The mesophilic inoculum was taken from a biowaste co-utilizing wastewater treatment plant in Zirl (Austria) [25 (link)] with a reactor capacity of 1350 m3, an operation temperature of 39 (±0.2) °C, pH of 7.4 (±0.21), and total solids content of 2.2 (±0.04) g 100 g−1 fresh weight. The thermophilic inoculum was derived from the outlet sampling port of a thermophilic anaerobic digestion plant in Roppen (Austria) where about 2.500 tons of green waste and 6.200 tons of biowaste are treated per year [26 (link)], with a total reactor capacity of 900 m3, an operation temperature of 53 (±0.3) °C, pH of 7.9 (±0.44), and a total solids content of 26.2 (±2.0) g 100 g−1 fresh weight. Additional information regarding digester conditions and characteristics can be looked up elsewhere [22 (link)]. Plastic bottles filled with sludge were tightly sealed and immediately brought to the laboratory. For liquid handling, the sludge was sieved and diluted as described previously [27 (link),28 (link)]. The headspace was exchanged with a N2/CO2 (70:30)-gas mixture. The prepared samples were incubated at 37 °C and 52 °C for 15 days (mesophiles) and 20 days (thermophiles), respectively, until the sum of volatile fatty acids (VFA) was <200 mg kg−1. Subsequently, the samples were stored at 4 °C until further use.
Straw from grain (straw) was air-dried, but otherwise not chemically, physically, or biologically (pre)treated. The straw was cut into pieces 4–7 cm long. The C/N ratio of the straw (ratio: 56) was analysed with a TruSpec® CHN analyser (Leco, Germany) according to the manufacturer’s protocol. The straw was filled into 120 mL serum flasks, functioning as batch reactors, in different carbon-load concentrations with 3 (defined as low carbon load, LCL), 34 (defined as medium carbon load, MCL), and 170 (defined as high carbon load, HCL) mmol carbon-C reactor−1, respectively.
A basal anaerobic broth based on previous investigations [29 (link)] was prepared and modified as follows (per litre): 0.4 g NaCl, 0.4 g MgCl2 × 6 H2O, 0.68 g KH2PO4, 0.18 g NaOH, 0.05 g CaCl2 × 2 H2O, 0.4 g NH4Cl, 0.5 g L-cysteine, 10.0 g sodium carboxymethylcellulose (CMC), 0.5 g yeast extract, 2.0 g sodium acetate, 1.0 g sodium formiate, 1 mL vitamin solution [29 (link)], 1 mL trace element solution SL-10 (German Collection of Microorganisms and Cell Cultures GmbH (DSMZ), Braunschweig, Germany), 2 mL sodium sulfide solution (120 g L−1 Na2S), and 1 mL resazurine solution (1.15 g L−1 resazurine). After the pH was adjusted with 0.1 M sodium hydroxide to 7.5 ± 0.2, 48 mL of the medium was filled into the 120 mL serum flasks which had previously been filled with straw (as described above). A control containing the anaerobic broth but no straw was also included and equally treated thenceforward. The sealing and headspace gas exchange took place according to previous protocols [22 (link)]. The flasks were subsequently autoclaved and cooled down before further use.
For each temperature regime, a volume of 12 mL diluted inoculum was injected into each reactor. Subsequently, the reactors were incubated at 37 °C and 52 °C, respectively, extending over an anaerobic incubation period of 28 days. All variations were prepared in triplicate. Samples were taken on day 2, 4, 7, 14, 21, and 28. Liquid samples for pH, VFA, phenyl acids, and C/Nliquid were processed immediately or frozen at −20 °C. The pH of the samples was measured with pH indicator strips 4.0–9.0 (Merck, Germany).
For each temperature regime, a PCR-DGGE approach [30 (link),31 (link)] was conducted with all variants of day 0 to check for the same microbial community structure at the beginning of the experiment. Moreover, control samples of day 0, as well as samples of day 14 and 28 were used for next-generation sequencing (NGS) analyses.
VFA, total carbon, total nitrogen (C/Nliquid ratio), as well as phenyl acid analyses were done according to previous studies [22 (link),28 (link),32 (link)]. The gas over-pressure was measured with a GHM Greisinger GDH 200 sensor and used to calculate biogas and methane production [NmL] as described previously [27 (link)].
Liquid samples (1 mL) from day 0, 14, and 28 were centrifuged at 20,000 g for 15 min and resuspended in 1 mL sterile ¼ Ringer solution. Subsequently, DNA extraction was done using the Soil Extract II Kit DNA (Macherey-Nagel). 700 µL of each sample were filled in bead-tubes and centrifuged at 11,000 g for 10 min. The supernatant was discarded and buffer SL-1 (700 µL) and the enhanced lysis buffer (50 µL) were added. Each further extraction step was done according to the manufacturer’s manual. The DNA was eluted in 50 µL elution buffer. DNA quantity and co-extraction of contaminants (absorbance ratio 260/280 and 260/230) was checked via the NanoDrop 2000c™ system.
For the quantification of methanogenic Archaea, the mlas-f/mcrA-r primer pair [33 (link),34 (link)] targeting the methyl coenzyme M reductase subunit A (mcrA) gene was used. Analyses were done on a Corbett Life Science (Qiagen, the Netherlands) Rotor-Gene Q system. The PCR procedure was conducted as follows: initial denaturation at 95 °C for 10 min, followed by 45 cycles of denaturation (95 °C for 30 sec), annealing (66 °C for 30 sec), and extension (72 °C for 30 sec). A PCR solution of 20 µL contained 9 µL PCR Mix (SensiFast™ SYBR No-Rox Kit (2×) (Bioline, UK), 380 nM of each primer, 1 mM MgCl2, 20% Betaine Enhancer Solution (5×) (VWR International, Germany), and PCR-grade water to reach a final volume of 18 µL, as well as a 2 µL template (5 ng DNA µL−1). An eight-point standard curve using gene copies of Methanosarcina thermophila and a melt-curve analysis were included in the approach.
The NGS library was prepared in-house. The small subunit (SSU) rRNA gene primers 515f and 806r [35 (link)], according to the Earth Microbiome Project [36 (link)], were used to target the V4 region. The first PCR step, including the 16S rRNA primers and the Illumina® adapter sequences, was performed as follows: initial denaturation at 95 °C for 5 min, followed by 30 cycles of denaturation (95 °C for 45 s), annealing (57 °C for 45 s), and extension (72 °C for 90 s). A final extension step of 72 °C for 10 min was set at the end of the PCR process. A PCR solution of 25 µL contained 12 µL PCR Mix (VWR Red Taq DNA Polymerase Master Mix Kit (2×)), 250 nM of each primer-adapter combination, 20% Betaine Enhancer Solution (5×), PCR-grade water to reach final volume of 24 µL, as well as 1 µL DNA template (5 ng DNA µL−1). The quality of the PCR products was checked with a 1.5% agarose gel using the dye GelGreen® Nucleic Acid Gel Stain (Biotium, Fremont, CA, USA). The PCR products of the first step were diluted 1:5 and used as a template for a second amplification to attach the Illumina® barcodes (i5 and i7). The same PCR procedure as in the first PCR step was used, except that only seven cycles were applied and the annealing temperature was set to 56 °C. The PCR products were again checked with a 1.5% agarose gel. Subsequently, final PCR products were quantified fluorometrically, as described previously [37 (link)]. The PCR products (15 ng) of each sample were pooled and purified with a Hi Yield® Gel/PCR DNA Fragment Extraction Kit (SLG®, Gauting, Germany) and eluted in 50 µL Tris-HCl buffer. The DNA quantity was again measured via QuantiFluor® dsDNA Dye (Promega, Madison, WI, USA). Co-extraction of contaminants was checked via the NanoDrop 2000c™ system. The final ready-to-load sample pool showed a DNA concentration of 19 ng µL−1 (260/280 absorbance ratio: 1.88) and was subsequently sent to Microsynth AG in Switzerland where the sequencing was done according to the company’s protocols.
In total, 27 mesophilic, 27 thermophilic, as well as nine MOCK samples were analysed. Raw sample reads were processed using the program mothur version 1.39.5 [38 (link)] and the MiSeq SOP (July 2019) [39 (link)]. A contig file was created with the paired-end reads (4,428,969 sequences in total, 70,301 ± 14,082 sequences sample−1). After quality filtering (approx. 24% of the sequences were discarded), unique sequences were aligned to the SILVA V132 database (Appendix A ). After another quality check and pre-clustering, chimeric amplicons were removed applying the VSEARCH algorithm (VSEARCH v2.3.4.). Sequence classification was done with the k-nearest neighbor (knn) algorithm. Sequences were binned to phylotypes based on their taxonomy. For a better comparability of samples while simultaneously ensuring an adequate coverage of the species richness, rarefaction curves were generated, and samples were normalised to 22,800 reads per sample [40 (link)]. The Mantel test showed that the similarity matrices prior to and after rarefaction did not differ significantly from each other (R > 0.99, p < 0.01, N = 9999). Quality-filtered sequences were uploaded to GenBank® via the submission tool, BankIt (Appendix B ). Information on the MOCK communities can be looked up in Appendix C .
After quality filtering and subsampling to 22,800 reads per sample, a FASTA file containing only representative sequences and an operational taxonomic unit (OTU) table was generated via mothur (version 1.42.1). The files were uploaded tohttps://piphillin.secondgenome.com (September 2019). The tool piphillin used the nearest-neighbor algorithm to pair 16S rRNA gene sequences to genomes [41 (link)]. The Kyoto Encyclopedia of Genes and Genomes (KEGG) database [42 (link)] of October 2018 was applied. The identity cut-off was set at 97%. The analyses focused on general biochemical pathways and on pathways regarding anaerobic degradation/turnover of aromatic compounds: degradation of aromatic compounds (KEGG orthology ko01220), phenylpropanoid biosynthesis (KEGG orthology ko00940), benzoate degradation (KEGG orthology ko00362), and aminobenzoate degradation (KEGG orthology ko00627).
After rarefaction analyses, meso- and thermophilic data were analysed separately, using only OTUs with a total abundance of ≥35 for each temperature regime. In mothur, the get.coremicrobiome command was applied to gain information on the microorganisms being present in every variant of the respective temperature regime [38 (link),39 (link)]. For characterising microorganisms important for explaining the variation between the C-load samples (class) of each temperature regime (biomarker discovery), the LEfSe command was applied [43 (link)]. For an interactive visualisation of relative sequence abundances of meso- and thermophilic samples, respectively, the tool KRONA was used [44 (link)]. The significance cut-off was set at α = 0.05 for all analyses. Significant genera were shown with the program STAMP 2.1.3 (Parks et al., 2014). For that purpose, White’s non-parametric t-test (two-sided) was used to distinguish between variants [45 (link)]. Confidence intervals were provided via percentile bootstrapping (1000 permutations). The false discovery rate was controlled with the Benjamini-Hochberg procedure (B-H adjustment) [46 (link)]. Via the program PAST® 3 [47 ], Spearman’s rank correlation analyses (Spearman rs) were done for all samples of day 28 for each temperature regime: Genera with a standard deviation below 3 over all samples of day 28 of each temperature regime were excluded; phenyl acids were log (x+1), and the OTU data box-cox (x+1) transformed. The false discovery rate was controlled with the B-H adjustment in Microsoft® Excel®. Moreover, the Mantel test (Gower Similarity Index) was applied in PAST® 3. For piphillin and biochemical analyses, the Mann–Whitney U test (M-W, two-sided) and the Friedman ANOVA (time series) were applied, respectively (Statistica™ 13 (TIBCO® Software Inc.)). Graphical presentations of correlation analyses and methanogenic properties were done with SigmaPlot™ 14 (Systat® Software Inc.), of general microbial properties with STAMP 2.1.3, and of biochemical and piphillin analyses with Statistica™ 13.
Straw from grain (straw) was air-dried, but otherwise not chemically, physically, or biologically (pre)treated. The straw was cut into pieces 4–7 cm long. The C/N ratio of the straw (ratio: 56) was analysed with a TruSpec® CHN analyser (Leco, Germany) according to the manufacturer’s protocol. The straw was filled into 120 mL serum flasks, functioning as batch reactors, in different carbon-load concentrations with 3 (defined as low carbon load, LCL), 34 (defined as medium carbon load, MCL), and 170 (defined as high carbon load, HCL) mmol carbon-C reactor−1, respectively.
A basal anaerobic broth based on previous investigations [29 (link)] was prepared and modified as follows (per litre): 0.4 g NaCl, 0.4 g MgCl2 × 6 H2O, 0.68 g KH2PO4, 0.18 g NaOH, 0.05 g CaCl2 × 2 H2O, 0.4 g NH4Cl, 0.5 g L-cysteine, 10.0 g sodium carboxymethylcellulose (CMC), 0.5 g yeast extract, 2.0 g sodium acetate, 1.0 g sodium formiate, 1 mL vitamin solution [29 (link)], 1 mL trace element solution SL-10 (German Collection of Microorganisms and Cell Cultures GmbH (DSMZ), Braunschweig, Germany), 2 mL sodium sulfide solution (120 g L−1 Na2S), and 1 mL resazurine solution (1.15 g L−1 resazurine). After the pH was adjusted with 0.1 M sodium hydroxide to 7.5 ± 0.2, 48 mL of the medium was filled into the 120 mL serum flasks which had previously been filled with straw (as described above). A control containing the anaerobic broth but no straw was also included and equally treated thenceforward. The sealing and headspace gas exchange took place according to previous protocols [22 (link)]. The flasks were subsequently autoclaved and cooled down before further use.
For each temperature regime, a volume of 12 mL diluted inoculum was injected into each reactor. Subsequently, the reactors were incubated at 37 °C and 52 °C, respectively, extending over an anaerobic incubation period of 28 days. All variations were prepared in triplicate. Samples were taken on day 2, 4, 7, 14, 21, and 28. Liquid samples for pH, VFA, phenyl acids, and C/Nliquid were processed immediately or frozen at −20 °C. The pH of the samples was measured with pH indicator strips 4.0–9.0 (Merck, Germany).
For each temperature regime, a PCR-DGGE approach [30 (link),31 (link)] was conducted with all variants of day 0 to check for the same microbial community structure at the beginning of the experiment. Moreover, control samples of day 0, as well as samples of day 14 and 28 were used for next-generation sequencing (NGS) analyses.
VFA, total carbon, total nitrogen (C/Nliquid ratio), as well as phenyl acid analyses were done according to previous studies [22 (link),28 (link),32 (link)]. The gas over-pressure was measured with a GHM Greisinger GDH 200 sensor and used to calculate biogas and methane production [NmL] as described previously [27 (link)].
Liquid samples (1 mL) from day 0, 14, and 28 were centrifuged at 20,000 g for 15 min and resuspended in 1 mL sterile ¼ Ringer solution. Subsequently, DNA extraction was done using the Soil Extract II Kit DNA (Macherey-Nagel). 700 µL of each sample were filled in bead-tubes and centrifuged at 11,000 g for 10 min. The supernatant was discarded and buffer SL-1 (700 µL) and the enhanced lysis buffer (50 µL) were added. Each further extraction step was done according to the manufacturer’s manual. The DNA was eluted in 50 µL elution buffer. DNA quantity and co-extraction of contaminants (absorbance ratio 260/280 and 260/230) was checked via the NanoDrop 2000c™ system.
For the quantification of methanogenic Archaea, the mlas-f/mcrA-r primer pair [33 (link),34 (link)] targeting the methyl coenzyme M reductase subunit A (mcrA) gene was used. Analyses were done on a Corbett Life Science (Qiagen, the Netherlands) Rotor-Gene Q system. The PCR procedure was conducted as follows: initial denaturation at 95 °C for 10 min, followed by 45 cycles of denaturation (95 °C for 30 sec), annealing (66 °C for 30 sec), and extension (72 °C for 30 sec). A PCR solution of 20 µL contained 9 µL PCR Mix (SensiFast™ SYBR No-Rox Kit (2×) (Bioline, UK), 380 nM of each primer, 1 mM MgCl2, 20% Betaine Enhancer Solution (5×) (VWR International, Germany), and PCR-grade water to reach a final volume of 18 µL, as well as a 2 µL template (5 ng DNA µL−1). An eight-point standard curve using gene copies of Methanosarcina thermophila and a melt-curve analysis were included in the approach.
The NGS library was prepared in-house. The small subunit (SSU) rRNA gene primers 515f and 806r [35 (link)], according to the Earth Microbiome Project [36 (link)], were used to target the V4 region. The first PCR step, including the 16S rRNA primers and the Illumina® adapter sequences, was performed as follows: initial denaturation at 95 °C for 5 min, followed by 30 cycles of denaturation (95 °C for 45 s), annealing (57 °C for 45 s), and extension (72 °C for 90 s). A final extension step of 72 °C for 10 min was set at the end of the PCR process. A PCR solution of 25 µL contained 12 µL PCR Mix (VWR Red Taq DNA Polymerase Master Mix Kit (2×)), 250 nM of each primer-adapter combination, 20% Betaine Enhancer Solution (5×), PCR-grade water to reach final volume of 24 µL, as well as 1 µL DNA template (5 ng DNA µL−1). The quality of the PCR products was checked with a 1.5% agarose gel using the dye GelGreen® Nucleic Acid Gel Stain (Biotium, Fremont, CA, USA). The PCR products of the first step were diluted 1:5 and used as a template for a second amplification to attach the Illumina® barcodes (i5 and i7). The same PCR procedure as in the first PCR step was used, except that only seven cycles were applied and the annealing temperature was set to 56 °C. The PCR products were again checked with a 1.5% agarose gel. Subsequently, final PCR products were quantified fluorometrically, as described previously [37 (link)]. The PCR products (15 ng) of each sample were pooled and purified with a Hi Yield® Gel/PCR DNA Fragment Extraction Kit (SLG®, Gauting, Germany) and eluted in 50 µL Tris-HCl buffer. The DNA quantity was again measured via QuantiFluor® dsDNA Dye (Promega, Madison, WI, USA). Co-extraction of contaminants was checked via the NanoDrop 2000c™ system. The final ready-to-load sample pool showed a DNA concentration of 19 ng µL−1 (260/280 absorbance ratio: 1.88) and was subsequently sent to Microsynth AG in Switzerland where the sequencing was done according to the company’s protocols.
In total, 27 mesophilic, 27 thermophilic, as well as nine MOCK samples were analysed. Raw sample reads were processed using the program mothur version 1.39.5 [38 (link)] and the MiSeq SOP (July 2019) [39 (link)]. A contig file was created with the paired-end reads (4,428,969 sequences in total, 70,301 ± 14,082 sequences sample−1). After quality filtering (approx. 24% of the sequences were discarded), unique sequences were aligned to the SILVA V132 database (
After quality filtering and subsampling to 22,800 reads per sample, a FASTA file containing only representative sequences and an operational taxonomic unit (OTU) table was generated via mothur (version 1.42.1). The files were uploaded to
After rarefaction analyses, meso- and thermophilic data were analysed separately, using only OTUs with a total abundance of ≥35 for each temperature regime. In mothur, the get.coremicrobiome command was applied to gain information on the microorganisms being present in every variant of the respective temperature regime [38 (link),39 (link)]. For characterising microorganisms important for explaining the variation between the C-load samples (class) of each temperature regime (biomarker discovery), the LEfSe command was applied [43 (link)]. For an interactive visualisation of relative sequence abundances of meso- and thermophilic samples, respectively, the tool KRONA was used [44 (link)]. The significance cut-off was set at α = 0.05 for all analyses. Significant genera were shown with the program STAMP 2.1.3 (Parks et al., 2014). For that purpose, White’s non-parametric t-test (two-sided) was used to distinguish between variants [45 (link)]. Confidence intervals were provided via percentile bootstrapping (1000 permutations). The false discovery rate was controlled with the Benjamini-Hochberg procedure (B-H adjustment) [46 (link)]. Via the program PAST® 3 [47 ], Spearman’s rank correlation analyses (Spearman rs) were done for all samples of day 28 for each temperature regime: Genera with a standard deviation below 3 over all samples of day 28 of each temperature regime were excluded; phenyl acids were log (x+1), and the OTU data box-cox (x+1) transformed. The false discovery rate was controlled with the B-H adjustment in Microsoft® Excel®. Moreover, the Mantel test (Gower Similarity Index) was applied in PAST® 3. For piphillin and biochemical analyses, the Mann–Whitney U test (M-W, two-sided) and the Friedman ANOVA (time series) were applied, respectively (Statistica™ 13 (TIBCO® Software Inc.)). Graphical presentations of correlation analyses and methanogenic properties were done with SigmaPlot™ 14 (Systat® Software Inc.), of general microbial properties with STAMP 2.1.3, and of biochemical and piphillin analyses with Statistica™ 13.
Full text: Click here