CO-dependent hydrogen production by the facultative anaerobe Parageobacillus thermoglucosidasius

Background The overreliance on dwindling fossil fuel reserves and the negative climatic effects of using such fuels are driving the development of new clean energy sources. One such alternative source is hydrogen (H2), which can be generated from renewable sources. Parageobacillus thermoglucosidasius is a facultative anaerobic thermophilic bacterium which is frequently isolated from high temperature environments including hot springs and compost. Results Comparative genomics performed in the present study showed that P. thermoglucosidasius encodes two evolutionary distinct H2-uptake [Ni-Fe]-hydrogenases and one H2-evolving hydrogenases. In addition, genes encoding an anaerobic CO dehydrogenase (CODH) are co-localized with genes encoding a putative H2-evolving hydrogenase. The co-localized of CODH and uptake hydrogenase form an enzyme complex that might potentially be involved in catalyzing the water-gas shift reaction (CO + H2O → CO2 + H2) in P. thermoglucosidasius. Cultivation of P. thermoglucosidasius DSM 2542T with an initial gas atmosphere of 50% CO and 50% air showed it to be capable of growth at elevated CO concentrations (50%). Furthermore, GC analyses showed that it was capable of producing hydrogen at an equimolar conversion with a final yield of 1.08 H2/CO. Conclusions This study highlights the potential of the facultative anaerobic P. thermoglucosidasius DSM 2542T for developing new strategies for the biohydrogen production. Electronic supplementary material The online version of this article (10.1186/s12934-018-0954-3) contains supplementary material, which is available to authorized users.


Background
In the next 30 years, the global energy demand will expand by ca. 30% and the vast majority (ca. 85%) of the energy resources required to offset the rising demand will come from non-renewable sources such as natural gas and crude oil [1]. This will result in increased pressure on the dwindling fossil fuel reserves and in greater emission of greenhouse gases into the Earth's atmosphere. There is thus an urgent need for further development and implementation of clean and renewable alternative energy sources [2,3].
Hydrogen (H 2 ) has recently become prominent as a very attractive clean and sustainable energy source, especially when generated via 'eco-friendly' strategies. In comparison to other fuels, H 2 has the highest energy content (141.9 MJ/kg higher heating value) [2]. Additionally, its complete combustion with pure oxygen produces only water (2 H 2 + O 2 → 2 H 2 O) as a by-product. The majority of current industrial H 2 production strategies, such as coal gasification, steam reformation and partial oxidation of oil, are unsustainable, harmful to the environment, energy intensive and expensive [4,5]. As such, over the past few years, the production of H 2 via microbial catalysis has drawn increasing interest. Several different strategies to produce biohydrogen, such as photofermentation of organic substances

Open Access
Microbial Cell Factories *Correspondence: teresa.mohr@kit.edu 4 Section II: Technical Biology, Institute of Process Engineering in Life Science, Karlsruhe Institut für Technologie (KIT), Kaiserstrasse 12,76131 Karlsruhe, Germany Full list of author information is available at the end of the article by photosynthetic bacteria, bio-photolysis of water by algae and dark fermentation of organic substances by anaerobic microorganism, have been explored [6]. These strategies have the advantage of lower energy expenditure, lower cost and higher yields than the industrial methods [7]. Another advantage is the potential to use cheap feedstocks, such as lignocellulosic waste biomass, which can be converted into a gas mixture termed 'synthesis gas' . This gas consists primarily of carbon monoxide (CO), carbon dioxide (CO 2 ) and H 2 [8]. In a further step, the CO can react with water to generate H 2 via a biologically-or chemically-mediated water-gas shift (WGS) reaction: CO + H 2 O → CO 2 + H 2 . During the biologically mediated reaction, a carbon monoxide dehydrogenase (CODH) oxidizes CO and electrons are released. Subsequently, a coupled hydrogenase reduces the released electrons to molecular hydrogen [9]. Several mesophilic, anaerobic prokaryotic taxa, including Rhodospirillum rubrum and Rhodopseudomonas palustris, are known for the ability to perform the WGS reaction [10].
It has been observed that higher yields of H 2 can be obtained in higher temperature fermentations [11]. Thus, there has been increasing interest in the use of thermophilic anaerobic bacteria, such as Carboxydothermus hydrogenoformans and Thermosinus carboxydivorans [12,13], as well thermophilic archaea (e.g. Thermococcus onnurineus) [7].
An industrial process utilizing CO-oxidizing bacteria for biohydrogen production has not yet been realized, although many CO-using hydrogenogenic species have been isolated. This may largely be attributed to the sensitivity of both the hydrogenase and CODH enzymes to oxygen [6,14]. For example, the hydrogenase of Thermotoga maritima lost 80% of its activity after flushing with air for 10 s [15]. Removal of O 2 from industrial waste gases or from bioreactors is prohibitively expensive, making the use of strictly anaerobic CO-oxidizing hydrogenogens unfeasible [16]. Here, we have analysed the hydrogenogenic capacity of the facultative anaerobe Parageobacillus thermoglucosidasius. Comparative genomics revealed the presence of three distinct hydrogenases, two uptake hydrogenases as well as one H 2 -evolving hydrogenase, which is linked to an anaerobic CODH. Evolutionary analysis showed that this combination of hydrogenases is unique to P. thermoglucosidasius and suggests that H 2 plays a pivotal in the bioenergetics of this organism. Furthermore, fermentations and downstream GC analysis showed that this facultative anaerobe is capable of utilizing CO in the WGS reaction to generate an equimolar amount of H 2 once most of the oxygen in the medium has been exhausted.

Microorganisms
The production of H 2 by P. thermoglucosidasius when grown in the presence of CO was tested using P. thermoglucosidasius DSM 2542 T . Two related strains, Geobacillus thermodenitrificans DSM 465 T and P. toebii DSM 14590 T , which lack orthologues of the three hydrogenase loci as well as the CODH locus, were included as controls. All strains were obtained from the DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Braunschweig, Germany).

Culture conditions and media
Pre-cultures and cultures were grown aerobically in mLB (modified Luria-Bertani) medium containing tryptone (1% w/v), yeast extract (0.5% w/v), NaCl (0.5% w/v), 1.25 ml/l NaOH (10% w/v), and 1 ml/l of each of the filter-sterilized stock solutions: 1.05 M nitrilotriacetic acid, 0.59 M MgSO 4 ·7H 2 O, 0.91 M CaCl 2 ·2H 2 O and 0.04 M FeSO 4 ·7H 2 O. The first pre-culture was inoculated from glycerol stock (20 µl in 20 ml mLB) and cultivated for 24 h at 60 °C and rotation at 120 rpm in an Infors Thermotron (Infors AG, Bottmingen, Switzerland). A second pre-culture was inoculated from the first one to an OD 600 of 0.1 and incubated as above for 12 h. For the experiments, 250 ml serum bottles were prepared with 50 ml mLB and a gas phase of 50% CO and 50% atmospheric air at 1 bar atmospheric pressure, which were inoculated with 1 ml from the second preculture. The experiments were conducted in quadruplicate for a total duration of 84 h.

Analytical methods
Samples were taken at different time points during the experimental procedure. Before and after the sampling the pressure was measured using a manometer (GDH 14 AN, Greisinger electronic, Regenstauf, Germany). To monitor the growth of the cultures, 1 ml of the culture was aspirated through the stopper and absorbance was measured at OD 600 using an Ultrospec 1100 pro spectrophotometer (Amersham Biosciences, USA). The determination of the gas compositions at different time points was conducted using a 3000 Micro GC gas analyzer (Inficon, Bad Ragaz, Switzerland) with the columns Molsieve and PLOT Q. A total of 3 ml was sampled from the head space and injected into the GC. A constant temperature of 80 °C was maintained during the total analysis time of 180 s. The gas compositions at the different sampling points were calculated using the formulas in Additional file 1.

Comparative genomic analyses
The large hydrogenase subunits were identified from the annotated genome of P. thermoglucosidasius DSM 2542 T (CP012712.1) by comparison against the Hydrogenase DataBase (HyDB) [17]. The full hydrogenase loci were identified by searching the genome up-and downstream of the large subunit gene, extracted and mapped against the genome using the CGView server [18]. The proteins encoded on the genome were compared by BlastP against the NCBI non-redundant (nr) protein database to identify orthologous loci. Full loci were extracted from the comparator genomes and all loci were structurally annotated using Genemark.hmm prokaryotic v.2 [19]. The resultant protein datasets were compared by local BlastP with Bioedit v 7.2.5 [20] to identify orthologues, where orthology was assumed for those proteins sharing > 30% amino acid identity over 70% sequence coverage.
A maximum likelihood (ML) phylogeny was constructed based on the amino acid sequences of three commonly used housekeeping markers: translation initiation factor IF-2 (InfB), DNA recombination and repair protein RecN RNA polymerase subunit B (RpoB). The proteins were individually aligned using M-Coffee [21], the alignments concatenated and poorly aligned regions were trimmed using Gblocks [22]. Finally, the trimmed alignment was used to generate a ML phylogeny using PhyML-SMS, using the optimal amino acid substitution model as predicted by the Smart Model Selection tool [23,24]. Similarly, ML phylogenies were constructed on the basis of the concatenated orthologous proteins encoded on the Pha, Phb, Phc and CODH loci.

Results
The genome of P. thermoglucosidasius encodes three distinct hydrogenases Analysis of the complete, annotated genome sequence of P. thermoglucosidasius DSM 2542 T showed the presence of three putative [Ni-Fe]-hydrogenase loci on the chromosome. Two of these hydrogenases are encoded on the forward strand, while the third is located on the reverse strand ( Fig. 1). Given the convoluted nomenclature of hydrogenase genes, we have termed these loci as Parageobacillus hydrogenase a, b and c, in accordance with their chromosomal locations), to distinguish between them.
The Pha locus (chromosomal position 2,456,963-2,469,832; 12.9 kb in size) comprises eleven protein coding sequences (NCBI Accession ALF10692-10702; PhaA-PhaK) ( Fig. 1; Additional file 2). Comparison of the amino acid sequence of the predicted catalytic subunit (ALF10727-PhaB) against HydDB [17] classifies the hydrogenase produced by this locus as a [Ni-Fe] group 1d uptake hydrogenase (E-value = 0.0). This unidirectional, membrane-bound, O 2 -tolerant hydrogenase is present in a broad range of obligately aerobic and facultatively anaerobic soil-borne, aquatic and host-associated taxa such as Ralstonia eutropha, Escherichia coli and Wolinella succinogenes [25,26]. The H 2 molecules consumed by group 1d hydrogenases are coupled to aerobic respiration (O 2 as electron acceptor) or to respiratory reduction of various anaerobic electron acceptors including NO 3− , SO 4 2− , fumarate and CO 2 . The P. thermoglucosidasius DSM 2542 T hydrogenase locus incorporates genes coding for both small (PhaA; ALF10692; 324 aa) and large (PhaB; ALF10693; 573 aa) catalytic hydrogenase subunits. The strain also encodes seven additional proteins involved in hydrogenase formation, maturation and incorporation of the Ni-Fe metallocenter, including a third hydrogenase subunit (PhaC) which is predicted to serve as cytochrome b orthologue and links the hydrogenase to the quinone pools of the respiratory chains ( Fig. 1; Additional file 2) [26]. The pha genes are flanked at the 5′ end by two genes coding for orthologues of the Twin-arginine translocation (Tat) pathway proteins TatA and TatC ( Fig. 1; Additional file 2). These have been shown to form part of the membrane targeting and translocation (Mtt) pathway which targets the fully folded hydrogenase heterodimer to the membrane [27].
The Phb locus (chromosomal position 2,488,614-2,503,714; 15.1 kb in size), located ~ 19 kb downstream of the Pha locus, comprises 16 protein coding sequences (NCBI Accession ALF10723-738; PhbA-PhbP) ( Fig. 1; Additional file 2). The predicted catalytic subunit (ALF10727-PhbE) compared against HydDB classifies the product of this locus as a [Ni-Fe] group 2a uptake hydrogenase (E-value = 0.0) [17]. Members of this group of uptake hydrogenases are widespread among aerobic soil bacteria and Cyanobacteria and play a role in recycling H 2 produced by nitrogenase activity and fermentative pathways [28,29]. The recycled H 2 is used in hydrogenotrophic respiration with O 2 serving as terminal electron acceptor, and thus group 2a hydrogenases are often O 2 -tolerant [26]. This locus encodes both large (PhbE; ALF10727; 544 aa) and small (PhbD; ALF10726; 317) [Ni-Fe] hydrogenase subunits and eight additional proteins with predicted roles in hydrogenase formation, maturation and incorporation of the Ni-Fe metal center in the large subunit ( Fig. 1; Additional file 2) [26]. Furthermore, this locus encodes six proteins whose role in hydrogenase biosynthesis and functioning remains unclear. These include a tetratricopeptide-repeat (PhbH) and NHL repeat (PhbK) containing protein, which also occur in the [Ni-Fe] group 2a hydrogenase loci of Nostoc punctiforme ATCC 29133 and Nostoc sp. PCC 7120, where they are co-transcribed with the hydrogenase genes and have been suggested to play a role in proteinprotein interactions and Fe-S cluster biogenesis (PhbJ) which may mediate electron transport to redox partners in downstream reactions [30].
BlastP and tBlastN analyses of the protein sequences encoded in the P. thermoglucosidasius DSM 2542 T hydrogenase loci showed that the Pha, Phb and Phc loci are universally present in eight other P. thermoglucosidasius strains for which genomes are available (Additional file 2: Table S1). These loci are highly syntenous and the encoded proteins share average amino acid identities of 99.73% ( 22 and 86.08% amino acid identity, respectively, and are predicted to play a role in the incorporation of nickel into the hydrogenase enzyme [32]. Limited orthology is observed between the hydrogenase catalytic subunits or other hydrogenase formation and maturation proteins, suggesting distinct evolutionary histories for the two uptake and one H 2 -evolving hydrogenases in P. thermoglucosidasius.

P. thermoglucosidasius contains a unique profile of hydrogenases with distinct evolutionary histories
The proteins encoded by the Pha, Phb and Phc loci were used in BlastP comparisons against the NCBI nonredundant (nr) protein database and HydDB (catalytic subunits) to identify orthologous loci in other bacterial taxa. This revealed that, aside from the α-proteobacteria Azospirillum halopraeferens DSM 3675 T and Rhodopseudomonas palustris BAL398, the combination of [Ni-Fe] group 1-2-4 hydrogenases appears to be unique to P. thermoglucosidasius (Fig. 2). In these two proteobacterial taxa the group 2a uptake hydrogenase is, however, replaced by a group 2b uptake hydrogenase. Group 2b uptake hydrogenases do not have a direct role in energy transduction but are flanked by a PAS domain protein which accepts the hydrogenase-liberated electrons, modulating the activity of a two-component regulator that upregulates the expression of other uptake hydrogenases, thereby serving as H 2 -sensing system [33,34]. The Pha uptake hydrogenase locus is relatively well conserved among members of the Firmicutes, including a number of taxa belonging to the Classes Bacilli, Clostridia and Negativicutes, as well as the phyla Proteobacteria and Bacteroidetes ( Fig. 2; Additional file 3). However, the more distantly related taxa retain little synteny with the Pha locus in P. thermoglucosidasius (Fig. 3a). Orthologues of the Pha are present in one other Parageobacillus spp., namely genomosp. NUB3621, with an average amino acid identity of 92.37% (13 proteins) with the DSM 2542 T Pha proteins. A phylogeny on the basis of nine conserved Pha proteins (PhaABCDGHIJK) showed a similar branching pattern (Fig. 3a) as observed for the phylogeny housekeeping protein (InfB-RecN-RpoB) phylogeny, suggesting that this is an ancestral locus that has been vertically maintained. This is supported by the low level of discrepancy in G+C content for the P. thermoglucosidasius strains, which are on average 0.87% above the genomic G+C content. Larger discrepancies are, however, evident among the Bacteroidetes, where G+C contents for the locus are on average 4.43% above that of the genome, and the absence of Pha loci in other Parageobacillus spp. including P. toebii (five genomes available) and P. caldoxylosilyticus (four genomes available) and Geobacillus spp. suggest a more complex evolutionary history for the group 1d hydrogenase.
Orthologous [Ni-Fe] group 2a uptake hydrogenase (Phb) loci are also common among the Firmicutes, but show a more restricted distribution within the family Bacillaceae, with only Aeribacillus pallidus 8m3 and Hydrogenibacillus schlegelii DSM 2000 T containing orthologues outside the genus Parageobacillus. Highly conserved and syntenous loci are, however, present in three non-thermoglucosidasius strains: Parageobacillus sp. NUB3621, Parageobacillus sp. W-2 and P. toebii DSM 18751 ( Fig. 3b; Additional file 3). Orthologous loci are present across a much wider range of phyla than the Pha locus, including members of the Chloroflexi, Gemmatimonadetes, Actinobacteria, Proteobacteria, Nitrospirae and Deinococcus-Thermus (Fig. 2). The latter is of interest as Thermus thermophilus SG0.5JP17-16 clusters with the Firmicutes in a phylogeny of ten conserved proteins (PhbBCDEFHJLMN-72.76% average amino acid identity with P. thermoglucosidasius DSM 2542 T ) (Fig. 3b), but is phylogenetically disparate from the Firmicutes. The T. thermophilus locus is present on the plasmid pTHTHE1601 (NC_017273) suggesting that this locus forms part of the mobilome. Furthermore, the G+C content of the Phb locus differs by an average of 4.55% from the average genomic G+C among the eight compared P. thermoglucosidasius strains, suggesting recent horizontal acquisition of this locus.    locus shows the most restricted distribution of the three loci among the Firmicutes, with orthologous loci only present in the eight compared P. thermoglucosidasius strains and members of the clostridial family Thermoanaerobacteraceae (Fig. 2). Further, Phc-like loci appear to be restricted to members of the Proteobacteria. High levels of synteny and sequence conservation can be observed among the Phc loci in both phyla, with the exception of the PhcK and PhcL proteins, which are only present in the P. thermoglucosidasius and Moorella thermoacetica DSM 21394 Phc loci (Fig. 3c). BlastP analyses indicate that PhcK and PhcL show highest orthology with PhbB and PhbC in the Phb locus and may have been derived through gene duplication events.
It is notable that the P. thermoglucosidasius Phc locus clusters with a subset of the Thermoanaerobacteraceae in the concatenated Phc protein phylogeny, including Moorella glycerini NMP, M. thermoacetica DSM 21394, Thermoanaeromonas toyohensis DSM 14490 T , Caldanaerobacter subterraneus subsp. tencogensis DSM 15242 T and subsp. yonseiensis DSM 13777 T and Thermoanaerobacter sp. YS13 (Fig. 3c). These differ from the remaining Thermoanaerobacteraceae taxa and the proteobacterial orthologous loci in that they are flanked by three genes, cooCSF, coding for an anaerobic carbon monoxide (CO) dehydrogenase, rather than genes coding for a formate dehydrogenase (FdhH) as is typical for the [Ni-Fe] group 4a hydrogenases [25]. These are generally accompanied by flanking genes coding for the formate dehydrogenase accessory sulfurtransferase protein FdhD, electron transporter HydN, transcriptional activator FhlA and formate transporters FdhC and FocA, which together with FdhH drive the anaerobic oxidation of formate (Fig. 3c) [26,[35][36][37].
BlastP analysis with the FdhH protein of M. thermoacetica DSM 2955 T (AKX95035) shows that an orthologue is present in P. thermoglucosidasius DSM 2542 T (ALF09582). The latter protein, however, shares limited orthology (39% amino acid identity; Bitscore: 497; E-value: 6e−614) with its M. thermoacetica counterpart and is furthermore localised ~ 1.5 Mb upstream of the Phc locus, suggesting the P. thermoglucosidasius FdhH protein does not function together with the [Ni-Fe] group 4a hydrogenase. Instead, the P. thermoglucosidasius Phc hydrogenase may form a novel complex with the adjacent anaerobic CODH locus.

The P. thermoglucosidasius [Ni-Fe] group 4a hydrogenase forms a novel complex with the anaerobic (Coo) CO dehydrogenase, with a distinct evolutionary history
The three genes located just upstream of the Phc hydrogenase locus, cooC, cooS and cooF code for a CO dehydrogenase maturation factor (Figs. 3c, 4), a CO dehydrogenase catalytic subunit and CO dehydrogenase Fe-S protein, respectively. Together these proteins catalyse the oxidation of CO to generate CO 2 (CO + H 2 O → CO 2 + 2 H + + 2ē). The electrons are then used in reduction reactions, including sulphate reduction, heavy metal reduction, acetogenesis, methanogenesis and hydrogenogenesis [38,39].
The CODH locus is also co-localised with the Phc hydrogenase locus and highly conserved among the eight other P. thermoglucosidasius genomes (99.36% average amino acid identity with CooCSF in P. thermoglucosidasius DSM 2542 T ), while no CODH orthologues are encoded on the genomes of any other Parageobacillus or Geobacillus spp. A phylogeny on the basis of the conserved CooS and CooF proteins (Fig. 4) showed that, as with the Phc locus phylogeny (Fig. 3c), those taxa where cooCFS flanks the Phc hydrogenase locus cluster together and show extensive synteny in both the coo and phc gene clusters. This would suggest the co-evolution of the anaerobic CODH and Phc [Ni-Fe] group 4a hydrogenase loci. However, differences in the G+C contents could be observed among the P. thermoglucosidasius coo (average G+C content 46.97%) and phc (average G+C content 49.02%) loci. This is even more pronounced among the Thermoanaerobacteraceae with this CODH-Phc arrangement, where the G+C contents of the two loci differs by an average of 6.62% and is particularly evident in C. subterraneus subsp. tencongensis where the G+C contents of the coo and phc loci differ by 11.77%, suggesting independent evolution of these two loci. This is further (See figure on previous page.) Fig. 3 Prevalence and synteny of the P. thermoglucosidasius-like [Ni-Fe] hydrogenases. a [Ni-Fe] group 1d orthologues. The ML phylogeny was determined on the basis of the trimmed alignment of nine Pha locus proteins (PhaABCDGHIJK-2206 amino acids in length). Hydrogenase genes are coloured in light blue (dark blue for large and small catalytic subunits), tatAE genes in purple and flanking genes in yellow in the synteny diagrams. b [Ni-Fe] group 2a orthologues. The ML phylogeny was determined on the basis of the trimmed alignment of 10 Phb locus proteins (PhbBCDEFHJLMN-2348 amino acids in length). Hydrogenase genes are coloured in red (dark red for large and small catalytic subunits), genes of no known function in biosynthesis and functioning of the hydrogenase in white and flanking genes in yellow in the synteny diagrams. c [Ni-Fe] group 4a orthologues. The ML phylogeny was determined on the basis of the trimmed alignment of nine Phc locus proteins (PhcABCDFGHIJ-2744 amino acids in length). Hydrogenase genes are coloured in light green (dark green for large and small catalytic subunits), anaerobic CODH genes in purple, formate dehydrogenase-related genes in blue and flanking genes in yellow in the synteny diagrams. Values on all trees reflect bootstrap analyses (n = 500 replicates) and all trees were rooted on the midpoint support by the phylogeny (Fig. 4), where the CODH-Phc loci cluster with CODHs which appear on their own and those flanked by an NAD/FAD oxidoreductase are thought to play a role in oxidative stress response [40]. The Energy Conserving Hydrogenase (ECH-[Ni-Fe] group 4c hydrogenase)-CODH complex, which has been shown to couple CO oxidation to proton reduction to H 2 in C. hydrogenoformans and Rhodospirillum rubrum, clusters more distantly from the CODH-[Ni-Fe] group 4a complex [41,42]. Overall, the results suggest that the CODH and [Ni-Fe] group 4a hydrogenase have evolved independently, but may form a complex linking CO oxidation to reduction of protons to produce CO 2 and H 2 .

The CODH-[Ni-Fe] group 4a hydrogenase complex effectively couples CO oxidation to hydrogenogenesis
The predicted function of the co-localized genes encoding the anaerobic CODH and H 2 -evolving hydrogenase (Fig. 3c) was tested using P. thermoglucosidasius DSM 2542 T . Two related strains, Geobacillus thermodenitrificans DSM 465 T and P. toebii DSM 14590 T , which lack both orthologues of the three hydrogenases and the anaerobic CODH, were included as controls. The cultivation of P. thermoglucosidasius DSM 2542 T in serum bottles with a gas atmosphere consisting of 50% CO and 50% air showed that this strain was able to effectively grow in the presence of 50% CO, reaching a maximum absorbance of 0.82 after 6 h of cultivation (Fig. 5). A fractional amount of CO was consumed at the beginning of the experiment, when oxygen was still available, by P. toebii DSM 14590 T (0.37 mmol) and G. thermodenitrificans DSM 465 T (0.216 mmol), respectively. This suggests that these strains may possess an alternative mechanism, such as an aerobic CO dehydrogenase, where CO oxidation is coupled to an electron transport chain which finally reduces oxygen [38]. For instance, a predicted aerobic CODH is present (CoxMSL-OXB91742-744) in P. toebii DSM 14590 T but is absent from G. thermodenitrificans DSM 465 T .
While the two control strains tolerated the presence of CO, no H 2 production was observed for either strain (Figs. 6 and 7). By contrast GC analyses revealed the production of H 2 by P. thermoglucosidasius DSM 2542 T after ~ 36 h (Fig. 8). This corresponds with O 2 reaching a

Discussion
The redox potential and diffusion coefficient of molecular H 2 make it a key component of metabolism and a potent energy source for many microbial taxa [25]. The ability to utilize this energy source relies on the production of various hydrogenase enzymes, which power both the consumption and production of H 2 and inextricably couple H 2 to energy-yielding pathways such as acetogenesis, methanogenesis and respiration [26,43]. Our comparative genomic analysis revealed that P. thermoglucosidasius contains a unique hydrogenase compliment comprised of two uptake hydrogenases (group 1d and 2a) and one H 2 -evolving hydrogenase (group 4a). Evolutionary analysis showed that these hydrogenases are derived through three independent evolutionary events. This indicates that H 2 is likely to play a pivotal role in P. thermoglucosidasius metabolism and bioenergetics in the ecological niches it occupies. By contrast, members of the sister genus Geobacillus lack orthologous hydrogenase loci and, aside from P. thermoglucosidasius, only the group 1d and 2a uptake hydrogenases share orthology in one and three Parageobacillus spp., respectively, even though they are frequently isolated from the same environments. The group 4a H 2 -evolving hydrogenase of P. thermoglucosidasius is not found in any other members of the class Bacilli and is most closely related to those found in members belonging to the class Clostridia, particularly the family Thermoanaerobacteraceae. Furthermore, it forms an association with a CODH, which is found in common with a more restricted subclade of strict anaerobes within the family Thermoanaerobacteraceae. Our fermentation studies with P. thermoglucosidasius in the presence of CO showed that P. thermoglucosidasius grows efficiently when exposed to high concentrations of CO and that the CODHgroup 4a hydrogenase complex can effectively couple CO oxidation to H 2 evolution, P. thermoglucosidasius can do so at a near-equimolar conversion. Furthermore, unlike other CO oxidizing hydrogenogenic bacteria, which are strict anaerobes, P. thermoglucosidasius is a facultative anaerobe capable of first removing residual oxygen from CO gas sources prior to producing H 2 via the water-gas shift reaction. The combination of these features makes P. thermoglucosidasius an attractive target for potential incorporation in industrial-scale production strategies of biohydrogen.