In preparation for routine microbiology and molecular biology experiments, such as cloning, strains were grown on LB-agar plates or in LB-broth at 37°C (Miller, 1972). Anaerobic growth for Hyd enzyme assays and enzyme activity-staining after native polyacrylamide gel electrophoresis (PAGE) was performed at 37°C as standing liquid cultures in the buffered rich medium TGYEP (1% w/v tryptone, 0.5% w/v yeast extract, 0.8% w/v glucose, 100 mM potassium phosphate, pH 6.5) (Begg et al., 1977). The growth medium was supplemented with trace element solution SLA (Hormann and Andreesen, 1989). When required, the antibiotics ampicillin, chloramphenicol, or kanamycin were added to a final concentration of 100, 25, or 50 μg ml^–1, respectively. Cells were harvested anaerobically when cultures had reached an OD₆₀₀ nm of between 0.8 and 1.2 by centrifugation at 5,000 g for 15 min at 4°C. Cell pellets were either used immediately or stored at –20°C until use.

For anaerobic protein overproduction experiments, E. coli strain SHH228 (ΔhypC ΔhybG) was transformed with the indicated plasmids using standard procedures (Sambrook et al., 1989). Cultivation of cells was performed in modified TB medium (2.4% w/v yeast extract, 1.2% w/v peptone from casein, 0.04% w/v glycerol, 0.4% w/v glucose and 0.003% w/v magnesium sulfate heptahydrate) (Soboh et al., 2012). Depending on the plasmid used, the medium also included either 100 μg ml^–1 ampicillin or 15 μg ml^–1 chloramphenicol to maintain plasmid selection. Cultures were incubated anaerobically without shaking at 37°C until an optical density at 600 nm (OD₆₀₀) of 0.4 was reached. To increase the amount of protein for purification purposes, the expression of the plasmid pASK-IBA3-borne hypC and hybG genes was induced by the addition of 0.2 μg ml^–1 anhydrotetracycline (AHT), whereas for pT7-7- and pACYC-Duet1-based plasmids, gene expression was induced by the addition of 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). Incubation of the cultures was continued at 30°C for 3 h, after which the cells were harvested by centrifugation at 5000 × g for 15 min at 4°C. The filling of bottles and tubes for the centrifugation of cultures and cell suspensions was performed under anoxic conditions in an anaerobic chamber (Coy Laboratories, Grass Lake, United States). Cell pellets, derived by centrifugation, were either used immediately or stored at – 20°C until use.

Two methods were used to determine H₂ production. One involved measuring cumulative H₂ production after anaerobic cultivation of strains, while the other determined continuous H₂ production in cell suspensions. Cumulative H₂ content was determined by growing strains in 15 ml Hungate tubes (initially filled with N₂) containing 8 ml of culture medium. The cultures were incubated for 20 h at 30°C and the H₂ concentration was measured by removing 200 μl aliquots from the headspace and analyzing the gas-phase using gas chromatography with a GC2010 Plus Gas Chromatograph (Shimadzu, Kyõto, Japan) as described (Pinske et al., 2015). Pure nitrogen was used as the carrier gas, and the amount of H₂ produced was calculated based on a standard curve prepared with pure H₂ gas. The experiment was repeated three times and each assay was performed in triplicate.

Continuous H₂ production by whole cells was determined using a modified Clark-type electrode equipped with an OXY/ECU module (Oxytherm, Hansatech Instruments, Norfolk, United Kingdom) to reverse the polarizing voltage to –0.7 V, essentially as described (Lindenstrauß et al., 2017). Cells were grown as described above, but only until the late-exponential phase was attained, and after the anaerobic centrifugation of cells to remove culture medium, the cell pellet was suspended in degassed 50 mM Tris, pH 7.0 and the centrifugation step was repeated. Subsequently, the cell pellet was suspended in 1 ml of degassed 50 mM Tris, pH 7.0 and 50 μl aliquots were added to the chamber of the electrode, which contained 1.95 ml of degassed 50 mM Tris, pH 7.0, equilibrated at 30°C. The reaction was started by adding 14 mM glucose, which was converted to formate intracellularly to act as a substrate of the FHL reaction and the amount of H₂ produced was determined using pure H₂ gas as described (Sargent et al., 1999). The assay was performed in triplicate using three biological replicates for each strain analyzed.

Cell paste was suspended in 2 ml of 50 mM MOPS, pH 7, including 5 μg DNase I ml^–1 and 0.2 mM phenylmethylsulfonyl fluoride per 1 g wet weight. Cells were disrupted by sonication (20 W for 2 min with 0.5 s pulses). Cell debris and unbroken cells were removed by centrifugation for 20 min at 21,000 × g at 4°C. The supernatant (crude extract) was carefully decanted into a fresh tube and was used immediately. Protein concentration was determined as described before (Lowry et al., 1951).

The total Hyd enzyme activity of the crude extracts was determined as H₂-dependent reduction of benzyl viologen (BV) as described (Ballantine and Boxer, 1985), except that the buffer used was 50 mM MOPS, pH 7.0. The wavelength used for the absorbance measurement was 600 nm and an ε_M value of 7,400 M^–1 cm^–1 was assumed for reduced BV. One unit of enzyme activity corresponded to the reduction of 1 μmol of substrate min^–1. Enzyme assays were performed in triplicate using three biological replicates.

Non-denaturing PAGE was performed according to Ballantine and Boxer (1985) using crude extracts (25 μg of protein). Prior to application onto the gel, crude extracts were incubated with a final concentration of 4% (v/v) Triton X-100 at 4°C for 15 min. Separating gels included 7.5% (w/v) polyacrylamide and 0.1% (w/v) Triton X-100. To visualize the activity of Hyd-1, Hyd-2, and Hyd-3, activity-staining after native PAGE was performed according to Pinske et al. (2012) using 50 mM MOPS, pH 7 buffer, which included 0.5 mM BV and 1 mM 2,3,5-triphenyltetrazolium chloride. Gels were incubated overnight at 25°C in an atmosphere of 9% N₂: 5% H₂. Experiments were repeated several times with the same results and a representative gel is shown.

All steps for protein purification were carried out in an anaerobic chamber (Coy Laboratories, Grass Lake, United States). Wet cell paste was suspended in 2 ml buffer W (50 mM Tris, pH 8, containing 150 mM NaCl) per 1 g cell paste. Phenylmethylsulfonyl fluoride (PMSF) was added to a final concentration of 0.8 mM and DNAse I to a final concentration of 10 μg/ml to the cell suspension. Cells were disrupted by sonication (Sonotrode, 35 W with 0.5 s pulses for 5 min) on ice. Cell debris and unbroken cells were removed by centrifugation at 21,000 × g for 20 min at 4°C. The supernatant obtained after centrifugation was used immediately for anaerobic protein purification. Strep-tag-II-tagged HypC and HybG were purified individually or in complex with HypD, using Strep-tactin sepharoseXT Sepharose (IBA Lifesciences, Göttingen), exactly as described (Soboh et al., 2013; Arlt et al., 2021). His-tagged pre-HycE was purified using cobalt-charged TALON Superflow agarose (Cytiva), following the manufacturer’s instructions.

Purified proteins were buffer-exchanged into anaerobic 50 mM Tris, pH 8, using 5 ml PD-10 columns containing G-25 matrix (Cytiva). The resulting protein samples were concentrated using Amicon centrifugal concentration filters (cut-off of 5 kDa for HypC and HybG proteins and 50 kDa for pre-HycE samples). Purified protein samples were stored at –80°C.

To examine the interaction between Strep-tagged HybG_WT, HybG_H52A, HypC_WT, or HypC_H51A with His-tagged pre-HycE 150 μg of each protein (5:1 mol excess of chaperone) was mixed and incubated at 30°C under anoxic conditions for 2 h. After incubation, the mixture was loaded onto either a 0.5 ml Strep-tactin sepharoseXT column or a 0.5 ml cobalt-charged TALON Superflow agarose column for the enrichment of the interaction partners. Columns were pre-equilibrated with buffer W (50 mM Tris/HCl, pH 8, containing 150 mM NaCl). After loading of sample, the columns were washed with 10 column volumes of buffer W to remove unbound proteins. Bound proteins were subsequently eluted with buffer W containing 50 mM biotin (Strep-tactin sepharoseXT columns), or with buffer W containing 300 mM imidazole (cobalt-charged NTA columns). Fractions of 0.5 ml were collected and aliquots from these were analyzed by electrophoresis on 12.5% (w/v) or 1% (w/v) denaturing polyacrylamide sodium dodecylsulfate (SDS)-PAGE (Laemmli, 1970). After the separation of polypeptides, they were transferred onto a nitrocellulose membrane as described (Towbin et al., 1979). After blocking the membrane, interaction partners were identified by challenging with polyclonal antiserum raised against HypC, HybG, or HycE (Pinske et al., 2015; Arlt et al., 2021). The detection was based on chemiluminescence using the Immuno-detection kit SuperSignal West Pico PLUS (Thermo Scientific, Brunswick, Germany) and an imager Amersham Imager 600 (GE Healthcare Bio-Sciences AB, Solingen, Germany).

The HypC chaperone preferentially matures pre-HycE, the Hyd-3 large subunit precursor, while its paralogue HybG preferentially introduces the Fe(CN)₂CO group of the [NiFe]-cofactor into pre-HyaB and pre-HybC, the respective precursors of Hyd-1 and Hyd-2 (Blokesch et al., 2001; Arlt et al., 2021). As the aim of this study was to determine the significance of the conserved His residue in HypC and HybG for their function, we decided to exchange the large, charged histidine residue in both proteins for a small, non-polar alanine residue. First, a series of eight plasmids carrying either the native hypC gene, the native hybG gene, or carrying hypC + hypD, or hybG + hypD together, was constructed (Table 1). Derivatives of these four plasmids were also constructed, in which codon 51 in hypC and codon 52 in hybG were mutated to decode as an alanine residue (see section “Materials and Methods”). It is important to note that in all experiments described in the current study, HypC and HybG both carried a C-terminal Strep-tag II. The presence of this StrepII-tag does not interfere with the functionality of either protein with respect to Hyd precursor maturation (Blokesch et al., 2004; Soboh et al., 2013; Thomas et al., 2018). Consequently, in the interest of convenience, we will henceforth generally refer to these proteins without mentioning the tag.

As the first experiment to examine the potential effects of exchanging the conserved histidine residues in HypC and HybG to alanine on the maturation of the Hyd-1, -2, and -3 enzymes, we first determined the total H₂-oxidizing Hyd enzyme activity (Figure 2A). Therefore, strain SHH228 (see Table 1), lacking genomic copies of both hypC and hybG, but retaining hybD (Hartwig et al., 2015), was transformed with each of the eight plasmids individually. After the anaerobic growth of the strains and preparation of crude extracts (see section “Materials and Methods” for details), the total Hyd enzyme activity was determined for each (Figure 2A). An extract derived from the parental strain MC4100 had a total Hyd activity of approximately 1.5 U mg^–1 and served as a positive control, while an extract derived from SHH228 (hypC hybG) had no detectable activity and acted as the negative control (Figure 2A). The introduction of plasmid phypCstrep carrying the parental hypC gene restored approximately 50% of the parental total Hyd enzyme activity to the mutant. This is consistent with HypC being required for the maturation of Hyd-3 and with the fact that under these growth conditions, Hyd-3 constitutes the bulk of the total Hyd activity (Sawers et al., 1985; Pinske et al., 2011). In contrast, the isogenic plasmid phypC(H51A)strep with a mutation in codon 51 of hypC only restored the total Hyd activity to a level that was less than 5% of that measured for the positive control MC4100 (Figure 2A). This indicates that the H51A amino acid exchange in HypC severely compromised its ability to function in maturation of Hyd-3, which is also consistent with the reduced amount of Fe(CN)₂CO group detected after purification of HypD associated with an HypC_H51R variant reported earlier (Soboh et al., 2013).

As the hypC gene is typically found located adjacent to hypD in the genomes of microorganisms that synthesize [NiFe]-Hyd, and because HypC and HypD form the central scaffold complex during the maturation of these enzymes (Böck et al., 2006; Lacasse and Zamble, 2016), we wished to determine whether their multicopy co-expression might rescue the Hyd-deficient phenotype exhibited by SHH228 synthesizing HypC_H51A, despite this strain already possessing a genomic copy of hypD. Therefore, we determined the total Hyd enzyme activity of the strain co-expressing both genes from the same plasmid. When the native hypC gene was co-expressed with hypD, a similar total Hyd enzyme activity was determined compared to when hypC was expressed alone from plasmid phypCstrep (Figure 2A). When the same experiment was repeated using a plasmid co-expressing hypD (pT-hypDCstrep in Table 1) and the mutated hypC gene synthesizing HypC_H51A [pT-hypDC(H51A)strep, Table 1], the total Hyd enzyme activity was recovered to a level close to that measured for the wild-type strain. This result suggests that HypC_H51A is still functional in the maturation of Hyd-3, but is less efficient at completing maturation unless HypD and the HypC proteins are co-over-produced.

The hybG gene, encoding HybG, is located within the hyb operon (Menon et al., 1994) and does not have its own associated hypD gene. The introduction of a plasmid carrying only hybG, encoding Strep-tagged native HybG, into strain SHH228 (ΔhypC ΔhybG) failed to result in restoration of wild-type total Hyd activity to the mutant and activity was barely detectable (Figure 2A). It should be noted that the percentage contribution of Hyd-1 plus Hyd-2 activity to the total H₂-oxidizing Hyd activity under fermentative conditions is only between 10–20% (Sawers et al., 1985; Pinske et al., 2011). This does not, however, explain the poor complementation achieved by re-introducing plasmid-borne native hybG. Therefore, to determine whether the co-expression of hybG with hypD might improve complementation and increase H₂-oxidizing Hyd activity, plasmid pT-hybG-hypDEF was tested. After fermentative growth, crude extracts derived from SHH228/pT-hybG-hypDEF revealed an increase in the total Hyd activity to around 10% of the total activity measured for the wild-type MC4100 (Figure 2A), which approximately represents an activity consistent with that expected for Hyd-1 plus Hyd-2 under these growth conditions (Pinske et al., 2011). Note that plasmid pT-hybG-hypDEF also carries the hypE and hypF genes. The presence of these genes does not affect the total Hyd enzyme activity determined compared to when only hybG + hypD or only hypC + hypD is present on the plasmid (Haase and Sawers, unpublished results).

The anticipated transformation of SHH288 with plasmid phybG(H52A)strep, which has a mutation in codon 52 in the hybG gene resulting in exchange of histidine for alanine and delivering HybG_H52A, failed to yield measurable Hyd activity (Figure 2A). Surprisingly, however, when hypD was co-expressed with this mutated hybG gene on pT-hybG(H52A)-hypDEF, the total Hyd activity measured was similar to that measured for hypC-hypD on pT-hypDCStrep in SHH228 (Figure 2A). This result suggests that this approximate 10-fold increase in total Hyd activity, relative to what was measured when pT-hybG-hypDEF was introduced into SHH228, was either due to an unexpected increase in the combined Hyd-1 and Hyd-2 activity, or because Hyd-3 activity was activated by the variant HybG_H52A protein when it was co-over-produced with HypD.

In order to resolve this issue, we first tested the same set of strains for their ability to produce H₂ during glucose fermentation (Figures 2B,C). The amount of H₂ accumulated in stationary-phase cells after batch cultivation revealed that the positive control MC4100 accumulated approximately 25 μmol H₂ OD₆₀₀ nm^–1, while the negative control SHH228 (ΔhypC ΔhybG) accumulated no detectable H₂ gas (Figure 2B). SHH228 transformed with pT-hybG(H52A)-hypDEF (synthesizing HypD plus HybG_H52A) accumulated H₂ to a level that was slightly more than that of the parental MC4100 strain, indicating that Hyd-3 was active. The same experiment performed with a plasmid bearing hypD and the parental hybG genes (pT-hybG-hypDEF) showed essentially no H₂ accumulation, as did SHH228 transformed with plasmids carrying only the parental hybG or only the mutated hybG genes lacking the additional hypD gene (Figure 2B). This supports the conclusion that the mutant HybG_H52A chaperone, synthesized at a high level together with HypD, was capable of maturing Hyd-3 and thus accounted for the H₂ production by the strain. As further controls, when SHH228 co-over-produced either HypC or HypC_H51A together with HypD, H₂ also accumulated to wild-type levels (Figure 2B). However, when the strain only synthesized HypC_H51A, H₂ accumulated to approximately 50% of parental levels.

To verify these results, the ability of the same set of strains to evolve H₂ in cell suspensions derived from exponentially grown cells was assessed using a hydrogen-electrode (Figure 2C). The results essentially reflected those obtained by measuring cumulative H₂ production, with the exception that SHH228 synthesizing HypC_H51A from phypC(H51A)strep had an H₂-evolving activity that was only 10% of that of the wild-type strain MC4100 (Figure 2C). These results confirmed the poor complementation of the Hyd -deficient phenotype exhibited by this strain when the total Hyd activity was measured in extracts (compare Figure 2A). It is likely that in the cumulative H₂ assay, the cells had sufficient time to accumulate H₂ in the stationary phase, which probably accounts for the difference when the two assay methods for H₂ production by the strain are compared (compare Figures 2B,C).

Bands corresponding to Hyd-1, Hyd-2, and Hyd-3 can be readily distinguished after native PAGE followed by staining the gel specifically for Hyd enzyme activity (Pinske et al., 2012). Moreover, this technique also allows the identification of an H₂:benzyl viologen oxidoreductase activity associated with the formate dehydrogenases (Fdh) N and O, which is a side-reaction of these enzymes (Soboh et al., 2011), but is a useful control for these experiments. As HybG typically cannot facilitate the maturation of Hyd-3 (Blokesch et al., 2001; Böck et al., 2006), we first looked at the extracts derived from both exponential-phase (Figure 3A) and stationary-phase cells (Figure 3B) for evidence of HybG-dependent synthesis of active Hyd-3. Regardless of whether the native hybG gene was expressed alone, or co-expressed with hypD from a plasmid, no manifestation of active Hyd-3 after native PAGE and staining for Hyd enzyme activity could be observed (Figure 3). In contrast, when HybG_H52A was co-synthesized with HypD in strain SHH288 (ΔhypC ΔhybG), and in the absence of any HypC, Hyd-3 activity could be visualized (Figure 3B). Although the activity band was relatively weak, it was clearly visible, especially in the crude extract derived from stationary-phase cells. Moreover, the activity band had an intensity comparable to that in exponential-phase cells when the native hypC gene was expressed on its own from plasmid phypCstrep (Figure 3A). This result indicates that HybG_H52A can mature the large subunit precursor (pre-HycE) of Hyd-3.

Hydrogenase-1 is more active in anaerobic stationary-phase cells, while Hyd-2 is more active in exponential-phase cells (Ballantine and Boxer, 1985; Sawers et al., 1985). The examination of the activity-stained bands in the other extracts revealed that, although Strep-tagged native HybG was capable of restoring maturation of both Hyd-1 and Hyd-2 in exponential- and stationary-phase cells (Figures 3A,B), the intensity of the respective activity bands was low. This possibly explains the poor phenotypic complementation by plasmid-borne hybG when introduced into SHH228 and the very low total Hyd activity measured in Figure 2A. Strain SHH228, synthesizing the HybG_H52A variant, showed no activity bands corresponding to either Hyd-1 or Hyd-2 when the cognate gene was expressed on its own from plasmid phybG(H52A)strep (Figures 3A,B). Notably, however, when either hybG allele was co-expressed with hypD, extracts derived from the corresponding cells revealed wild-type levels of activity-staining bands for Hyd-1 and Hyd-2 (Figure 3). These results suggest that the co-synthesis of HypD either resulted in the stabilization of the respective HybG chaperones when the cognate genes were co-expressed or facilitated interaction of the proteins to allow scaffold complex formation, thus improving the efficiency of maturation of both the large-subunit precursors, pre-HyaB and pre-HybC.

In this regard, crude extracts derived from stationary-phase cells of SHH288/phypCstrep, which synthesized Strep-tagged native HypC, revealed only a weak activity band that migrated at the position of Hyd-1, but no such similar activity band was observed in an extract derived from SHH288 co-expressing hypC and hypD (Figure 3B); the levels of Hyd-3 remained similar for both strains, providing an internal loading control. This underscores the preferential maturation of pre-HycE over pre-HyaB by HypC carrying Fe(CN)₂CO (see also Figure 1A).

HypC-HypD and HybG-HypD complexes can be readily isolated from anaerobically cultivated cells (Blokesch et al., 2004; Soboh et al., 2012) and it has been shown using structural analyses (Watanabe et al., 2012) and by native mass spectrometry (Arlt et al., 2021) that both sets of complexes form (1:1) heterodimers. To determine whether the exchange of the conserved histidine residue that is predicted to be important for the interaction with HypD (see Figure 1A; Watanabe et al., 2012; Miki et al., 2020) has an impact on this interaction, we enriched, by affinity chromatography in a single step HybG-HypD, HybG_H52A-HypD, HypC-HypD, and HypC_H51A-HypD complexes from anaerobically grown SHH228 cells. To do this, we took advantage of the StrepII-tag on the chaperones (see section “Materials and Methods”). The resulting complexes were separated by SDS-PAGE and visualized using Coomassie Blue staining (Figure 4). While the native HybG and HypC proteins each could be enriched together with high amounts of HypD and minimal contaminating polypeptides, the mutated chaperone proteins clearly interacted more poorly with HypD. The complexes formed were less-well resolved, the affinity-enriched samples included considerably more contaminating proteins, and, based on densitometric analysis (ImageQuant TL program Cytiva), the apparent stoichiometries were significantly lower than those observed for the respective native chaperone (at least 50% lower for HybG_H52A:HypD and ∼85% lower for HypC_H51A:HypD).

Next, we purified separately Strep-tagged native HybG and HybG_H52A, as well as a His-tagged derivative of pre-HycE, the precursor of the Hyd-3 large subunit. These were then used in interactions experiments (see section “Materials and Methods”) to test whether HybG_H52A could form a complex with pre-HycE (Figure 5). The results show that when complexes were allowed to form between His-tagged pre-HycE and Strep-tagged native HybG and the mixture was subsequently separated on a cobalt-charged TALON Superflow agarose column, only low amounts of HybG co-eluted with pre-HycE (Figure 5A). When the same mixture was passed over a Strep-tactin sepharose column, no pre-HycE could be identified to co-elute with Strep-tagged HybG (Figure 5B). In contrast, when the same experiment was repeated with Strep-tagged HybG_H52A, more of the chaperone was shown to co-elute with His-tagged pre-HycE after cobalt-charged TALON agarose chromatography (Figure 5C), and pre-HycE could clearly be identified to co-elute with Strep-tagged HybG_H52A after Strep-tactin sepharose affinity chromatography (Figure 5D).

In a similar experiment performed with Strep-tagged HypC, Strep-tagged HypC_H51A, and His-tagged pre-HycE, native HypC co-eluted strongly with His-tagged pre-HycE (Figure 5E), while HypC_H51A interacted more poorly with pre-HycE (Figure 5F).