Water samples were filtered twice (see Methods), first through a large filter (0.45 µm, LF or “Large Filter”) and then the filtrate was passed through a small filter (0.05 µm, UF or “Ultrafine Fraction”). Using whole genome shotgun metagenomics from four water samples (LF92 and UF92 from the 92 m depth, LF99 and UF99 from the 99 m depth) as well as one sediment sample, we provide the first comprehensive whole genome shotgun metagenomics investigation of this section of the lake and highlight both the taxonomic composition and potential metabolic strategies for survival, as well as identify areas for deeper investigation.
Cell counts and dissolved nutrients
In order to determine the habitability of the anoxic basin, the cell counts were measured in the oxycline (75 m depth) and the anoxic region (92 and 99 m depth), where oxygen content is < 1 mg/L (Fig. 1). The cell counts were 84,129 cells/mL at 75 m, 895,516 cells/mL at 92 m, and 775,404 cells/mL at 99 m. In the anoxic depths, the cell counts were roughly an order of magnitude higher than the count at 75 m depth. Dissolved nutrient levels were also significantly higher in the anoxic section than the oxycline (Table 1). The ammonium content at 92 m was 1675.98 µmol/L, double the amount at 99 m, which was 758.53 µmol/L. The concentration was much lower in the oxycline at 4.11 µmol/L. The phosphate content in the oxycline was 0.06 µmol/L. The concentration was over three orders of magnitude higher in the anoxic region, measuring 35.19 µmol/L at the 92 m depth and 47.30 µmol/L at the 99 m depth. The silicate content in the oxycline was 86.76 µmol/L. The concentration was once again higher in the anoxic region, measuring 387.62 µmol/L at 92 m and 664.60 µmol/L at 99 m.
Taxonomic profiling
Unclassified organisms
Taxonomic annotations were performed using the Lowest Common Ancestor Star algorithm (LCA*) with the RefSeq database21. With this annotation tool, any organism that could not be assigned a taxonomic classification beyond the superphylum level (bacteria, archaea, prokaryotes, root) was designated as “unclassified”. LCA* was able to classify an average of 60% of the contigs across all samples. Kaiju22 and Kraken223 were able to classify an average of 18.75% and 19.5% of the reads from all samples respectively. Given that LCA* was able to annotate the greatest fraction of the reads, we focused on these results. The percentage of unclassified organisms increased with depth, from 37% at 92 m to 43% at 99 m to 45% of the community in the sediment layer. At the phylum level, 97–99% of the unclassified organisms were assigned to the bacterial superphylum. In each sample, fewer than three percent of the unclassified organisms belonged to the archaea superphylum (Figure S1).
Classified organisms
Archaea make up 2% of classified organisms at the 92 m depth and 4% of classified organisms at the 99 m depth and in the sediments. Bacteria make up 98% of classified organisms at the 92 m depth, and 96% of classified organisms at 99 m depth and in the sediment. Specifics of classified and unclassified organisms per sample as well as deeper taxonomic classifications may be found in the Supplementary Information (Figures S1 and S2).
Classified archaea
The total percentage of archaea in the samples doubles from 1% of the community at the 92 m depth to 2% at 99 m and in the sediment (Fig. 3). Archaea are almost entirely classified at or beyond the phylum level in these samples. Euryarchaea are the most abundant archaeal phylum in all samples. They make up 97% of archaea, or 1% of all organisms at the 92 m depth. At the 99 m depth and the sediment layer, they make up over 99% of archaea, or 2% of all organisms. Euryarchaea are well-known anoxic methanogens, and their higher abundance at 99 m and in the sediment could explain why the concentration of dissolved methane was found to be highest at the deep anoxic layer near the water–sediment interface (Fig. 1)13. More details on the archaeal community may be found in Table 2 and in the Supplementary Information.
The taxonomic composition of the communities at the phyla level. Classified organisms in the figure only include phyla that make up more than 1% of the community.
Classified bacteria
Proteobacteria are the most abundant classified bacteria in the system. The classes present include Alphaproteobacteria, which make up 1% of organisms in LF92, 2% in UF92, and 1% in LF99, UF99, and the sediment. Betaproteobacteria make up 8% of organisms in LF92, 10% of organisms in UF92, 5% in UF99, 4% in LF99, and 6% in the sediment. Deltaproteobacteria make up 16% of the community in LF92, 9% in UF92 and UF99, 13% in LF99, and 15% in the sediment. Gammaproteobacteria make up 2% of the organisms in LF92, 3% of organisms in UF92, 2% in UF99 and LF99, and 1% of the community in the sediment. Planctomycetes make up 2% of all organisms in LF92, 0.5% of organisms in UF92 and 0.3% in UF99, and 3% of organisms in LF99. The most abundant class belonging to this phylum is Planctomycetia. Actinobacteria make up 1% of organisms in the LF92, UF92 and LF99 samples, 2% of organisms in UF99, and a total of 3% of classified organisms. The only major annotated class within Actinobacteria is also named Actinobacteria and makes up more than 70% of this phylum in the water column and 48% of the phylum in the sediment. Bacteroidetes make up 10% of the community in LF92, 15% of the community in UF92, 12% of the community in UF99, 9% of the community in LF99 and 3% of the community in the sediment. The major annotated classes belonging to Bacteroidetes are Marinifilaceae, Cytophagaceae and Chitinophagaceae. Firmicutes constitute 8% of the organisms within LF92, 11% of UF92 and UF99, 9% of LF99, and 5% of the sediment. The major annotated classes within this phylum are Clostridia and Bacili. A summary can be found in Table 2 and more detail is included in the Supplementary Information.
Taxonomic differences between fraction sizes
To ascertain if cell size played a role in survival capabilities in this extreme environment, we assessed the metabolic potential within the size fractions. While the percentage of Proteobacteria in the community is more related to depth than filter size in the water column, within the Proteobacteria, we found the large fraction samples had 13–16% of the community classified as Deltaproteobacteria. In contrast, only 9% belong to ultrafine fractions. At 15%, the fraction of Deltaproteobacteria in the sediment is similar to that of the LF samples. Planctomycetes make up 2–3% of the LF samples, while they make up only 0.3–0.5% of the UF communities. Planctomycetes make up 4% of the community within the sediment. More details on these differences can be found in Table 2. The size of the organisms did not seem to affect survival.
Functional analysis and metabolic pathways of interest
To determine how organisms use energy sources available to them, MetaPathways V2.524 was used to assign metabolic pathways to the open reading frames (ORFs) and predicted metabolic pathways using pathwaytools25. An average of 830 (standard deviation of 170) pathways per sample were identified. 310 pathways were shared among all samples while an average of 27 (standard deviation of 15) pathways were unique to each sample. An average of 43 pathways were unique to unclassified organisms among the samples, and 101 pathways were unique to classified organisms among all samples. Overall, an average of 1.3% of all the pathways were responsible for biosynthesis, 14.2% of pathways were responsible for degradation, and an average of 0.2% of all the pathways belonged to metabolite precursor generation. The greatest pathways abundance, with an average of 84.1% of all pathways, belonged to energy metabolism. To link function and taxonomy, we used the LCA annotation (obtained using the RefSeq database26) assigned to the ORF that corresponded to a pathway.
Next, we focused on pathways relevant to the survival on Enceladus. First, the metabolic pathways related to methanogenesis were examined; we then investigated pathways related to nitrogen and sulfur compounds that have either been detected or could potentially exist on Enceladus.
Methane metabolic pathways
Because the high concentration of dissolved methane in the anoxic water is believed to be due to biotic methanogenesis13, the pathways for methanogenesis in the environment were explored and linked to the organisms annotated to those pathways (Fig. 4). Methanogenesis from (x)methylamine resulted in three times fewer methanogenesis pathways in LF92 (21.2%) and LF99 (26.40%) than in UF92 (70.40%). In the sediments, however, it made up only 9% of methanogenesis pathways. In the water column, these pathways were present in unclassified organisms, Firmicutes, Deltaproteobacteria, and Euryarchaea. In the sediments, (x)methylamine methanogenesis pathways were only present in unclassified organisms and Deltaproteobacteria.
An overview of the methanogenesis pathways. (A) Shows the methanogenesis pathway abundances in the phyla. (B) Indicates the methanogenesis pathway abundances in each sample. Since only 44% of the reads mapped back to the UF99 sample, not much information can be inferred from the data belonging to this sample. Here, we view methane metabolism from carbon dioxide to be the main pathway present in Euryarchaea (the only known anoxic methanogens). Methanogenesis from acetate is another pathway seen abundantly in non-methanogens. This could indicate acetate as a source of energy in this anoxic environment.
Methanogenesis from CO2 was most abundant in the UF92 (19.8%) and LF92 (16.3%) samples, at almost double what was found in LF99 (9.7%). In the sediments, 16% of methanogenesis pathways belonged to methanogenesis from CO2. Methanogenesis from CO2 was the main methanogenesis pathways present in Euryarchaea in every sample. Except for LF92, pathways for methanogenesis from CO2 were only present in Euryarchaea and unclassified organisms.
Pathways belonging to methanogenesis from acetate were similar among LF92 (60.5%), LF99 (54.9%), and the sediment (74.9%). In all samples, methanogenesis from acetate was the most abundant in unclassified organisms, Deltaproteobacteria, Betaproteobacteria, and Firmicutes.
Nitrogen
Nitrogen metabolism plays an important role in every ecosystem and nitrogen-bearing organic compounds have been detected in the plumes of Enceladus. In order to determine how nitrogen is cycled in this environment, we investigated the presence and abundance of nitrogen metabolic pathways and found that nitrate reduction pathways were the most abundant in our samples, including assimilatory, dissimilatory, and denitrification pathways. While assimilatory and denitrification pathways were mainly present in the water column (and were the most abundant in the water), other nitrogen pathways were present in both water and within the sediment, including assimilatory nitrate reduction, nitrogen fixation, ammonia assimilation, urea degradation, and denitrification (Fig. 5).
An overview of the nitrogen metabolism pathways. (A) Shows the nitrogen metabolism pathway abundances in the phyla. (B) Shows the nitrogen metabolism pathways abundance in each sample. Since only 44% of the reads mapped back to the UF99 sample, not much information can be inferred from the data belonging to this sample. Nitrate reduction, the most abundant nitrogen metabolism reaction, can couple with methane oxidation and facilitate the usage of methane as an energy source.
Assimilatory nitrate reduction was the most abundant nitrogen pathway, making up between 39–64.9% of nitrogen metabolism pathways in the water column and only 4.7% within the sediment community (Fig. 5). This pathway was mainly found in unclassified organisms, but was also present in large abundance in Deltaproteobacteria, Bacteroidetes, and Firmicutes.
The ammonia assimilation cycle made up between 35.4 and 49.2% of nitrogen metabolism pathways in the water column, however, unlike assimilatory nitrate reduction, in the sediment this pathway represented 60.4% of the nitrogen metabolism pathways. This pathway was present in nearly all major phyla in UF92, LF99, and the sediment. It was most abundant in unclassified organisms.
Denitrification was present in unclassified organisms and Bacteroidetes. This pathway made up 9% of nitrogen metabolism pathways in the water column but was absent in the sediment, suggesting that the environmental conditions in the sediment are not conducive to denitrification.
Sulfur
Sulfate is an effective electron acceptor in anoxic environments and an important metabolic compound in anoxic environments. Additionally, current evidence suggests that a possible biosphere within the oceans of Enceladus may not be limited by the availability of sulfur31. Sulfur reducing bacteria and methane oxidizers have been found to work alongside each other. Given the connection of sulfur pathways with methane metabolism and the high concentration of methane in the anoxic basin, sulfur metabolism pathways in this environment were also examined. In the anoxic basin, the pathways responsible for sulfur metabolism were sulfate activation for sulfonation, sulfate reduction, and hydrogen sulfide biosynthesis (Fig. 6).
An overview of the sulfur metabolism pathways. (A) Shows the sulfur metabolism pathway abundances in the phyla. (B) Shows the sulfur metabolism pathways abundance in each sample. Since only 44% of the reads mapped back to the UF99 sample, not much information can be inferred from the data belonging to this sample. Sulfate reduction, the most abundant sulfur metabolism reaction, can couple with methane oxidation and facilitate the usage of methane as an energy source.
The highest abundance of sulfur pathways belonged to sulfate reduction. This pathway made up 64.5% of sulfur metabolism pathways in the LF92 sample. It made up 66% of sulfur metabolism pathways in the UF92 sample and 56.2% of sulfur metabolism pathways in the LF99 sample. Only 8.1% of sulfur metabolism pathways in the sediments belonged to sulfate reduction. Within the sediments, only unclassified organisms had this pathway. In the water column, this pathway was found in unclassified organisms, Deltaproteobacteria, Betaproteobacteria, and Bacteroidetes.
The hydrogen sulfide biosynthesis pathway was only present in the LF99 (6.2%) and sediment samples (33%). This pathway is present in unclassified organisms, Actinobacteria, Betaproteobacteria, and Deltaproteobacteria.
Source: Ecology - nature.com
