Bacterium isolation was carried out in LB medium supplemented with increasing concentrations of mercury and tellurite (Rodríguez-Rojas et al., 2015). Briefly, P. putida ATH-43 was grown aerobically in LB medium supplemented with 40 μM HgCl₂ at 25°C for 48 h. DNA extraction was performed using the Wizard® Genomic DNA Purification Kit (Promega). The quality and quantity of genomic DNA was determined by 0.8% agarose gel electrophoresis and by 260/280 nm absorbance ratio using the microplate multireader Tecan Infinite® 200 PRO.

Tree was constructed using Maximum Likelihood algorithm by MEGA 6.0 software. Best model was calculated revealing that Jukes Cantor model was the best fit for this nucleotide data set. The option “use all sites” for gaps treatment was also applied. Node numbers represent the per cent of bootstrap replicates of 1000 resamplings (values below 50% are not shown). Sequence alignments was from nucleotide position 24 to 1488 as compared to E. coli K12. Scale bar represents 0.01 substitutions per nucleotide positions. Arrow points to the outgroup E. coli K12 (accession number AP012306). Accession numbers for all Pseudomonas strains included in the study are given in parentheses. Average Nucleotide Identity (ANI) was performed with pyANI using ANIm, which is based on hidden Markov models. For pangenome analysis, GET_HOMOLOGS was used with three clustering algorithms (bidirectional best-hit, COGtriangles, and OrthoMCL). Only congruent results between the three algorithms were used for the final analysis. Strains of P. putida and their respective genome accession numbers were W619 (CP000949), SQ1 (JTCJ00000000), F1 (CP000712), KT2440 (AE015451), SF1 (LDPF00000000), BIRD-1 (CP002290), DLL-E4 (CP007620), H8234 (CP005976), and NBRC (AP013070).

P. putida ATH-43 DNA was submitted to Macrogen® (Seoul, Korea) for next generation whole-genome shotgun sequencing using Illumina Hiseq 2000 platform (January, 2015). Read mapping, de novo assembly and genome annotation was performed at the Fraunhofer Chile Research Foundation (Santiago, Chile). DNA sequence was determined by a whole-genome shotgun strategy with a mate pair library of 3 kb (Macrogen®). A total of 10.05 million reads were obtained with an average length of 101 nucleotides. All reads were quality filtered and assembled using the A5 pipeline, an integrated pipeline for de novo assembly of microbial genomes (Tritt et al., 2012). The assembled genome of P. putida ATH-43 consists of 5.8 Mbp distributed over 260 contigs and organized in 64 scaffolds with fold coverage of 172X.

Open reading frame prediction and annotation was carried out using standard operational procedures (Tanenbaum et al., 2010). Gene models were predicted using Glimmer 3.02 (Salzberg et al., 1998). Predicted coding sequences were annotated by comparison with public databases, BLAST 2.2.31 (Altschul et al., 1990) was used to find homologous sequences with COG, UNIPROT, and NR-NCBI databases, and Hmmer 3.1 was used against PFAM and TIGRFAM. Automatic metabolic reconstruction was carried out using PRIAM software (Claudel-Renard et al., 2003).

P. putida ATH-43 is a Gram negative, non-sporulating, motile, aerobic, rod-shaped (average bacterium dimensions were 6.5 μm length and 2.3 μm width) and psychrotrophic bacterium (Figure 1A) that was isolated from Antarctic sediments at anthropogenic settlements visibly contaminated with oxide compounds. Sequencing of complete 16S rRNA gene revealed phylogenetic affiliation of strain ATH-43 as a member of the Pseudomonadaceae family from the Gammaprotebacteria class, strongly related to P. putida species (Figure 1B). Interestingly, strain ATH-43 forms a distinctive clade with isolate P. putida W619, which are clearly differentiated from the rest of Pseudomonas isolates, suggesting a different evolutionary ancestral divergence (Figure 1B). In concordance, the Average Nucleotide Identity (ANI) analysis performed with nine selected strains, revealed a close relationship with P. putida strains W619 and SQ1 (Supplementary Figure 1A), further contributing to defining the relationship between these bacterial strains. In addition, a pangenome analysis was performed with the same strains that were used for ANI analysis. In this line, ATH-43 strain presented an important number (741) of unique accessory genes, that may contribute to the multi-stress resistance phenotype observed in this strain. Also, a core genome of 1370 genes was shared among the P. putida strains, given by their threshold of 90–99% nucleotide identity (Supplementary Figure 1B).

The assembled genome of P. putida ATH-43 consists of 5,830,220 bp, with an average G+C content of 61.5% (Table 1). A set of 79 tRNA genes and two clusters of rRNA genes were identified. From a total of 5124 predicted protein-coding sequences (CDSs), 4436 (86.5%) open reading frames (ORFs) matched coding sequences available in public databases, and 3372 (65.8%) were assigned in clusters of orthologous groups (COG) categories (Figure 1C).

Open reading frame prediction revealed the presence of multiple genetic determinants related to heavy metals and some metalloids tolerance. This prediction included resistance genes and operons to Hg²⁺, Cu²⁺, Zn²⁺, Ni²⁺, Cd²⁺, Co²⁺, Pb²⁺, CrO42−, AsO43−, SeO32−, and TeO32− (Figure 1D and Supplementary Table 1), and also genes regarding resistance to antibiotics and drugs such as tetracycline, macrolides, penicillins, aminoglycosides, and streptomycin, among others (Supplementary Table 2). In addition, computational prediction showed the presence of genes encoding seven cold-shock proteins, 773 hypothetical proteins, and an important battery of genes participating in the oxidative stress response, including the unusual mycothiol synthase found exclusively in Gram positive bacteria (Rawat et al., 2007).

The ATH-43 genome harbors more tRNA gene sequences than other known P. putida genomes, a trait that may reflect the cell's adaptation to extreme conditions (Wu et al., 2011). In fact, higher tRNA genes content seem to be related with specific cold adaption mechanisms, as determined by comparative genomics analysis (Dutta and Chaudhuri, 2010). In addition, the number of genes involved in signal transduction mechanisms and inorganic ion transport and metabolism (two COG categories) in this bacterium's genome, is similar to other P. putida strains but exceeds that displayed by other Gammaproteobacteria (Wu et al., 2011). These data may suggest the presence of complex systems of molecular mechanisms controlling gene expression in microorganisms that thrive under highly variable environments.

On the other hand, 13 IS elements were found in the ATH-43 genome sequence using the IS finder tool (www-is.biotoul.fr), along with 21 transposases and 17 integrases (not shown), all elements routinely associated with horizontal gene transfer providing advantage in metal and antibiotic resistance, general stress tolerance, and aromatic compound degradation, among others (Vos et al., 2015; Koonin, 2016). As with tRNA genes, the genome of ATH-43 displays a higher number of genetic determinants involved in metal, antibiotic, and oxidative stress resistance as compared with other reference P. putida genomes (Wu et al., 2011), which may be a reflect of the dramatic selective pressure occurring in the Antarctic continent.

The whole-genome shotgun project was deposited in GeneBank and is publicly available since July, 2015 under the accession number LBME00000000 (Direct link: http://www.ncbi.nlm.nih.gov/nuccore/LBME00000000.1).

The 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.