| Sumario: | Figure S1. Geographical distribution of the six locations studied here and their inter-location distances (given in the table). The distances given in each location refer to Mallorca. The hypersaline here are: Mallorca (Es Trenc and S’Avall and separated by 3.5 km) and Santa Pola in Spain, Fără Fund in Romania, Great Salt Lake (USA), Pampa (Laguna Colorada Chica and Laguna Colorada Grande, separated by 23 Km), and Lake Grassmere (New Zealand). Table S1. Statistics of Salinibacter MAGs recovered using different bioinformatic tools. MAGs selected for further analysis and taxonomic classification are marked in red. Text S1: In addition to the taxonomic study, here we assessed the different tools SPADES and MegaHIT assemblers, and MetaBAT binning tools selecting contigs with length > 2,000 pbs or > 5,000 pbs to retrieve the best quality MAG (Sup. Table S1). We observed that different tools and combinations rendered different qualities of the MAGs. The selection of contigs with length > 5,000 pbs binning rendered MAGs with a lower genome size, lower completeness and similar contamination. From the FF metagenome (Romania), the highest quality MAG was retrieved using the assembler SPADES and MetaBAT binning tool using a minimum contig length of 2,000 pbs. From CCH and CG metagenomes (Argentina), the highest quality MAGs were retrieved using the assembler MegaHIT and MetaBAT binning tool using a minimum contig length of 2,000 pbs (Sup. Table S1). Figure S2. Genomic clustering based on Average Amino-acid Identity (AAI) of genomes included in the study. Figure S3. Pangenome hierarchical clustering based on the presence (grey) or absence (black) of orthologous genes. Table S2. Pangenome statistics between genomes belonging to the same species. Table S3: CRISPR-Cas systems found in the Salinibacter pepae and Salinibacter grassmerensis genomes. Table S4. Cell morphologies of the new isolates. All cells have been cultivated onto MA agar. for 10 days at 12ºC. The morphology was observed under an optical Microsope (Zeiss Axio Imager 100X). Photomicrograph showing cells of strain observed by oil-immersion differential interference contrast (DIC) microscopy (Nomarski). Figure S4. Hierarchical clustering of all cellular and supernatant metabolomes analyzed with electrospray-negative mode ICR-FT/MS (upper panel). Loading plots in where single molecules are colored based on the specific classes that they belong to: Grey: Core Metabolome. Yellow: discriminative for Sal. ruber. Red: discriminative for Sal. pepae. Orange: discriminative for Sal. altiplanensis (middle panel). Core metabolomes of the genus (lower panel). (A) water-soluble (left figures) and (B) water-insoluble (but methanol soluble; right figures) fractions. Figure S5. Count of saturated, mono-, di-, tri-, tetra-, and more unsaturated fatty acids classified as core metabolites in water-soluble (orange color) and water-insoluble (blue color) fractions versus their degree of unsaturation. Insert Venn diagram show count of unique and common fatty acids in each case. Figure S6. Count of discriminating fatty acids classified as saturated, mono-, di-, tri-, tetra-, and more unsaturated compounds in Sal. pepae, Sal. ruber and Sal. altiplanensis according to their degree of unsaturation in the water-soluble fractions or in the water-insoluble fractions. Figure S7. Van Krevelen plots showing shared compounds assigned in Sal. grassmerensis NZ140T with the discriminating molecular compositions of Sal. ruber, Sal. pepae and Sal. altiplanensis when comparing sample pairs (see also Figure 5B, 5C and 5D). Here we looked at the molecular compositions assigned in the case of Sal. grassmerensis NZ140T and checked whether they are present in the discriminating molecular compositions of Sal. ruber, Sal. pepae and Sal. altiplanensis. Insert histograms represent the molecular series based on CHO (blue), CHOS (green), CHNO (orange), and CHNOS (red) atom combinations. Insert percentages represent the numbers (in %) of shared molecular compositions of Sal. grassmerensis NZ140T with respect to the total discriminant compounds of Sal. ruber, Sal. pepae as well as Sal. altiplanensis (see also Figure 5B, 5C and 5D). Text S2: From the Laguna Colorada Chica, we detected 38 contigs encoding for a 16S rRNA gene and 3 affiliated with the Salinibacteraceae family (Sup. Figure S6). We identified one almost complete 16S rRNA sequence (1,555 bp) in a contig with a sequencing depth of 77.3X, and 2 partial sequences (< 945 bp) showing a sequencing depth < 12.7X. The largest 16S rRNA gene sequence with the highest coverage affiliated with the Salinibacter genus with an identity of 97.3% with Sal. pepae ESAV49Ts, 96.9% with Sal. ruber M31T, 96.3% with Sal. altiplanensis AN15T, 96% with Sal. grassmerensis NZ140T and 96.2% with Sal. abyssi ROFFTs, respectively (Sup. Figure S6 and Sup. Spreadsheet S2). The shorter assembled 16S rRNA sequences showed a percentage of similarity with any of the Salinibacter sequences < 95.4% (Sup. Figure S6). ARCCHTs represented the most abundant Salinibacter population in the metagenome of origin (with 1.9% relative abundance and 56X coverage), we confidently assigned the contig encoding the largest Salinibacter 16S rRNA gene to ARCCHTs. The sequence was deposited under the accession number GCA_947077715Ts. Supporting the assignation, from the Laguna Colorada Grande we detected 28 contigs encoding a 16S rRNA gene, 3 affiliated with the Salinibacteraceae family. The single complete 16S rRNA gene sequence (accession number GCA_947077705) showed 100% identity with the one assigned to ARCCHTs (Sup. Figure S3). In agreement, ARCG coverage 12X and the 16S rRNA encoding contig 17X. The agreement between the coverages and both affiliations based on 16S rRNA gene sequence reconstruction (Figure 1A) and the core-genome reconstruction (Figure 1B) confidently assigned the 16S rRNA genes to their respective MAGs in the samples of origin. Figure S8. Phylogenetic reconstruction based on the 16S rRNA gene sequence analysis of all Salinibacter species available in the LTP_01_2022, the MAGs recovered from metagenomes and the Salinibacteraceae 16S rRNA sequences recovered from assembled metagenomes from Colorada Chica (CCH) and Colorada Grande (CG), both located in Argentina. The tree was reconstructed using the maximum likelihood algorithm and is the result of the consensus of different approaches using distinct filters and datasets. The multifurcations indicate a branching order that could not be resolved. Bar indicates 10% sequence divergence. In brackets the accession number of each sequence is given. Table S5. MAGs metabolic reconstruction based on KEGG database annotations.
|
|---|
