ABSTRACT
In this study, we established a rapid multiplex method to detect the relative abundances of amplified 16S rRNA genes from known cultivatable methanogens at hierarchical specificities in anaerobic digestion systems treating industrial wastewater and sewage sludge. The method was based on the hierarchical oligonucleotide primer extension (HOPE) technique and combined with a set of 27 primers designed to target the total archaeal populations and methanogens from 22 genera within 4 taxonomic orders. After optimization for their specificities and detection sensitivity under the conditions of multiple single-nucleotide primer extension reactions, the HOPE approach was applied to analyze the methanogens in 19 consortium samples from 7 anaerobic treatment systems (i.e., 513 reactions). Among the samples, the methanogen populations detected with order-level primers accounted for >77.2% of the PCR-amplified 16S rRNA genes detected using an Archaea-specific primer. The archaeal communities typically consisted of 2 to 7 known methanogen genera within the Methanobacteriales, Methanomicrobiales, and Methanosarcinales and displayed population dynamic and spatial distributions in anaerobic reactor operations. Principal component analysis of the HOPE data further showed that the methanogen communities could be clustered into 3 distinctive groups, in accordance with the distribution of the Methanosaeta, Methanolinea, and Methanomethylovorans, respectively. This finding suggested that in addition to acetotrophic and hydrogenotrophic methanogens, the methylotrophic methanogens might play a key role in the anaerobic treatment of industrial wastewater. Overall, the results demonstrated that the HOPE approach is a specific, rapid, and multiplexing platform to determine the relative abundances of targeted methanogens in PCR-amplified 16S rRNA gene products.
INTRODUCTION
The anaerobic digestion process has become a method increasingly used for the recycling of organic substances in waste and wastewater into methane as a renewable energy source (1). The process can be operated using different reactor formats under specific environmental conditions with various microbial groups involved in a multistep reaction. Because of their key position in the terminal step that produces the gaseous final products methane and carbon dioxide, the reactor performance ultimately relies on the activities of the methanogen population. Therefore, increased understanding of the relevant population structures and their ecological significance inside the reactor could provide key information with which to increase the efficiency of organic matter decomposition and methane recovery and achieve better reactor output.
The methanogenic archaea (methanogens) are a group of strict anaerobes that use a limited number of substrates, such as H2/CO2 (hydrogenotrophic methanogens), acetate (acetotrophic methanogens), and methyl group-containing compounds (methylotrophic methanogens) (2). They can be detected through the analysis of their unique cellular features (3, 4), as well as, more effectively, analysis using molecular biological tools (5). For example, to reveal the structure of methanogen populations in environmental samples, the community DNA is first recovered. Then, the nucleotide sequences of the 16S rRNA gene or the gene coding for the α subunit of methyl coenzyme M reductase (mcrA) relevant to methanogens is PCR amplified and analyzed with molecular techniques, such as cloning and sequencing (6), denaturing gradient gel electrophoresis (DGGE) (7), terminal restriction fragment length polymorphism (8), single-stranded conformation polymorphism (SSCP) (9), amplicon length PCR (10), or a combination of these. Alternatively, hybridization-based methods, such as membrane hybridization (11, 12) and fluorescence in situ hybridization (FISH) (13), and the quantitative PCR (qPCR) method (14, 15) combined with oligonucleotide primers/probes specifically targeting the selected methanogen groups can be used for quantitative analysis. Narihiro and Sekiguchi recently reported a comprehensive set of probes targeting the methanogens at various taxonomic specificities (16), enabling the rapid evaluation of the relative abundances of the methanogen populations at different hierarchies (17). However, because of limitations in their throughput capabilities, the implementation of the described methods requires a large number of primers/probes and tends to be laborious and time-consuming. Therefore, a method that facilitates the high-throughput determination of the relative abundance of methanogens in the complex microbial consortia is desired.
In a previous study, we developed the hierarchical oligonucleotide primer extension (HOPE) method for the rapid and multiplexed analysis of the relative abundance of multiple microbial targets in PCR amplicons from complex environmental samples (18). This technique arranges multiple oligonucleotide primers that target the regions of the 16S rRNA genes at various phylogenetic specificities in a single-nucleotide strand extension reaction, allowing the DNA polymerase to incorporate a single fluorescently labeled dideoxynucleotide triphosphate (e.g., the dye terminators) at the 3′ ends of the primers. An automatic DNA sequencer then analyzes the size and color of the extended primers. According to the hierarchical design concept, the results from the correctly extended primers provide phylogenetic information on the detected targets. The relative abundance of the targeted group relative to its higher taxonomic rank can also be determined according to the ratios of the fluorescence intensities obtained using 2 primers that have an associated hierarchical relationship. The HOPE method can achieve single-mismatch discrimination and high multiplex capabilities. With preamplification of the 16S rRNA gene by PCR, HOPE reached an excellent detection sensitivity (18) and has been applied to the study of the Bacteroides and Prevotella groups within the Bacteroidales (19, 20), cyanobacteria (21), and Dehalococcoides (22).
In this study, we constructed a set of 27 16S rRNA-based primers with specificities spanning the domain, order, and genus levels for methanogens and combined them using the HOPE approach. After optimizing the method with reference strains, we used the developed HOPE technique to evaluate the relative abundances of methanogens inside the anaerobic digestion systems treating industrial wastewater and sewage sludge. By using this approach, we were able to provide insights into the structure, dynamics, and distributions of the methanogen populations among the evaluated microbial consortia and successfully identify the key groups associated with different anaerobic treatment systems. These findings could facilitate the development of improved methods for the management of methane production in anaerobic treatment systems.
MATERIALS AND METHODS
Methanogen reference strains.Twenty-four methanogen cultures obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ; Braunschweig, Germany) or the Bioresource Collection and Research Center (BCRC; Hsinchu, Taiwan) were used as the source of reference organisms (see Fig. S1 in the supplemental material). A Methanocella paludicola culture (strain SANAE) was kindly provided by Hiroyuki Imachi.
Environmental samples.In this study, we analyzed 19 consortium samples collected from different types of reactors and wastewater, sampling heights in a reactor, or sampling times during reactor operations using the HOPE method. As shown in Table 1, granule samples 1 and 2 were collected from full-scale up-flow anaerobic sludge bed (UASB) reactors treating purified terephthalic acid (PTA) wastewater, in which acetate, benzoate, terephthalate, and p-toluate account for the major components in the wastewater stream (23). The terephthalate-degrading biofilm sample (sample 3) was taken from a thermophilic (50°C) fixed film reactor (24), and the p-toluate-degrading granule sample (sample 4) was obtained from a mesophilic (35°C) fed-batch reactor in the laboratory. Both reactors inoculated with sludge sample 2 were operated for at least 6 months. In a 20-m-high full-scale anaerobic fluidized bed (AFB) reactor treating phenolic resin wastewater (25), 2 samples that were biofilm attached on granular activated carbon were obtained from sampling ports at 3-m and 19-m heights (samples 5 and 6, respectively). Sludge sample 7 was taken from a full-scale UASB reactor treating pulp and mill wastewater. Five sludge samples (samples 8 to 12) were taken from the sampling ports at different heights (0.3 m, 1 m, 2 m, 3 m, and 4 m, respectively) of a full-scale UASB reactor treating thin-film transistor liquid crystal display (TFT-LCD) manufacturing wastewater, in which tetramethyl ammonium hydroxide constituted the main wastewater ingredient (26). Samples 13 to 15 and samples 16 to 19 were obtained from 2 different full-scale anaerobic digesters for the treatment of waste sludge in a sewage treatment plant from September 2009 to January 2010. Samples were collected and stored at −80°C prior to analysis.
Methanogenic consortia analyzed by the HOPE method in this study
DNA preparation and PCR amplification of the archaeal 16S rRNA sequence.Total DNA from reference cultures and environmental samples was prepared using different methods. In brief, to quickly obtain DNA templates from reference cultures, the methanogen cells were harvested using centrifugation, washed with sterile Milli-Q water, and resuspended in a buffer solution containing 0.5% Triton X-100, 5 mmol/liter dithiothreitol, and 500 mmol/liter Tris-HCl (pH 7.5). Genomic DNA was prepared after disrupting the cell suspension using a microwave irradiation treatment, as described previously (27). For the recovery of DNA from various microbes in environmental samples, the microbial cells were subjected to freeze-thaw, enzymatic, and chemical treatments, and their community DNAs were purified and precipitated using phenol-chloroform-isoamyl alcohol and ethanol precipitation procedures (28), followed by cleanup with a Wizard DNA cleanup kit (Promega, Madison, WI). The archaeal 16S rRNA gene fragments were obtained by PCR amplification with forward primer A1F (29) or A109F and reverse primer 1492R (30). Approximately 5 to 50 ng genomic DNA was used in a 100-μl PCR solution and subjected to thermocycling: 95°C for 5 min; 20 to 25 cycles of DNA denaturation at 95°C for 45 s, annealing at 52°C for 30 s, and primer extension at 72°C for 45 s; and a final extension at 72°C for 5 min in a thermocycler (Biometra, Germany). The amplicons were concentrated and purified using a QIAquick PCR purification kit (Qiagen), and their concentrations were quantified using a NanoVue spectrophotometer (Fisher Scientific, Uppsala, Sweden). Partial 16S rRNA sequences were obtained for validating the purity of the reference methanogens.
Design and arrangement of hierarchical primers.The 16S rRNA-targeted oligonucleotide primers (Table 2) were newly designed or modified from primers/probes for methanogens reported previously. These primers were assessed in silico according to key criteria, including primer characteristics (direction, extended nucleotides, length, GC content, and melting temperature) and the number and position of the mismatch that occurred in the nontargets from the study by Wu et al. (31) using ARB software (32) and an aligned 16S rRNA sequence database (SSU Ref version 96; http://www.arb-silva.de) that was updated with 420 almost-full-length archaeal sequences (∼1,400 nucleotides [nt]) reported from 2009 to 2011. The stabilities of the primer-template duplex and other possible duplex structures (hairpin, self-dimer, and heterodimer) were evaluated using the tools provided with OligoAnalyzer version 3.1 software (Integrated DNA Technologies). Multiple primers were arranged in a single tube on the basis of the phylogenetic affiliations of the methanogen targets (Table 2 and Fig. 1), so that the relative abundance of the individual archaeal target detected by each specific primer relative to the abundances of the targets detected by higher-level primers could be determined. The melting temperature and the labeled dye terminator (color) of the primers were also considered to ensure consistency in the fluorescence signals among the primers in a tube. To clearly differentiate extended primers in the same HOPE reaction mixture, primers that were extended with the same dye terminator were modified using different lengths of tails at their 5′ ends.
Oligonucleotide primers used for HOPE analysis in this study
Phylogenetic tree of known methanogens based on neighbor-joining analysis of 16S rRNA gene sequences. A total of 27 oligonucleotide primers were designed to target the methanogens at the specificities of domain (1 primer), order (4 primers), and genus (22 primers). The associated coverage (name) of each primer and the corresponding combinations (MTPX0 to MTPX4) used for multiplexing are indicated at the right of the tree. Aquifex pyrophilus was used as an outgroup. The bar indicates 5% base substitution.
HOPE analysis.For HOPE analysis, a single-nucleotide primer extension reaction mixture was prepared in a 5-μl solution containing 2.5 μl of the ready-to-use premix from a SNaPshot multiplex kit (Applied Biosystems), 5 to 20 ng purified DNA amplicons, and 1.5 μM (each) the unlabeled oligonucleotide primers (Mission Biotech, Taiwan). The reaction was conducted in a thermocycler (Biometra, Germany) with 20 cycles of 96°C for 10 s, 62°C for 20 s, and 72°C for 15 s. The 3′ ends of the primers were extended with a single dye terminator upon the primer-template duplex formed, because the premix contained AmpliTaq DNA polymerase, salt, and buffer. The resulting products were then treated with shrimp alkaline phosphatase (Promega, Germany) at 37°C to minimize the possible background fluorescence caused by unincorporated dye terminators (21), prior to their subjection to capillary electrophoresis with a GeneScan Liz 120 standard using an ABI Prism 3130 genetic analyzer (Applied Biosystems). The electrophoretic program included a denaturation step (60°C), an injection step in which a voltage of 2.1 kV was applied for 40 s, and a separation step. The electrophoretic size and fluorescent intensity of each of the extended primers were recorded for calculation of calibration factors and the relative abundances of the methanogen populations, as described previously (20). To ensure data reproducibility, each sample was analyzed in triplicate.
PCA.A principal component analysis (PCA) was used to evaluate the structural similarities among the methanogen populations quantified in the samples. Prior to the PCA, the relative abundances of methanogens detected using genus-specific primers were expressed as the percentages of the total archaeal population detected using an Archaea-specific primer. The data matrix (19 cases × 11 variables), based on the covariances, was analyzed using Statistica version 8.0 software (StatSoft).
Construction of archaeal 16S rRNA gene clone libraries and phylogenetic analysis.The archaeal 16S rRNA gene clone libraries were constructed using the PCR amplicons and a TOPO TA cloning kit (Invitrogen, CA). The cloned inserts that shared a banding profile in restriction fragment length polymorphism analysis with the 2 tetramer restriction enzymes HaeIII and TaqI (New England BioLabs, MA) were grouped into one phylotype, and the sequences of each phylotype were analyzed using an ABI 3130 genetic analyzer (Applied Biosystems, Foster City, CA). The detected archaeal phylotypes were checked for chimeric artifacts using the Pintail program (33), and then the sequence were compared with the sequences in the entire database using a BLAST search. The topology of the phylogenetic tree with 1,000 replicates of bootstrap analysis was constructed with the neighbor-joining method and the Jukes-Cantor correction using the MEGA5 program (34).
RESULTS
Hierarchical primers and specificities.To analyze the methanogens occurring in the methanogenic consortia from wastewater treatment ecosystems using HOPE, we developed 27 16S rRNA-targeted oligonucleotide primers at different taxonomic specificities. The primers were designed with a binding length of 17 to 24 nt and a predicted melting temperature of 47.9°C to 66.8°C (Table 2). To achieve the highest discrimination capability with the fidelity of the Taq polymerase, the primers were usually designed with at least 2 mismatches against the nontargeted sequences (17) or a single mismatch at the terminal or penultimate position from the 3′ end of the primer (31).
As shown in Fig. 1, one Archaea-specific primer, ARC911fm, extended with a G nucleotide could detect most cultivated and uncultivated methanogens. To detect methanogen populations at the order level, four primers, including MSMX859f, MG1197fm, MB1167fm, and MCMT1008fm, specifically extended with a C, C, C, and G nucleotide in Methanosarcinales, Methanomicrobiales, Methanobacteriales, and Methanococcales, respectively, were used. For analyzing the acetotrophic, methylotrophic, and hydrogenotrophic groups of methanogens down to the genus level, 22 specific primers were developed. Among them, 9 primers were newly designed and the remaining 13 were adopted from previous studies, with modifications (Table 2). The MS537fm and F3SC979fm primers could be extended with various nucleotides for different groups within Methanosarcina and Methanofollis (Table 2), respectively, conferring their abilities for multiplex detections per primer. Under the stringent thermal conditions (i.e., a 62°C annealing temperature), all primers could be correctly extended with the dye terminators for prediction of the targeted methanogens but could not be extended when using the DNA of nontargeted methanogens as the templates (see Fig. S1 in the supplemental material). The single-mismatch discrimination capability of the HOPE technique was confirmed using the primer GMG266f, which could not be extended when testing DNAs with a single mismatch from Methanoplanus petrolearius (DSM 11571), Methanofollis liminatans (DSM 4140), and Methanolacinia paynteri (DSM 2545) (see Fig. S2 in the supplemental material).
Primer combination for multiplexing.The primers were arranged into 7 sets, with each combination achieving 4 to 6 multiplex detections (Table 2 and Fig. 1). The tube with multiplex primer set 0 (MTPX0) combined an Archaea-specific primer, ARC911fm, with 4 order-level primers to detect the methanogens at higher taxonomic ranks, while the other sets (MTPX1 to MTPX4), which each contained one order-level primer and 3 to 5 corresponding genus-specific primers, could detect the targeted methanogens within the respective orders. Therefore, the HOPE approach using 7 primer sets could provide phylogenetic information on the targeted methanogens at the domain, order, and genus ranks, as well as determine their relative abundances in relation to the total amplified 16S rRNA genes at the order and domain levels. To facilitate the separation from the primers extended with the same dye terminators, 9 primers were modified by adding poly(T) tails or nonsense sequences (4 to 24 nt) at the 5′ ends of the primers (Table 2). Figure S3 in the supplemental material, for example, displays an electrophoretic illustration showing the separation of 5 primers in MTPX2b.
Detection limit.To evaluate the lowest detection limit, we prepared a series of PCR-amplified 16S rRNA gene mixtures with relative abundances ranging from 2.5% to 0.16% Methanoculleus palmolei (DSM 4273) in a total template amount of 10 ng to test primers F2SC660fm and MG1197fm, which displayed low fluorescence responses. As shown in Fig. 2a, the results indicated that the fluorescence intensities of both extended primers decreased linearly (r2 > 0.999), corresponding to a decrease in the amount of DNA of the target M. palmolei (DSM 4273), and suggested an approximate abundance of at least 0.16%. We then evaluated the detection limit of the HOPE approach using the PCR-amplified 16S rRNA genes from sample 3. As shown in Fig. 2b, HOPE analysis with primer set MTPX0 could detect methanogens from the orders Methanosarcinales and Methanomicrobiales, which accounted for 94.3% ± 4.5% and 6.6% ± 0.1% of the amplified Archaea 16S rRNA genes, respectively. Using the tube MTPX2b, as shown in Fig. 2c, the results further indicated the presence of Methanospirillum targeted by F7SC1257f under all tested analytical conditions, with a consistent abundance of 10.7% to 11.9% within the amplified Methanomicrobiales 16S rRNA genes. Methanoculleus targeted by F2SC660fm, with a relative abundance of 3.0% ± 0.5% within the amplified Methanomicrobiales 16S rRNA genes, was detectable with ≥10 ng of the template only. Consequently, the HOPE technique could successfully detect specific methanogens with relative abundances of as low as 0.2% of the total archaeal 16S rRNA genes amplified from a complex environmental sample. This detection sensitivity is comparable to that of the previous HOPE method developed for the detection of Bacteroides and Prevotella within the Bacteroidales (18–20). To achieve improved detection performance, we applied 15 ng DNA per environmental sample as a template for the HOPE analysis.
Evaluation of the detection limits of HOPE analysis. (a) Fluorescence response in terms of the peak area of the extended primers ddGTP-F2SC660fm and ddCTP-MG1197fm in the HOPE analysis using the tube MTPX2b and known concentrations (0.16% to 2.5%) of the 16S rRNA gene amplified from targeted Methanoculleus palmolei (DSM 4273) pooled with nontargeted Methanocella paludicola strain SANAE as templates. (b) Relative abundances of order-level methanogens with respect to the total archaeal populations detected using the tube MTPX0 and amplified 16S rRNA gene products from a terephthalate-degrading biofilm (sample 3) as the template. The methanogens affiliated with the orders Methanobacteriales targeted by primer MB1167fm and Methanococcales targeted by primer MCMT1008fm were not detected. (c) Relative abundances of genus-level methanogens with respect to the total Methanomicrobiales populations detected using the tube MTPX2b and varied amounts of amplified 16S rRNA genes from terephthalate-degrading biofilm (sample 3) as the template. The methanogens affiliated with the genera Methanogenium targeted by primer GMG266f and Methanofollis targeted by primer F3SC979fm were not detected.
Relative abundances of methanogens in consortia analyzed by HOPE.Figure 3 shows the results for the anaerobic sludge samples analyzed using the HOPE approach with 7 sets of primers under the optimized conditions. Among the 27 hierarchical primers, 12 primers exhibited positive primer extension reactions for the archaeal 16S rRNA genes amplified from the 19 sludge samples. Figure 3a displays a color-coded array showing the distribution of the relative abundances of the methanogens detected at different hierarchical levels. At the order level, the results showed that the Methanosarcinales-related methanogens detected using the MSMX859f primer occurred in all samples in proportions ranging from 19.5% to 102.9% (standard deviations [SDs], 0.6% to 7.3%) of the PCR-amplified archaeal 16S rRNA genes. The proportion of >100% detected in some cases was likely the consequence of technical variability. In 9 and 12 samples, respectively, methanogens in the orders Methanobacteriales (ddCTP-MB1167fm) and Methanomicrobiales (ddCTP-MG1197fm) were detected at relative abundances of 3.3% to 80.7% (SDs, 0.7% to 8.4%) and 3.6% to 52.5% (SDs, 0.1% to 5.9%) in the archaeal 16S rRNA genes amplified from the samples, respectively. The MCMT1008fm primer did not detect the methanogens related to the order Methanococcales in any of the sludge samples. The HOPE analysis suggested high detection coverage of methanogens (>92.8% of amplified Archaea 16S rRNA genes) in 14 sludge samples. However, about 12.2% to 22.8% of the archaeal 16S rRNA genes amplified from 5 samples were not covered by the 4 primers. These most likely represented other methanogens or nonmethanogen archaeal archaeons.
(a) Color-coded representation of the relative abundances of methanogens detected using HOPE analysis with the 7 primer tubes and amplified 16S rRNA gene products obtained from 19 sludge samples as the templates. (Top) Relative abundances of order-level methanogens with respect to the total archaeal populations analyzed with the tube MTPX0; (bottom) relative abundances of the genus-level methanogens with respect to the corresponding taxonomic orders analyzed using the tubes MTPX1 to MTPX4. The percentages were transformed into a gray scale ranging from 0% (not detected [ND], white) to 100% (black) at intervals of 5%. Abbreviations: U, up-flow anaerobic sludge bed reactor; I, anaerobic fixed film reactor; F, anaerobic fluidized bed reactor; T, fed-batch reactor; A, anaerobic digester; G, granular sludge; B, attached biofilm; S, sludge floc. (b) Spatial distribution of Methanosarcinales-related (■) and Methanobacteriales-related (□) methanogens with respect to the total archaeal populations; (c) spatial distribution of Methanosarcina-related (■), Methanosaeta-related (□), and Methanomethylovorans-related (
) methanogens with respect to the order Methanosarcinales; (d) spatial distribution of Methanothermobacter-related methanogens (■) with respect to the order Methanobacteriales in a full-scale up-flow anaerobic sludge bed reactor treating TFT-LCD wastewater. The bar represents the standard deviation from triplicate analyses.
Figure 3a further shows the distribution of the relative abundances detected using genus-level primers in relation to their corresponding order-level specificities. It was observed that 8 of the 23 genus-specific primers provided positive dye terminator-extending reactions in the 19 sludge samples. These primers detected 2 or 3 methanogen genera for each order and covered total proportions of 14.3% to 85.7%, 0% to 50.7%, and 0% to 1.2% within the amplified 16S rRNA genes related to the orders Methanosarcinales, Methanomicrobiales, and Methanobacteriales, respectively. Within the order Methanosarcinales, the tube MTPX1a detected the MX802f-targeted members of the acetotrophic Methanosaeta to be predominant in the 18 sludge samples. Their relative abundance in amplified genes was typically relatively high (60.4% to 90.9% within the Methanosarcinales) but could decrease to <4.3% in the sludge (samples 8 and 12) from a full-scale UASB reactor treating TFT-LCD wastewater. The MTPX1b tube frequently detected Methanosarcina, the other acetotrophic methanogen, with the ddGTP-extended primer MS537fm in the samples from full-scale reactors treating TFT-LCD wastewater (samples 8 to 12) and sewage sludge (samples 13 to 19). Their relative abundances in amplified genes ranged from 0.8% to 25.1% within the Methanosarcinales. In addition, the methylotrophic Methanomethylovorans spp., accounting for a relative abundance of ≤71.5% in the amplified Methanosarcinales 16S rRNA genes, were also frequently detected in 7 sludge samples using the newly designed primer MML419r.
Within the order Methanomicrobiales, the tubes MTPX2a and MTPX2b could detect the methanogens from 3 genera (Methanoculleus, targeted by F2SC660fm; Methanospirillum, targeted by F7SC1257f; and Methanolinea, targeted by ML398fm) in the samples obtained from the reactors treating sewage sludge (samples 13 to 19) and PTA wastewater-related substrates (samples 1 to 4). According to the obtained data, in the amplified Methanomicrobiales 16S rRNA gene products, the Methanolinea spp. (10.6% to 101%; SDs, 0.1% to 3.4%) were more abundant than the Methanoculleus spp. (1.7% to 27.6%; SDs, <5.0%) and the Methanospirillum spp. (3.3% to 11.5%; SDs, 0.2% to 0.6%) in these samples. In addition, 2 genera within the order Methanobacteriales could be detected using the tube MTPX3. One was Methanothermobacter, targeted by GMTB645f, which occurred with a relative abundance of 0.9% to 15.9% (SDs, <5.0%) within the amplified Methanobacteriales 16S rRNA genes from samples 5, 6, 9, 11, 15, and 19, and the other was Methanobrevibacter, targeted by GMB376rm, which occurred with a relative abundance of 0.9% to 7.1% within the amplified Methanobacteriales 16S rRNA genes (SDs, 0.1% to 0.2%) from the anaerobic digesters (samples 15 and 19). Our observations on these 5 methanogen genera suggested the occurrence of hydrogenotrophic methanogenesis in the reactors.
Because the methanogen communities in the sludge samples collected from different heights in a full-scale UASB reactor treating TFT-LCD wastewater (samples 8 to 12) showed marked variations, we conducted further analyses. As shown in Fig. 3b, at the order level, Methanosarcinales-related methanogens displayed a high abundance within the Archaea (59.8% to 73.5% of the amplified 16S rRNA genes; SDs, 1.9% to 2.9%) at the bottom of the sludge bed and higher spaces of the reactor, whereas those from the order Methanobacteriales were abundant (75.2% to 88.7% of the amplified 16S rRNA genes; SDs, 0.1% to 8.5%) in the center of the reactor (1 to 2 m high). The relative abundance of the Methanomicrobiales-related methanogens in the 5 samples tested was below the detection level (<0.2%), suggesting a low density of the population in the ecosystem. At the genus level, as shown in Fig. 3c, the relative abundance of the Methanomethylovorans spp. was 60.6% to 71.5% (SDs, 1.4% to 2.2%) within the Methanosarcinales 16S rRNA genes or 44.5% to 49.2% within the Archaea 16S rRNA genes amplified at the top and bottom of the reactor. Within the amplified Methanosarcinales 16S rRNA genes, the Methanosaeta spp. were relatively more abundant (61.0% to 72.4%; SDs, 0.6% to 3.5%) than the Methanomethylovorans spp. (∼18.2%; SDs, 0.6% to 3.2%) and Methanosarcina spp. (0.8% to 4.2%; SDs, <0.1%) at reactor heights of 1 to 3 m, and the Methanobacteriales-related hydrogenotrophic populations, which accounted for 32.9% to 80.7% (SDs, 0.1% to 8.4%) of the amplified 16S rRNA genes, were highly predominant. However, aside from the Methanothermobacter spp., approximately 96.4% of the amplified 16S rRNA genes associated with Methanobacteriales were unknown (Fig. 3d).
PCA.The results from PCA of the obtained data revealed that the 19 analyzed consortium samples could be classified into 3 distinct clusters, which indicated the degree of population similarity among the samples (Fig. 4). Cluster I consisted of samples 1 to 7, collected from the reactors treating wastewater predominantly containing aromatic compounds, and could be differentiated by a dominant acetotrophic Methanosaeta population (67.1% ± 11.6% of the amplified Archaea 16S rRNA genes). Cluster II consisted of samples 8 to 12, collected from a full-scale UASB reactor treating TFT-LCD wastewater, which displayed a wide distribution in the coordination space. The members of the methylotrophic Methanomethylovorans (46.9% ± 3.3% of the amplified Archaea 16S rRNA genes) or the unknown Methanobacteriales-related populations (77.6% ± 4.3% of the amplified Archaea 16S rRNA genes) were dominant in this cluster. Cluster III was represented by samples 13 to 19, which were collected from the 2 full-scale anaerobic digesters on different dates, and was associated with a high abundance of the hydrogenotrophic Methanolinea spp. in the archaeal communities (39.1% ± 6.0% of the amplified Archaea 16S rRNA genes).
PCA for grouping the genus-level methanogens distributed among the 19 sludge samples analyzed using HOPE. Symbols: ○, sample number; □, genus-level methanogens detected; Δ, unknown groups related to a known order.
Clone library analysis of the 16S rRNA gene sequences.To validate the HOPE analysis results, 2 archaeal 16S rRNA gene clone libraries were constructed using the DNA from samples 2 and 18. As shown in Fig. 5, all of the sequences obtained in the clone library of sludge sample 2 were phylogenetically associated with the known methanogens. Among them, 92.4%, 4.4%, and 3.8% of the total clones were related to Methanosarcinales, Methanomicrobiales, and Methanobacteriales, respectively. The most predominant methanogens were the acetotrophic Methanosaeta (87% of the total clones), followed by the methylotrophic Methanomethylovorans (5.4% of the total clones). The hydrogenotrophic methanogens detected included Methanolinea, Methanobacterium, and Methanoculleus, each accounting for <4.3% of the total clones. The other clone library (sample 18) consisted of the sequences retrieved from Methanosarcinales (75% of the total clones), Methanomicrobiales (18% of the total clones), and uncultured Euryarchaeota (6% of the total clones). The most predominant phylotype was the acetotrophic methanogen Methanosaeta (75% of the total clones), followed by the 2 hydrogenotrophic populations Methanospirillum (9% of the total clones) and Methanoregula (3% of the total clones).
Phylogenetic tree of archaeal 16S rRNA gene sequences retrieved from the sludge of a full-scale UASB reactor treating PTA wastewater (sample 2) and a full-scale anaerobic digester (sample 18). The tree was reconstructed with the 16S rRNA sequences (>1,200 bp) of the clones and their close relatives obtained in this study and from the NCBI database on the basis of the neighbor-joining method with 1,000 bootstraps and rooted with Aquifex pyrophilus. Black circles at the branch nodes represent the bootstrap values greater than 75%. The bar shows 5% base substitutions. Boldface indicates the cloned representatives and their percentages in the clone library analysis in this study, and those labeled with PTA granule and anaerobic digester (AD) sludge clones were obtained from samples 2 and 18, respectively. The predicated phylogeny is indicated in brackets at the right. TCE, trichloroethylene.
The results from the clone libraries were generally in accordance with those from HOPE analysis. Clone library results indicated that the predominant Methanosaeta spp. accounted for 75% to 87% of the total clones, and HOPE analysis provided a result of 43% to 64.1% of the amplified Archaea 16S rRNA genes. The clone library and HOPE analysis results showed the hydrogenotrophic Methanolinea spp. (4.3% versus 6.5%), Methanoculleus spp. (1.1% versus 2.6%), and Methanospirillum spp. (9% versus 1.5%) and the methylotrophic Methanomethylovorans spp. (5.4% versus 19.4%) to have comparable abundances. However, we observed that certain differences were displayed between the 2 approaches. For example, for sludge sample 18, the Methanolinea spp. and Methanosarcina spp. with an observable abundance (13.3% to 38.0% of amplified Archaea 16S rRNA genes) could be detected using HOPE with primers ML398fm and MS537fm, respectively, whereas the clone library did not retrieve the relevant sequences. The clone library could identify the relative abundance (3.3%) of a Methanobacterium-related phylotype (R27-A39) within the Methanobacteriales; however, the HOPE analysis using the MB1167fm primer did not detect the methanogens from Methanobacteriales.
DISCUSSION
In this study, we successfully developed the HOPE approach to detect the relative abundances of numerous methanogen populations in the pool of archaeal 16S rRNA genes amplified from microbial consortia. After optimizing for specificity, sensitivity, and multiplexing capability, the approach was able to detect known methanogens with a total abundance of up to 77.2% to 107.9% in amplified 16S rRNA genes from total archaeal populations in anaerobic reactors used for the treatment of various industrial wastewaters and sewage sludge at order and genus specificities. The obtained results strongly suggest that HOPE analysis can improve understanding of the relative abundance of methanogens in complex bioreactor ecosystems.
According to the results obtained by HOPE, the analyzed methanogenic consortia predominantly consisted of methanogen populations from the order Methanosarcinales (ddCTP-extended MSMX859f), Methanomicrobiales (ddCTP-extended MG1197fm), and/or Methanobacteriales (ddCTP-extended MB1167fm). The acetotrophic and methylotrophic methanogens in the order Methanosarcinales generally dominated the archaeal communities. The acetotrophic Methanosaeta spp. were the most commonly detected methanogens in the samples, with a high abundance in sample cluster I (67.1%; SD, 11.6%) and a medium abundance in sample cluster III (38.7%; SD, 4.1%) (Fig. 3 and 4). The results from clone library analysis supported these observations, showing that 87% and 75% of the total clones were associated with the Methanosaeta in sludge samples 2 and 18, respectively (Fig. 5). These findings were in high accordance with those from several previous studies (25, 28, 35–37). The dominance of the Methanosaeta population indicated its success within the niche for the production of methane from acetate, a common constituent in industrial wastewater (23) as well as a central metabolite resulting from the anaerobic fermentation of substances. In addition to direct use by the acetotrophic methanogens, acetate can be converted by way of a syntrophic interaction between the acetate-utilizing bacteria and hydrogenotrophic methanogens in anaerobic sludge digesters (38, 39). This process might be reflected by our observations of a high abundance of the hydrogenotrophic methanogens and a slightly lower abundance of the acetotrophic methanogens (Methanosaeta and Methanosarcina) in the anaerobic digesters.
In further analyses, we observed Methanosaeta spp. and Methanosarcina spp. in sample clusters II and III, suggesting competition for acetate. The Methanosaeta population has a high affinity (lower Ks) for acetate and, thus, typically displays a competitive advantage over the Methanosarcina spp. at low acetate concentrations (40). However, Methanosarcina could produce methane using all three known methanogenesis pathways (2) and would be more competitive than Methanosaeta in high acetate concentrations (41) and high temperatures (55 to 60°C) (17). In this study, the Methanosarcina population showed an approximately 16.5% higher abundance than Methanosaeta in sample 12, which was collected from a sampling port in proximity to the effluent of the UASB reactor. This observation suggests that it might confer high competitive capabilities for using substrates such as acetate and methylated amines under more oxidizing redox conditions.
Although several studies have detected 16S rRNA sequences related to the Methanomethylovorans-like archaeons in freshwater sediments (42) as well as in anaerobic reactors (9, 43, 44) using DGGE, cloning/sequencing, and SSCP techniques, their quantity is poorly understood. To our knowledge, no previous study has designed probes/primers for quantifying the relative abundance of Methanomethylovorans spp. in environmental samples (17). In this study, we were able to effectually evaluate the relative abundance of Methanomethylovorans 16S rRNA gene sequences amplified from the samples of 4 reactors using the HOPE technique and a newly designed primer, MML419r. The finding indicated an important role of Methanomethylovorans in the ecosystem of anaerobic reactors and revealed that for the anaerobic treatment of industrial wastewater, methylotrophic methanogenesis contributes a proportion of the methane formed. This process was particularly prominent in the treatment of TFT-LCD wastewater (i.e., cluster II sludge samples), because the principal wastewater constituent was methylated compounds.
The known members within the orders Methanomicrobiales and Methanobacteriales are hydrogenotrophic. These groups of methanogens were commonly detected inside the anaerobic reactors, and their abundances could also be clearly differentiated among the 3 clusters of samples. For example, the amplified 16S rRNA genes of hydrogenotrophic methanogens from the order Methanomicrobiales exhibited a high abundance (34.2% to 48.5%) in sample cluster III (samples 13 to 19) and a medium-level abundance (∼7.0%) in sample cluster I (samples 1 to 7) but were undetectable in sample cluster II. Closer examination at the genus level revealed the prevalence of 3 genera. The Methanospirillum spp. and Methanoculleus spp. are important hydrogen scavengers in anaerobic digesters (9, 45) and constituted 0.2% to 1.6% and 0.1% to 3.2% of the amplified Archaea 16S rRNA genes, respectively, in the reactors treating sewage sludge, terephthalate, and PTA wastewater. Imachi et al. recently identified the Methanolinea genus, obtaining the first species, Methanolinea tarda, from an anaerobic digester (46). The HOPE analysis findings suggested that the Methanolinea spp. were relatively abundant (31.6% to 45.0% of the amplified Archaea 16S rRNA genes) in anaerobic digesters (cluster III sludge samples), and their abundances could be distinctively discriminated from other sources of sludge samples (0.7% to 6.5% of the amplified Archaea 16S rRNA genes, clusters I and II). In addition to reactor ecosystems, this methanogenic archaeon was present in natural environments (47–49). Its relatively extensive distribution suggests that this novel archaeon plays an important role in various methanogenic ecosystems.
In this study, the HOPE method also detected members in the order Methanobacteriales, hydrogenotrophic methanogens from the genera Methanobrevibacter and Methanothermobacter that have frequently been observed in methanogenic processes (9, 45, 50), using 2 new primers (GMB376rm and GMTB645f, respectively). Their individual abundances were usually lower than 1.2% in the amplified Archaea 16S rRNA genes, with no distinct distribution pattern among the 3 sample clusters (Fig. 3 and 4). In a previous study with the RNase H method, it was found that the 2 populations were active and represented the major groups within the Methanobacteriaceae in the reactors treating sugar- and clear-liquor-processing wastewaters (17). However, among the 6 samples in this study, the Methanobrevibacter spp. and Methanothermobacter spp. accounted for <19.2% of the amplified Methanobacteriales 16S rRNA genes, suggesting the presence of a large proportion of other methanogens within the same order. As shown in Fig. 4, the high predominance (∼80.7% of the amplified Archaea 16S rRNA genes) of the unknown Methanobacteriales-related groups suggested their significance in the full-scale UASB reactor treating TFT-LCD wastewater. The low percentages covered by the genus-specific primers within the order Methanobacteriales were likely attributed to a lack of primers to detect the genus Methanobacterium, which was actually detected using the cloning library (Fig. 5), and/or the mysterious groups. It should be particularly mentioned that no primers with a good coverage for the genus Methanobacterium could be developed. We actually excluded a primer, GMBA755, designed by Narihiro and coworkers (17) from the present work, because the specificity would be compromised in the HOPE analysis. To solve this, several primers with individual specificities below the genus level can be designed to collectively achieve a better coverage of various groups within the genus Methanobacterium in the future, and HOPE can easily detect a target at the species level (18). With the advantages in detection sensitivity and throughput, HOPE can complement other available molecular methods, such as the cloning library method, to further study the mysterious groups.
Our study results further demonstrated the ability of the HOPE approach to identify the associations between reactor operations and methanogen diversity. As shown in Fig. 3, the population structures of the methanogens observed in sludge samples 13 to 19 were relatively constant, suggesting the stable dynamics of the archaeal community in the 2 anaerobic digesters within the sampling periods. In contrast, as shown in Fig. 3b to d, the abundances of the methanogen populations in a full-scale UASB reactor treating TFT-LCD wastewater showed marked changes along the height of the reactor. Similarly, in a full-scale AFB reactor, the Methanomethylovorans spp. and Methanomicrobiales-related populations, including Methanoculleus spp., were detected in the biofilm sampled at a 3-m height in the reactor but were undetectable in the biofilm sampled at the top of the reactor (19 m). It is likely that this differential distribution pattern reflects differences in environmental conditions, which might be associated with the vertical configuration of the reactors, operational conditions (organic loading, up-flow velocity, and effluent recirculation), and wastewater characteristics.
As demonstrated, HOPE represents a hierarchical approach to analyze the entire set of PCR-amplified products from the archaeal community. Using the optimized conditions obtained in this study, the method is considerably straightforward for estimating the relative abundance of known methanogen targets at a sensitivity corresponding to 0.2% of the amplified 16S rRNA genes (Fig. 2). The reliability of the HOPE results obtained in this study was validated using a clone library approach (Fig. 5), in which the abundance of a specific phylotype is dependent on the ultimate number of clones analyzed. Similar to the cloning library, the use of HOPE to determine the relative abundance of PCR-amplified products can be subject to PCR biases (51), while it is occasionally difficult to obtain a single primer with specificity for a diverse group. These are the likely causes of the dissimilarities in the results between the HOPE method and cloning library analysis. Therefore, special care, such as collection of DNA products from sludge samples at the exponential phase of PCR amplification (19) and careful evaluation of primer specificity (31), should be taken to achieve the accuracy of relative abundance quantitation.
It is clear that through comparison of the analytical throughput in a single reaction, the HOPE method can be in a niche between the low-throughput methods, such as FISH, qPCR, and the RNase H method, and the high-throughput microarray platform method. The developed HOPE assay can detect 22 targeted methanogens at different taxonomic levels (513 reactions = 27 primers × 19 samples) within 8 to 9 h, a time which is comparable to that needed for the RNase H method (17). To have a complete set of methanogen targets, more primers should be designed for the recently recognized Methanocellales order and 11 methanogen genera not yet included in this study. The total number of methanogen targets can be flexibly increased by adding more HOPE reactions or by adding a primer(s) to individual reactions (Fig. 1). When combined with a qPCR to analyze the absolute quantity of total archaeal 16S rRNA genes, the HOPE analysis can be more quantitative for methanogens and has high applicability to the samples from natural environments.
In summary, the HOPE method with a large number of primers can successfully determine the relative abundances of known methanogens in amplified 16S rRNA gene products at order and genus specificities in a rapid and high-throughput manner. The overall results provide insight into the diversity, temporal dynamics, and spatial distributions of the acetotrophic, hydrogenotrophic, and methylotrophic methanogens and suggest that in addition to the acetotrophic Methanosaeta, the recently identified hydrogenotrophic Methanolinea and methylotrophic Methanomethylovorans play important roles inside anaerobic reactors treating industrial wastewater and sewage sludge. Considering its technical simplicity and robust multiplexing capabilities, the HOPE approach can be used as a platform for the routine monitoring of methanogen populations for improved management of process operations and could facilitate relevant studies of the methanogen populations in various methanogenic ecosystems, especially whenever analysis of a large number of samples is needed.
ACKNOWLEDGMENTS
We thank I-An Tsai and Tsuyoshi Yamaguchi for their experimental assistance and Hiroyuki Imachi for kindly providing the reference strain. The comments of W.-T. Liu and H. Tamaki were deeply appreciated, too.
This research was supported by a grant (NSC97-2221-E006-038-MY3) from Taiwan's National Science Council.
FOOTNOTES
- Received 21 July 2013.
- Accepted 23 September 2013.
- Accepted manuscript posted online 27 September 2013.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.02450-13.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
