Spirochaetes are frequently detected in anoxic hydrocarbon- and organohalide-polluted groundwater, but their role in such ecosystems has remained unclear. To address this, we studied a sulfate-reducing, naphthalene-degrading enrichment culture, mainly comprising the sulfate reducer Desulfobacterium N47 and the rod-shaped Spirochete Rectinema cohabitans HM. Genome sequencing and proteome analysis suggested that the Spirochete is an obligate fermenter that catabolizes proteins and carbohydrates, resulting in acetate, ethanol, and molecular hydrogen (H2) production. Physiological experiments inferred that hydrogen is an important link between the two bacteria in the enrichment culture, with H2 derived from fermentation by R. cohabitans used as reductant for sulfate reduction by Desulfobacterium N47. Differential proteomics and physiological experiments showed that R. cohabitans utilizes biomass (proteins and carbohydrates) released from dead cells of Desulfobacterium N47. Further comparative and community genome analyses indicated that other Rectinema phylotypes are widespread in contaminated environments and may perform a hydrogenogenic fermentative lifestyle similar to R. cohabitans. Together, these findings indicate that environmental Spirochaetes scavenge detrital biomass and in turn drive necromass recycling at anoxic hydrocarbon-contaminated sites and potentially other habitats.
The bacterial phylum Spirochaetes is well known for its commensal and pathogenic species, among them the causative agents of syphilis and Lyme disease . However, 16S rRNA gene sequences of environmental Spirochaetes are also frequently detected in anoxic environments, including sites contaminated with organic pollutants [2, 3]. Environmental Spirochaetes are also present in many enrichment cultures that degrade hydrocarbons or chlorinated compounds under anoxic conditions, including toluene , naphthalene , alkanes , terephthalate , and trichloroethene [7, 8]. For example, the relative abundances of Spirochaetes range from 1.4 to 20.8% in 10 reported sulfate-reducing or methanogenic hydrocarbon-degrading enrichment cultures originating from hydrocarbon-contaminated groundwater, aquifer sediments, and oil sands tailings ponds [2, 9]. In these cultures, sulfate-reducing or fermenting bacteria primarily mediate the anaerobic transformation of hydrocarbon compounds . In contrast, the functional role of the Spirochaetes has remained unresolved.
The recent isolation of Sphaerochaeta pleomorpha and Sphaerochaeta globosa from trichloroethene-dechlorinating enrichment cultures provided evidence that some environmental Spirochaetes are morphologically and metabolically distinct from host-associated representatives from the phylum . Contrary to other Spirochaetes, which are motile and spiral-shaped, the two strains are rod-shaped, non-motile microorganisms that lack flagella [7, 8]. It was proposed that these environmental Spirochaetes might serve as autotrophic acetogens [11, 12], but homoacetogenic growth with H2–CO2 was not observed . Genome-based inferences of the metabolic capacity of these isolates suggested a fermentative and saccharolytic lifestyle . Nevertheless, their functional role was untested experimentally and remains unclear.
We have shown that, under sulfate-reducing conditions, naphthalene can be anaerobically mineralized to CO2 by the stable enrichment culture N47 derived from sediments of a contaminated aquifer [5, 13]. Based on 16S rRNA gene sequence analysis, ~93% of the culture comprised a hitherto uncultured sulfate reducer Desulfobacterium strain N47 that has been shown to be responsible for naphthalene degradation [5, 14]. The remainder of the enrichment culture mainly comprised a Spirochete, Rectinema cohabitans strain HM, which is neither able to degrade naphthalene nor reduce sulfate or other sulfur compounds in pure culture . Other possible species might exist in this enrichment with unknown functions, but were not detected by terminal restriction fragment length polymorphism fingerprinting  because they represented less than 1% of the total DNA .
Here, we intended to elucidate the functional role of environmental Spirochaetes using the enrichment culture N47 as an example. By means of physiological, proteogenomic, and comparative genome analyses, we show that environmental Spirochaetes are involved in necromass recycling and provide hydrogen and nutrients to associated hydrocarbon degraders. Through reconstructing the genomes of two other putative Rectinema species, we also provide evidence that environmental Spirochaetes may generally contribute to necromass recycling in anoxic hydrocarbon- and organohalide-contaminated environments as well as other anoxic habitats.
Materials and methods
Bacterial cultures and cultivation
The naphthalene-degrading, sulfate-reducing enrichment culture N47 was cultivated in bicarbonate-buffered freshwater medium as described previously . After autoclaving, 3 mL of 1.5% (w/v) naphthalene dissolved in 2,2,4,4,6,8,8-heptamethylnonane (Sigma-Aldrich, Steinheim, Germany) were added to each cultivation bottle. Enrichment culture N47 mainly contained a sulfate reducer, candidate Desulfobacterium strain N47, and a spirochete, Rectinema cohabitans .
Attempts to isolate Desulfobacterium strain N47 in pure culture by serial dilution in agar roll tubes resulted in a highly enriched Desulfobacterium culture that lacked the Spirochete R. cohabitans and was 96% dominated by Desulfobacterium strain N47 (Figure S2, details in Supporting Information). The culture performed sulfate reduction and naphthalene oxidation and was cultivated in the same way as the original enrichment culture. This culture is described as “highly enriched Desulfobacterium culture” hereafter. R. cohabitans strain HM  was isolated from the enrichment culture N47 and cultivated in bicarbonate-buffered freshwater medium (pH 7.0), which was reduced with a final concentration of 0.5 mM sodium sulfide and amended with 10 mM glucose and 0.1% yeast extract. All culture bottles were incubated stationary at 30 °C in the dark.
To test whether R. cohabitans can utilize dead biomass of Desulfobacterium N47 as a carbon source, cells of the highly enriched Desulfobacterium culture were harvested from 150 mL cultures by centrifugation at 3214 × g (30 min). The supernatant, together with the heptamethylnonane phase containing the naphthalene, was discarded. The cell pellets were suspended in 2.5 mL anoxic fresh medium (without substrate or yeast extract), and autoclaved three times in a closed 100 mL serum bottle at 121 °C for 20 min. Thereafter 0.5 mL dead biomass was transferred to serum bottles filled with 30 mL fresh medium and the bottles were inoculated with 3 mL cultures of R. cohabitans. Due to the carryover from the previous cultures, glucose, yeast extract, and fermentation products could have been transferred with the inoculum. Thus, fresh media with only starter cultures of R. cohabitans or only dead biomass were incubated as control experiments. To investigate whether R. cohabitans can utilize putative carbon sources from the supernatant of the highly enriched Desulfobacterium culture, 30 mL centrifuged supernatant were filtered through 0.22 μm pore-size membranes (Sarstedt, Germany), transferred to fresh serum bottles, and inoculated with 3 mL starter cultures of R. cohabitans. Non-inoculated bottles served as controls.
To further test whether R. cohabitans has the capability to ferment necromass, different substrates relevant to subsurface habitats were added to 50 mL bicarbonate-buffered freshwater medium and inoculated with 5 mL starter cultures of R. cohabitans. These substrates included 5 mM sucrose, 5 mM valine, 5 mM pimelic acid, or 0.1% tryptone. Furthermore, living cells of the highly enriched Desulfobacterium culture grown in 75 mL cultures were harvested by centrifugation and added as substrate as described above. Fresh media with only starter cultures of R. cohabitans were incubated as carryover controls.
Hydrogen was measured with a Shimadzu GC-8A Gas Chromatograph  or a Compact GC (Global Analyser Solutions, Breda, The Netherlands) . For the former, high concentrations of hydrogen (>100 parts per million by volume (ppmv)) were measured with a thermal conductivity detector, whereas low concentrations (<100 ppmv) were measured with a reduction gas detector (RGD2 detector, Trace Analytical, Stanford, CA, USA). The measurements were calibrated with standard H2 gases (Messer Schweiz) (2 and 50 ppmv for samples lower than 100 ppmv, 1000 ppmv for samples higher than 100 ppmv). For the latter, the GC was equipped with a Molsieve 5A pre-column and Porabond column (CH4, O2, H2, and N2). The carrier gases were N2 for the H2 channel and He for the rest of the gases (CH4, O2 and N2). The concentrations (>100 ppmv) were determined using a thermal conductivity detector.
Sequencing and reconstruction of Spirochete genomes
For comparative genomic analyses, three high-quality draft genomes were obtained, namely those of R. cohabitans, uncultured Spirochete bacterium bdmA 4, and uncultured Spirochete bacterium SA-8. The latter two genomes contained 16S rRNA gene sequences with 95% identity to R. cohabitans. The draft genome of the uncultured Spirochete bacterium SA-8 was derived from the metagenome of the terephthalate-degrading community TA, which is available on IMG/M (Integrated Microbial Genomes & Microbiomes) as taxon object ID 3300001095 . The draft genome of uncultured Spirochete bacterium bdmA 4 was reconstructed from the metagenome of another sulfate-reducing, naphthalene-degrading enrichment culture (designated Sob). Culture Sob was enriched from a creosote-contaminated wood preservation facility located in Sobĕslav, south of Bohemia (Czech Republic), as described by Kummel et al. . Information on the shotgun sequencing, assembly, and binning processes  for the genome of R. cohabitans and the metagenome of enrichment culture Sob are described in the Supporting Information. The draft genomes of R. cohabitans, uncultured Spirochete bacterium bdmA 4, and uncultured Spirochete bacterium SA-8 were loaded into the MicroScope genome annotation platform and automatically annotated . The annotations were manually curated and edited using the Magnifying Genome (MaGe) environment , and metabolic pathways were predicted using the integrated pathway tools of MaGe that are based on the KEGG (Kyoto Encyclopedia of Genes and Genomes) and MicroCyc databases [22, 23]. The presence of glycoside hydrolases and extracellular peptidases was evaluated using dbCAN (E < 1e−10 and cover fraction >0.4)  and MEROPS (E < 1e−10) , respectively.
Potential genes encoding hydrogenase catalytic subunits were identified in the genomes of R. cohabitans and Desulfobacterium N47 by screening translated sequences through the web database HydDB . Sequences were validated as hydrogenases if they encoded cysteine-containing motifs characteristic of hydrogenases  and branched with hydrogenase reference sequences on phylogenetic trees. To infer evolutionary relationships, the source and reference sequences were aligned using the MUSCLE algorithm  and visualized on neighbor-joining phylogenetic trees  constructed with MEGA7  and bootstrapped with 500 replicates. For genetic organization maps, protein/domain function was predicted using the CDD (Conserved Domain Database)  and iron–sulfur clusters were predicted by identifying conserved cysteine residues as previously described .
The phylogeny of the Spirochaetes was inferred using 16S rRNA gene sequences retrieved from the NCBI (National Center for Biotechnology Information) Genome Database. Sequences were aligned using the MUSCLE algorithm included in MEGA7 and trimmed to uniform length . A maximum likelihood phylogenetic tree was constructed in MEGA7 using a general time reversible substitution model and uniform rates among sites. The tree was bootstrapped with 500 replicates.
Mass spectrometry-based proteome analyses
For proteomics, the enrichment culture N47 and the highly enriched Desulfobacterium culture (without R. cohabitans) were grown with naphthalene (four parallel cultures each). In addition, four cultures of R. cohabitans were grown with 10 mM glucose and 0.1% yeast extract. The cultures were harvested after 6 weeks for the enrichment culture or 10 days for R. cohabitans and then immediately frozen at −20 °C. The frozen cell pellets were molten at room temperature and taken up in 1 mL phosphate-buffered saline (3 mM Na2HPO4, 1.1 mM KH2PO4, and 155.2 mM NaCl; pH 7.4). The samples were lysed by ultrasonification in a Bioruptor® (Diagenode) for 5 min to break the cells, release the proteome, and shear the genomic DNA (experimental settings: water bath temperature, 4 °C; power setting, H position (High); sonication cycle, 30 sec ON, 30 sec OFF; total sonication time, 5 min). The turbid solution was cleared by centrifugation (12,000 × g, 4 °C, 5 min) and the supernatant was transferred to a fresh 1.5 mL Eppendorf tube. The protein concentration was determined by using a modified Bradford assay (RotiNanoquant; Roth) following the manufacturer’s instructions. A detailed protocol for the subsequent sample preparation for Liquid chromatography tandem-mass spectrometry, as well as the settings during data acquisition and data analysis, can be found in the Supporting Information. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE  partner repository (https://www.ebi.ac.uk/pride/archive/) with the dataset identifier PXD005624.
The genome of R. cohabitans HM suggests it is an obligate fermenter and hydrogen producer
We sequenced and assembled a draft genome of R. cohabitans HM (2.82 Mb, estimated to be 98% complete, see Supporting Information) to develop hypotheses about its ecophysiological role in the enrichment culture N47. Analysis of the metabolic capacity from the genome suggests that R. cohabitans performs a fermentative lifestyle. It is predicted to be capable of degrading proteins and carbohydrates, resulting in the production of the endproducts ethanol, acetate, and H2 (Fig. 1 and Supporting Information). In contrast, key genetic determinants of aerobic and anaerobic respiration, the tricarboxylic acid cycle, and carbon fixation via the reductive acetyl-CoA pathway are absent from the genome. This is consistent with previous substrate utilization tests .
Furthermore, we identified putative hydrogenase genes in the genomes of both R. cohabitans and Desulfobacterium N47. To gain insight into their potential ecophysiological roles, we classified them based on their amino acid sequence phylogeny and genetic organization according to a recently developed hydrogenase classification scheme [26, 27]. Seven putative [FeFe]-hydrogenase genes were identified in the genome of the Spirochete. Three are monomeric [FeFe] Group A1 or B hydrogenases (Fig. 2a, c), which are known to couple oxidation of reduced ferredoxin to evolution of H2 during carbohydrate and protein fermentation in diverse bacteria [27, 33]. Two others encode trimeric electron-bifurcating [FeFe] Group A3 hydrogenases (Fig. 2a, c), which reversibly couple the oxidation of ferredoxin and NADH to the evolution of H2, thereby providing a novel mechanism to balance redox state and adenosine triphosphate demand . The genome also encodes [FeFe] Group C1 and C3 hydrogenases (Figure S1), which are candidate subgroups involved in H2 sensing .
Re-analysis of the genome of the sulfate-reducing Desulfobacterium N47  revealed the presence of two Group 1a and Group 1b [NiFe]-hydrogenases (Fig. 2b). These are known to couple H2 oxidation to cytochrome c3 reduction in the respiratory chains of sulfate-reducing deltaproteobacteria. Whereas the former sequence is incomplete due to sequence coverage gaps, a complete operon encoding the 1b enzyme and its cytochrome c3 partner was detected (Fig. 2d) .
The genome-based predictions indicate that the association of the two predominant bacteria in the enrichment culture may be driven by the transfer of H2; we predict that H2 is produced during fermentation by R. cohabitans using [FeFe]-hydrogenases and subsequently consumed by the sulfate-reducing Desulfobacterium N47 using [NiFe]-hydrogenases.
Hydrogen is exchanged between R. cohabitans and Desulfobacterium N47
To confirm that R. cohabitans is a fermentative H2 producer, we performed pure culture growth experiments with 10 mM glucose as the electron and carbon source. Gas chromatography demonstrated that H2 rapidly accumulated in the headspace during cultivation with glucose, resulting in H2 concentrations up to 400,000 ppmv after 2 weeks (Fig. 3a). Proteomic analysis of cells grown with glucose confirmed the expression of all hydrogenase-related genes with high peptide abundances (Table S2). Additionally, when grown with valine, sucrose, and tryptone, respectively, R. cohabitans produced almost threefold more hydrogen compared to controls after 1 week (Fig. 3b). This further confirms that R. cohabitans performs a proteolytic and saccharolytic lifestyle that results in hydrogen production.
The highly enriched Desulfobacterium culture was cultivated under sulfate-reducing conditions with naphthalene as carbon source. Meanwhile, the headspace was supplemented with a limited amount of H2 gas (~8285 ppmv) as an additional electron donor for the reduction of sulfate (20 mM). Consistent with genome-based predictions of hydrogenases for Desulfobacterium N47, we observed a gradual decrease in H2 concentration (Fig. 3a). Consumption of H2 by Desulfobacterium N47 was inhibited when 5 mM sodium molybdate (Na2MoO4), a potent inhibitor of sulfate reduction , was added at day 31 (Fig. 3a). This suggests that, in common with classical sulfate-reducing isolates , Desulfobacterium N47 directly couples H2 oxidation to sulfate reduction through an anaerobic respiratory chain.
Finally, in order to reveal a potential relationship between the two members, we monitored H2 concentrations in the enrichment culture N47 (R. cohabitans + Desulfobacterium N47) with naphthalene and sulfate. The concentration of H2 remained constant (~2 ppmv) for 1 month (Fig. 3a). However, a gradual increase in H2 concentration was observed when sulfate reduction by Desulfobacterium N47 was inhibited by the addition of sodium molybdate. This suggests that there is a tight coupling between Spirochete-mediated fermentative H2 evolution and Deltaproteobacterium-mediated respiratory H2 consumption during the growth of Desulfobacterium with naphthalene as the sole supplemented carbon source and sulfate as the electron acceptor.
Organic carbon derived from necromass fuels fermentation by R. cohabitans
While the above results demonstrate that H2 is an important link between R. cohabitans and Desulfobacterium N47, the question remained about the origin of the reducing equivalents (i.e., electron sources) used for H2 production by R. cohabitans. Considering that R. cohabitans cannot utilize naphthalene and no other exogenous organic sources were added, the potential carbon sources for R. cohabitans include metabolites formed during naphthalene degradation and/or dead biomass from the deltaproteobacterium.
To address this question, we compared the proteomes of the highly enriched Desulfobacterium culture (without R. cohabitans, Figure S2) and the enrichment culture N47. The two proteomes revealed that only 5 out of 1266 proteins detected in total and assigned to Desulfobacterium N47 were differentially expressed between the two cultures, none of which could be assigned to a distinct ecophysiological function (Fig. 4). However, 53 proteins belonging to R. cohabitans were identified only in the enrichment culture N47 but not in the highly enriched Desulfobacterium culture (Table S3). Only a few proteins of R. cohabitans were detected in enrichment culture N47, most likely due to its relatively low abundance compared to the Desulfobacterium. Most of the R. cohabitans proteins expressed in the enrichment culture N47 were related to transporters for oligopeptides, dipeptides, amino acids, sugars, and dicarboxylic acids (Fig. 4 and Table S3). Other enriched enzymes include protease M41, glutamate dehydrogenase (amino acid catabolism), pyruvate-ferredoxin oxidoreductase (ferredoxin reduction), and components of the Sec translocon (secretion of hydrolases, insertion of transporters) (Table S3). This further supports that the Spirochaete performed fermentative degradation of polysaccharides and proteins in the enrichment culture N47.
As we did not add complex carbohydrates or peptides to the medium, these compounds are likely to originate from dead biomass of Desulfobacterium N47. We verified that R. cohabitans could use organic carbon sources derived from Desulfobacterium N47 to support fermentative H2 production by cultivating the Spirochaete with biomass of the highly enriched Desulfobacterium culture. Initially, the biomass was firstly autoclaved to produce artificial necromass and then added to culture bottles of R. cohabitans. After 3 days, the H2 production by R. cohabitans was twofold higher in fresh media incubations supplemented with either heat-killed cells or filtered supernatant of the highly enriched Desulfobacterium culture (Fig. 3c). Given that autoclaved biomass does not reflect the biomass present in the natural environment, we tested if biomass without any treatment could support growth of R. cohabitans. As Desulfobacterium N47 cannot survive under the defined growth conditions of R. cohabitans, it will die automatically and thus produce necromass due to lack of energy. After 7 days, the H2 production by R. cohabitans was 2.4 times higher compared to a control with inoculum only and no substrate (Fig. 3b). The two experiments show that R. cohabitans can use necromass and potentially exudates in the supernatant derived from Desulfobacterium N47 to support its obligately fermentative lifestyle.
Pimelate and other dicarboxylic acids are important downstream metabolites of anaerobic naphthalene degradation and have been detected in the supernatant of the enrichment culture N47 . However, the genomic search did not support that R. cohabitans could utilize them as carbon sources, as we could not find genes for the β-oxidation of dicarboxylic acids . Further substantiating this, substrate testing showed that pimelate does not stimulate hydrogen production (Fig. 3b). Together, the genome predictions (Fig. 1), substrate tests (Fig. 3), and proteomic data (Fig. 4) suggest that R. cohabitans primarily feeds on proteins and sugars derived from detrital biomass in the enrichment culture N47.
Spirochaetes may generally mediate necromass recycling in anoxic hydrocarbon- and organohalide-contaminated habitats
To further illuminate the ecological niche of Spirochaetes in polluted habitats, we compared the genome of R. cohabitans with the draft genomes of two uncultured environmental Spirochaetes that were reconstructed from metagenomics sequences: uncultured Spirochete bacterium SA-8, originating from a terephthalate-degrading, methane-producing bioreactor , and uncultured Spirochete bacterium bdmA 4, derived from a naphthalene-degrading, sulfate-reducing culture enriched from a creosote spill .
Genome-based reconstructions suggest that, akin to R. cohabitans, both organisms are non-spiral Spirochaetes and obligate fermenters (Fig. 1 and Table S1). In addition, both genomes encode [FeFe] Groups A and B hydrogenases that are phylogenetically related to those of R. cohabitans (Fig. 2a and S1), suggesting that they also fermentatively produce H2, potentially sustaining associated hydrogenotrophic methanogens and sulfate reducers. Previously reported transcriptomic analysis supports that the uncultured Spirochete bacterium SA-8 participates in proteolytic amino acid degradation similar to R. cohabitans . Revisiting the published genomes of the nonspiral Spirochaetes S. pleomorpha and S. globosa from trichloroethene-degrading cultures also indicated a potential for H2 production through saccharolytic mixed-acid fermentation (Fig. 2a and S1) . In these cases, H2 probably serves as an electron donor to enhance reductive dechlorination activities of organohalide-respiring bacteria.
The 16S rRNA gene sequences of R. cohabitans, uncultured Spirochete bacterium bdmA 4, and uncultured Spirochete bacterium SA-8 are 95% identical to each other, and form a common clade distinct from Treponema species in the 16S rRNA-based phylogenetic tree (Fig. 5 and S3). In the same clade, we could also identify numerous 16S rRNA gene sequences derived from contaminated aquifers or bioreactors containing hydrocarbons or organohalides as well as oil reservoirs. A small selection is shown in Figure 5. This indicates that Rectinema phylotypes are widely distributed in hydrocarbon- and organohalide-contaminated anoxic sites. The two strains S. globosa and S. pleomorpha are only distantly related to the three members of the genus Rectinema , suggesting that nonspiral morphology occurs in several lineages and that a similar metabolism is present in distinct groups of environmental Spirochaetes (Figure S3). Altogether, comparative genomic analyses suggest that these environmental nonspiral Spirochaetes fulfill ecophysiological functions similar to R. cohabitans in their communities: recycling necromass and providing hydrogen for hydrogenotrophic degraders of organohalides and hydrocarbons in these communities.
Anaerobic degradation by indigenous microorganisms is the most important process that contributes to removal of anthropogenic organic pollutants in subsurface sites (e.g. contaminated groundwater) . For some time, research on anaerobic degradation processes mostly focused on key players that are responsible for initial degradation of pollutants and/or reduction of electron acceptors (Fig. 6). However, it has become clear in recent years that the complete oxidation of organic pollutants to CO2 is based on labor sharing: participation of both key players for initial degradation of pollutants and secondary degraders for complete mineralization of catabolic by-products [6, 39, 40]. Additionally, molecular surveys of anaerobic enrichment cultures or contaminated sites have revealed that many microbial species are not clearly correlated to degradation processes. They seem to be neither key players nor secondary degraders and are therefore unlikely to be involved in direct degradation of pollutants [6, 41]. Reflecting this, environmental Spirochaetes are common and abundant in contaminated sites and in enrichment cultures that degrade hydrocarbons and organohalides, but there are no indications that they are linked to degradation of either these compounds or by-products . Our study of the enrichment culture N47 now provides experimental proof of one functional trait to these frequently detected organisms.
In the enrichment culture N47, the chemo-organo-heterotrophic Desulfobacterium strain N47 degrades naphthalene as energy and carbon source with sulfate as electron acceptor, leading to primary biomass production in the absence of sunlight. Under nutrient-rich conditions, if there is sufficient sulfate and naphthalene in the aqueous system (e.g., >~50 μM) , the sulfate reducer can live alone without the need of R. cohabitans to supply additional electron donors. In contrast, R. cohabitans relies on organic carbon sources derived from detrital biomass of Desulfobacterium N47 to sustain its growth. In turn, necromass degradation via hydrolysis and fermentation results in production of hydrogen and potentially short-chain fatty acids and alcohols. These products are then utilized by Desulfobacterium N47, thereby resulting in the recycling of necromass.
R. cohabitans is only one example out of many fermentative microorganisms that might mediate equivalent functions in necromass recycling. Even sulfate reducers themselves are quite versatile microorganisms and some perform hydrogenogenic fermentation when lacking electron acceptors such as sulfate or nitrate . However, a fermenter that is more specialized for proteolysis and can secrete extracellular proteases and sacharolytic enzymes will most likely outcompete such sulfate reducers or other fermenters for necromass degradation. Consistently, the genomes of environmental Spirochaetes like R. cohabitans have genes encoding several extracellular peptidases (Merops family M23B, S8A, S26A, and S33) and carbohydrate-converting enzymes including hydrolases (Fig. 1 and Supporting Information). This is because necromass usually occurs in the form of biopolymers that must be hydrolyzed outside the cell before metabolizing by microorganisms. Hence, exoenzymes play an essential role in initial processing by hydrolyzing complex organic matter derived from detrital compounds of other community members (e.g., sulfate reducers and fermenters) in the system . Similar physiological properties are also well known for other organisms, such as Clostridia, that fulfill similar ecological roles in anoxic habitats.
In anoxic hydrocarbon- and organohalide-contaminated habitats, the biomass turnover rates are potentially high  because hydrocarbons (e.g., naphthalene and other aromatic compounds) or organohalides (e.g., trichloroethene) are usually present in relatively high concentrations in such habitats compared to pristine sites . Their presence reshapes the abundances and interactions of indigenous microbes. In addition, they provide exogenous energy and carbon inputs, allowing for the buildup of significant biomass especially at contaminant plume fringes [38, 45]. For example, in the terephthalate-degrading bioreactor mentioned above, up to 10% of the terephthalate-derived carbon was converted to biomass which is a typical yield for anaerobic cultures . In some benzene and toluene-degrading enrichment cultures, heterotrophic CO2 assimilation is also supposed to contribute greatly to biomass accumulation as evidenced from protein-based stable isotope probing (protein-SIP) and DNA-SIP experiments [41, 46]. Furthermore, at hydrocarbon- and organohalide-contaminated sites, high concentrations of organic solvents could challenge the integrity of cell membranes, inducing cell leakage or lysis [40, 47, 48].
Given the potential high biomass turnover, recycling of necromass is therefore important for ecosystem function and prevents loss of nutrients in contaminated ecosystems (Fig. 6). Most notably, it may contribute to enhancing biodegradation rates at contaminant plume fringes in groundwater or oil reservoirs where electron donors (e.g., hydrocarbons) and acceptors (e.g., sulfate or nitrate) meet in opposing gradients (Fig. 6) . Firstly, during necromass recycling, the production of H2 and other substrates such as acetate and ethanol has frequently been reported to stimulate microbial blooms (the so-called “priming”)  or serve as secondary substrates for co-metabolism, enhancing biodegradation rates . Secondly, environmental Spirochaetes or similar ecotypes remineralize the necromass, which may also be important for recycling of nutrients such as nitrogen, phosphorus, and trace elements within a nutrient-poor subsurface ecosystem, increasing overall rates of biodegradation [51, 52].
At this point, we can only speculate on the importance of necromass recycling in the subsurface because quantitative data are not available. In our enrichment culture N47, R. cohabitans constituted ~7% of the microbial population. Typical growth yields for anaerobic microorganisms are around 5–10% of the consumed substrates, as exemplified by the terephthalate-degrading culture mentioned above . If necromass is the only electron and carbon source for R. cohabitans, at least a tenfold number of Desulfobacterium cells (i.e., 70% of total cell numbers) must have died to form necromass and serve as substrate for R. cohabitans to produce the observed 7% of the total cell numbers. Considering the actual proportion of 93% Desulfobacterium cells, this means that 163% of the total Desulfobacterium cell numbers (93% living cells plus 70% necromass) must have grown. Thus, approximately 43% of all Desulfobacterium cells that have grown in the culture died and were recycled by R. cohabitans. The necessary difference in growth rate to account for the necromass recycling is quite small here. For example, if the culture grows exponentially from a 10% parent inoculum to full density (i.e., the cell numbers increased tenfold) in 6 weeks, the doubling time is approximately 13 days. Taken the described necromass recycling calculated above into consideration, it only shortens to ~11 days for the Desulfobacterium cells (i.e., the actual cell numbers increased 16.3-fold in 6 weeks). However, if such values of cell death and necromass recycling are transferable to subsurface habitats, and assuming that cell death is much higher under non-optimal conditions in the environment, we may speculate that fermentative organisms like R. cohabitans might constitute a subsurface microbial loop [53, 54] that can explain biomass recycling in the underground without a food chain (Fig. 6).
We thank Masaru K. Nobu for supplying genome sequences of uncultured Spirochete SA-8, Melanie Klose for hydrogen measurement, Karine Labadie and Eric Mahieu for Illumina sequencing, and the Genoscope’s LABGeM team for providing the MicroScope platform. We acknowledge partial funding from the FP7-KBBE MAGIC-PAH (project 245226) for (meta)genome sequencing. Xiyang Dong acknowledges a fellowship from the China Scholarship Council (File No. 201306060075). Chris Greening acknowledges funding from the ARC DECRA Fellowship (DE170100310) and ARC Discovery Project (DP180101762). Markus Kaiser acknowledges partial funding from the ERC starting grant No. 258413 and from the DFG grant No. INST 20876/127-1 FUGG. Rainer U. Meckenstock acknowledges funding from the ERC advanced grant EcOILogy (No. 666952).