Abstract: In this study, filamentous bacteria in the activated sludge of a WWTP were investigated throughout a one-year period using high-throughput short-read (Illumina) and full-length (PacBio) 16S rRNA gene amplicon sequencing. The results showed that a total of 28 filamentous bacteria genera
were identified using Illumina sequencing. Also, we found 25 species using PacBio sequencing, belonging to Curvibacter, Mycobacterium, Haliscomenobacter, Defluvicoccus, Sphaerotilus, Thiothrix, Leptothrix, Gordonia and Tetrasphaera genera. Active Volatile Suspended Solids (AVSS) were
calculated from ATP data contained in living microorganisms, this parameter represents the living biomass concentration, and the food/microorganisms ratio (F/M ratio) was calculated using AVSS instead of MLVSS. To assess the contribution of the F/M ratio to the variability observed in the filamentous bacteria structure we carried out distance-based linear models (DISTLM) and distancebased redundancy analysis (dbRDA).
2. Profiling of filamentous bacteria in activated sludge by 16S
RNA amplicon-based sequencing
⁕
Alonso JL, ⁕
Zornoza A, ⁕
Barbarroja P., ⁕
Amoros I.,**
Claros J., **Pastor L, ***Lardin C.
⁕
Instituto de Ingeniería del Agua y Medio Ambiente, Universitat Politècnica de València, Spain
⁕⁕
Depuración de Aguas del Mediterráneo, Paterna (Valencia), Spain
⁕⁕⁕
Entidad de Saneamiento y Depuración de Aguas de la Región de Murcia (ESAMUR),
Murcia, Spain
Abstract: In this study, filamentous bacteria in the activated sludge of a WWTP were investigated
throughout a one-year period using high-throughput short-read (Illumina) and full-length (PacBio) 16S
rRNA gene amplicon sequencing. The results showed that a total of 28 filamentous bacteria genera
were identified using Illumina sequencing. Also, we found 25 species using PacBio sequencing,
belonging to Curvibacter, Mycobacterium, Haliscomenobacter, Defluvicoccus, Sphaerotilus, Thiothrix,
Leptothrix, Gordonia and Tetrasphaera genera. Active Volatile Suspended Solids (AVSS) were
calculated from ATP data contained in living microorganisms, this parameter represents the living
biomass concentration, and the food/microorganisms ratio (F/M ratio) was calculated using AVSS
instead of MLVSS. To assess the contribution of the F/M ratio to the variability observed in the
filamentous bacteria structure we carried out distance-based linear models (DISTLM) and distance-
based redundancy analysis (dbRDA).
Keywords: filamentous bacteria, activated sludge, 16S amplicon sequencing, AVSS
Filamentous bacteria are normally present in the activated sludge in small amounts.
Under specific conditions they proliferate to such an extent that they markedly affect
the treatment plant performance causing sludge bulking or foaming. The
food/microorganisms ratio (F/M ratio) is adopted in many plants as the key
operational parameter. However MLVSS measures neither the active biomass, nor is
necessarily proportional to it, this parameter needs to be viewed critically (Seviour et
al., 2010). Active Volatile Suspended Solids (AVSS), calculated from the cellular
ATP contained in living microorganisms, represents the living biomass concentration.
In this study, filamentous bacteria in the activated sludge of a WWTP were
investigated throughout a one-year period using high-throughput short-read (Illumina)
and full-length (PacBio) 16S rRNA gene amplicon sequencing. To assess the
contribution of the F/M ratio to the variability observed in the filamentous structure
we carried out distance-based linear models (DISTLM). Distance-based redundancy
analysis (dbRDA) was used to visualize the DISTLM. The reformulated F/M ratio
was calculated using AVSS instead of MLVSS.
Samples from activated sludge system were collected every fifteen days during one
year from a bioreactor belonging to a WWTP located in Spain. The plant treats 25 000
m3 day-1 of mainly municipal sewage and adopts an anoxic∕aerobic (AO) process. In
this work PhotonMasterTM
Luminometer from Luminultra® and QG21WasteTM
kit were
used to measure the total-ATP (tATP) and the dissolved-ATP (dATP). The tATP
included both cellular and extracellular ATP, while the dATP only included
extracellular ATP (released from dead or lysis cell). AVSS were calculated from ATP
data. Filamentous bacteria communities of 24 and 7 activated sludge samples were
analysed by Illumina MiSeq high-throughput sequencing method and PacBio single
molecule real-time (SMRT) technology, respectively. Bacterial DNA was extracted
using the FastDNA® SPIN kit for soil. The V3 and V4 regions of the bacterial 16S
3. rRNA were amplified by PCR using primers PRO341F and PRO805R. The full-
length hypervariable regions (V1-V9) of the 16S rRNA gene were amplified using
primers 27F and BS-R1407. To assess the contribution of the F/M ratio to the
variability observed in the filamentous bacteria community structure, we carried
DISTLM and dbRDA. Multivariate analyses was performed with PRIMER v7.
A total of 28 filamentous bacteria genera were identified in activated sludge
samples. Among those genera, Thiothrix (5.06%), Defluvicoccus (1.98%), Candidatus
Microthrix (3.85%), Mycobacterium (1.02%), Sphaerotilus (0.91%), Candidatus
Alysiosphaera (0.79%), and Ornatilinea (0.66%) were the most abundant (Fig. 1).
Thiothrix are typically found in activated sludge systems with filamentous bulking
problems (Nielsen et al., 2000). However, in the present study, sludge bulking was
not observed (SVI 34-82 ml/g).
Fig. 6. Heatmap and cluster analysis displaying the abundance (square root) and association of filamentous
bacteria communities and samples.
The dbRDA bubble plot results revealed two groups of filamentous bacteria
associated with different F/M ratio. Group I was associated mainly with F/M ratio
above 0.46. In contrast, group II was associated with F/M ratio below 0.46. The
construction of DISTLM allows associating the filamentous bacteria to F/M ratio,
obtaining valuable information to the knowledge of these dynamic populations.
REFERENCES
Nielsen, P.H., de Muro, M.A., Nielsen, J.L., 2000. Studies on the in situ physiology of
Thiothrix spp. present in activated sludge. Environ. Microbiol. 2, 389-398.
Seviour, J., Lindrea, K. & Oehmen, A. 2010. The activated sludge process. In:
Microbial Ecology of Activated Sludge (R. Seviour & P.H. Nielsen, eds.). IWA
Publishing, London, pp. 57-94
4. Profiling of filamentous bacteria in activated
sludge by 16S RNA amplicon-based sequencing
Alonso JL1*, Zornoza A1, Barbarroja P1, Claros J2, Pastor L2, Lardín C3
1 Instituto de Ingeniería del Agua y Medio Ambiente, Universitat Politècnica de València, Valencia, Spain. *Email: jalonso@ihdr.upv.es
2 Depuración de Aguas del Mediterráneo (DAM), Paterna (Valencia), Spain
3 Entidad de Saneamiento y Depuación de Aguas de la Región de Murcia (ESAMUR), Murcia, Spain
Introduction
Filamentous bacteria are normally present in the activated sludge in small amounts. Under specific conditions they proliferate to such an
extent that they markedly affect the treatment plant performance causing sludge bulking or foaming (Fig. 1). The organic loading rate
(OLR) is adopted in many plants as the key operational parameter. However, mixed liquor volatile suspended solids (MLVSS) measures
neither the active biomass, nor is necessarily proportional to it, this parameter needs to be viewed critically (Seviour et al., 2010). Active
volatile suspended solids (AVSS), calculated from the cellular ATP contained in living microorganisms, represents the living biomass
concentration (Fig. 2). In this study, filamentous bacteria in the activated sludge of a wastewater treatment plant (WWTP) were
investigated throughout a one-year period using high-throughput short-read (Illumina) and full-length (PacBio) 16S rRNA gene amplicon
sequencing. To assess the contribution of the OLR to the variability observed in the filamentous structure we carried out multivariate
analysis. The reformulated OLR was calculated using AVSS instead of MLVSS (Kg BOD5/Kg AVSS.day)
Material & Methods
Sample site: Samples from activated sludge system were collected every fifteen days during one year from a bioreactor belonging to a
WWTP located in Spain. The plant treats 25 000 m3 day-1 of mainly municipal sewage and adopts an anoxic⁄aerobic (AO) process.
ATP analysis: PhotonMasterTM Luminometer from Luminultra® was used to measure ATP of 24 activated sludge samples. Volumes of
1.0 mL and 100 μL of activated sludge were used to quantify the total-ATP (tATP: intracellular plus extracellular released from dead cell)
and the dissolved-ATP (dATP: extracellular ATP) using QG21WasteTM kit. The results were determined in RLU (Related Luminescence
Units) and were converted to ng ATP·mL-1 using a sandard ATP solution (UltraCheck™) and LumiCaptureTMLite software. AVSS were
calculated from ATP data.
DNA extraction, Illumina and PacBio sequencing:. One ml of 24 sludge samples were extracted in duplicate using the FastDNA® SPIN
kit for soil (MP Biomedicals, OH, USA) (Fig. 3). The hypervariable V3–V4 regions of bacterial 16S rRNA gene were amplified by Fundación
FISABIO sequencing service (Valencia, Spain) using the primers PRO341F and PRO805R. The subsequent amplicon sequencing on the
Illumina Miseq platform was also performed by FISABIO sequencing service (Valencia, Spain) using a 2 × 300 nucleotide paired-end
reads protocol. Raw Illumina sequences were analysed using QIIME™ version 1.8.0. The most abundant sequence of each OTU was
picked as its representative, which was used for taxonomic assignment against 16S SILVA_128 database at 97% identity (cut-off level of
3%) using default parameters. The full-length hypervariable regions (V1-V9) of the 16S rRNA gene were amplified of 7 activated sludge
samples using primers 27F and BS-R1407. The library preparation of the V1-V8 region was performed using PacBio platform at GATC
Biotech (Konstanz, Germany).
Statistical analysis: To assess the contribution of the F/M ratio to the variability observed in the filamentous bacteria community structure,
we carried a distance-based linear models (DISTLM-dbRDA). All multivariate analyses were performed with PRIMER v7 (Clarke & Gorley,
2015) and PERMANOVA+ (Anderson et al., 2008).
Results & Discussion
A total of 28 filamentous bacteria genera were identified in activated sludge samples. Among those genera, Thiothrix (5.06%) (Fig. 4a), Defluvicoccus (1.98%), Candidatus Microthrix (3.85%), Mycobacterium (1.02%), Sphaerotilus
(0.91%), Candidatus Alysiosphaera (0.79%), and Ornatilinea (0.66%) were the most abundant (Fig. 5). Thiothrix are typically found in activated sludge systems with filamentous bulking problems (Nielsen et al., 2000). However, in
the present study, sludge bulking was not observed (SVI 34-82 ml/g). Analysis of the bioreactor using PacBio platform showed the presence of 4 Curvibacter species (C. delicatus, C. fontanus, C. gracilis and C. lanceolatus) (Fig.
4b), 1 Haliscomenobacter species (H. hydrossis), 1 Defluvicoccus species (D. vanus), 1 Alysisosphaera species (Candidatus Alysiosphaera europeae), 1 Sphaerotilus species (S. natans), 5 Thiothrix species (T. flexilis, T.
eikelbomii, T. caldifontis, T. fructosivorans and T. ramosa), 3 Leptothrix species (L. ginsengisoli, L. discophora and L. cholodnii), 2 Tetrasphaera species (T. elongate and T. japonica), 5 Trichococcus species (T. collinsii, T.
flocculiformis, T. palustris, T. pasterurii, T. patagoniensis) (Fig. 4d), 1 Gordonia species (G. alkaliphila) and 1 Saccharimonas species (Candidatus Saccharimonas aalborgensis). As shown in the canonical analysis of principal
coordinates plot (CAP), the results revealed some differences in filamentous bacteria population between groups months (temporal pattern), showing three groups of bacteria correlated with different temperature ranges (Fig. 6).
On the other hand, the results also revealed significant differences between OLR (BOD/MLVSS) and OLR (BOD/AVSS), which indicates the interest in exploring both variables for monitoring the biological process (Fig. 7).
Figure 1. Activated sludge separation problems. (a) Foaming. (b) Bulking. (c) Type 021N.
phase contrast microscopy, 1000x.
Conclusions
The construction of models has allowed us to verify that the concentration of AVSS is more suitable for the calculation of the OLR than MLVSS (traditional), obtaining valuable information to the knowledge of these dynamic
populations. Therefore, we propose the reformulation of this variable for process control in WWTPs.
Figure 7. Comparison between the values of the ORL
(BOD/MLVSS) and OLR (BOD/AVSS), during the study.
Figure 6. CAP based on filamentous bacteria abundance data,
according to the temporal factor. The circles and the bubbles
illustrate the cluster analysis (80% of similarity) of biological
variables and temperatura value, respectively.
The figure 8 shows the best ecological interpretation model (DISTLM) of filamentous bacteria population dynamics from MLVSS, AVSS, tATP, dATP, cATP, BSI (biomass stress index), TSS (total suspended solids), ABR (active
biomass ratio), OLR (BOD/MLVSS) and OLR (BOD/AVSS) (predictor variables). The results revealed a greater influence of OLR (BOD/AVSS) (axis dbRDA2) in the population dynamics of filamentous bacteria, contrary to the
OLR (BOD/MLVSS), which was not selected in the model.
References
Nielsen, P.H., Daims, H., y Lemmer, H. (eds) (2009b) FISH Handbook for Biological Wastewater Treatment. London: IWA Publishing.
Eikelboom, D. (2000) Process Control of Activated Sludge Plant by Microscopic Investigations. London: IWA Publishing.
Clarke, K.R, y Gorley, R.N. (2015) PRIMER v7: User Manual/Tutorial. PRIMER-E, Plymouth, 296pp.
Anderson, M.J., Gorley R.N., y Clarke, K.R. (2008) PRIMER + for PERMANOVA: Guide to Software and Statistical Methods. PRIMER-E. Ltd, Plymouth. United Kingdom.
Figure 5. Cluster analysis of the filamentous bacteria and samples, according to temporal factor. The heatmap illustrates the relative abundance
(transform: square root) of the bacteria, using a taxonomy indicator Figure 8. Distance-based redundancy analysis (dbRDA)
bubble plot illustrating the best model. The overlay
vectors illustrate the multiple partial correlations of the
base variables.
Figure 4. Filamentous bacteria. Phase contrast, 1000x. (a)
Thiothrix sp. (b) Cuvibacter (type 0041/0675). (c) Isosphaera sp.
(Nostocoida limícola III) (d) Trichococcus (Nostocoida limícola I).
a b
c d
a b ,
c
Figure 2. Cellular ATP representing the living biomass concentration
Figure 3. Fast-Prep DNA extraction method with the DNA purification step