Skip to main content

The effects of synthesis gas feedstocks and oxygen perturbation on hydrogen production by Parageobacillus thermoglucosidasius



The facultatively anaerobic thermophile Parageobacillus thermoglucosidasius is able to produce hydrogen gas (H2) through the water–gas shift (WGS) reaction. To date this process has been evaluated under controlled conditions, with gas feedstocks comprising carbon monoxide and variable proportions of air, nitrogen and hydrogen. Ultimately, an economically viable hydrogenogenic system would make use of industrial waste/synthesis gases that contain high levels of carbon monoxide, but which may also contain contaminants such as H2, oxygen (O2) and other impurities, which may be toxic to P. thermoglucosidasius.


We evaluated the effects of synthesis gas (syngas) mimetic feedstocks on WGS reaction-driven H2 gas production by P. thermoglucosidasius DSM 6285 in small-scale fermentations. Improved H2 gas production yields and faster onset towards hydrogen production were observed when anaerobic synthetic syngas feedstocks were used, at the expense of biomass accumulation. Furthermore, as the WGS reaction is an anoxygenic process, we evaluated the influence of O2 perturbation on P. thermoglucosidasius hydrogenogenesis. O2 supplementation improved biomass accumulation, but reduced hydrogen yields in accordance with the level of oxygen supplied. However, H2 gas production was observed at low O2 levels. Supplementation also induced rapid acetate consumption, likely to sustain growth.


The utilisation of anaerobic syngas mimetic gas feedstocks to produce H2 and the relative flexibility of the P. thermoglucosidasius WGS reaction system following O2 perturbation further supports its applicability towards more robust and continuous hydrogenogenic operation.


Increasing energy demands and dwindling fossil fuel reserves are driving the development and implementation of environmentally friendly energy production technologies, using renewable resources [1]. Hydrogen (H2) gas is regarded as an ideal energy carrier, with carbon neutral combustion characteristics and high energy density [2]. While H2 gas is currently produced almost exclusively through the reformation of non-renewable fossil fuels, various alternative renewable H2 gas production processes have and are currently being developed [3].

Numerous microbial taxa are capable of producing H2 gas from a wide spectrum of organic and inorganic feedstocks in an environmentally friendly and energetically favourable manner [4, 5]. These are subdivided into light dependent and independent processes with those incorporating dark-, photo-, and hybrid-fermentative processes having dominated mainstream research [4]. Despite yielding promising results, these processes are hindered by limitations such as cost, feedstock availability/suitability, turnover rate and scalability, that restrict their industrial application [3]. Less well-explored hydrogenogenic processes such as the biological water–gas shift (WGS) reaction, are promising alternatives to the aforementioned systems.

The reversible WGS reaction involves the oxidation of carbon monoxide to carbon dioxide and the reduction of water to molecular hydrogen at an equimolar ratio (CO + H2O → CO2 + H2 ∆GO′ =  − 20 kJ/mol CO) [6]. The potential to conduct the WGS reaction has been identified in more than twenty-five bacterial species spanning the phyla Proteobacteria, Firmicutes and Dictyoglomi and among four archaeal species in the phyla Crenarchaeota and Euryarchaeota [7]. The biological WGS reaction has been best characterised in the mesophilic proteobacterium Rhodospirillum rubrum, the thermophilic Firmicutes Carboxydothermus hydrogenoformans and Parageobacillus thermoglucosidasius and the hyperthermophile Thermococcus onnurineus in the phylum Euryarchaeota [8,9,10,11]. The reaction proceeds via a carbon monoxide dehydrogenase (CODH)/H2 evolving hydrogenase enzyme complex [12,13,14]. Both the CODH and H2 evolving hydrogenase enzymes are highly sensitive to O2, greatly restricting their industrial applicability [15, 16]. Although several hydrogenases demonstrate greater O2-tolerance, they exhibit poorer H2 gas turnover efficiencies [16]. As such, the majority of bacterial taxa capable of conducting the WGS reaction are strict anaerobes. One exception, however, is the facultatively anaerobic thermophile P. thermoglucosidasius, which carries a Ni-CODH/NiFe group 4a hydrogenase complex and is capable of the WGS reaction following oxygen depletion [10, 17,18,19]. By optimising growth medium composition and pH, fermentation temperatures and CO concentration, H2 gas production yields have been improved [20], but further developments towards industrial application are necessary. Key foci include progression of the process towards more continuous operation and evaluation of system efficiency when utilising gas feedstocks more reflective of industrially/commercially available waste gases. Industrial waste gasses or synthesis gases (syngas) are highly variable, typically comprising CO, H2 and CO2. Impurities including tars, particulate carbon, light hydrocarbons, inorganic compounds (sulfur-, nitrogen-, halogen- and metal-incorporated compounds) and O2 are also typically found at low levels [21, 22].

Here we investigated the effects of gas feedstocks, with compositions more reflective of industrial waste gases, on H2 gas production by P. thermoglucosidasius through bioassays in conjunction with GC and HPLC analysis of select gases and metabolites. Furthermore, as optimal P. thermoglucosidasius growth is observed under aerobic atmospheres, balancing aerobic conditions conducive to sustained biomass accumulation and anaerobic conditions obligatory for H2 production is necessary for more advanced continuous/semi-continuous hydrogenogenic operation. Therefore, the impacts of temporal and volumetric O2 perturbations (as would occur with industrial and syngas) on P. thermoglucosidasius DSM 6285 hydrogenogenesis were further evaluated in this study.


Strains and media

P. thermoglucosidasius DSM 6285 was maintained and cultured routinely in modified Luria Bertani medium (mLB) [10]. Experimental cultures were grown in modified ammonium sulphate medium (mASM) [23].

Experimental setup

Pre-cultures were grown in 100 mL mLB in 500 ml baffled shake flasks, inoculated with 40 \(\mu {\text{L}}\) of glycerol stock and incubated aerobically at 55 °C for 14 h at 120 rpm in an Infors HT Thermotron (Infors AG, Bottmingen, Switzerland). Following incubation, appropriate volumes of pre-culture were used to inoculate experimental cultures to an initial OD600 of ~ 0.1. Experimental cultivations were conducted in 250 ml serum bottles sealed with a rubber stopper, containing 50 mL ASM with differing gas headspace compositions at atmospheric pressure.

To investigate the effects of variable industrial syngas mixtures on the P. thermoglucosidasius DSM 6285 WGS reaction, the strain was cultured in several gas headspace atmospheres (Table 1). Four sets of experimental cultures were cultivated in triplicate. The first set served as an aerobic control grown under an optimal gas atmosphere composed of 50% air and 50% CO. The second set served as an anaerobic control, with a gas headspace composed of 50% N2 and 50% CO. This was achieved by flushing bottle headspaces with N2 for 5 min at a flow rate of 0.2 L/minute, followed by injection of the bottles with CO to an overpressure of 1 bar, with subsequent equilibration to atmospheric pressure. The third culture set was cultured under an initial gas headspace comprising 50% N2 and 50% synthetic syngas mixture (Table 1; Air Liquide, Düsseldorf, Germany). This was achieved by flushing bottle headspaces with N2 for 5 min at a flow rate of 0.2 L/minute, then injecting the bottles with the synthetic gas mixture to an overpressure of 1 bar, followed by subsequent equilibration to atmospheric pressure. The fourth culture set was cultivated under a 100% atmosphere of the synthetic syngas mixture. Following inoculation, bottles were incubated at 55 °C for 134 h at 120 rpm.

Table 1 Composition of synthetic syngas stock and initial gas headspace mixtures

To investigate the influence of O2 addition on WGS reaction-driven hydrogenogenesis, two experiments were conducted. In all cases, the serum bottle gas headspace of each stoppered replicate was reconstituted to an ideal initial atmosphere of 50% air and 50% CO prior to inoculation [23]. This was achieved by flushing bottles with CO to an overpressure of 1 bar, then equilibrating the headspace to atmospheric pressure. O2 was injected into the gas headspace of bottles containing actively fermenting cultures at various stages of the fermentation. Four time points (post-inoculation) were selected for O2 addition, reflective of distinct periods of the WGS reaction fermentation profile [18]. These included the mid aerobic phase (14 h), the commencement of the anaerobic phase (24 h), the early-mid hydrogenogenic phase (38 h) and the late hydrogenogenic phase (48 h). The experiment was carried out in two batches, comprising three sets of cultures, one control and two experimental cultures, respectively. In each batch, cultures were cultivated in triplicate with one set serving as a control where no O2 was added at any timepoint. The two remaining culture sets in each batch were designated a time point where a defined volume of O2 was injected. O2 volumes injected were ~ 5% of the initial serum bottle gas headspace volume, calculated as \({v}_{{O}_{2}}={v}_{t}-{v}_{m}-{v}_{i}\), where \({v}_{{O}_{2}}\) = volume O2 added, \({v}_{T}\) = total serum bottle volume, \({v}_{m}\) = volume of the liquid culture medium and \({v}_{i}\) = inoculum volume. Following inoculation, bottles were incubated at 55 ˚C for 120 h at 120 rpm, with O2 being added to the appropriate replicate set at the designated time points.

Finally, the influence of different volume O2 additions on the WGS system at a single time point was investigated. A 38-h post-inoculation timepoint, representative of the peak hydrogenogenic period, was selected based on the fermentation profiles of the previous experiments. Five sets of experimental cultures were cultivated in triplicate under a gas atmosphere of 50% air and 50% CO, with one set serving as a control where no O2 was added. Each of the four remaining culture sets were designated an O2 volume to be injected at 38 h, either ~ 1, 3, 5 or 10% of the initial serum bottle gas headspace volume, calculated as described above. Following inoculation, bottles were incubated at 55 ˚C for 120 h at 120 rpm, with the appropriate O2 volumes being added to the respective replicate sets at 38 h.

Sampling and analytics

Liquid culturing medium and gas headspace sampling was conducted at various time points for each experiment. For the experiment where different syngas mixtures were evaluated, samples were taken at 0, 14, 24, 38, 48, 62, 72, 86, 96, 110, 120 and 134 h post-inoculation. For the first oxygen supplementation experiment, where a fixed O2 volume was supplemented at various timepoints in the WGS reaction fermentation, sampling points were 0, 10, 14, 24, 34, 38, 48, 58, 72, 84, 96, 110 and 120 h post-inoculation. Sampling points for the final experiment, where variable O2 volumes were supplemented during the peak hydrogenogenic period were 0, 14, 24, 38, 41, 44, 48, 62, 72, 86, 96, 110, 120, and 134 h post-inoculation. In both O2 addition experiments, prior to the addition of O2 at applicable time points, both a liquid and gas sample were taken and immediately followed by a second gas sampling post-O2-addition. A volume of 3 mL gas headspace was drawn at each time point and analyzed using a 300 Micro GC gas analyzer (Inficon, Switzerland), with argon and helium as carrier gases connected with Molsieve 5 Å and PLOT Q columns for data acquisition. Gas composition was calculated based on the ideal gas law as previously described [10]. Headspace pressure was determined before sampling using a manometer (GDH 14 AN, Greisinger electronic, Regenstauf, Germany) and temperature was maintained at 60 ˚C. A volume of 1 mL of liquid culture medium was sampled at each sampling point for growth and used for subsequent HPCL analysis. Growth was estimated using absorbance (OD600) values using an Ultrospec 1100 pro spectrophotometer (Amersham Biosciences, Uppsala, Sweden) and pH was measured using a Profilab pH 597 (Xylem Analytics, Weilheim, Germany).

Following absorbance and pH measurement, samples were centrifuged at 16,000 × g for 10 min and supernatants transferred to a clean 1.5 mL microcentrifuge tube for storage at − 20 ˚C. Thawed supernatants (100 μL) were dispensed in 1.5 mL HPLC autosampler vials fitted with micro-inserts. Glucose and twelve central carbon metabolites (1-propanol, acetate, butyrate, ethanol, formate, glyoxylate, iso-butyrate, iso-valerate, lactate, propionate, succinate and valerate) were monitored over 0–2.0 g/L concentration ranges. Samples were analyzed using an Agilent 1100 series HPLC system (Agilent Technologies, Waldbronn, Germany) connected to a wavelength detector and refractive index detector with a 50 mm long pre-column (model Rezex ROA-Organic Acid H+ 8% Guard Column) and a 300 mm long separation column (model Rezex ROA-Organic Acid H+ 8%). Analysis was conducted using a 5 mM H2SO4 mobile phase, 50 ˚C column temperature, 0.6 mL/min flow rate for 40 min per sample and injection volume of 20 μL. Data acquisition and handling was conducted using Chemstation (Agilent Technologies) and datasets were summarized and figures plotted using R v. 4.3.2 [24, 25].

Statistical analysis

Statistical analysis for the Gas chromatography and HPLC was performed in R v. 4.3.2 [24, 25]. To determine significant changes in gas headspace and metabolite profile compositions between the controls and experimental cultures, one way analysis of variance (ANOVA) was conducted between the respective constituents within each experiment, followed by post-hoc Tukey Honest Significant Difference (HSD) tests. To determine significant changes in H2 following O2 addition in both the temporal and volumetric experiments, H2 levels at each sampling point within culture sets were subjected to one way ANOVA, followed by post-hoc Tukey HSD tests.


Anaerobic synthesis gas mimetic feedstocks promote earlier onset of hydrogenogenesis at the expense of biomass accumulation

Previous studies have characterised the fermentation profile of P. thermoglucosidasius DSM 6285 under various aerobic and anaerobic CO enriched headspaces, citing a 50% CO/50% air atmosphere as an ideal balance between suitable growth conditions and facilitating downstream WGS reaction conditions [18,19,20, 23]. Here we observed distinct fermentation profiles between the aerobic and anaerobic control, syngas (50%) and syngas (100%) atmospheres (Fig. 1).

Fig. 1
figure 1

Gas composition (1), metabolite (2) and growth and pH (3) profiles for P. thermoglucosidasius DSM 6285 cultured under 50% CO/50% Air- (a) 50% CO/50% N2- (b) 50% synthetic syngas- (c) and 100% synthetic syngas- (d) -gas headspaces

The aerobic control growth profile corresponded with previously observed patterns. This included an initial sharp increase in absorbance during the aerobic phase, peaking at 0.327 ± 0.008 by 48 h (coincidental with peak WGS reaction activity), followed by gradual decreases towards the end of fermentation (Fig. 1). All anaerobic atmospheres exhibited significantly poorer growth throughout the fermentations (Additional file 1: Table S1). Slight, albeit much less pronounced increases were observed for the anaerobic control (50%) and syngas (50%) atmospheres during the first 14 h of the experiment. This growth was not, however, shared by the syngas (100%) atmosphere replicates, which increased negligibly, plateauing similarly to the N2 and synthetic syngas (50%) atmosphere replicates at 0.111 ± 0.007. Growth of the syngas (100%) atmosphere replicates was also atypical, exhibiting a granular flocculant-type of growth compared to the turbid/cloudy appearance observed in the other atmospheres (results not shown).

In the early phases of each fermentation, pH values decreased across all atmospheres to varying extents (Fig. 1). The aerobic control reached a low of 5.82 ± 0.02 by 14 h, accompanied and likely accounted for by concerted acetate production observed in this period. Anaerobic atmospheres demonstrated similar pH profiles to the aerobic control, however, reaching significantly lower pH values of 5.31 ± 0.04 by 38 h (Additional file 1: Table S1). These atmospheres exhibited strong concurrent production of lactate, formate and acetate which plateaued in conjunction with pH stabilisation and were likely responsible for the observed pH fluctuations. Following initial decreases, pH stabilized across all atmospheres, with the aerobic control stabilizing at a final value of 5.87 ± 0.02 and the anaerobic atmospheres at 5.37 ± 0.03.

H2 production was observed within 14 h in the anaerobic atmospheres (Fig. 1), while the aerobic control only showed H2 production by 24 h. Notably, H2 production in the aerobic control was observed with concurrent O2 levels of 0.95 ± 0.06%. Peak H2 productivity for the aerobic control was much higher than the anaerobic atmospheres with peak productivity observed between 38 and 48 h at 0.0137 ± 0.0048 mmol h−1. For the anaerobic control and syngas (100%) atmospheres, lower peak productivity was observed earlier between 24 and 38 h at 0.0063 ± 0.0005 mmol h−1 and 0.0047 ± 0.0016 mmol h−1 respectively, while the syngas (50%) atmosphere exhibited the earliest peak productivity between 14 and 24 h at 0.0055 ± 0.0005 mmol h−1. Notably for the anaerobic atmospheres, H2 production appeared to coincide with mixed acid fermentation, suggesting concurrent operation of these processes (Fig. 1).

H2 mol yields per CO mol consumed (H2/CO) across the fermentations varied between the points where CO was completely consumed and the endpoints of the fermentations at 134 h. Each atmosphere, except for the syngas (100%) atmosphere showed complete CO consumption within 134 h. Unsurprisingly, due to its comparatively low initial CO levels, the syngas (50%) atmosphere demonstrated the earliest CO depletion (84 h), followed by the aerobic control (110 h), then the anaerobic control (120 h). By 134 h, 1.06 ± 0.63% CO remained in the syngas (100%) atmosphere gas headspace. At the points where CO was completely consumed, the aerobic and anaerobic controls and the syngas (50%) atmosphere exhibited significantly different H2/CO yields from one another (Additional file 1: Table S2), of 89.48 ± 0.83%, 85.40 ± 0.14% and 93.62 ± 0.31% respectively. However, by the end point of the fermentations, yields fluctuated further, with both the aerobic control and syngas (50%) atmospheres having exhibited decreases of 4.92 ± 0.47% and 6.73 ± 0.85% respectively. Conversely, the anaerobic control H2/CO yield increased by 8.48 ± 0.43%. By 134 h, despite not having depleted CO, the H2/CO yield of the syngas (100%) atmosphere stood at 83.83 ± 1.42%, similar to that of the aerobic control but significantly lower than those of the anaerobic control and syngas (50%) atmospheres (Additional file 1: Table S2).

Analysis of the metabolite profile with HPLC showed that lactate, formate and ethanol were, in descending order, the most prevalent metabolites produced under anaerobic atmospheres, while only negligible amounts of formate and ethanol were observed for the aerobic atmosphere (Fig. 1). Lactate, formate and ethanol levels across the anaerobic atmospheres stabilized at 9.3442 ± 0.2808 mmol, 7.6075 ± 0.4817 mmol and 5.2032 ± 1.0371 mmol after 38 h levels despite the syngas (100%) atmosphere having exhibited significantly lower productivity of each acid between 0–14 h (Additional file 1: Table S1).

Acetate was the central metabolite detected through HPLC in the aerobic atmosphere, rising rapidly and peaking at 12.1255 ± 0.2620 mmol by 38 h (Fig. 1). Slight decreases in acetate levels were observed between 38 and 62 h to 11.0224 ± 0.1368 mmol but remained relatively stable thereafter. Acetate accumulation in the anaerobic atmospheres was significantly lower than the aerobic atmosphere (Additional file 1: Table S1), accumulating to only a third of that of the aerobic control. The aerobic control also exhibited significant increases in succinate production between 38 and 72 h (Additional file 1: Table S1), reaching a final level of 2.5111 ± 0.0606 mmol, while the anaerobic atmospheres showed negligible succinate accumulation. Glucose was completely consumed within 14 h post-inoculation in the aerobic control. Low levels of glucose, below the 0.1 mg L−1 HPLC method detection limit, were observed throughout all three anaerobic fermentations. However, most of the initially available glucose (> 99%) was consumed by 24 h in the N2 control and syngas (50%) atmospheres and by 38 h in the syngas (100%) atmosphere.

Pre-hydrogenogenic phase O2 perturbations extend the lag phase prior to H2 production

Oxygen addition at the various timepoints resulted in distinct fermentation profiles. Oxygen additions at 14 h and 24 h (Fig. 2) exhibited markedly different influences on the various fermentation profiles compared to the control and 38 h and 48 h O2 additions (Fig. 3). For all O2 additions, significantly higher peak and final absorbance (OD600) values were observed compared to the controls following O2 addition (Additional file 1: Tables S3 and S4). In the cases of the 38 h and 48 h O2 additions, which coincided with the hydrogenogenic phase, severe reductions in H2 production were observed, implying the growth observed was primarily the result of O2 addition and not supplemented through the WGS reaction. While the control OD600 values declined following peak levels at 48 h, the 38 h and 48 h O2 additions exhibited further significantly elevated peak OD600 values of 0.332 ± 0.011 and 0.302 ± 0.006 at 48 h and 58 h respectively following oxygen addition (Additional file 1: Tables S3 and S4).

Fig. 2
figure 2

Gas composition (1) metabolite (2) and growth and pH (3) profiles for P. thermoglucosidasius DSM 6285 cultured under a 50% CO/50% Air gas headspace with no O2 addition (a), 14 h O2 addition (b) and 24 h O2 addition (c)

Fig. 3
figure 3

Gas composition (1), metabolite (2) and growth and pH (3) profiles for P. thermoglucosidasius DSM 6285 cultured under a 50% CO/50% Air gas headspace with no O2 addition (a), 38 h O2 addition (b) and 48 h O2 addition (c)

Steep declines in pH to lows of 5.78 ± 0.01 (Fig. 2) and 5.93 ± 0.02 (Fig. 3) within 14 h were observed across the fermentation batches in the aerobic phase, reflective of the observed acetate, formate and lactate accumulation. In the controls, these declines were followed by slight increases alongside near complete and complete consumption of formate and lactate, stabilizing at final pH values of 5.88 ± 0.02. Similar profiles were observed until either the addition of O2 in the 14 h and 24 h O2 addition replicates or 34–38 h in the 38 h and 48 h O2 additions replicates. O2 addition resulted in the significant increase in pH compared to the controls (Additional file 1: Tables S3 and S4), until the final sampling point, likely the result of decreases in observed acetate levels.

H2 gas was first detected across all samples within 38 h. The highest O2 level detected with concurrent H2 production was 2.49 ± 0.06% O2. O2 additions elevated O2 levels an average of 4.51 ± 0.35%. Peak H2 productivities were experienced between 34 and 38 h at 0.0297 ± 0.0004 mmol h−1 (Batch 1 control), 0.0205 ± 0.0045 mmol h−1 (Batch 2 control) and 0.0130 ± 0.0052 mmol h−1 (38 h O2 addition). The 48 h O2 addition replicates demonstrated peak H2 productivity between 38 and 48 h at 0.0143 ± 0.0006 mmol h−1, whilst both the 14 h and 24 h O2 additions showed more delayed peak productivities of 0.0133 ± 0.0004 mmol h−1 and 0.0123 ± 0.0002 mmol h−1, respectively, between 48 and 58 h. H2 productivity decreased gradually as CO was depleted, eventually resulting in the net consumption of H2, with a maximal H2 consumption rate of 0.0030 ± 0.0002 mmol h−1 across all replicate sets.

Significant increases in H2 levels were observed in both controls between 24 and 72 h (Additional file 1: Table S5). In each instance where O2 was added, significant disruptions to H2 production were observed (Additional file 1: Tables S3 and S4). H2 production was delayed following O2 addition at 14 h and 24 h and resulted in significantly lower final H2 levels (2.7351 ± 0.0058 mmol and 2.9094 ± 0.0127 mmol respectively) compared to the control (3.0260 ± 0.0112 mmol). H2 productivity was significantly lower between both O2 additions and the control between 24 and 38 h. Notably, the 24 h O2 addition replicates showed earlier and more intensive H2 production than the 14 h O2 addition replicates, with comparably higher H2 levels and H2 productivities during the majority of the fermentation (Fig. 2; Additional file 1: Tables S3 and S4). Following O2 addition for the 38 h and 48 h O2 addition replicate sets, both H2 levels and H2 productivities were significantly lower than the control for the subsequent 10 h period (Additional file 1: Tables S3 and S4). However, weak H2 production was still observed across this period, with H2 productivities of 0.0011 ± 0.0004 mmol h−1. Following this period, however, significant increases in H2 levels were observed for the 38 h and 48 h O2 addition replicates between 48–84 h and 58–84 h respectively (Additional file 1: Table S5). However, H2 levels for both O2 additions remained significantly lower than the control following O2 addition.

Both controls and pre-hydrogenogenic phase 14 h and 24 h O2 additions showed complete utilisation of CO, with trace amounts of CO remaining in the 38 h and 48 h O2 addition replicates (Fig. 2). H2/CO yields at the points of complete CO depletion were 83.26 ± 1.19% (Controls), 74.35 ± 0.07% (14 h O2 addition) and 79.10 ± 0.26% (24 h O2 addition), respectively. These were all significantly different from one another (Additional file 1: Table S2). Yields of 80.35 ± 0.13% and 81.33 ± 0.30% were observed for the 38 h and 48 h O2 addition replicates by 120 h. Both yields were significantly lower than the respective control but not significantly different from one another (Additional file 1: Table S2).

HPLC analysis detected fluctuations in acetate and succinate profiles following oxygen addition. Acetate levels across all samples increased dramatically from the onset of the fermentations and peaked either prior to O2 addition or within 34 h post-inoculation (Fig. 2). In the controls, after reaching peak acetate levels of 11.4257 ± 0.5833 mmol, a low-level net consumption of acetate was observed, gradually reducing acetate levels to a final concentration of 10.4132 ± 0.6191 mmol. In each case where O2 was added, significantly lower acetate levels were observed throughout the remainder of the fermentations compared to the respective controls (Additional file 1: Tables S3 and S4). Correspondingly, significantly higher acetate consumption levels were observed between all O2 addition replicates and the controls immediately following O2 addition. This was observed between 14–34 h (14 h O2 addition), 24–38 h (24 h O2 addition), 38–82 h (38 h O2 addition) and 38–72 h (48 h O2 addition), respectively post-O2 addition.

Although succinate levels were observed to increase gradually from the start of all fermentations, prominent increases were only observed after 24 h in the controls, peaking at 2.544 ± 0.1405 mmol (Fig. 2). O2 addition at 14 h and 24 h delayed succinate accumulation, presenting significantly lower succinate levels than the control between 34–58 h and 38–58 h respectively (Additional file 1: Table S3). Despite this, the 14 h O2 addition replicates accumulated significantly higher final succinate levels compared to both the control and 24 h O2 addition replicates, peaking at 3.1261 ± 0.1344 mmol. Following O2 addition at 38 h and 48 h, significantly lower levels of succinate were observed compared to the control between 48–58 h (Additional file 1: Table S4). Despite this delay in succinate production, there were no significant differences between the final accumulated succinate concentrations of the control, 38 h and 48 h O2 additions, which averaged at 2.7774 ± 0.9710 mmol.

Glucose levels across all samples were almost completely consumed (> 97%) 14 h post-inoculation and were completely consumed within 34 h. Large formate 12.9833 ± 1.0624 mmol and lactate 6.5760 ± 0.5515 mmol peaks were detected at 14 h post-inoculation across each replicate set. Both compounds were almost completely consumed thereafter with only minor increases in formate below the detection limit (< 0.1 g/L) of the HPLC method.

Minor O2 perturbation during peak-hydrogenogenesis minimally disrupts H2 production in P. thermoglucosidasius DSM 6285

Prior to oxygen addition, the 38 h varied O2 volume addition fermentations demonstrated the same trends in growth, pH, gas composition and HPLC metabolite profiles as observed previously (Figs. 1, 2). Following varied oxygen additions, however, disruptions in these profiles were scaled accordingly with the volumes of oxygen supplemented (Fig. 4). O2 levels post-O2 addition were 0.90 ± 0.22% (1% O2 addition), 2.83 ± 0.07% (3% O2 addition), 4.70 ± 0.14% (5% O2 addition) and 9.08 ± 0.14% (10% O2 addition), respectively.

Fig. 4
figure 4

Gas composition (1) metabolite (2) and growth and pH (3) profiles for P. thermoglucosidasius DSM 6285 cultured under a 50% CO/50% Air gas headspace with no O2 addition (a), 1%- (b), 3%- (c), 5%- (d) and 10%- (e) -O2 addition at 38 h

Following O2 addition OD600 values increased in accordance with volumes of O2 added, with larger O2 volume additions resulting in greater peak OD600 values achieved (Fig. 4). The 1% O2 addition prompted the weakest response, but exhibited significantly higher values compared to the control from 48 h onwards (Additional file 1: Table S5). The 3% O2 addition showed a slightly more delayed increase in OD600 values, with significantly elevated OD600 values compared to the control from 62 h onwards. The 5% and 10% O2 additions both exhibited a slight decrease in OD600 values following O2 addition between 38 and 41 h, however, showed a steep increase thereafter. From 44 h onwards, the 5% O2 addition maintained significantly higher OD600 values compared to the control. The 5% O2 addition also maintained significantly higher OD600 values than those of the 1% and 3% O2 additions over the majority of the fermentation and was similar to the 10% O2 addition. Although the 10% O2 addition exhibited the greatest increase in OD600 values, peaking at 0.395 ± 0.006 at 72 h, OD600 values appeared to decrease sooner than those of the 5% O2 addition.

Larger O2 volume additions correspondingly resulted in larger sustained increases in pH, likely the result of the observed reduction in acetate levels following O2 addition (Fig. 4). Apart from the 1% O2 addition, which displayed no significant difference from the control, all other O2 additions prompted significant increases in pH values 6 h post-O2 addition (Additional file 1: Table S5). pH values stabilized in each case after ~ 62 h, alongside plateaus in acetate consumption and OD600 peaks, with the 10% O2 addition almost returning to starting pH values, reaching as high as 6.35 ± 0 (Fig. 4).

H2 was first detected at 38 h across all replicate sets at elevated levels (Fig. 4). The control exhibited a H2 productivity of 0.0460 ± 0.0013 mmol h−1 between 38 and 41 h. Comparatively, H2 productivities of the O2 addition replicates were all significantly lower than the control in this period (1%: 0.0137 ± 0.0062 mmol h−1; 3%: 0.0002 ± 0.0003 mmol h−1; 5%: 0.0005 ± 0.0012 mmol h−1; 10%: 0.0009 ± 0.0003 mmol h−1; Additional file 1: Table S5). In the case of the 1% O2 addition, no significant difference in H2 productivity with the control was detected after 41 h (Additional file 1: Table S6), with H2 productivity between 38–48 h (0.0101 ± 0.0013 mmol h−1), the peak H2 productivity period, only slightly lower than that of the control (0.0150 ± 0.0006 mmol h−1). However, unlike the control, no significant increase in H2 levels were observed between 38 and 41 h (Additional file 1: Table S5). Significant increases in H2 levels were only detected in the 1% O2 addition replicates between 41 and 62 h (Additional file 1: Table S5), when O2 levels were below 0.79 ± 0.13%. Despite this, no significant difference in H2 concentration was detected between the 1% O2 addition and the control over this period (Additional file 1: Table S6). However, following complete CO consumption, the 1% O2 addition replicates exhibited significantly lower final H2 levels (2.8999 ± 0.0788 mmol) than the control (3.1277 ± 0.0874 mmol; Additional file 1: Table S6).

Following O2 addition, significantly lower levels of H2 were detected between the 3%, 5% and 10% O2 addition replicates and the control until the endpoints of the fermentations (Additional file 1: Table S6). Lower final H2 levels were observed accordingly with higher volumes of O2 added (Control: 3.0525 ± 0.0110 mmol, 1% O2 addition: 2.8290 ± 0.0097 mmol, 3% O2 addition: 2.8038 ± 0.0084 mmol, 5% O2 addition: 2.7613 ± 0.0085 mmol, 10% O2 addition: 2.7657 ± 0.0096 mmol). However, no significant differences in final H2 levels were observed between the 1% and 3%- and between the 5% and 10% -O2 addition replicates (Additional file 1: Table S6). Following the reduction in H2 productivity experienced by the 3%, 5% and 10% O2 addition replicates between 38 and 41 h, H2 consumption was observed for each of these O2 additions between 41 and 44 h at rates of 0.0002 ± 0.0005 mmol h−1, 0.0019 ± 0.0002 mmol h−1 and 0.0018 ± 0.0005 mmol h−1 respectively (Fig. 4). H2 productivity for these replicates remained significantly lower than the control between 38 and 48 h for 3% and 5% O2 addition replicates and 38–62 h for the 10% O2 addition replicates (Additional file 1: Table S6). However, H2 productivity increased following 44 h with the 10% O2 addition replicates demonstrating significantly higher (P < 0.01) H2 productivity than the control between 62 and 96 h (Additional file 1: Table S6). Consequently, significant increases in H2 levels were only detected between 48 and 62 h for the 3% and 5% O2 addition replicates and between 62 and 72 h for the 10% O2 addition replicates (Additional file 1: Table S5). Following O2 addition, O2 levels at the first point where a significant increase in H2 was observed (Additional file 1: Table S6), were 0.03 ± 0.06% (3% O2 addition), 0.17 ± 0.09% (5% O2 addition) and 0.97 ± 0.1% (10% O2 addition).

CO was completely consumed across each condition by 134 h (Fig. 4). The control exhibited CO depletion first (110 h) alongside the 1% O2 addition replicates, then the 3% O2 addition replicates (120 h) and the 5% and 10% O2 addition replicates (134 h). The control exhibited the highest H2/CO yield both at the first point of CO depletion (88.56 ± 0.62%) and by the end of the fermentation (83.86 ± 0.62%). In both instances yields were significantly higher than all O2 addition yields (Additional file 1: Table S2), despite exhibiting a decrease in yield following CO depletion. Although producing significantly different yields at the first points of CO depletion (82.26 ± 0.25% and 85.26 ± 0.30% respectively), the 1% and 3% O2 additions showed no significant difference in H2/CO yield by the endpoint of the fermentation (Additional file 1: Table S2), with final yields of 77.82 ± 0.00%. In both cases, however, reductions in H2/CO yields of 4.70 ± 0.41% and 8.30 ± 0.94%, respectively, were observed between the two points. The 5% followed by 10% O2 additions exhibited the lowest H2/CO yields at 134 h, with yields of 76.51 ± 0.20% and 75.89 ± 0.24%, respectively.

Acetate levels across all samples increased dramatically from the onset of the fermentations and peaked prior to O2 addition (Fig. 4). Peak acetate levels of 12.5025 ± 0.1713 mmol were observed by 24 h, followed by net reduction, with the control reaching a final acetate concentration of 11.1013 ± 0.6522 mmol. The 1% O2 addition showed almost no differences in acetate levels compared to the control (Additional file 1: Table S6). In the cases of the 3%, 5% and 10% O2 additions, significantly lower acetate levels were observed after 48 h, 44 h and 48 h, respectively, with final acetate levels of 8.5019 ± 0.5227 mmol (3% O2 addition), 5.3856 ± 0.3675 mmol (5% O2 addition) and 2.2098 ± 1.0733 mmol (10% O2 addition). Greater degrees of acetate consumption were observed in accordance with greater volumes of O2 added, with the most noteworthy decreases in acetate observed by the 10% O2 addition, followed by the 5%, 3% and finally 1% O2 additions, with the latter showing no significant difference from the control (Additional file 1: Table S6).

Succinate levels increased gradually from the start of all fermentations (Fig. 4), although prominent increases in succinate were staggered. Whilst there was no significant difference in succinate productivity immediately following O2 addition between the 1%, 3% and 5% O2 additions and the control, the 10% O2 addition exhibited net succinate consumption following O2 addition, between 38 and 48 h and significantly lower succinate productivity until 62 h post-inoculation (Additional file 1: Table S6). All O2 additions exhibited significantly higher succinate productivity compared to the control between 62 and

–84 h. Compared to the control which exhibited final succinate levels of 2.7534 ± 0.1376 mmol, the 5% O2 addition replicates accrued significantly higher levels of 3.2737 ± 0.2034 mmol by the end of the fermentation. Glucose levels across all samples were almost completely consumed (> 97%) within 14 h and were completely consumed within 34 h. Formate and lactate peaks were detected at 14 h post-inoculation across each replicate set, however, both compounds were almost completely consumed thereafter with only minor increases in formate below the detection limit (< 0.1 g/L) of the HPLC method.


Syngas fermentation provides the opportunity for microbial conversion to various value-added products from industrial waste gases [26]. Here we explored the effects of synthetic syngas mixtures on the P. thermoglucosidasius WGS reaction. Our analyses showed that P. thermoglucosidasius grown in anaerobic syngas atmospheres, with highly varied CO and CO2 compositions and impurities including methane, ethane, propane and butane, was able to produce H2, being limited only by CO availability. However, cultivation in these anaerobic atmospheres severely hampered biomass accumulation.

Poor biomass accumulation was likely due to the distinct bioenergetics associated with aerobic and anaerobic respiration as well as fermentation. Anaerobic atmospheres exhibited distinct mixed acids fermentation profiles favouring lactate, formate, ethanol and acetate production, which could account for the more marked reductions in pH observed across these conditions [27,28,29,30]. The aerobic control exhibited pronounced acetate production prior to commencement of H2 production, alongside concerted succinate production which coincided with O2 depletion and onset of the WGS reaction. As well as serving as a means for balancing cell redox levels and potentially as an alternative carbon pool, it has been suggested that elevated acetate levels increase the CO uptake capacity of CO consumers [23, 31]. Notably, two-fold greater H2 production rates were observed under an aerobic atmosphere, where acetate levels were also several fold higher than anaerobic/microaerobic atmospheres.

In the syngas atmospheres, varying CO levels influenced growth and metabolism to a large extent, impacting downstream H2 production. H2 production in these atmospheres showed considerably shorter lag phases between inoculation and onset of the WGS reaction, similar to the anaerobic control and previous anaerobic cultivations that have exhibited reductions to within 4 h post-inoculation [19]. The 50% syngas, atmosphere, with the lowest CO level (~ 17%), exhibited pronounced earlier onset of fermentative metabolism and H2 production compared to the anaerobic control (CO level ~ 46%) and the syngas (100%) atmosphere (CO level ~ 38%). From previous aerobic studies the opposite was anticipated as atmospheres with enriched CO levels facilitated more prolific growth responses and correspondingly greater hydrogenogenic activity [20, 23]. We observed distinct conversion efficiencies between the first observed points of CO depletion and the fermentation end-points, with general yield decreases towards the end. Given the absence of known CO producing pathways, this was most likely the result of H2 consumption by one or both of two putative NiFe-uptake hydrogenases encoded on the genome of P. thermoglucosidasius DSM 6285 [17]. The consumption of H2 by group 1d uptake hydrogenases is coupled with aerobic respiration or respiratory reduction of anaerobic oxidants like fumarate, NO3−, SO42− and CO2 [12]. Group 2a uptake hydrogenases are involved in aerobic respiration and recycling H2 produced through nitrogenase activity or fermentative pathways in aerobic soil and Cyanobacteria [12]. It is plausible that these enzymes couple H2 oxidation with an anaerobic electron transport chain, however, this requires further validation.

Despite showing accelerated commencement of H2 productivity, anaerobic cultivation greatly reduced biomass accumulation, inadvertently reducing H2 production potential. It can be reasoned that sustained active biomass accumulation and consequently O2 availability are important considerations for developing this system towards more continuous operation. It would also be beneficial to balance maximally anaerobic conditions to minimally disrupt or delay WGS reaction onset, but maintain O2 supply to sustain catalytically active biomass. As such we further explored the impacts, both temporally and volumetrically, of O2 perturbation of the P. thermoglucosidasius DSM 6285 WGS reaction system at different key phases.

O2 addition across the different phases of the P. thermoglucosidasius DSM 6285 fermentation negatively impacted H2 production at all stages. O2 addition prior to commencement of H2 production delayed H2 production to a greater extent than O2 addition during the anaerobic phase, likely by extending more energetically efficient aerobic respiration [32]. O2 addition in the anaerobic phase likely prompted resumption of aerobic respiration, although, with a more limited pool of reducing agents in higher oxidation states, required more rapid commitment to the WGS reaction for energy conservation, as observed for the WGS system in other taxa [13, 33,34,35,36].

Later O2 additions in the hydrogenogenic phase showed stark reductions in H2 production alongside strong and extended gains in biomass accumulation, indicative of redirection of P. thermoglucosidasius towards aerobic metabolism, which offers greater potential energy yields. Expectedly, increasing volume O2 additions saw more pronounced disruptions to H2 production, however H2 production resumed, albeit at lower rates, following O2 depletion. As CO availability was not a limiting factor, this was likely the combined result of the extended effects of O2 induced inhibition of the WGS components and the shift towards maximizing use of the oxidative respiratory chain [15, 35, 37, 38]. The Ni-CODH and group 4a NiFe-hydrogenase responsible for conducting the P. thermoglucosidasius DSM 6285 WGS are both classed as highly O2 sensitive [12, 15, 35], although the tolerance of the P. thermoglucosidasius WGS system to O2 disruptions have not previously been investigated. Here we observed the initial commencement of WGS activity in the presence of 2.49 ± 0.06% O2 in the gas headspace and resumption of H2 production following O2 disruption once O2 levels dropped below 0.97 ± 0.1%.

Although 1% O2 addition initially reduced H2 productivity between 38 and 41 h, there were only minor observable differences in productivity of the 1% O2 addition condition (0.0101 ± 0.0013 mmol h−1) compared to the control (0.0150 ± 0.0006 mmol h−1) over the peak period of H2 productivity (38–48 h). Productivity remained positive throughout, despite O2 addition, demonstrating a final H2 yield with no significant difference to the control. Conversely, reductions in H2 levels were observed in the 3–10% O2 additions, most likely due to the activity of one or both of the two putative O2-tolerant NiFe-uptake hydrogenases, which potentially couple H2 oxidation with aerobic respiration.

The lag in growth experienced post 5% and 10% O2 addition coincided with relatively low increases in CO2, despite rapid consumption of O2, likely the result of the sudden pronounced oxidative stress. This phenomenon has been observed previously in Klebsiella aerogenes and Escherichia coli, where similar stimuli resulted in apparent uncoupling between growth and energy conservation, resulting in a switch towards energetically wasteful oxidation pathways [39, 40]. In conjunction with the O2 susceptibility of the WGS machinery, this may have effected lower H2 production and yields following both temporally and volumetrically diverse O2 perturbations.

Large increases in acetate, formate and lactate were detected early in all fermentations (Figs. 1, 2, 3, 4), resulting in pH decreases. These compounds likely served as overflow metabolites, acting as electron sinks to reduce the backlog of NADH generated from CO disrupted aerobic metabolism, at the expense of energetic efficiency [23, 28, 41]. Subsequently, shortfalls in ATP generation were likely compensated by acetate kinase-mediated acetate production. Accumulation of these acids have also been attributed to the development of microoxic conditions developed in the media during O2 consumption [42, 43]. Subsequent consumption of lactate was likely through the putative LutA-C protein complex encoded by P. thermoglucosidasius, proteins associated with aerobic lactate utilisation in Bacillus subtilis [23, 44]. Presently the fate of the carbon and electrons derived from the lactate remains unclear. Notably, formate was almost completely consumed by the onset of the WGS. Formate consumption is associated with concomitant CO2 and H2 production, however, no concomitant H2 production was observed in this period, suggesting diversion of electrons towards another oxidant or electron transport chain. It remains unclear as to how formate was metabolised in these circumstances and represents an area for further investigation.

Strong acetate production in the aerobic phase resembled previous P. thermoglucosidasius fermentations and likely compensated for detraction of potential ATP production by lactate and formate production [23, 28]. Acetate production was likely also bolstered through mixed acids fermentation to the detriment of ethanol-directed fermentation, which aligns with the insubstantial levels of ethanol detected here and in a previous study [28]. Following O2 addition at all timepoints strong acetate consumption accompanied by slightly delayed growth responses, increasingly so for larger volume O2 additions. Notably, greater degrees of CO2 than H2 were evolved between O2 addition and concerted resumption of WGS activity. Furthermore, following O2 addition, increases in succinate were observed towards the tail-end of acetate consumption, where O2 levels were approaching depletion. Together these factors suggest that a portion of the acetyl-CoA derived from the acetate consumed may have integrated into the TCA cycle and been incorporated into biomass via the glyoxylate shunt, as suggested previously [23, 45] however confirmation of this requires further validation. Following O2 depletion, acetate consumption ceased, indicating a necessity for O2 for concerted acetate consumption in P. thermoglucosidasius. Further investigation into the metabolism of P. thermoglucosidasius under these conditions may further yield a greater understanding of the dynamics and interplay between the WGS reaction and alternative metabolic pathways.


Cultivation of P. thermoglucosidasius DSM 6285 under anaerobic synthetic syngas atmospheres induced faster onset of H2 production than previously identified ideal aerobic gas mixtures, however, this was at the expense of biomass accumulation. This in turn severely hampered H2 turnover. O2 supplementation at all stages of the fermentation hindered H2 production to greater extents accordingly with the volume of O2 supplied, however, also supported reinvigorated biomass accumulation. Furthermore, upon depletion of the supplemented O2, hydrogenogenic activity was rapidly resumed following a negligible lag period. Supplementation with 1% O2 negligibly impacted H2 production, however, still gave some support to growth. Notably, with elevated (10%) O2 supplementation, H2 consumption was observed briefly. This was potentially due to the two putative uptake hydrogenases encoded by P. thermoglucosidasius DSM 6285, which may present appealing targets for knockout mutagenesis. H2 conversion efficiency was also reduced accordingly with the level of O2 administered. The tolerance of the system to O2 perturbations may be linked with the metabolism of accumulated acetate, where O2 addition induces strong acetate consumption to support cell growth at all stages of the fermentation, including during the aerobic/microaerobic phase. It may be possible to incorporate this capacity for O2 induced acetate consumption and the relative flexibility of the P. thermoglucosidasius DSM 6285 WGS reaction to O2 perturbation to facilitate cyclical/fed batch hydrogenogenic fermentations on syngas at a larger scale. Additionally, focusing the strain towards concerted acetate production through inactivation of formate or lactate production pathways may facilitate concentrated acetate production to further supplement downstream acetate dependent growth.

Availability of data and materials

The dataset supporting the conclusions of this article is included within the article and its additional file.


  1. IRENA. Global energy transformation: a roadmap to 2050. Abu Dhabi, 2018.

  2. Elbeshbishy E, Dhar BR, Nakhla G, Lee HS. A critical review on inhibition of dark biohydrogen fermentation. Renew Sustain Energy Rev. 2017;79:656–68.

    Article  CAS  Google Scholar 

  3. Abdalla AM, Hossain S, Nisfindy OB, Azad AT, Dawood M, Azad AK. Hydrogen production, storage, transportation and key challenges with applications: a review. Energy Convers Manage. 2018;165:602–27.

    Article  CAS  Google Scholar 

  4. Hosseini SE, Wahid MA. Hydrogen production from renewable and sustainable energy resources: promising green energy carrier for clean development. Renew Sustain Energy Rev. 2016;57:850–66.

    Article  CAS  Google Scholar 

  5. van Niel EW. Biological processes for hydrogen production. In: Hatti-Kaul R, Mamo G, Mattiasson B, editors. Anaerobes in biotechnology. Springer: Cham; 2016. p. 155–93.

    Google Scholar 

  6. Alfano M, Cavazza C. The biologically mediated water–gas shift reaction: structure, function and biosynthesis of monofunctional [NiFe]-carbon monoxide dehydrogenases. Sustain Energy Fuels. 2018;2:1653–70.

    Article  CAS  Google Scholar 

  7. Fukuyama Y, Inoue M, Omae K, Yoshida T, Sako Y. Anaerobic and hydrogenogenic carbon monoxide-oxidizing prokaryotes: versatile microbial conversion of a toxic gas into an available energy. Adv Appl Microbiol. 2020;110:99–148.

    Article  CAS  PubMed  Google Scholar 

  8. Haddad M, Cimpoia R, Guiot SR. Performance of Carboxydothermus hydrogenoformans in a gas-lift reactor for syngas upgrading into hydrogen. Int J Hydrogen Energy. 2014;39:2543–8.

    Article  CAS  Google Scholar 

  9. Hosseini SS, Aghbashlo M, Tabatabaei M, Younesi H, Najafpour G. Exergy analysis of biohydrogen production from various carbon sources via anaerobic photosynthetic bacteria (Rhodospirillum rubrum). Energy. 2015;93:730–9.

    Article  CAS  Google Scholar 

  10. Mohr T, Aliyu H, Küchlin R, Polliack S, Zwick M, Neumann A, Cowan D, De Maayer P. CO-dependent hydrogen production by the facultative anaerobe Parageobacillus thermoglucosidasius. Microb Cell Fact. 2018;17:1–12.

    Article  Google Scholar 

  11. Lee SH, Kim MS, Kang SG, Lee HS. Biohydrogen production of obligate anaerobic archaeon Thermococcus onnurineus NA1 under oxic conditions via overexpression of frhAGB-encoding hydrogenase genes. Biotechnol Biofuels. 2019;12:1–8.

    Article  Google Scholar 

  12. Greening C, Biswas A, Carere CR, Jackson CJ, Taylor MC, Stott MB, Cook GM, Morales SE. Genomic and metagenomic surveys of hydrogenase distribution indicate H2 is a widely utilised energy source for microbial growth and survival. ISME J. 2016;10(3):761–77.

    Article  CAS  PubMed  Google Scholar 

  13. Schoelmerich MC, Müller V. Energy conservation by a hydrogenase-dependent chemiosmotic mechanism in an ancient metabolic pathway. Proc Natl Acad Sci. 2019;116(13):6329–34.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Adachi Y, Inoue M, Yoshida T, Sako Y. Genetic engineering of carbon monoxide-dependent hydrogen-producing machinery in Parageobacillus thermoglucosidasius. Microbes Environ. 2020;35(4):ME20101.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Merrouch M, Hadj-Saïd J, Domnik L, Dobbek H, Léger C, Dementin S, Fourmond V. O2 inhibition of Ni-containing CO dehydrogenase is partly reversible. Chem Eur J. 2015;21(52):18934–8.

    Article  CAS  PubMed  Google Scholar 

  16. Lu Y, Koo J. O2 sensitivity and H2 production activity of hydrogenases—a review. Biotechnol Bioeng. 2019;116:3124–35.

    Article  CAS  PubMed  Google Scholar 

  17. Mohr T, Aliyu H, Küchlin R, Zwick M, Cowan D, Neumann A, De Maayer P. Comparative genomic analysis of Parageobacillus thermoglucosidasius strains with distinct hydrogenogenic capacities. BMC Genomics. 2018;19:1–10.

    Article  Google Scholar 

  18. Aliyu H, Mohr T, Cowan D, De Maayer P, Neumann A. Time-course transcriptome of Parageobacillus thermoglucosidasius DSM 6285 grown in the presence of carbon monoxide and air. Int J Mol Sci. 2020;21(11):3870.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Imaura Y, Okamoto S, Hino T, Ogami Y, Katayama YA, Tanimura A, Inoue M, Kamikawa R, Yoshida T, Sako Y. Isolation, genomic sequence and physiological characterization of Parageobacillus sp. G301, an isolate capable of both hydrogenogenic and aerobic carbon monoxide oxidation. Appl Environ Microbiol. 2023;89(6):e00185-e223.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Mohr T, Aliyu H, Biebinger L, Gödert R, Hornberger A, Cowan D, De Maayer P, Neumann A. Effects of different operating parameters on hydrogen production by Parageobacillus thermoglucosidasius DSM 6285. AMB Express. 2019;9(1):1–10.

    Article  Google Scholar 

  21. Chiche D, Diverchy C, Lucquin AC, Porcheron F, Defoort F. Synthesis gas purification. Oil Gas Sci Technol-Revue d’IFP Energies Nouvelles. 2013;68:707–23.

    Article  CAS  Google Scholar 

  22. Ramachandriya KD, Kundiyana DK, Sharma AM, Kumar A, Atiyeh HK, Huhnke RL, Wilkins MR. Critical factors affecting the integration of biomass gasification and syngas fermentation technology. AIMS Bioeng. 2016;3(2):188–210.

    Article  CAS  Google Scholar 

  23. Aliyu H, Kastner R, Maayer PD, Neumann A. Carbon monoxide induced metabolic shift in the carboxydotrophic Parageobacillus thermoglucosidasius DSM 6285. Microorganisms. 2021;9(5):1090.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, Grolemund G, Hayes A, Henry L, Hester J, Kuhn M. Welcome to the Tidyverse. J Open Source Softw. 2019;4(43):1686.

    Article  Google Scholar 

  25. R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing. 2023.

  26. Neumann A, Dörsam S, Oswald F, Ochsenreither K. Microbial production of value-added chemicals from pyrolysis oil and syngas. In: Xian M, editor. Sustainable production of bulk chemicals: integration of bio-, chemo-resources and processes. Springer: Dordrecht; 2015. p. 69–105.

    Google Scholar 

  27. Cripps RE, Eley K, Leak DJ, Rudd B, Taylor M, Todd M, Boakes S, Martin S, Atkinson T. Metabolic engineering of Geobacillus thermoglucosidasius for high yield ethanol production. Metab Eng. 2009;11(6):398–408.

    Article  CAS  PubMed  Google Scholar 

  28. Tang YJ, Sapra R, Joyner D, Hazen TC, Myers S, Reichmuth D, Blanch H, Keasling JD. Analysis of metabolic pathways and fluxes in a newly discovered thermophilic and ethanol-tolerant Geobacillus strain. Biotechnol Bioeng. 2009;102(5):1377–86.

    Article  CAS  PubMed  Google Scholar 

  29. Najar IN, Thakur N. A systematic review of the genera Geobacillus and Parageobacillus: their evolution, current taxonomic status and major applications. Microbiology. 2020;166(9):800–16.

    Article  CAS  PubMed  Google Scholar 

  30. Liu J, Han X, Tao F, Xu P. Reprogramming a high robust Geobacillus thermoglucosidasius for efficient synthesis of polymer-grade lactic acid under extremely high temperature (60℃). bioRxiv. 2023:2023–04.

  31. Allaart MT, Diender M, Sousa DZ, Kleerebezem R. Overflow metabolism at the thermodynamic limit of life: How carboxydotrophic acetogens mitigate carbon monoxide toxicity. Microb Biotechnol. 2023;16(4):697–705.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Unden G. Transcriptional regulation and energetics of alternative respiratory pathways in facultatively anaerobic bacteria. Biochim Biophys Acta BBA Bioenerget. 1998;1365(1–2):220–4.

    Article  CAS  Google Scholar 

  33. Maness PC, Huang J, Smolinski S, Tek V, Vanzin G. Energy generation from the CO oxidation-hydrogen production pathway in Rubrivivax gelatinosus. Appl Environ Microbiol. 2005;71(6):2870–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Pisa KY, Huber H, Thomm M, Müller V. A sodium ion-dependent A1AO ATP synthase from the hyperthermophilic archaeon Pyrococcus furiosus. FEBS J. 2007;274(15):3928–38.

    Article  CAS  PubMed  Google Scholar 

  35. Vignais PM, Billoud B. Occurrence, classification, and biological function of hydrogenases: an overview. Chem Rev. 2007;107(10):4206–72.

    Article  CAS  PubMed  Google Scholar 

  36. Welte C, Krätzer C, Deppenmeier U. Involvement of Ech hydrogenase in energy conservation of Methanosarcina mazei. FEBS J. 2010;277(16):3396–403.

    Article  CAS  PubMed  Google Scholar 

  37. Fernandez VM, Hatchikian EC, Cammack R. Properties and reactivation of two different deactivated forms of Desulfovibrio gigas hydrogenase. Biochim Biophys Acta BBA Protein Struct Mol Enzymol. 1995;832(1):69–79.

    Google Scholar 

  38. Kim SM, Lee J, Kang SH, Heo Y, Yoon HJ, Hahn JS, Lee HH, Kim YH. O2-tolerant CO dehydrogenase via tunnel redesign for the removal of CO from industrial flue gas. Nat Catal. 2022;5(9):807–17.

    Article  CAS  Google Scholar 

  39. Harrison D, Loveless J. Transient responses of facultatively anaerobic bacteria growing in chemostat culture to a change from anaerobic to aerobic conditions. Microbiology. 1971;68(1):45–52.

    CAS  Google Scholar 

  40. Partridge JD, Scott C, Tang Y, Poole RK, Green J. Escherichia coli transcriptome dynamics during the transition from anaerobic to aerobic conditions. J Biol Chem. 2006;281(38):27806–15.

    Article  CAS  PubMed  Google Scholar 

  41. Campbell K, Herrera-Dominguez L, Correia-Melo C, Zelezniak A, Ralser M. Biochemical principles enabling metabolic cooperativity and phenotypic heterogeneity at the single cell level. Curr Opin Syst Biol. 2018;8:97–108.

    Article  Google Scholar 

  42. Liang J, Roberts A, van Kranenburg R, Bolhuis A, Leak DJ. Relaxed control of sugar utilization in Parageobacillus thermoglucosidasius DSM 2542. Microbiol Res. 2022;256: 126957.

    Article  CAS  PubMed  Google Scholar 

  43. Metcalfe GD, Sargent F, Hippler M. Hydrogen production in the presence of oxygen by Escherichia coli K-12. Microbiology. 2022;168(3): 001167.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Chai Y, Kolter R, Losick R. A widely conserved gene cluster required for lactate utilization in Bacillus subtilis and its involvement in biofilm formation. J Bacteriol. 2009;191:2423–30.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Wolfe AJ. The acetate switch. Microbiol Mol Biol Rev. 2005;69(1):12–50.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references


Open Access funding enabled and organized by Projekt DEAL. MM received funding from the South African CSIR-DSI Interprogramme Bursary scheme. This research was funded by Bundesministerium für Bildung und Forschung (BMBF) Bioökonomie International (Grant No. 031B1056).

Author information

Authors and Affiliations



MM, MSA, HA, AN and PDM planned the experiments. MM and MSA collected and analysed the experimental data. MM and BAM conducted the statistical analysis and MM prepared all figures, tables and additional files. MM, MSA, HA, AN and PDM drafted the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Anke Neumann or Pieter de Maayer.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Table S1.

Significant differences observed between different initial gas headspace compositions. Table S2. Significant differences observed between CO/H2 yields. Table S3. Significant differences observed between pre-hydrogenogenic phase O2 addition conditions. Table S4. Significant differences observed between hydrogenogenic phase O2 addition conditions. Table S5. Significant differences in H2 levels observed between sampling points. Table S6. Significant differences observed between variable O2 additions during peak hydrogenogenic period.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Mol, M., Ardila, M.S., Mol, B.A. et al. The effects of synthesis gas feedstocks and oxygen perturbation on hydrogen production by Parageobacillus thermoglucosidasius. Microb Cell Fact 23, 125 (2024).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: