Metagenomic survey of methanesulfonic acid (MSA) catabolic genes in an Atlantic Ocean surface water sample and in a partial enrichment

Abstract

Methanesulfonic acid (MSA) is a relevant intermediate of the biogeochemical cycle of sulfur and environmental microorganisms assume an important role in the mineralization of this compound. Several methylotrophic bacterial strains able to grow on MSA have been isolated from soil or marine water and two conserved operons, msmABCD coding for MSA monooxygenase and msmEFGH coding for a transport system, have been repeatedly encountered in most of these strains. Homologous sequences have also been amplified directly from the environment or observed in marine metagenomic data, but these showed a base composition (G + C content) very different from their counterparts from cultivated bacteria. The aim of this study was to understand which microorganisms within the coastal surface oceanic microflora responded to MSA as a nutrient and how the community evolved in the early phases of an enrichment by means of metagenome and gene-targeted amplicon sequencing. From the phylogenetic point of view, the community shifted significantly with the disappearance of all signals related to the Archaea, the Pelagibacteraceae and phylum SAR406, and the increase in methylotroph-harboring taxa, accompanied by other groups so far not known to comprise methylotrophs such as the Hyphomonadaceae. At the functional level, the abundance of several genes related to sulfur metabolism and methylotrophy increased during the enrichment and the allelic distribution of gene msmA diagnostic for MSA monooxygenase altered considerably. Even more dramatic was the disappearance of MSA import-related gene msmE, which suggests that alternative transporters must be present in the enriched community and illustrate the inadequacy of msmE as an ecofunctional marker for MSA degradation at sea.

Keywords: Bacteria; Biogeochemical cycle; Gene; Metagenomics; Methanesulfonic acid; Ocean; Sulfur.

Figure 1. Phylogenetic composition (phyla) of the two metagenomes, SCD0 and SCDE (in percentage).

Only phyla with abundance ≥1% in either sample are shown (based on EBI Metagenomics analysis of the data).

Figure 2. Taxonomic composition of the two metagenomes (in percentage).

Only taxa with abundance ≥2% in either sample are shown. Taxonomic classification as provided by the phylogenetic analysis of the EBI Metagenomics pipeline (which explains some apparent taxonomic inconsistency such as “Alpha Rickettsiales Pelagibacteraceae,” referring to sequences that could be classified down to the family level, and further down “Alphaproteobacteria,” referring to sequences that could be classified just at the class level). Alpha, and Gamma are abbreviations for the corresponding classes within the Proteobacteria.

Figure 3. Significant shifts in phylogenetic composition observed due to the enrichment.

Shown are percentages of the abundance of each taxon in each sample (SCD0 or SCDE) over the total taxon abundance (SCD0 + SCDE). Only statistically significant differences are shown. Taxa are as provided by the phylogenetic analysis of the EBI Metagenomics pipeline (which explains some apparent taxonomic inconsistency, see note in Fig. 2). Alpha, Beta, Gamma, Delta and Epsilon are abbreviations for the corresponding classes within the Proteobacteria.

Figure 4. Comparison of the rarefaction curves constructed with sequencing data from the amplicon survey experiment.

SCD0-A and SCDE-A refer to gene msmA before and after the enrichment with MSA, respectively. SCD0-E refers to gene msmE before the enrichment.

Figure 5. Conservation analysis of the predicted MsmA sequences from joined samples SCD0-A and SCDE-A.

Displayed on top is the consensus sequence. Amino acids in blue correspond to PCR primers. Amino acids in red correspond to the cysteine and histidine residues typical of the Rieske-associated motif. Amino acids in green represent the characteristic long spacer found in the Rieske motif in MsmA. Low conservation of the beginning and end of the sequence (corresponding to PCR primers) are artifacts explainable by the presence of short reads in the dataset.

Figure 6. Conservation analysis of the predicted MsmE sequences from sample SCD0-E.

Displayed on top is the consensus sequence. Amino acids in blue correspond to the PCR primers. Low conservation of the beginning and end of the sequence can be explained as in Fig. 5.

Figure 7. Phylogenetic tree of msmA sequences.

Clusters with sequences from this study are in blue: CLA stands for msmA clusters. Between brackets is the total number of sequences in each cluster. Horizontal bars indicate the relative frequency of sequences of each cluster (SCD0 in blue and SCDE in red). The accession numbers of the sequences previously published are between brackets. The omitted branch is constituted mostly by sequences from genomes of cultured strains. The accession numbers of these sequences are the following: AF354805 (Marinosulfonomonas methylotropha str. TR3), KJ789392 (Methylobacterium sp. str. P1), KJ789392 (Hyphomicrobium sp. str. P2), GOS sequence JCVI_READ_2101946, KM879220 (C. Filomicrobium marinum str. Y), NC_011892 (Methylobacterium nodulans ORS 2060), NC_011894.1 (Methylobacterium nodulans ORS 2060), KJ789395 (Methylobacterium sp. str. RD41), NZ_KB375270 (Afipia felis str. ATCC 53690), EF459501 (Afipia felis str. 25E1), AF091716 (Methylosulfonomonas methylovora str. M2), CP001751 (C. Puniceispirillum marinum str. IMCC1322), NZ_AKCV01000022 (Ralstonia sp. str. PBA), AP014581 (Burkholderia sp. str. RPE67), and CP003775 (Burkholderia cepacia str. GG4), CCYE01000041 (Pseudomonas xanthomarina str. S11). Nucleotide sequences corresponding to cluster-representative MsmA sequences obtained at 0.15 distance cutoff were used to infer the phylogenetic tree by Maximum Likelihood with 100 bootstrap iterations. Bootstrap values <50% are omitted.

Figure 8. Phylogenetic tree of msmE sequences.

Clusters with sequences from this study are in blue: CLE stands for msmE clusters. Between brackets is the total number of sequences in each cluster. Sequences from cultured strains are in orange. The accession numbers of the sequences previously published are between brackets. The omitted branch is constituted by sequences from genomes of cultured strains, with the following accession numbers: NZ_AZUP00000000.1 (Methyloversatilis discipulorum str. FAM1), NZ_AFHG01000044 (Methyloversatilis universalis str. FAM5), NZ_ARVV01000001 (Methyloversatilis discipulorum str. RZ18-153), NZ_AKCV01000024 (Ralstonia sp. str. PBA), CCAZ020000001 (Afipia felis genospecies A str. 76713), NZ_JNIJ01000008 (Bradyrhizobium sp. str. URHD0069), NZ_KB891326 (Thiobacillus thioparus str. DSM 505), NZ_AQWL01000003 (Thiobacillus denitrificans str. DSM 12475), AZSN01000017 (Methylibium sp. str. T29-B), NC_008825 (Methylibium petroleiphilum str. PM1), NZ_JADL01000017 (Rhodospirillales bacterium str. URHD0088), KP025766 (Methylobacterium sp. str. P1), AF091716 (Methylosulfonomonas methylovora str. M2), and KP025767 (Marinosulfonomonas methylotropha str. TR3). Nucleotide sequences corresponding to cluster-representative MsmE sequences obtained at 0.10 distance cutoff were used to infer the phylogenetic tree by Maximum Likelihood with 100 bootstrap iterations. Bootstrap values <50% are omitted.