Introduction

Soda lakes are formed in environments where high rates of evaporation lead to the accumulation of soluble carbonate salts due to the lack of dissolved divalent cations. Consequently, soda lakes are defined by their high salinity and stable highly alkaline pH conditions, making them dually extreme environments. Soda lakes occur throughout the American, European, African, Asian and Australian continents and host a wide variety of Archaea and Bacteria, specialized at surviving under such high salt and high pH conditions [1]. These haloalkaliphiles drive a number of biogeochemical cycles essential to their survival, most notably; the sulfur cycle is very active in these unique habitats [2,3,4]. The most noteworthy taxa associated with the reductive sulfur cycle are the Deltaproteobacteria and the Firmicutes . Recently, a number of Gram-positive Firmicutes genomes have been analyzed and published describing their metabolic potential and environmental adaptations, including the polyextremophile Natranaerobius thermophilus [5], and species belonging to the Desulfotomaculum spp. [6,7,8] and the Desulfosporosinus spp. [9]. Here we give an extended insight into the first known genome of a haloalkaliphilic Gram-positive sulfur disproportionator within the phylum Firmicutes : Dethiobacter alkaliphilus AHT1T.

Organism information

Classification and features

The haloalkaliphilic anaerobe D. alkaliphilus AHT1T was isolated from hypersaline soda lake sediments in northeastern Mongolia [10]. D. alkaliphilus AHT1T cells are Gram-positive and the motile rod-shaped cells form terminal ellipsoid endospores (Fig. 1). The strain tolerates salt concentrations ranging from 0.2–0.8 M Na+ with an optimum at 0.4 M and is an obligate alkaliphile, growing within a pH range from 8.5–10.3 with an optimum at 9.5 [10]. Phylogenetic analysis showed that strain AHT1T is a member of the phylum Firmicutes and the order Clostridiales (Fig. 2). Its closest relative is an acetate-oxidizing syntrophic alkaliphile, described as “Candidatus Contubernalis alkalaceticum” which was isolated from a soda lake [11] (Fig. 2). The 16S ribosomal RNA of D. alkaliphilus AHT1T (EF422412) is 88% identical to the 16S rRNA of “Candidatus Contubernalis alkalaceticum” (DQ124682) [12].

Fig. 1
figure 1

Morphology of D. alkaliphilus AHT1T. a Phase contrast micrograph of cells. b Electron microscope image of a D. alkaliphilus AHT1T cell

Full size image
Fig. 2
figure 2

Neighbour-joining tree based on 16S rRNA gene sequences showing the phylogenetic position of D. alkaliphilus AHT1T to other species within the phylum Firmicutes. The Deltaproteobacteria were used as an outgroup, but were pruned from the tree. The dots indicate bootstrap values between 80 and 100%. The scale bar indicates a 2% sequence difference. The tree was constructed with the ARB software package [48] and the SILVA database [29]. The bootstrap values were calculated using MEGA-6 [49]

Full size image

Extended feature descriptions

D. alkaliphilus AHT1T is an obligate anaerobe that can produce sulfide by using elemental sulfur and polysulfides as electron acceptor [10]. Additionally, it has been shown to incompletely reduce thiosulfate to sulfide and sulfite with hydrogen or formate as electron donor [10]. Strain AHT1T is the first representative from the Firmicutes with the metabolic capacity to grow by elemental sulfur disproportionation [13] and, therefore, is a very interesting organism to compare to the typical sulfur disproportionators from the Deltaproteobacteria . This species may play an important role in the reductive sulfur cycle in soda lake environments [2] and possibly also in other alkaline anaerobic habitats, such as serpentinization “cement springs”, where sequences closely related to Dethiobacter have been found [14, 15]. Also, its affiliation with the syntrophic Clostridia “Candidatus Contubernalis alkalaceticum” (Fig. 2) implies that D. alkaliphilus AHT1T could be involved in syntrophic anaerobic metabolic activity. More classifications and features of this species are listed in Table 1.

Table 1 Classification and general features of D. alkaliphilus AHT1T
Full size table

Genome sequencing information

Genome project history

This organism was selected for sequencing at the JGI (http://jgi.doe.gov) based on its potential for bioremediation and biotechnological applications. It is part of the Community Science Program: Haloalkaliphilic sulfate-, thiosulfate- and sulfur-reducing bacteria (CSP_788492). The project is registered in the Genomes OnLine Database (Ga0028528) [16] and the permanent draft genome sequence is deposited in GenBank (RefSeq: NZ_ACJM00000000.1). Draft sequencing and assembly were performed at the JGI using state of the art sequencing technology [17]. The project information is summarized in Table 2.

Table 2 Project information
Full size table

Growth conditions and genomic DNA preparation

Strain AHT1T was grown anaerobically at 30 °C in Na-carbonate buffered mineral medium (22 g/L Na2CO3, 8 g/L NaHCO3, 6 g/L NaCl, 1 g/L K2HPO4) with a pH of 10 and 0.6 M total Na+. Additionally, 4 mM NH4Cl, 1 mM MgCl2 x 6H2O and 1 mlL−1 trace element solution were added [18]. After sterilization, acetate serving as carbon source (2 mM) and thiosulfate (20 mM) the electron-acceptor, were also added to the medium. The culture (2 L) was grown in a 10 L bottle mounted on a magnetic stirrer whereby the headspace (8 L) was replaced by 100% (v/v) H2, at 0.5 Bar overpressure, acting as the electron-donor. Half the culture volume (1 L) was centrifuged at 13,000 g for 30 min, the pellet was washed with 1 M NaCl and frozen at -80 °C until further downstream processing. DNA was extracted from the pellet by the phenol-chloroform method after pre-treatment with SDS-proteinase K according to Marmur [19]. The concentration and molecular weight of the DNA were checked by UV spectroscopy and gel electrophoresis, respectively.

Genome sequencing and assembly

The size of the assembled D. alkaliphilus AHT1T genome sequence was 3.12 Mbp. The draft genome was generated at the JGI using a combination of Sanger, Solexa/Illumina [20] and 454 DNA sequencing technologies [21]. An 8 Kb Sanger library was constructed that provided 2.5 x coverage of the genome (15,321 reads generated) and a Solexa shotgun library and a 454 Titanium standard library, which provided 25× genome coverage totalling 110.0 Mbp of 454 data. The 454 Titanium data were assembled with Newbler. The Newbler consensus sequences were computationally shredded into 2 Kb overlapping fake reads (shreds). Illumina sequencing data was assembled with VELVET, version 1.0.13 [22], and the consensus sequences were computationally shredded into 1.5 Kb overlapping fake reads (shreds). We then integrated Sanger reads, the 454 Newbler consensus shreds and the Illumina VELVET consensus shreds using the PGA assembler [23], to combine sequence data from all three platforms for a most contiguous assembly. The software Consed [24] was used in the computational finishing process as described previously [25]. The final draft assembly contained 34 contigs in 5 scaffolds.

Genome annotation

The assembled sequence was automatically annotated with the JGI prokaryotic annotation pipeline [26] with additional manual review using the IMG-ER platform [27]. Genes were predicted using Prodigal [28], ribosomal RNAs were detected using models built from SILVA [29] and tRNAs were predicted with tRNAScanSE [30]. The predicted CDs were translated and used to search the NCBI non-redundant database UniProt, TIGRFam, Pfam, KEGG, COG and InterPro databases. The final annotated genome is available from the IMG system [31]. We performed a CheckM analysis [32] and assessed that the genome is 95.8% complete.

Genome properties

The genome is 3,116,746 bp long with a GC content of 48.46%. A total of 3213 genes were found, of which 3163 coded for proteins and 50 genes encoded only RNA. From the total genes, 69.19% was assigned a putative function. The IMG taxon ID is 643,886,183. The different functional gene groups are summarized in Table 3. Furthermore, the number of genes assigned to functional COG categories is displayed in Table 4.

Table 3 Nucleotide content and gene count levels of the genome
Full size table
Table 4 Number of genes associated with general COG functional categories
Full size table

Insights from the genome sequence

Extended insights: Metabolic potential

Hydrogen metabolism requires a number of hydrogenase operons, including the hyd operon, and a Ni-Fe metallocenter assembly (hyp) [33]. The first part of the hydrogenase hyd operon is the small hydrogenase subunit hydA located at DealDRAFT_1217, the closest NCBI BLAST hit [12] of this protein is the hydA gene in Desulfotomaculum gibsoniae (Desgi_1397) with 70.4% similarity in a pair-wise alignment [34]. Directly adjacent to hydA, is the large subunit hydB (DealDRAFT_1218) in the D. alkaliphilus AHT1T genome. This subunit is most similar (75.9%) to the hydB subunit in Dehalobacter sp. UNSWDHB (UNSWDHB_1527) [12, 34]. DealDRAFT_1219 is a cytochrome B561 of 198 amino acids and could therefore be the interacting partner and gamma subunit hydC in the hyd operon. The 6-gene hyp operon hypABCDEF is responsible for the assemblage of the Ni-Fe uptake hydrogenases [35]. The last 5 proteins of the hyp operon are annotated in the D. alkaliphilus AHT1T genome (DealDRAFT_0838-DealDRAFT_0842) and follow the organization hypBFCDE, as has been seen before in Rhizobium [36]. The first gene in the operon (DealDRAFT_0843) is a hypothetical protein of 88 nucleotides length and is assigned to pfam01155 hypA, which is 42.6% identical to the hypA gene in Moorella thermoaceticum. Therefore, this hypothetical protein is most likely hypA in D. alkaliphilus AHT1T. Using hydrogen as electron donor, D. alkaliphilus AHT1T can grow autotrophically by fixing inorganic carbon through the Wood Ljungdahl pathway, the key genes are all present in the genome (Fig. 3a), including the acs gene cluster (Fig. 3b). Heterotrophic growth by D. alkaliphilus AHT1T can be achieved with glucose and fructose [10], the entire glycolysis pathway is present in the genome (Fig. 4). Carbohydrate metabolism in D. alkaliphilus AHT1T also includes oxidation of short chain organic acids; the tetrameric pyruvate oxidoreductase is present in the conformation porBADC (DealDRAFT_1244 – DealDRAFT_1247). Lactate dehydrogenases could not be found, although there is an L-lactate permease (DealDRAFT_0239), an L-lactate transport protein (DealDRAFT_1845) and a large and small subunit acetolactate synthase (DealDRAFT_2169 and 2170). For assimilation of acetate, strain AHT1T has an acetyl coenzyme A synthetase (DealDRAFT_1887).

Fig. 3
figure 3

a KEGG orthologs annotated in the gene pathway encoding Wood Ljungdahl inorganic carbon fixation in D. alkaliphilus strain AHT1T. b The acs gene cluster with locus tags. All locus tag numbers are indicated and preceded by DealDRAFT_

Full size image
Fig. 4
figure 4

KEGG orthologs annotated in the Embden-Meyerhof pathway of organic carbon assimilation in D. alkaliphilus strain AHT1T. The numbers of the locus tags of the genes catalyzing each reaction are indicated and must be preceded by DealDRAFT_

Full size image

D. alkaliphilus AHT1T might play a role in the reductive sulfur cycle in alkaline habitats since it grows as a thiosulfate and sulfur/polysulfide reducer or by sulfur disproportionation in laboratory cultures [10]. The genome sequence contains a thiosulfate sulfurtransferase (DealDRAFT_1917), which is located directly adjacent to another sulfur transferase (Rhodanese domain DealDRAFT_1918). Both alpha and beta subunits of the adenylylsulfate reductase apr operon were also found (DealDRAFT_1379, DealDRAFT_1380). The qmo electron transfer complex, which usually accompanies the apr operon [37], is not found. Key sulfur reduction genes such as sat (sulfate reduction), dsr (sulfite reduction) and psr (sulfur reduction) were also not found in this draft genome. As D. alkaliphilus AHT1T can reduce and disproportionate elemental sulfur/polysulfide in laboratory cultures [10, 13], the absence of these genes is surprising. It is conceivable however, that the sequencing quality of the permanent draft is insufficient to recover complete pathways. Indeed, CheckM analysis revealed that the genome was only 95.8% complete. Unfortunately, we can therefore not explain the key dissimilatory disproportionation mechanism from this genomic data. The genome also contains some assimilatory sulfate reduction genes, such as cysND (DealDRAFT_1193 and DealDRAFT_1192).

Extended insights: Haloalkaliphilic adaptations

In order to generate ATP, D. alkaliphilus AHT1T has an ntp gene operon encoding a vacuolar ATP synthase (V0V1-type) (DealDRAFT_1677 – DealDRAFT_1685) (Fig. 5a). This operon structure is conserved among the Clostridia (Fig. 5b). The ntp operon encodes the ATP synthase for ATP generation and follows the GILEXFABD organization in the Deinococcus-Thermus phylum [38]. In the Firmicutes , the gene organization is slightly different at GIKECFABD (Fig. 5a, b). In D. alkaliphilus AHT1T these genes are located from DealDRAFT_1685 (ntpG) to DealDRAFT_1677 (ntpD). The ntpD subunit within the operon is annotated as being of the V-type. In order to confirm that the ATP synthase is indeed V-type [39], we constructed a phylogenetic tree of the transmembrane c/K subunits of Firmicutes known specifically to be V- or F-type [40] and NCBI annotation] and aligned the D. alkaliphilus AHT1T ntpC sequence (DealDRAFT_1683) with these other sequences (Fig. 6a) [41]. As seen before, there was a clear separation between V-type and F-type ATP synthase, where the AHT1T sequence clustered together with the V-type ATP synthase. In addition, the sequences are tentatively clustered into separate H+ or Na+ coupled ATPase branches. The AHT1T sequence was positioned within a Na+ coupled V-type ATP synthase group, indicating that this organism’s ATP synthase is coupled specifically to Na+ translocation across the membrane. In order to explore this further, we looked at specific Na+ binding residues and ligands on the transmembrane c/K subunit [40], and created a Weblogo for the Na+ specific Firmicutes V-type ATP synthase (Fig. 6b) [42, 43]. When we aligned the ntpC sequence of D. alkaliphilus AHT1T we found that it contains all the conserved five amino acids (Ser26, Leu57, Thr60, Gln61 and Tyr64) specific for Na+ translocation [40] (Fig. 6c). Thus, the D. alkaliphilus AHT1T genome contains a Na+ coupled V-type ATP synthase.

Fig. 5
figure 5

a The ntp Vacuole-type ATP synthase operon structure. b 93 ntpD homologs (DealDRAFT_1677) within the genus Clostridia were aligned in Clustal Omega [34] and an unrooted neighbour-joining tree was generated in MEGA-6 [49]. From this tree, we picked the branch that contained the D. alkaliphilus AHT1T ntpD sequence and computed a new neighbourjoining tree with gene DCR20291_1119 as an outgroup. The scale bar indicates a 0.5% sequence difference and conserved gene neighbourhoods of those genes were investigated using MGcV [50]. Large dots at the tree nodes indicate a bootstrap value of >85 (1000 replicates)

Full size image
Fig. 6
figure 6

a Phylogeny of the F- vs. V-type ATPase within the Firmicutes. Numbers on the tree nodes indicate bootstrap values (1000 replicates). Scale bar indicates 0.2% sequence difference. b Weblogo of conserved region within the ntpC/K Firmicu subunit [42, 43]. c Weblogo of aligned D. alkaliphilus AHT1T subunit ntpC (DealDRAFT_1683) where conserved Na+ binding regions (in B and C) are indicated with black arrows

Full size image

In order to import protons to retain the intracellular pH, the genome contains the multi-subunit electrogenic sodium/proton antiporter mrp (DealDRAFT_2487–2497), that pumps protons into the cell and sodium out of the cell [44]. To retain osmotic balance, D. alkaliphilus AHT1T has numerous substrate binding regions and transporters for glycine betaine (e.g. DealDRAFT_2378, _2380 and DealDRAFT2842, _2844), leading to the conclusion that osmoprotectants are used to maintain cellular turgor pressure, instead of the salt-in strategy. Another necessity for alkaliphilic bacteria is to prevent proton leakage from cells, which they can achieve through structural membrane adaptations [1]. The genome contains the genes to synthesize the squalene precursors dimethylallyl diphosphate and isopentenylallyl diphosphate through the non-mevalonate pathway [45]. The accompanying locus tags within the KEGG non-mavalonate pathway (M00096) are dxs (DealDRAFT_0731), dxr/ispC (DealDRAFT_2409), ispD (DealDRAFT_2331), ispE (DealDRAFT_2584), ispF (DealDRAFT_2332), ispG (DealDRAFT_2411) and ispH (DealDRAFT_0659). However, we did not find genes similar to hpnCDE, which function in the formation of squalene from its precursors [46]. Thus, D. alkaliphilus AHT1T does not seem to have this membrane adaptation to haloalkaline environments, although it could also be due to the incompleteness of the genome. Nevertheless, it has been shown that Bacillus lentus C-125, also a Firmicute, survives in the haloalkaline environment by increased levels of acidic polymers in its cellular membrane resulting in a cell wall negative charge [47]. It is possible that D. alkaliphilus AHT1T supports a similar mechanism to survive the alkaline pH values of its environment.

Conclusions

In this manuscript we globally characterize the genome of D. alkaliphilus AHT1T, which was isolated from hypersaline soda lakes sediment in north-eastern Mongolia. Investigation of the genome of this anaerobic sulfidogen identified genes for the Wood Ljungdahl pathway (autotrophic growth, Fig. 3) and the Embden-Meyerhof pathway (heterotrophic growth Fig. 4). Thus the carbon metabolism of this microbe is fairly versatile. D. alkaliphilus AHT1T is capable of disproportionation in laboratory cultures, thus future genomic analyses with qPCR may provide insights into the disproportionation of sulfur compounds. D. alkaliphilus AHT1T is well adapted to the haloalkaline environment, we found genes for active energy generation with a sodium V-type ATP synthase (Fig. 6). In addition, transporters for the osmoprotectants glycine and betaine were found to maintain cellular homeostasis and protection from the saline external environment. Further research will extend our knowledge on the ecophysiology of haloalkaliphiles, their role in nutrient cycling in extreme environments and their adaptations to this polyextreme environment. Moreover, insight in the genome sequence and subsequent transcriptomic or proteomic analysis will be helpful to infer the potential role of D. alkaliphilus AHT1T in the biotechnological removal of sulfur compounds from wastewater and gas streams.