Culture-Independent Analyses Reveal Novel Anaerolineaceae as Abundant Primary Fermenters in Anaerobic Digesters Treating Waste Activated Sludge

1Anaerobic digestion (AD) involves the conversion of organics to valuable methane, which is facilitated by the tightly coupled synergistic activities of complex microbial communities. The process essentially consists of four sequential microbial-mediated processes: hydrolysis, fermentation (acidogenesis), acetogenesis (dehydrogenation) and methanogenesis (acetoclastic or hydrogenotrophic). Members of the phylum Chloroflexi are widespread in full-scale ADs, constituting up to 50% of the bacterial community, and are largely confined to the family Anaerolineaceae. Surprisingly, although their abundance indicates they must play a considerable role in these systems, their physiology and ecology is largely unknown. Most species of the Anaerolineaceae were isolated from anaerobic digester systems and have a fermentative metabolism, utilizing carbohydrates and proteinaceous carbon sources under anaerobic conditions. A role in fermentation in AD systems is additionally supported by the annotation of available genomes derived from metagenomes and with in situ evidence for the Chloroflexi phylum. An in-depth understanding of the ecology and function of the Chloroflexi in biogas systems requires the characterisation of the abundant genera of the phylum.

2A recent large scale amplicon sequencing survey of Danish full-scale AD communities revealed the A6 phylotype, a member of the Anaerolineaceae known only by their 16S rRNA gene sequence, to be among the most abundant genus-level-taxa in these systems; at times being present in excess of 20% of the amplicon reads. Advances in sequencing and metagenomic analyses enable the attainment of full genomes from the uncultured majority of microorganisms. In the absence of a pure culture, the aim of this study was to apply metagenomics to obtain a genome from a representative of the A6 phylotype, giving the first insight into their physiology.

3Metagenome sequencing and genome binning was performed essentially as detailed by. Metagenomes were prepared from sludge obtained from the anaerobic digester tank at Fredericia wastewater treatment plant in Denmark. The Fredericia plant has 2 × 2000 m3 mesophilic digester tanks with CambiTM thermal high-pressure (THP) pre-treatment of influent sludge and treats 8000 metric tonnes dry sludge per year. Sludge was obtained from the digester at two different time points 36 weeks apart (March and December, 2011) enabling later differential coverage binning of genomes. The DNA was extracted from AD sludge using the FastDNA® Spin kit for soil (MP Biomedicals, Santa Ana, CA, United States) following the standard protocol with minor modifications, as recommended by. These differences included a four times increase in the duration of bead-beating and a sludge input volume of 50 μl.

4Illumina TruSeq PCR free libraries and Nextera mate pair libraries were prepared according to the manufacturers’ protocol and paired-end sequenced (2 × 150 bp) on the Illumina HiSeq 2000 platform. The metagenomic reads were co-assembled using default settings in CLC Genomics Workbench (CLC Bio v. 7.5.1., Aarhus, Denmark). Reads were mapped to the assembly separately for each sample using default settings in CLC Genomics Workbench. The assembly and mapping information was exported as.fasta and.csv files, respectively, which were processed with the mmgenome workflow script ‘data.generation.2.1.0.sh’ to generate the files necessary for the binning process. Binning was carried out in the R environment using the mmgenome package1. The raw metagenome reads and the annotated genome sequence data have been submitted to the European Nucleotide Archive (ENA) under the study accession number PRJEB19949.

5Genome annotation was performed in the ‘MicroScope’ annotation pipeline. Automatic annotations were validated manually for the genes involved in metabolic pathways of interest with the assistance of the integrated MicroCyc and KEGG (Kyoto Encyclopedia of Genes and Genomes) databases. The genome annotations are available on the ‘MicroScope’ website2.

6Phylogenetic analysis and probe design was performed with the ARB software. Potential probes were assessed in silico with the mathFISH software. The Ribosomal Database Project (RDP) PROBE MATCH function was used to screen for non-target sequences with single base indels. In order to improve the fluorescence in situ hybridization (FISH) signal to background ratio, both the 5′ and 3′ ends of oligonucleotide FISH probes were labeled with either the sulfoindocyanine dyes Cy3 or Cy5, or the 5(6)-carboxyfluorescein-N-hydroxysuccinimide ester (FLUOS) [DOPE-FISH]. Probe validation and optimisation was based on generated formamide dissociation curves. With no pure cultures available, both probes were optimized using biomass with high abundance of the target organism from the anaerobic digester located at Ejby Mølle, Denmark (sampled August, 2013). In addition, CFX-A6-1278 was validated against Pelolinea submarina MO-CFX1T which has a single base mismatch. Pure cultures with a single base mismatch to the CFX-A6-450 probe were not available. Probes designed in this study were deposited into the probeBase database. The ARCH915 and the MX825mix (MX825; MX825b; MX825c) probes were applied to target the domain Archaea and the genus Methanosaeta, respectively. Quantitative FISH (qFISH) values were calculated as a percentage area of the total biovolume, stained with the 4′,6-diamidino-2-phenylindole (DAPI) DNA stain (50 μM, 1 h, at 4°C), which hybridized with the specific probe. The qFISH analyses were based on 25 fields of view taken at 630× magnification using the Daime image analyses software (DOME, Vienna, Austria). Microscopy was performed with either an Axioskop epifluorescence microscope (Carl Zeiss, Oberkochen, Germany) or a White Light Laser Confocal Microscope (Leica TCS SP8 X) fitted with a 405 nm diode laser (Leica Microsystems, Kista, Sweden).

7Amplicon sequencing survey data of full-scale ADs at wastewater treatment plants in Denmark showed a high abundance of the A6 phylotype in many of the mesophilic anaerobic digester tanks, but not in the primary or secondary sludge fed into these systems, suggesting that they are growing and well-adapted to mesophilic digester environment (Figure 1). In order to obtain genomes for the A6 taxon, metagenomes were generated for the Fredericia AD plant due to the observed high abundance of the target phylotype (representing up to 10% of the metagenome reads). A complete circular genome (CAMBI-1), classified to the novel MiDAS taxonomy defined A6 genus based on its 16S rRNA gene sequence, was successfully assembled from the metagenomes (see Table 1 for details). Phylogenetic analysis of the 16S rRNA gene revealed that CAMBI-1 clusters together with isolates of the Anaerolineaceae, sharing 85–90% 16S rRNA gene sequence identity (Figure 2). Based on the recommendations of, this indicates that CAMBI-1 should be considered to represent a novel genus within the family.

8Examination of the CAMBI-1 genome for PFAM proteins related to archetypic mono- and diderm cell envelopes, revealed a monoderm cell envelope architecture consistent with other Chloroflexi (Figure 3). The genome annotation and specialized searches using the PilFind program did not reveal any genes associated with flagella, fimbriae or pili, suggesting a non-motile lifestyle. Putative genes associated with spore coat polysaccharide biosynthesis protein SpsC (CFX1CAM_0088; 1106) were annotated, although definitive candidates for other spore related genes were not found and their ability to form spore like structures is unclear.

9The CAMBI-1 genome lacks a cytochrome oxidase, electron transport chain complexes and several key enzymes required for a complete TCA cycle, indicating a strict anaerobic metabolism. Annotation of a catalase (CFX1CAM_0578) and superoxide dismutase (CFX1CAM_2274) indicates some resistance to oxidative stress. Genes for the dissimilarity reduction of sulfate, nitrate or nitrite were also not annotated. Although an ability for denitrification was not annotated, the organism possesses a putative nitric oxide reductase (norV) (CFX1CAM_0414) and a putative hydroxylamine reductase (hcp) (CFX1CAM_0418), which both have suggested involvement in protection against nitrosative stress. Key genes for the Wood-Ljungdahl pathway and the Calvin-Benson-Bassham cycle were not annotated, indicating an inability to fix carbon for autotrophy. Potential for the pentose phosphate and Embden-Meyerhof-Parnas glycolysis pathways were present. Several annotated genes suggest a fermentative physiology consistent with other members of the family Anaerolineaceae (Figure 4 and Table 2). Pyruvate can be converted to acetyl-CoA by a pyruvate: ferredoxin oxidoreductase (CFX1CAM_0326), pyruvate dehydrogenase (CFX1CAM_1724-1726) or a pyruvate formate lyase (CFX1CAM_0333), with formate released from activity of the latter potentially oxidized to CO2 by an annotated formate dehydrogenase (CFX1CAM_1212). Potential fermentation by-products from acetyl-CoA include acetate, mediated by an acetyl-CoA synthetase (CFX1CAM_0825; 1292), and ethanol, facilitated by acetaldehyde (CFX1CAM_1715) and alcohol dehydrogenases (CFX1CAM_0055). The annotation of putative genes associated with the methylmalonyl-CoA pathway (CFX1CAM_1019; 1020; 2064–2067) indicates that propionate could be produced as a metabolic by-product from the fermentation of amino acids. Annotated tungsten-containing aldehyde ferredoxin oxidoreductases (AORs) (CFX1CAM_1238; 2051) may function to oxidize aldehydes derived from amino acid oxidation. Several described members of the Anaerolineaceae (Table 2) produce hydrogen as a fermentation by-product. However, definitive evidence for a hydrogenase was not found in the CAMBI-1 genome.

10Fluorescence in situ hybridisation probes were designed to visualize the morphology of the A6 in situ (Table 3). The CFX-A6-450 and CFX-A6-1278 probes were designed to cover the phylotype. These can be applied together with different fluorochromes, where the overlap gives a higher confidence in specificity, or with the same fluorochrome to give a higher coverage of the group and to increase the signal to background ratio that can be problematic with AD samples. Application of these probes to the Fredericia AD biomass, and several additional full-scale digesters, revealed that these organisms form short filaments that are typically approximately 0.3 μm thick and 5–10 μm long (Figure 5C), but were occasionally observed at lengths of >100 μm. Unlabelled helper probes were designed for the CFX-A6-1278 and CFX-A6-450 probes, but only CFX-A6-1278_H1 gave an increase in fluorescence and is recommended for use (Table 3). Competitor probes were designed to cover un-validated single base mismatches in non-target sequences (Table 3). Stringency of the CFX-A6-1278 probe was supported by its application to P. submarina MO-CFX1T – a non-target isolate with a single mismatch to the probe – which gave no positive fluorescent signal. Unlike some prominent wastewater-related Chloroflexi, the A6 are covered by the EUBmix FISH probe set routinely applied to cover most members of the domain bacteria.

11The A6 were distributed throughout the flocs and were sometimes observed co-located with the filamentous Archaea Methanosaeta spp. (Figure 5), indicating the potential for synergistic relationships. Parallel growth of these two filamentous genera was observed in multiple plants but was only frequent in the Ejby Mølle AD (23% of Methanosaeta filaments having associated A6; see Table 4). Unfortunately, the low FISH signal to background ratio for the thin A6 filaments did not permit qFISH studies or statistical co-location analyses; noting that visual assessment often indicated a higher abundance of A6 relative to the Methanosaeta spp., with most of both genera growing independent of contact with the other. Amplicon sequencing surveys indicate that the Methanosaeta are the most abundant methanogenic archaeal genus in mesophilic ADs located at WWTPs in Denmark, which was supported by qFISH in this study where they made up approximately half of the archaeal biovolume (Table 4). As such, the potential synergistic relationship with the A6 filaments may have important implications for methane production in some full-scale AD systems (e.g., Ejby Mølle). Methanosaeta are widely known to be acetoclastic methanogens, which would therefore benefit from the use of any acetate theoretically produced by the fermentative A6. It has also been demonstrated that some Methanosaeta species in ADs can utilize electrons sourced through direct interspecies electron transfer (DIET), from ethanol oxidizing Geobacter spp., to reduce CO2 to methane. Thus, the A6 may transfer excess electrons from the oxidation of organic carbon to the Methanosaeta via DIET, facilitated by the observed close proximity of the two organisms. It is thought that cytochromes and electrically conductive pili structures facilitate electrons flow to the recipient organism. A single membrane associated polyheme cytochrome c (CFX1CAM_1800) was annotated which had low homology to known DIET cytochromes (22% amino acid sequence similarity with OmcS from Geobacter sulfurreducens (Acc. No. AAR35877)), though no pili associated genes were identified in the CAMBI-1 genome. It may be that novel conductive structures are utilized by these organisms, or the CAMBI-1 genome may not represent the A6 species co-located with the Methanosaeta. Further work into the possibility for DIET and other potential interactions between the two genera is required.

12This study provides the first insight into the ecology of the A6 phylotype. Genomic evidence, coupled with their high relative abundance, suggests that members of the phylotype are important fermenters in mesophilic AD systems. The annotation of the representative CAMBI-1 genome gives a basic understanding of their physiology, noting that most of the coding sequence was not associated with any function (Table 1). A detailed understanding of the ecology of the A6 will be achieved with in situ and gene expression studies and by obtaining axenic cultures for representatives of the genus. The attainment of a genome and the optimisation of FISH probes in this study provides an important foundation for these approaches. Importantly, having complete genomes representing the abundant members of the community is an essential reference for metatranscriptomic and metaproteomic gene expression studies that will together allow organism-based metabolic networks to be developed for anaerobic digester systems – giving a comprehensive view of the ecology of these biotechnologically important systems.

13In the absence of a pure culture, and with only partial characterisation, organisms have previously been given provisional Candidatus status names. However, in accordance with the recent recommendations of, regarding the naming of uncultured organisms where the genome sequence is known, we propose the name Brevefilum fermentans gen. nov. sp. nov. with the CAMBI-1 genome as a provisional type species reference.

14Bre’veh.fi.lum fer.men’tans. L. adj. brevis short; L. neutral. n. filum line; L. part. adj. fermentans fermenting. Brevefilum fermentans a fermenting short filament.

15SM, RK, and PN planned the experimental work. SM and EF performed the FISH-based analyses. RK performed the DNA-based experimental work. RK, SK, and MA performed the metagenomic analyses. SM and MD performed the genome annotation. The manuscript was written by SM and drafted by all authors.

16The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.