong® Gold Antifade Reagent (Thermo Fisher Scientific, USA). Detailed procedures were performed according to the protocols described by Nielsen (Nielsen et al. 2009b) and stained images were captured with an epifluorescence microscope (BX51 Olympus, Japan) equipped with the cooled charged-coupled device (CCD) camera (AxioCam MRm, Zeiss, Germany) in true color using suitable filter set and exposure times of a few milliseconds. At least 20 random fields were chosen, and the area ratio of M. parvicella to the total biomass was calculated by image analyzing software (Image J, http://rsbweb.nih.gov/ij/). The percentage of FISH detectable cells was determined as the fraction of bacterial cells positive to probe EUBmix over the total cells stained by 4′6-diamidino-2-phenylindole (DAPI).

Statistical analysis

Statistical analysis to determine the Pearson correlation between M. parvicella abundance and performance parameters were performed using the SPSS16.0 software package (SPSS®, Chicago, IL, USA). The p values of 0.05 and 0.01 were considered as the cutoff for statistical significance between two data sets. Principal component analysis (PCA) to evaluate the effects of main parameters (temperature, sludge load, etc.) on M. parvicella abundance was carried out by R version 3.3.3 (Statistical Computing Software) using vegan, permute and lattice packages.

Isolation and maintenance of M. parvicella

Sludge samples for M. parvicella isolation were also collected from the end of aerobic zone in Xianhongmen WWTP on 10th March, 2016 (SVI = 181.2 ± 12.7 mL g−1). Isolation attempts were performed by micromanipulation according to Skerman (Skerman 1968) on different growth media, including R2A (Reasoner & Geldreich 1985), SLIJKHUIS_A, SLIJKHUIS_F, MSV (Williams & Unz 1985), MSV + acetate, MSV + peptone, MSV + acetate + bicarbonate, MSV + peptone + bicarbonate. These media were selected and modified based on the previous successes in isolating filamentous bacteria. The micromanipulated filaments were incubated at 20 °C and checked daily. The visible colonies formed by single filaments were transferred onto a fresh agar plate for further enrichment. After purified multiple times, pure cultures of the isolated filaments could be achieved.

Identification and phylogenetic analysis

The 16S rRNA genes were amplified by colony PCR, using primers 27f and 1492r for bacteria domain using the PCR Perfect Taq DNA Polymerase (5 PRIME GmbH – Germany) as previously described (Rossetti et al. 2003). PCR products were purified using the PCR Extract Mini purification kit (5 PRIME GmbH – Germany). 16S rRNA gene sequences of the PCR amplificates were obtained using the following primers: 530f, 907r, 1055f, and 1392r. The almost full-length 16S rRNA gene sequences identified in this study were deposited in the GenBank under accession numbers MG283267 (strain NFA), MG283268 (strain NFB), MG283269 (strain NFC), MG283270 (strain NFD), MG283271 (strain NFE), MG283272 (strain NFH), respectively.

RESULTS

Bulking problems and process performance

SVI, temperature and process performance were monitored over 15 months (Figure 1; Figure S2, available with the online version of this paper). A significant negative correlation (r = –0.877; p < 0.01) was observed between SVI and temperature, which was in accordance with previous studies. When temperature was higher than 20 °C, SVI level was instead below 150 mL g−1, commonly considered as the threshold of no-bulking sludge (Jenkins et al. 1993). Overall, good average removal efficiencies for chemical oxygen demand (COD), biochemical oxygen demand (BOD), total nitrogen (TN), total phosphorus (TP) and PO43–P were achieved with average values of 95 ± 1.3%, 98 ± 0.7%, 77 ± 5.2%, 94 ± 3.3% and 92 ± 2.6%, respectively. However, the ammonium (NH4+-N) removal efficiency slightly decreased to 90.8 ± 5.8% during the sludge bulking period. Simultaneously, nitrate in the effluent decreased dramatically, accompanied with the accumulation of nitrite. Similar fluctuations during sludge bulking had also been observed in another full-scale WWTP in Beijing with the same treatment process (Wang et al. 2014a).

Figure 1

The inlet (blue triangle) and outlet (red circle) concentrations of COD, BOD, NH4+-N, TN, TP, PO43–P, NO3N and NO2N during the experiment period. Please refer to the online version of this paper to see this figure in color: http://dx.doi.org/10.2166/wst.2019.136.

Figure 1

The inlet (blue triangle) and outlet (red circle) concentrations of COD, BOD, NH4+-N, TN, TP, PO43–P, NO3N and NO2N during the experiment period. Please refer to the online version of this paper to see this figure in color: http://dx.doi.org/10.2166/wst.2019.136.

Identification and quantification of filamentous bacteria

The dynamic change of filamentous bacteria community was monitored by microscopic observations and the specific abundance of bulking filaments was further determined by qPCR and FISH analysis. Different filamentous morphotypes including M. parvicella, Eikelboom Type 0041, Thiothrix, Candidatus Promineofilum, and Nostocoida limicola were observed in the activated sludge (Figure 2). Based on the microscopic observation (Figure S3, available online), Ca. Promineofilum (Type 0092) and Type 0041 were dominating in the summer (from June to August, with the average temperature of 25.2 ± 0.9 °C) with no evident impact on the sludge settleability. During the cold season (from December to March, with the average temperature of 15.7 ± 0.6 °C), the abundance of filamentous microorganisms increased significantly and M. parvicella became the dominant filamentous bacteria, confirming its important role in bulking problems. Simultaneously, Type 0041 and Ca. Promineofilum abundance decreased.

Figure 2

Main filamentous bacteria in activated sludge observed under microscopy.

Figure 2

Main filamentous bacteria in activated sludge observed under microscopy.

The morphological identification of M. parvicella was further confirmed, and the dynamic changes of M. parvicella abundance in the transition from non-bulking to bulking sludge were monitored using different bio-molecular methods (qPCR and FISH) (Figure 3). Remarkably, both techniques generated similar trends of the M. parvicella relative abundance over the WWTP operation. In particular, qPCR data showed the sharp increase of M. parvicella from 5 to 10% (the percentage of total 16 s rRNA gene copies) up to around 40% of the total bacteria during bulking period. In line with qPCR data, FISH confirmed 10–15% (the percentage of EUBmix fluorescence area) of M. parvicella relative abundance as a ‘gross’ threshold value for an activated sludge with good settleability properties. The relative abundance of M. parvicella has been extensively quantified to be 0–18% in worldwide WWTPs with various processes, and above 3% M. parvicella in activated sludge could cause foaming problems in WWTP (Kaetzke et al. 2005; Kumari et al. 2009). However, this value was unsuitable for the prediction or determination of bulking events in Northern China, because the abundances of M. parvicella in some WWTPs were higher than 3%, even in the non-bulking period (Wang et al. 2014b).

Figure 3

The relative abundance of M. parvicella quantified by microscopic observation, qPCR, and FISH.

Figure 3

The relative abundance of M. parvicella quantified by microscopic observation, qPCR, and FISH.

Correlations between M. parvicella and process parameters

The correlations between M. parvicella abundance and main process parameters were evaluated. Significant positive correlations (p < 0.05) was found between M. parvicella abundance and TP (r = 0.71) removal efficiencies. On the contrary, sludge load and M. parvicella abundance were negatively correlated (r = –0.695, p < 0.05). Based on the PCA results (Figure S4, available online), the accumulated variance contribution rate of first principle (PC1) and second principle (PC2) was 73.8%. The correlations between M. parvicella abundance and key parameters were in line with the above t-test results. The samples (purple circles) taken during the low temperature period (from December to April, 14–18.8 °C) with the average value of 16.6 °C were clustered in the first and fourth quadrants, and the rest of the samples (green circles) in relatively high temperature (from May to November, 19.2–26.9 °C) were clustered in the second and third quadrants. In addition, a strong negative correlation between sludge load and M. parvicella abundance was indeed observed in this large plant during more than one year of monitoring (Figure S5, available online). The average sludge load in bulking period was 0.11 ± 0.005 kg COD kg MLSS−1·d−1, while that of non-bulking was 0.14 ± 0.024 kg COD kg MLSS−1·d−1. When the sludge load was lower than 0.12 kg COD kg MLSS−1·d−1, M. parvicella abundance dramatically increased and SVI was over 150 mL g−1.

Isolation and phylogenetic analysis of M. parvicella

A large effort was made for the isolation of M. parvicella by micromanipulation (Table 1). Totally, 545 M. parvicella filaments, identified by microscopic observation at 320 magnification, were micromanipulated on 6 different growth media, which were previously reported to be suitable for the isolation of this filamentous morphotype (Blackall et al. 1995; Rossetti et al. 1997; Levantesi et al. 2006). Based on the successful isolation cases, more isolation attempts were initially performed on R2A medium. According to the microscopic observations, there were several stages for the growth of M. parvicella, including straight-line growth, thickening and bending, elongation and twining growth, colony forming (Figure 4). Most filaments were eliminated because of contamination, growing stop, or physical damage. Frequently, the growth of fast-growing bacteria, accidentally micromanipulated together with the filaments, hampered the successful isolation of the slow-growing filaments. Moreover, the lack of potential growth factors and physical damages during the transfer and spread plate probably caused cells to stop growing. Pure strains were finally obtained on R2A medium, which had also been widely adopted in previous experience of M. parvicella isolation. Although some initial growth was observed on Slijkuhis's medium A and MSV + ACT + B, a visible colony was never produced on these media. Overall, only 1% of the filaments grew enough for further enrichment and preservation, which confirmed the paucity of the growth of this microorganism in axenic conditions.

Table 1

Summary of M. parvicella isolation attempts

Medium Micromanipulated filaments Contaminated filamentsa Growing filamentsb Stop-growing filamentsc Isolates 
R2391 296 (76%) 89 (23%) 
SLIJK_A 38 16 (42%) 22 (58%) 
SLIJK_F 35 8 (23%) 27 (77%) 
MSV-ACT 18 8 (44%) 10 (55%) 
MSV-PEP 20 10 (50%) 10 (50%) 
MSV-ACT-B 17 11 (65%) 6 (35%) 
MSV-P-B 26 6 (23%) 20 (77%) 
Medium Micromanipulated filaments Contaminated filamentsa Growing filamentsb Stop-growing filamentsc Isolates 
R2391 296 (76%) 89 (23%) 
SLIJK_A 38 16 (42%) 22 (58%) 
SLIJK_F 35 8 (23%) 27 (77%) 
MSV-ACT 18 8 (44%) 10 (55%) 
MSV-PEP 20 10 (50%) 10 (50%) 
MSV-ACT-B 17 11 (65%) 6 (35%) 
MSV-P-B 26 6 (23%) 20 (77%) 

aContamination due to the fast growth of bacteria transported accidentally with the filament during micromanipulation.

bInitially growing micromanipulated filaments on each growth medium.

cMicromanipulated filaments that did not grow or stopped growing (in brackets the % of the micromanipulated filaments that stopped growing).

Figure 4

Isolation process of M. parvicella. Left: Schematic diagram of M. parvicella growth stages. Right: Purification and enrichment.

Figure 4

Isolation process of M. parvicella. Left: Schematic diagram of M. parvicella growth stages. Right: Purification and enrichment.

In total, six new strains of M. parvicella (strains NFA, NFB, NFC, NFD, NFE, and NFH) were isolated from activated sludge samples originating from Xiaohongmen WWTP. In line with previous studies (Blackall et al. 1996; Rossetti et al. 1997), the isolated strains grew poorly and slowly under axenic conditions. Filamentous growth states of these six strains were observed under microscopy and recorded every two days. Based on the length changes, their growth rates were roughly estimated to be 0.191 μm d−1, 0.109 μm d−1, 0.151 μm d−1, 0.059 μm d−1, 0.108 μm d−1 and 0.134 μm d−1. Therefore, several weeks were necessary to obtain the visible growth on agar plates. Colony morphology was suborbicular, ivory, and showed a filamentous margin. In phase contrast microscopy, the filaments were unbranched with a slightly thinner diameter (0.3–0.6 μm) and a length of 50–150 μm. This microorganism was strong Gram-positive and Neisser positive granules were visible inside.

The almost complete 16S rDNA sequences for the newly isolated M. parvicella strains (about 1,400 nucleotides) were retrieved and their taxonomic affiliation defined. Phylogenetic analysis showed that all strains belong to the same phylogenetic group within a deep-branching cluster of the Actinomycetes subphylum sharing 98–99% 16S rRNA gene similarity with CandidatusMicrothrix parvicella (Figure 5). The newly isolated strains were generally in line with what has been described for other identified isolates, while these strains grew poorly with a relatively lower growth rate of 0.125 ± 0.045 μm d−1 (Table 2).

Table 2

The comparison of main characteristics among M. parvicella strains

Strain Source Medium Growth temperature (°C) Colony diameter (mm) Filaments diameter (μm) Filaments length (μm) Growth rate (d−1Reference 
New isolates China R2A 20 0.4–0.8 0.3–0.6 50–150 – This study 
Bio 17 The Netherlands MSV + peptone 20 Visible 0.3–0.7 – – Levantesi et al. (2006)  
RN 1 Italy R2A 7–25 ≤1 0.5–0.8 50–300 0.37–0.66 Tandoi et al. (1998)  
EU 18 Italy GS medium 20 – 0.3–0.7 – – Levantesi et al. (2006)  
Slijkhuis strain The Netherlands Slijkhuis A 25 – 0.4–0.6 300–500 0.38 Slijkhuis (1983b)  
Ben 43 Australia R2A 20–22 Visible 0.5–0.9 – – Blackall et al. (1996)  
DAN 1-3 Australia Modified NTM 20–22 Visible 0.4–0.85 – – Blackall et al. (1995)  
TNO 1-1 The Netherlands MSV A + T + B 10–36.5 – 0.3–0.8 – 0.36 ± 0.05 Levantesi et al. (2006)  
Strain Source Medium Growth temperature (°C) Colony diameter (mm) Filaments diameter (μm) Filaments length (μm) Growth rate (d−1Reference 
New isolates China R2A 20 0.4–0.8 0.3–0.6 50–150 – This study 
Bio 17 The Netherlands MSV + peptone 20 Visible 0.3–0.7 – – Levantesi et al. (2006)  
RN 1 Italy R2A 7–25 ≤1 0.5–0.8 50–300 0.37–0.66 Tandoi et al. (1998)  
EU 18 Italy GS medium 20 – 0.3–0.7 – – Levantesi et al. (2006)  
Slijkhuis strain The Netherlands Slijkhuis A 25 – 0.4–0.6 300–500 0.38 Slijkhuis (1983b)  
Ben 43 Australia R2A 20–22 Visible 0.5–0.9 – – Blackall et al. (1996)  
DAN 1-3 Australia Modified NTM 20–22 Visible 0.4–0.85 – – Blackall et al. (1995)  
TNO 1-1 The Netherlands MSV A + T + B 10–36.5 – 0.3–0.8 – 0.36 ± 0.05 Levantesi et al. (2006)  
Figure 5

Phylogenetic tree of M. parvicella filamentous microorganisms and their closest relatives based on comparative analysis of 16S rRNA gene sequences.

Figure 5

Phylogenetic tree of M. parvicella filamentous microorganisms and their closest relatives based on comparative analysis of 16S rRNA gene sequences.

DISCUSSION

According to the analysis on the changes in main parameters, sludge bulking only had an obvious impact on the nitrogen removal performance, while most water quality parameters remained relatively stable. In particular, the BOD, COD and TP removal efficiencies were almost the same for both bulking and non-bulking conditions. Similar situations had been reported in many WWTPs. For example, COD, the total phosphorus and soluble orthophosphate removals during the sludge bulking period were comparable to those in the non-bulking period (Wang et al. 2014b). The data collected from another large Chinese plant (Jiuxianqiao WWTP in Beijing) with an oxidation ditch process showed that COD, BOD and TP removals were still achieved in spite of a filamentous sludge bulking that lasted for about half a year (Guo et al. 2010). The maintenance of nutrients removal performance during bulking period probably depends on the potential function of dominant filamentous bacteria. A previous study suggested that M. parvicella might be responsible for phosphorus removal during sludge bulking period when ‘Candidatus Accumulibacter phosphatis’ (key polyphosphate-accumulating bacteria) was excluded from the systems (Wang et al. 2014b). A strong positive correlation was also found in our study between M. parvicella abundance and TP removals efficiency, confirming a previous hypothesis. Moreover, it could be also supported by the evidence of polyphosphate granules in M. parvicella filaments microscopically visualized after Neisser staining (data not shown).

The negative correlations between temperature and SVI/M. parvicella abundance were illustrated in this study. The seasonal change of M. parvicella abundance and corresponding bulking events had occurred in many countries (Table S4, available with the online version of this paper). Sludge bulking caused by M. parvicella usually occurred in a low-temperature (around 15 °C) period and lasted for several months (Seviour et al. 1990; Lacko et al. 1999; Madoni et al. 2000). Low temperature of 13 °C has been verified to be a favorable condition allowing M. parvicella to maintain a relatively higher abundance in the mixed culture system (Fan et al. 2017). In this study, however, no growth was observed on the 23 micromanipulated filaments cultured at 15 °C, and another five filaments stopped growing after transfer from 20 °C to 15 °C. These discrepancies on the suitable temperature for M. parvicella growth are probably related to the different growth environments. In the in situ conditions, M. parvicella with high intracellular storage capacity and low maintenance energy requirement outcompete floc-forming bacteria during a cold period. While in pure culture without competition, low temperature becomes a limiting factor to decrease or even inhibit its metabolic and proliferation rates. Additionally, unknown growth-promoting factors in wastewater was not supplemented in culture media.

Operational parameters would have impacts on microbial community in activated sludge, among which sludge load showed a significantly negative correlation with filaments abundance in this study. Many in situ investigations showed that sludge load was a closely related with filamentous bulking (Noutsopoulos et al. 2002; Zhang et al. 2014), and low sludge loading conditions could stimulate the competitive growth of M. parvicella (Wagner 1982; Knoop & Kunst 1998). Catching the different stages of bulking events might be a possible reason for the differences in threshold of sludge load between the bulking and the non-bulking periods. Results of our recent study show that M. parvicella growth can be quickly suppressed when the sludge load was higher than 0.1 kg COD kg MLSS−1·d−1, and high-sludge-load strategy has been verified in a full-scale plant to be an effective approach for filamentous sludge bulking control (Fan et al. 2018).

Several hundreds of filaments had been micromanipulated from activated sludge, but only six strains were finally achieved on R2A medium in pure culture. Although R2A was recognized as a suitable medium for M. parvicella, the growth rates of its strains were still extremely low. In the full-scale plant, however, M. parvicella abundance could rapidly increase to a high level during a short period, indicating the absence of some potential growth factors in this medium. It had been extensively reported that long chain fatty acids (LCFAs) were favorable substrates for M. parvicella growth (Andreasen & Nielsen 2000; Muller et al. 2012; Fan et al. 2017), so adding LCFAs in pure culture medium might stimulate its growth, and more studies on the physiology of M. parvicella are necessary and meaningful. Sequencing analysis revealed these newly isolated stains were highly similar with CandidatusMicrothrix parvicella, most of which were isolated from European WWTPs. Such long distance between the source areas reconfirmed the wide distribution of M. parvicella, and also indicated that no geographic isolation or barriers formed within the group during its evolution. As another typical filament in bulking sludge, Nocardioform bacteria have been identified in different genera of Skermania (Eales et al. 2006) and Gordonia (Seviour & Nielsen 2010). By contrast, all isolated M. parvicella strains belong to the same genus of Candidatus Microthrix.

CONCLUSIONS

In conclusion, this study investigated the dynamics of M. parvicella in a large full-scale activated sludge plant and its correlations between the changes of this filamentous bacterium and the main process parameters. A long-term monitoring for 15 months showed significant negative correlations between temperature, sludge load and M. parvicella abundance. In addition, this is the first study using multiple quantitative approaches to provide the threshold for the rapid onset of M. parvicella bulking. A relative abundance of 10–15% was considered as the warning M. parvicella level, indicating the need of proper corrective treatments (e.g. polyaluminium chloride addition or sludge load control).

Using micromanipulation, six M. parvicella strains were firstly isolated from a Chinese WWTP sharing 98–99% 16S rRNA gene similarity with CandidatusM. parvicella. Due to these strains belonging to the viable but non-culturable (VBNC) bacteria, it is still difficult to obtain enough biomass for further investigations or culture collection. Considering the few actively growing isolates of M. parvicella so far available worldwide, these newly isolated strains represent a valuable source to validate the metabolic model recently developed from the sequencing and annotation of M. parvicella strain RN1 genome (McIlroy et al. 2013). However, further enrichment for more biomass is necessary for studying physiochemical properties and building specific bulking control (e.g. bacteriophage-based control strategy).

ACKNOWLEDGEMENTS

We are grateful for the financial support from the National Science and Technology Major Project on Water Pollution Control and Treatment (No. 2015ZX07203-005-03) and the Social Development Program of Hangzhou (No. 20191203B11). This work was supported by the Free Exchange Programs between the University of Chinese Academy of Sciences and the Italian National Research Council.

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

ETHICAL APPROVAL

This article does not contain any studies with human participants or animals performed by any of the authors.

REFERENCES

REFERENCES
Amann
R. I.
,
Binder
B. J.
,
Olson
R. J.
,
Chisholm
S. W.
,
Devereux
R.
&
Stahl
D. A.
1990
Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations
.
Appl. Environ. Microbiol.
56
(
6
),
1919
1925
.
APHA
1998
Standard Methods for the Examination of Water and Wastewater
.
American Public Health Association, Water Environment Federation
,
Washington, DC
,
USA
.
Blackall
L. L.
,
Seviour
E. M.
,
Cunningham
M. A.
,
Seviour
R. J.
&
Hugenholtz
P.
1995
‘Microthrix parvicella’ is a novel, deep branching member of the actinomycetes subphylum
.
Syst. Appl. Microbiol.
17
(
4
),
513
518
.
Blackall
L. L.
,
Stratton
H.
,
Bradford
D.
,
Del Dot
T.
,
Sjörup
C.
,
Seviour
E. M.
&
Seviour
R. J.
1996
Candidatus Microthrix parvicella,’ a filamentous bacterium from activated sludge sewage treatment plants
.
Int. J Syst. Bacteriol.
46
(
1
),
344
346
.
Eales
K. L.
,
Nielsen
J. L.
,
Seviour
E. M.
,
Nielsen
P. H.
&
Seviour
R. J.
2006
The in situ physiology of Skermania piniformis in foams in Australian activated sludge plants
.
Environ. Microbiol.
8
(
10
),
1712
1720
.
Eikelboom
D.
&
Van Buijsen
H.
1983
Microscopic Sludge Investigation Manual
.
TNO, Delft
,
The Netherlands
.
Fan
N.
,
Qi
R.
,
Rossetti
S.
,
Tandoi
V.
,
Gao
Y.
&
Yang
M.
2017
Factors affecting the growth of Microthrix parvicella: batch tests using bulking sludge as seed sludge
.
Sci. Total. Environ.
609
,
1192
1199
.
Fan
N.
,
Wang
R.
,
Qi
R.
,
Gao
Y.
,
Rossetti
S.
,
Tandoi
V.
&
Yang
M.
2018
Control strategy for filamentous sludge bulking: bench-scale test and full-scale application
.
Chemosphere
210
,
709
716
.
Guo
J.
,
Peng
Y.
,
Peng
C.
,
Wang
S.
,
Chen
Y.
,
Huang
H.
&
Sun
Z.
2010
Energy saving achieved by limited filamentous bulking sludge under low dissolved oxygen
.
Bioresource Technol.
101
(
4
),
1120
1126
.
Jenkins
D.
,
Richard
M.
&
Daigger
G.
1993
Manual on the Causes and Control of Activated Sludge Bulking and Foaming
.
Lewis Publishers, Chelsea, MI, USA
.
Jenkins
D.
,
Richard
M. G.
&
Daigger
G. T.
2004
Manual on the Causes and Control of Activated Sludge Bulking, Foaming and Other Solids Separation Problems
.
Lewis Publishers
,
Washington, DC
,
USA
.
Lacko
N.
,
Bux
F.
&
Kasan
H.
1999
Survey of filamentous bacteria in activated sludge plants in KwaZulu-Natal
.
Water S.A.
25
,
63
68
.
Levantesi
C.
,
Rossetti
S.
,
Thelen
K.
,
Kragelund
C.
,
Krooneman
J.
,
Eikelboom
D.
,
Nielsen
P. H.
&
Tandoi
V.
2006
Phylogeny, physiology and distribution of ‘Candidatus Microthrix calida’, a new Microthrix species isolated from industrial activated sludge wastewater treatment plants
.
Environ. Microbiol.
8
(
9
),
1552
1563
.
McIlroy
S. J.
,
Kristiansen
R.
,
Albertsen
M.
,
Karst
S. M.
,
Rossetti
S.
,
Nielsen
J. L.
,
Tandoi
V.
,
Seviour
R. J.
&
Nielsen
P. H.
2013
Metabolic model for the filamentous ‘Candidatus Microthrix parvicella' based on genomic and metagenomic analyses
.
ISME J.
7
(
6
),
1161
1172
.
Muller
E. E.
,
Pinel
N.
,
Gillece
J. D.
,
Schupp
J. M.
,
Price
L. B.
,
Engelthaler
D. M.
,
Levantesi
C.
,
Tandoi
V.
,
Luong
K.
&
Baliga
N. S.
2012
Genome sequence of ‘Candidatus Microthrix parvicella’ bio17-1, a long-chain-fatty-acid-accumulating filamentous actinobacterium from a biological wastewater treatment plant
.
J. Bacteriol.
194
(
23
),
6670
6671
.
Nielsen
P. H.
,
Daims
H.
,
Lemmer
H.
,
Arslan Alaton
I.
&
Olmez Hanci
T.
2009a
FISH handbook for biological wastewater treatment
.
IWA Publishing
,
London, UK
.
Nielsen
P. H.
,
Kragelund
C.
,
Seviour
R. J.
&
Nielsen
J. L.
2009b
Identity and ecophysiology of filamentous bacteria in activated sludge
.
FEMS Microbiol. Rev.
33
(
6
),
969
998
.
Reasoner
D. J.
&
Geldreich
E.
1985
A new medium for the enumeration and subculture of bacteria from potable water
.
Appl. Environ. Microbiol.
49
(
1
),
1
7
.
Rossetti
S.
,
Christensson
C.
,
Blackall
L.
&
Tandoi
V.
1997
Phenotypic and phylogenetic description of an Italian isolate of ‘Microthrix parvicella’
.
J Appl. Microbiol.
82
(
4
),
405
410
.
Rossetti
S.
,
Blackall
L. L.
,
Levantesi
C.
,
Uccelletti
D.
&
Tandoi
V.
2003
Phylogenetic and physiological characterization of a heterotrophic, chemolithoautotrophic Thiothrix strain isolated from activated sludge
.
Int. J. Syst. Evol. Microbiol.
53
(
5
),
1271
1276
.
Seviour
R. J.
&
Nielsen
P. H.
2010
Microbial Ecology of Activated Sludge
.
IWA Publishing
,
London
,
UK
.
Seviour
E.
,
Williams
C.
,
Seviour
R.
,
Soddell
J.
&
Lindrea
K.
1990
A survey of filamentous bacterial populations from foaming activated sludge plants in eastern states of Australia
.
Water Res.
24
(
4
),
493
498
.
Skerman
V.
1968
A new type of micromanipulator and microforge
.
Microbiol.
54
(
2
),
287
297
.
Slijkhuis
H.
1983a
Microthrix parvicella, a filamentous bacterium isolated from activated sludge: cultivation in a chemically defined medium
.
Appl. Environ. Microbiol.
46
(
4
),
832
839
.
Slijkhuis
H.
1983b
The physiology of the filamentous bacterium Microthrix parvicella
.
The University of Wageningen, Wageningen
,
The Netherlands
.
Tandoi
V.
,
Rossetti
S.
,
Blackall
L. L.
&
Majone
M.
1998
Some physiological properties of an Italian isolate of “Microthrix parvicella”
.
Water Sci. Technol.
37
(
4–5
),
1
8
.
Wagner
F.
1982
Study of the Causes and Prevention of Sludge Bulking in Germany
.
Bulking of activated sludge (eds. Chambers and Tomlinson)
.
Ellis Horwood Ltd
,
Chichester, UK
, pp.
29
46
.
Williams
T. M.
&
Unz
R. F.
1985
Isolation and characterization of filamentous bacteria present in bulking activated sludge
.
Appl. Microbiol. Biotechnol.
22
(
4
),
273
282
.
Zhang
A. L.
,
Zhang
X.
&
Wang
S.
2014
The influence of low organic loading on sludge bulking and papermaking wastewater treatment effect
.
J. Shaanxi Uni. Sci. Technol.
32
(
2
),
26
30
.
35
.

Supplementary data