The Genome of Varunaivibrio sulfuroxidans Strain TC8^T, a Metabolically Versatile Alphaproteobacterium from the Tor Caldara Gas Vents in the Tyrrhenian Sea

1. Introduction

Hydrothermal and gas vents are manifestations of volcanism on earth. Both environments are characterized by venting of gas and/or hydrothermal fluids, enriched in reduced, inorganic chemical species into the oxygenated water column. This creates a redox disequilibrium, which is then harnessed by lithotrophic microrganisms that convert chemical energy into ATP. Tor Caldara is a shallow-water gas vent located in the Mediterranean Sea, near the town of Anzio, Italy. The composition of gases released by the Tor Caldara vents is mostly CO₂ (avg. 77 mol%) and H₂S (avg. 23 mol%) [1]. The Tor Caldara gas vents are generated by degassing of CO₂ of deep magmatic/mantle origin due to the alkali-potassic volcanoes present in the region, in particular the Colli Albani volcanic complex [2]. Since the gas venting at Tor Caldara is a peripheral manifestation of the quiescent Colli Albani volcano, whose activity began 600 ka ago [2], the system is thought to be very stable over time in comparison to the more dynamic venting associated with active hydrothermal vents.

The continuous venting of CO₂ and H₂S gas from the seafloor at Tor Caldara supports substrate-attached chemosynthetic microbial biofilms mostly composed of members of the Camplylobacterota (aka Epsilonproteobacteria) and Gammaproteobacteria [1,3], in line with previous studies of coastal and deep-sea marine hydrothermal environments e.g., [4,5,6,7,8,9] as well as more diverse microbial communities associated with the vent-associated sediments [10]. Alphaproteobacteria have also been observed to contribute abundantly to shallow-water vent microbial communities [11,12,13,14]. Consistent with these observations, Alphaproteobacteria were relatively abundant at Tor Caldara (4.5 to 15.1 % of the vent-associated sediment microbial community and 6 to 10 % of the substrate-attached biofilms) [1,10], and a novel genus of sulfur-oxidizing, chemolithoautotrophic alphaproteobacterium, Varunaivibrio sulfuroxidans strain TC8^T, was isolated and characterized from the sediments associated with the gas vents of this site [15]. Currently, V. sulfuroxidans belongs to the family Thalassospiraceae (although a recent study proposed its reassignment to the new family Magnetovibrionaceae) [16], which includes four genera of marine bacteria: Magnetospira, Magnetovibrio, Thalassospira and Varunaivibrio [17]. In particular, Magnetospira thiophila, Magnetovibrio blakemorei and Varunaivibrio sulfuroxidans are facultative chemilthoautotrophs that conserve energy by anaerobic (nitrate) and/or microaerobic respiration, with reduced sulfur species as electron donors [15,18,19].

In this study we obtained the genome sequence of V. sulfuroxidans strain TC8 ^T and we analyzed it in light of its known physiological and metabolic traits, we reconstructed its central metabolism, its heavy metal detoxification and DNA uptake mechanism. This is one of the few chemolithoautotrophic alphaproteobacterial genomes that have been sequenced from a shallow-water hydrothermal vent, another being a Rhodovulum sp. from the coastal hydrothermal system of Kueishantao Island, Taiwan [20].

2. Materials and Methods

2.1. Genome Project History

The genome of V. sulfuroxidans strain TC8^T was sequenced in August 2016 as part of a genome sequencing effort of newly characterized bacterial strains isolated at the Deep-Sea Microbiology Lab at Rutgers University (http://marine.rutgers.edu/deep-seamicrobiology/). The draft genome of V. sulfuroxidans was sequenced at the MicrobesNG facility (http://microbesng.uk/), and it was closed in house using the Oxford Nanopore MinION platform [21].

2.2. Growth conditions and genomic DNA preparation

Varunaivibrio sulfuroxidans strain TC8^T was grown in modified medium 1011 [22] supplemented with with 0.2% (w/v) KNO₃ under a gas phase of N₂/CO₂ (80:20 v/v; 200 kPA) according to Patwardhan and Vetriani, 2016 [15]. The genomic DNA was extracted from 0.5 – 1 g of cell biomass following a phenol:chloroform extraction protocol. In short, the bacterial pellet was added to 850 ul extraction buffer (50 mM Tris-HCl, 20mM EDTA, 100mM NaCl; pH 8.0) supplemented with100 ul of lysozyme (100 mg/ml) and incubated at 37°C for 30 min. This mix was then supplemented with 5 ul of proteinase K (20mg/ml) and incubated at 37°C for 30 min followed by the addition of 50 ul SDS (20%) and further incubated at 65°C for 1 hr. The genomic DNA was extracted by performing two consecutive phenol:chloroform:isoamylalcohol (25:24:1) extractions. The supernatant was precipitated in 3 M sodium-acetate and isopropanol, washed twice with 70% ice cold ethanol and re-suspended in ultra-pure water.

2.3. Genome sequencing and assembly

A hybrid approach involving short Illumina reads and long Nanopore MinIon reads was used to sequence the genome of V. sulfuroxidans. Genomic DNA was sequenced using a paired end 2x250 bp MiSeq library which generated a total of 1,004,710 quality checked short reads. A total of 40,150 long reads were also obtained on the Naonopre MinIon platform using the SQK-RAD004 library prep kit. Obtained short and long reads were assembled successively using Unicycler [23] and Geneious [24]. Assembly yielded 1 complete contig of a total length of 3,066,297 bp, with a 138x sequencing coverage. The genome sequencing project of V. sulfuroxidans strain TC8^T is available under accession number: CP119676.

2.4. Genome annotation

The genome assembly was annotated using the Joint Genome Institute IMG system [25] and the RAST (Rapid Annotation using Subsystem Technology) server. The predicted proteins of Varunaivibrio sulfuroxidans TC8^T involved in key metabolic pathways were searched using the KEGG database [26]. CRISPERs and tandem repeats were identified using CRISPRFinder [27]. Prophages in the genome were searched using Phaster [28]. Genomic islands were identified using IslandViewer [29].

2.5. Phylogenetic, average nucleotide identity and digital DNA-DNA hybridization analyses

Phylogenetic analyses were carried out as previously described [7,30]. In brief, 16S rRNA gene sequences were aligned with ClustalO [31] and the alignment was manually refined using SEAVIEW [32]. The maximum likelihood phylogeny tree was inferred from the alignment using PhyML [33], with the Jukes and Cantor substitution model [34] and 500 bootstrap resamplings. The aminoacid sequences of NifH, ComEC and CpaC from strain TC-8^T and closest homologues were retrieved from GenBank, aligned in SEAVIEW [32] and the maximum likelihood phylogeny trees were inferred from the alignment using PhyML [33], with the Li and Gasquel (LG) aminoacid replacement matrix [35] and 500 bootstrap resamplings. The average nucleotide identity (ANI) between strain TC8^T and its closest relative, Magnetovibrio blakemorei, was calculated using the EZBioCloud tool [36]. The digital DNA-DNA hybridization (dDDH) value between strain TC8^T and M. blakemorei was obtained using the genome-to-genome distance calculator (GGDC) [37].

3. Results and Discussion

3.1. Overview of Varunaivibrio sulfuroxidans strain TC8^T

V. sulfuroxidans strain TC8^T is a mesophilic, facultatively anaerobic, facultatively chemolithoautotrophic bacterium isolated from a sulfidic shallow-water marine gas vent located at Tor Caldara, in the Tyrrhenian Sea, Italy [15]. Cells are Gram-stain-negative curved rods with one or more polar flagella. Cells were approximately 1–1.5 µm in length and 0.6 µm in width. Strain TC8^T grows optimally at 30 °C, with 15-20 g NaCl l/1iter and optimum pH between 6.0 and 7.0. The generation time under optimal conditions is 8 h. Strain TC8^T is a facultative chemolithoautotroph also capable of using organic substrates (tryptone, peptone, Casamino acids, pyruvate and glycerol). Chemolithoautotrophic growth occurs with sulfur and thiosulfate as the electron donors, CO₂ as the carbon source, and nitrate or oxygen (5 %, v/v) as the electron acceptors. Strain TC8 ^T is able to grow in the absence of a source of fixed nitrogen (nitrate or ammonium) under a headspace of N₂/CO₂/O₂ (53:45:2 v/v), which implies that it is able to fix gaseous nitrogen.

Phylogenetic analysis of the 16S rRNA gene sequence of strain TC8^T placed this organism in a lineage within the family Thalassospiraceae, with Magnetovibrio blakemorei strain MV1^T (91.25% sequence identity) as its closest cultured relative (Figure 1). When uncultured microorganisms known solely by their 16S rRNA gene are included in the analysis, the closest sequence to V. sulfuroxidans is clone 7M24_051 (98.89% sequence identity; Figure 1), obtained from a iron-rich, inactive sulfide associated with the deep-sea hydrothermal vent system of the East Pacific Rise at 9° North [38].

3.2. Genome Structure

The genome of V. sulfuroxidans strain TC8^T contains 3,066,297 bp, with coding regions taking 88.1 % of the genome and a G+C content of 54.4 % (Table 1). The most abundant gene categories assigned to the clusters of orthologous genes (COGs) database code for amino acid transport and metabolism (10.3 %), energy conversion (8.9 %), translation and ribosomal structure (7.6 %) and inorganic ion transport and metabolism (7.5 %; Table 2). 27.1 % of the genes of strain TC8^T were not in the COG database, while 4.7 % were classified as unknown function (Table 2).

The genome structure of V. sulfuroxidans strain TC8^T, including the gene distribution, predicted genomic islands, prophage and genes involved in membrane transport, carbon and energy metabolism is summarized in Figure 2. Nine regions of the genome of strain TC8T were predicted as possible genomic islands (Figure 2A, line 2), while one region was found to be a complete bacteriophage (Figure 2B). The intact prophage found in the genome of TC8T has a size of 33.4Kb, with a total of 48 proteins and a GC content of 65.53% (Figure 2B). It consists of 22 different phage-like proteins, with the majority (31%) of proteins similar to prophage vB_RhkS_P1, which was induced in Rhodovulum sp. strain P5, an alphaproteobacterium isolated from a shallow-water hydrothermal vent off Kueishantao Islands [20,39].

The genome of strain TC8^T encodes one short CRISPR array, which is 431 bp in length and has 6 spacers. The spacers did not show sufficient alignment to any sequences in the Genbank Phage or RefSeq plasmid databases.

The average nucleotide identity (ANI) between the genomes of V. sulfuroxidans and its closest cultured relative, M. blakemorei, is 68.86%, while the estimated digital DNA-DNA hybridization (dDDH) between the two genomes is 18.7%, confirming their assignment to different genera.

3.3. Carbon Metabolism and Microaerobic Respiration

The genome of strain TC8^T encodes for all the enzymes necessary for carbon fixation using the Calvin-Benson-Bassham (CBB) cycle, including the key enzymes ribulose 1,5- bisphosphate carboxylase (RubisCO) and phosphoribulokinase (Figs. 2A and 3 and Table S1). The RubisCO enzyme catalyzes the carboxylation of ribulose 1,5- bisphosphate, which results in two molecules of 3-phosphoglycerate. Presently, there are three forms of RubisCO known that catalyze the CO₂ fixation reaction, with form I being the most dominantly occurring [40]. From II, encoded by the cbbM gene, was first isolated from the alphaproteobacterium, Rhodospirillum rubrum. Similarly to other members of the family Thalassospiraceae, the genome of V. sulfuroxidans TC8^T encodes a form II RubisCO closely related to the enzyme from M. blakemorei (sequence identity 88.45%). The carbon concentrating mechanisms of strain TC8^T includes a type γ carbonic anhydrase, which catalyzes the interconversion between the dissociated ions of carbonic acid and carbon dioxide and water (Figure 3 and Table S1). The amino acid sequence of the carbonic anhydrase of V. sulfuroxidans is 69.4 % identical to that of M. blakemorei.

The genome of strain TC8^T also encodes for all enzymes involved in the breakdown of organic carbon via the glycolysis pathway as well as for all the enzymes involved in the tricarboxylic acid (TCA) cycle, which provides reducing power for oxidative phosphorylation in heterotrophs (Figure 2 and Table S1). Thus, the genome of V. sulfuroxidans strain TC8^T encodes for a heterotrophic as well as autotrophic metabolism, in line with its experimentally determined metabolic characteristics [15].

The genome of strain TC8^T encodes the entire microaerobic respiratory chain, including the high-affinity cbb3 type cytochrome oxidase and cytochrome c oxidase, consistent with the experimental evidence for microaerophilic growth (Figure 3) [15].

3.4. Sulfur Metabolism

V. sulfuroxidans TC8^T was originally described as sulfur-oxidizing chemolithoautotroph [15]. Accordingly, the genome encodes the SoxABXYZ pathway that is responsible for thiosulfate oxidation (Figure 3 and Table S1). Strain TC8^T lacks the SoxCD component which, in Paracoccus pantotrophus, is designated as a sulfur dehydrogenase [41]. Without SoxCD_, thiosulfate is likely oxidized to elemental sulfur and not all the way to sulfate, as demonstrated experimentally in Allochromatium vinosum [42]. The aminoacid sequence of V. sulfuroxidans TC8^T SoxA is 65.83% identical to that of the M. blakemorei enzyme. Besides the Sox gene cluster, the genome of strain TC8^T encodes the sulfide:quinone oxidoreductases (SQR, 79.6 % identity to the M. blakemorei enzyme) and a putative flavocytochrome c-sulfide dehydrogenase (Fcc; Figure 3 and Table S1). The presence of these enzymes suggests that strain TC8^T has the potential to oxidize sulfide to elemental sulfur [43]. Current knowledge indicates that, in chemolithotrophic sulfide-oxidizing bacteria, SQR-mediated sulfide oxidation is more energetically favorable than sulfide oxidation by Fcc, and that overall SQR is a more efficient enzyme [44]. However, due to its high affinity for sulfide, Fcc is hypothesized to be selectively expressed at low sulfide concentrations [3,45]. The dsr gene cluster, encoding several enzymes involved in oxidation of stored sulfur, is also present in the genome of strain TC8^T (Figure 3 and Table S1). According to current understanding, sulfur is brought into the cytoplasm in a persulfidic form (e.g., glutathione persulfide) [43]. DsrL releases the sulfide from the persulfide, which then passes on to the reversely operating sulfite reductase DsrAB via DsrC and DsrEFH [46]. The DsrAB oxidizes sulfide to sulfite. Next, DsrAB interacts with the trans-membrane DsrMKJOP complex, which passes electrons in to the respiratory electron chain. Genes encoding for all the aforementioned proteins are present in the genome of strain TC8^T. The sulfite formed is likely further oxidized to sulfate via AprAB and SAT enzymes, both encoded by the genome of strain TC8^T. The presence of several sulfur oxidation pathways in the genome of strain TC8^T indicates its metabolic flexibility and its ability to oxidize a variety of reduced sulfur compounds to conserve energy, which likely provide this bacterium with an advantage to thrive in its sulfidic habitat.

3.5. Nitrogen Metabolism

Dissimilatory nitrate reduction to ammonia (DNRA) and denitrification are two key pathways involved in nitrogen metabolism that occur commonly in vent organisms [47,48]. V. sulfuroxidans conserves energy via the oxidation of reduced sulfur species coupled to the reduction of nitrate to dinitrogen gas [15]. In line with this experimental evidence, the genome V. sulfuroxidans encodes all four proteins involved in the complete denitrification pathway (Figure 3 and Table S1). The first step of this pathway, i.e., the reduction of nitrate to nitrite, is catalyzed by a nitrate reductase, either the membrane bound NarG or the periplasmic NapAB. All four genes encoding the membrane bound nitrate reductase complex, namely narG, narH,narJ and narI are present in the genome of strain TC8^T. The napA and napB genes encoding the periplasmic nitrate reductase NapAB are also present [47]. The aminoacid sequence of the periplasmic nitrate reductase (NapA) of V. sulfuroxidans is 80.26% identical to that of M. blakemorei. The genome has the entire nap gene cluster consisting of the quinol dehydrogenase NapC, the maturation chaperone NapD, the membrane bound dehydrogenases NapG and NapH and subunit NapF, whose role remains unclear (Figure 3). Several microorganisms, including Escherichia coli, encode both the membrane bound (Nar) and periplasmic (Nap) nitrate reductases. Potter et al. (1999) experimentally demonstrated that, in nitrate limited conditions, the presence of only the high-affinity Nap enzyme afforded E. coli a selective advantage over mutant strains that only encoded the Nar enzyme [49]. We hypothesize that, in the dynamic habitat of the Tor Caldara gas vents, the expression of the two nitrate reductases in V. sulfuroxidans is modulated by the availability of nitrate. The reduction of nitrite to nitric oxide (NO) is carried out by cytochrome cd1 protein NirS, encoded by the nirS gene. NorBC further reduces two molecules of NO to nitrous oxide (N₂O). NosZ is involved in the reduction of N₂O to N₂, albeit is not present in all organisms [50]. However, NosZ is present in the genome of strain TC8^T, which is consistent with its ability to generate dinitrogen gas during growth under nitrate reducing conditions [15]. Along with dissimilatory nitrate reduction, the genome also shows potential for assimilatory nitrate reduction, as the nasA and nasB genes that encode for cytoplasmic nitrate reductase NasAB are present (Figure 3). Genes for the assimilatory nitrite reductases NIT-6 and NirA were absent. However, the genome encoded for the NirBD protein which has also been shown to be involved in assimilatory nitrite reduction [51,52].

Nitrogen fixation, namely the reduction of N₂ gas to two molecules of ammonia, is an important part of the marine nitrogen cycle, since it provides a source of newly fixed nitrogen to the ecosystem. The nitrogenase complex catalysing this reaction is made up of two proteins: dinitrogenase reductase and dinitrogenase encoded by three genes, nifH, nifD and nifK [53]. All three genes are found in the genome of V. sulfuroxidans, consistent with its ability to grow in the absence of fixed nitrogen (Figure 3 and Table S1) [15]. The genome of strain TC8^T also encodes a nitrogenase-associated Fe-S ferredoxin (fdx gene; Table S1), which is 44.7 % identical to that of Rhodobacter capsulatus. We hypothesize that the fdx-encoded ferredoxin serves as electron donor to nitrogenase in V. sulfuroxidans, as experimentally demonstrated for R. capsulatus [54]. Phylogenetic analysis of the NifH placed the enzyme of V. sulfuroxidans in a cluster that includes several members of marine Alphaproteobacteria, including M. blakemorei, Cohaesibacter marisflavi, C. intestine and Rhodobium orientis, as well as Martelella endophytica, which was isolated from the rose rhizosphere (Figure 4). The NifH aminoacid sequences of V. sulfuroxidans and M. blakemorei are 88.8% identical. The genome of V. sulfuroxidans encodes several enzymatic complexes that catalyze the oxidoreduction of nitrogen (nitrate reduction, assimilation and nitrogen fixation) and sulfur species (thiosulfate, sulfur and possibly sulfide oxidation), effectively linking the nitrogen and sulfur cycles.

3.6. Heavy Metal Detoxification

Hydrothermal vent environments are known to have elevated concentrations of toxic heavy metals; hence, the resident microorganisms encode genes involved with heavy metal detoxification [55,56,57,58]. The genome of strain TC8^T contains both the ars and the mer operons involved in detoxification of arsenate and mercury, respectively (Figure 3 and Table S1). Arsenic in the form of arsenite or arsenate is toxic for most microorganisms, and resistance is associated with a three gene or five gene ars operon. The five gene ars operon in strain TC8^T consists of arsenate reductase ArsC, membrane efflux protein ArsB, two trans-acting repressors ArsR and ArsD, as well as another efflux protein ArsB that infers higher resistance to arsenic [59,60]. The mer operon of strain TC8^T encodes for the periplasmic mercury-binding protein MerP, the trans-membrane transport protein MerT and the mercury reductase MerA [61,62]. V. sulfuroxidans did not grow in the presence of arsenate as the electron acceptor [15]. The genome does not encode for a selenate reductase; however, it has been shown that bacterial nitrate reductases, which the genome encodes for, are also capable of reducing selenate [63,64]. The genome also encodes for the DedA protein, which has shown to be involved selenite resistance [65]. The resistance of V. sulfuroxidans to mercury and selenite was not tested.

3.7. DNA uptake system

Knowledge of extracellular DNA uptake in Alphaproteobacteria is limited; however, recent studies suggest the involvement of a type IVc pilus known as the Tad pilus system [66]. In the naturally competent plant pathogen, Liberibacter crescens strain BT1, the tad pilus aids in survival through surface attachment and DNA uptake for nucleotide metabolism [67,68]. L. crescens, a member of the family Rhizobiaceae within the Alphaproteobacteria, was experimentally proven to be naturally competent, and the genes involved in this system were identified and found to be conserved among Liberibacter spp. [66,68].

Therefore, we used L. crescens strain BT1 as a reference organism to identify the putative DNA uptake system of V. sulfuroxidans. We discovered that the genome of V. sulfuroxidans encodes genes homologous to the tad genes organized in a cluster similar in structure to that of M. blakemorei and L. crescens (Figure 5). The tad pilus system consists of a biogenesis ATPase (tadZ/cpaE), motor ATPase (tadA/cpaF), prepilin peptidase (tadV/cpaA), inner membrane staging complex (tadB/cpaG and tadC/cpaH), secretin (rcpA/cpaC), periplasmatic subunits (tadG/cpaB and rcpB/cpaD), inner membrane anchor (rcpC/tadG), and pilotin (tadD/cpaO) [66]. V. sulfuroxidans have homologs to genes in this system except for pilotin (Figs. 2). The genome of strain TC8^T also encodes the DNA translocation machinery homologous to L. crescens strain BT1, as well as core genes involved in DNA uptake homologous to those encoded by L. crescens. Phylogenetic tree reconstruction of CpaC, the enzyme secretin of the type IVc secretion system, revealed that the enzyme of V. sulfuroxidans was most closely related to that of M. blakemorei (44.79% aminoacid sequence identity; Figure 6). The CpaC from L. crescens was placed in a cluster with homologous enzymes from other members of the family Rhizobiaceae (33.48% aminoacid sequence identity with the CpaC from V. sulfuroxidans).

ComEC is also necessary for natural competence and DNA internalization in L. crescens, and a homolog was found in strain TC8^T [68]. Phylogenetic analysis showed that the competence protein ComEC of V. sulfuroxidans was related to the homologous protein from M. blakemorei (45.47% aminoacid identity; Figure 7), while ComEC from L. crescens clustered with a group of proteins from various strains of Rhizobium and was 32.3% identical to that of V. sulfuroxidans.

Based on these results, we hypotesize that V. sulfuroxidians and M. blakemorei are naturally competent via the Tad pilus/Com system. However, the use of the Tad pilus system might differ between L. crescens and V. sulfuroxidans in the fact that former depends on this system for the essential uptake of nucleotides while strain TC8^T can synthesize its own nucleotides.

4. Conclusions

In this study we sequenced and analyzed the genome of V. sulfuroxidans strain TC8^T, whose physiology and metabolism were previously experimentally characterized. The genome of strain TC8^T revealed that this facultative chemolithoautotroph fixes carbon via the CBB cycle, oxidize reduced sulfur species via the Sox, SQR and the oxidative Dsr pathways, encodes the nitrogenase gene cluster for nitrogen fixation and reduces nitrate via the periplasmic and/or cytoplasmic nitrate reductase complexes, Nap and Nar. Based on homology with the experimentally verified DNA uptake system of L. crescens, we further identified the Tad-mediated secretion system and competence gene cluster in the genome of V. sulfuroxidans and its closest relative, M. blakemorei. These findings suggest that both V. sulfuroxidans and M. blakemorei are naturally competent and can take up exogenous DNA.

Linking the genome sequence of V. sulfuroxidans with its observed metabolism establishes a direct association between genotype and phenotype, and enables genome-based metabolic inferences from environmental, sequence-based surveys, e.g., metagenome-assembled genomes (MAGs) of closely related strains (e.g., close relatives of V. sulfuroxidans identified by sequencing in hydrothermal and sedimentary habitats).

Overall, the genome of V. sulfuroxidans, along with its experimentally validated phenotype, reveals the metabolic versatility of this bacterium and corroborates its ability to thrive in the highly dynamic habitat of sulfidic gas vents.