Introduction

Salt flats or salares are one of the most interesting biomes on earth [1]. Based on their hypersaline conditions, prokaryotes have evolved to develop biochemical processes with potential applications in biotechnology [2], providing also opportunities for biosignature detection on other planets [3]. Due to geological, climatic and geomorphological factors, dozens of endorheic basins are located in Northern Chile including evaporitic bodies and saline lakes. Brines and crusts of these saline deposits are enriched in arsenic [4].

The Salar de Ascotán [5] is an athalassohaline environment located at the bottom of a tectonic basin surrounded by volcanic systems in east-west direction, including some active volcanoes with altitudes from 5000 to 6000 m.a.s.l. [6]. The saline crusts are mainly composed of chlorides (halite) and sulfates (gypsum) to economic boron-bearing minerals associated with significant amounts of arsenic sulfides [5], with the arsenic concentrations the highest found in the area [7].

In order to understand the bacterial role in the arsenic biogeochemical cycle at circumneutral pH, several sampling expeditions to Salar de Ascotán, in the Chilean High-Andes, have been taken since 2000. The microbial diversity of this salt flat was first analyzed [7], then enrichment [8], isolation and sequencing efforts [9] as well as geochemical in situ investigations wer performed [10]. In addition, the distribution of genes for the As (III) oxidation (aioA), As (V) detoxifying respiration (arrA), As detoxification (arsC), and As (III) extrusion (acr3) was explored in Salar de Ascotán and other natural environments in Northern Chile with arsenic concentrations spanning six orders of magnitude. The abundance of Firmicutes -like arsC genes compared to the Enterobacterial-like arsC genes in these environments suggested an important role of thioredoxin and the Firmicutes phylum in the local As biogeochemistry [11].

Fusibacter is a minor genus into the Clostridiales order within the Firmicutes phylum. Currently, it comprises four Gram-positive species with validly published names. This group started with the discovery of the thiosulfate-reducing bacterium Fusibacter paucivorans , being the most studied, isolated from oil-producing wells [12]. Fusibacter tunisiensis was isolated from an anaerobic reactor used to treat olive-mill wastewater [13]. Recently, Fusibacter bizertensis was identified from a corroded kerosene storage tank [14], and more recently, Fusibacter fontis was the first species of this genus isolated from a natural environment [15]. In general terms, the reported members of this genus are fermentative and halotolerant anaerobes. Moreover, these species share sulfur-reducing features capable of generating sulfide starting from elemental sulfur [13, 15] or thiosulfate [12, 13] sources. To date, a whole-genome sequence has not been reported for any species within this genus.

Here, we report the first draft genome of a strain of Fusibacter plus some microbiological properties of this halotolerant isolate, recovered from a saline environment in Northern Chile. The strain was deposited as Fusibacter sp. 3D3 as ATCC BAA-2418 because we are still running the necessary tests and deposits to describe the isolate as a new species and “ Fusibacter ascotence” is the proposed species name.

This report contributes to a better understanding of the ecophysiology of extreme halotolerant microorganisms inhabiting saline environments and their role in the arsenic biogeochemistry.

Organism information

Classification and features

Fusibacter sp. 3D3 is an indigenous strain of the Salar de Ascotán hypersaline sediments isolated at the Centro de Biotecnología, Universidad Católica del Norte, Antofagasta, Chile. Enrichment, isolation, and growth experiments were performed in a fresh Newman-modified minimal medium [7] containing, 1% (w/v) NaCl, 0.1% (w/v) yeast extract, and 1 mM cysteine adjusted to pH 7.0. After autoclaving, 10 mM lactate as electron donor, and 20 mM sodium sulfate and 2 mM sodium arsenate as electron acceptors were added in order to complete 20 mL of medium into 50 mL-anaerobic-bottles (Supelco). The strain was incubated in an anaerobic chamber (Airlock, Coydrive), in dark, at 30 °C, under N2:CO2:H2 gas atmosphere (80:15:5, v/v) up to 10 days. The pure colonies were obtained by inclined tubes of agar prepared with the mentioned Newman modified medium plus 2% (w/v) agar incubated under anaerobic conditions at 30 °C. Single yellow colonies were restreaked several times to obtain pure isolates and then were transferred to the liquid medium. Transmission electron microscopy revealed rod-shaped cells (0.4 μm × 3-10 μm) (Fig. 1a).

Fig. 1
figure 1

Isolation of Fusibacter sp. 3D3. a Transmission electron micrograph of bacterial cells filled with electron dense granules of variable density (Bar = 500 nm). b Arsenic speciation, lactate and acetate measurements of Fusibacter sp. strain 3D3 culture in Newman’s modified medium with 20 mM lactate, 10 mM sulphate, 2 mM arsenate, 0.1% (w/v) yeast extract, and 1 mM cysteine versus time. All error bars represent the standard error of the mean of triplicate cultures

Full size image

Arsenate reduction was tested by inoculation of 1 × 10−6 cells mL−1 into 20 mL of fresh Newman-modified medium under incubation conditions described above. An abiotic control was carried out in sterile medium without inoculum. Growth curves were performed in triplicate and monitored by counting chambers (0.01 mm × 0.0025 mm2, Neubauer, Marienfeld). Samples were acquired periodically, then centrifuged (15,000×g; 10 min) to remove cells, and finally filtered through a 0.2 μm cellulose filter. The filtered supernatant was sealed and refrigerated at 4 °C to preserve arsenic speciation until analysis. As (V) and As (III) concentrations were measured with a mobile phase of 10 mM acid phosphate at 6.25 pH by Millennium Excalibur HPLC System (PS Analytical, Orpington, UK). To quantify lactate and acetate, each filtrate was injected in a Dionex IonPac AS11-HC column to run a high-performance liquid chromatography (Thermo Scientific model 3200) with an isocratic concentration of KOH. Arsenate reduction and simultaneous lactate consumption were evidenced (Fig. 1b). Arsenate reduction has not been reported in the other members of the Fusibacter genus.

The Initial identification of strain 3D3 was performed by 16S rRNA gene amplification using a previously described method [7]. The 16S rRNA sequences of strain 3D3 clustered with type strains of Fusibacter species when those were aligned using Clustal W and manually corrected. A phylogenetic tree was constructed using neighbor-joining, maximum-parsimony, and maximum likelihood algorithms with bootstrap values of 500 replicates using the MEGA program version 6.22. Phylogenetic analysis of the 16S rRNA sequence indicated that the strain 3D3 belongs to the genus Fusibacter and exhibits a similarity of 98% with Fusibacter sp. Vns02, and 95% with both Fusibacter paucivorans and Fusibacter tunisiensis (Fig. 2).

Fig. 2
figure 2

Phylogenetic tree based on the 16S rRNA gene sequences highlighting the position of Fusibacter sp. strain 3D3 relative to other type and non-type strains of the genus Fusibacter. The Genbank database accession codes (in brackets) is indicated. Bootstrap values for 500 replicates are indicated at the nodes. Scale bar 0.02 substitutions per nucleotide position

Full size image

RapID™ NF Plus and RapID™ One (Thermo Scientific), two qualitative micromethods employing conventional and chromogenic substrates for the biochemical features identification were performed following the manufacturer’s instructions (Table 1). A single colony was inoculated into the given fluid and incubated for 24 h at 30 °C. Then, the inoculation fluid was transferred to the corresponding panel. The interpreted results were entered into the electronic RapID code database (ERIC electronic compendium, version 1.0.771, patch 0613). Comparing with the ERIC™ database, our results failed to identify our isolate (Table 1).

Table 1 Biochemical analyses of Fusibacter sp. strain 3D3
Full size table

Genome sequencing information

Genome project history

Based on its phylogenetic position and 16S rRNA similarity, Fusibacter sp. 3D3 (Taxonomy ID: 1,048,380) was previously submitted to NCBI in 2010 (Gene Bank 16S rRNA gene: FR873490.1) (Fig 2). Later, in 2013, it was deposited as Fusibacter sp. 3D3 in ATCC BAA-2418, being the first strain of this genus coming from an extreme arsenic bearing and saline biotope (Table 2). This organism was selected for genome sequencing based on its interesting phenotypic characteristics. Recently, in 2016, the submission of the whole shotgun project assembled as a draft genome was performed to the DNA Data Bank of Japan under the Bioproject accession number PRJDB4973 and Biosample number SAMD00055724 (ID 573014). This Whole Genome Shotgun project has been deposited at GenBank under the accession BDHH00000000. The version described in this paper is the first version, BDHH01000000 [16]. Table 3 presents the project information and its association with MIGS version 2.0 compliance [17].

Table 2 Classification and general features of Fusibacter sp. strain 3D3 [18]
Full size table
Table 3 Project information
Full size table

Growth conditions and genomic DNA preparation

Fusibacter sp. strain 3D3 was grown anaerobically on fresh Newman [18] modified medium and conditions of incubation described above. DNA was extracted using High Pure Template Preparation Kit (Roche, Germany), according to the manufacturer instructions. Both quantity and quality of the genomic DNA were measured using a NanoDrop ND-1000 spectrophotometer (Thermo-Fisher Scientific Inc.) and analyzed by DGGE (200 C. B. S. Scientific Company), respectively. The purity of strain 3D3 was confirmed by a single band in the DGGE profile.

Genome sequencing and assembly

The genome of Fusibacter sp. 3D3 was sequenced on an Illumina MiSeq platform at Molecular Research Laboratory (MR. DNA, Shallowater, TX). The library for each sample was prepared using a Nextera DNA Sample Preparation Kit (Illumina), following the manufacturer’s instructions. Sequencing of 2 × 300-bp paired-end reads allowed for an estimate of 20,000 output with an average coverage over 50 times fold. The assemblage of quality-filtered reads was executed by MR-DNA for the complete genome sequence. As a result, the draft genome of ~5.1 Mbp size was generated. Reads were assembled de novo using Newbler v2.0.01.14. The final draft assembly contained 57 contigs identifying 4780 genes using RAST [19].

Genome annotation

Genes were predicted using Glimmer 3.02 [20] as part of the RAST annotation pipeline using SEED platform for 4780 features identified. Whole RNA genes were also predicted by the same annotation platform [21]. The predicted protein coding genes were analyzed for the presence of signal peptides using SignalP 4.1 Server [22]. TMHMM Server v. 2.0 was utilized for prediction of transmembrane helices in proteins [23]. Geneious 7.1.9 (Biomatters) software was used to analyze COGs utilizing the BLAST COG database (Table 4). Pfam domains were computed using InterProScan 5.19-58.0 [24]. CRISPRs repeats were found submitting the contigs to the CRISPRs Finder web server [25].

Table 4 Number of genes associated with general COG functional categories
Full size table

Genome properties

The draft genome for Fusibacter sp. 3D3 contained 5,111,250 nucleotides with an average G + C content of 37.6% (Table 5). From 4780 genes, 4700 were predicted protein coding, and 80 RNA coding genes (12 rRNA, and 62 tRNA genes). The putative function was assigned to 63.1% of the genes, while the remaining genes were annotated as hypothetical proteins. The distribution of genes in COGs functional categories is presented in Table 4.

Table 5 Genome statistics
Full size table

Insights from the genome sequence

Similarity analysis of genes involved in the arsenic metabolism indicated that the closest available genome of strain 3D3 in the database was Clostridium sticklandii [26], which belongs to the Clostridiales family as well. The subsystem information approach to genome annotation performed by RAST/SEED [27] confirmed the relation to other members in the Clostridiales order (Table 6). Arsenic detoxification genes are clearly present in Fusibacter sp. 3D3 genome, however, genes coding for arsenate respiratory reductases (arr) and arsenite oxidases (aio) have a very low percentage of similarity with genes coding for the enzymes evidenced at protein level. The arsC gene sequence identified in the Fusibacter sp. genome was clustered inside the Firmicutes -like arsC gene clade whose predominance has been reported in Salar de Ascotán [11].

Table 6 BLAST results of predicted and best-scored proteins related to arsenic in the Fusibacter sp. strain 3D3 genome
Full size table

In the vicinity of the ArsC coding gene is the gen F3D3_RS05420. This piqued our interest and, in a deeper analysis, we found that F3D3_RS05420 codify for pyridine nucleotide-disulfide oxidoreductase NADH dehydrogenase (accession number: WP_069871897). The preliminary information indicates that the protein encoded by the gen F3D3_RS05420 is part of a new family of proteins of unknown function. However, the genomic context shows us some clues to formulate a hypothesis. By means of comparative genomics we identify two common components accompanying genes like F3D3_RS05420: A) genes codifying for transcription regulators and, B) genes codifying for arsenical transporters (Fig. 3c). This could be an indicative of a possible role in the response to stress by As. The multiple sequence alignment carried out using MUSCLE application [28] in CLC Genome Workbench 8.0 (Qiagen) shows that the protein is distributed in the Firmicutes bacteria and it is strongly conserved (Additional file 1: Figure S1).

Fig. 3
figure 3

Summary of subsystems identified by RAST/SEED. a Subsystem coverage. 41% of the whole CDS were identified in subsystems. b Subsystem category distribution of all non-hypothetical CDS. No photosynthesis and secondary metabolites CDS were assigned. c Assigned CDS for arsenic metabolism in the Fusibacter sp. strain 3D3 genome comparing to closely related species. Diagram of the chromosomal region (Contig 49) of the arsenic-related genes compared with five microorganisms. 1. Arsenical resistance operon repressor ArsR, 2. Pyridine nucleotide-disulfide oxidoreductase NADH dehydrogenase (EC 1.6.99.3), 3. Arsenical resistance protein ACR3, 4. Arsenate reductase ArsC (EC 1.20.4.1), 5. arsenical efflux pump membrane protein ArsB, 6. Arsenical resistance operon trans-acting repressor ArsD, 7. Arsenical pump-driving ATPase ArsA (EC 3.6.3.16)

Full size image

The protein architecture of WP_069871897 shows a CoA-disulfide reductase domain (TIGR03385) and a rhodanese domain (PFAM00581). A rhodanese domain is also present in the ACR2 protein of Saccharomyces cerevisiae which also has arsenate reductase activity [29]. The catalytic loop of the rhodanese domain has two known configurations, a short version with four residues to accommodate sulfur or selenium atoms and an extended version with five residues to accommodate an arsenic or phosphorus atoms [30]. The architecture of WP_069871897 and related proteins suggest a role similar to ACR2, namely arsenic reductase. However, the comparison between the catalytic loops of ACR2 (Q06597) “CTGSKNRG” with the “CNKGVTGN” of WP_069871897 does not show an apparent similitude, which makes it difficult to extrapolate the activity of ACR2 with WP_069871897. In addition, the presence of the arsC gene in the compared genomes (Fig. 3c), but not in the same genomic context, suggests that the WP_069871897 and related proteins does not substitute the ArsC activity. Then, it remains a challenge to the scientific community to answer if the proteins similar to WP_069871897 are a new kind of arsenic reductase or if they are in some way involved with the response to arsenic stress.

Conclusions

The 5.11 Mbp draft genome sequence of Fusibacter sp. 3D3 is arranged in 57 contigs, being the first Fusibacter draft genome published. It potentially includes 4700 protein-coding genes, 67.1% of which were assigned to function prediction. 80 RNA genes partitioned in 12 rRNA and 62 tRNAs were identified. The release of the genome sequence of this strain will provide new insights into arsenic reduction processes in hypersaline biomes and further understanding of the mechanisms used by halophile bacteria to endure high osmotic stress in natural and industrial saline environments.