Distinct Physiological Roles of the Three [NiFe]-Hydrogenase Orthologs in the Hyperthermophilic Archaeon Thermococcus kodakarensis

Abstract

Hydrogenases catalyze the reversible oxidation of molecular hydrogen (H₂) and play a key role in the energy metabolism of microorganisms in anaerobic environments. The hyperthermophilic archaeon Thermococcus kodakarensis KOD1, which assimilates organic carbon coupled with the reduction of elemental sulfur (S⁰) or H₂ generation, harbors three gene operons encoding [NiFe]-hydrogenase orthologs, namely, Hyh, Mbh, and Mbx. In order to elucidate their functions in vivo, a gene disruption mutant for each [NiFe]-hydrogenase ortholog was constructed. The Hyh-deficient mutant (PHY1) grew well under both H₂S- and H₂-evolving conditions. H₂S generation in PHY1 was equivalent to that of the host strain, and H₂ generation was higher in PHY1, suggesting that Hyh functions in the direction of H₂ uptake in T. kodakarensis under these conditions. Analyses of culture metabolites suggested that significant amounts of NADPH produced by Hyh are used for alanine production through glutamate dehydrogenase and alanine aminotransferase. On the other hand, the Mbh-deficient mutant (MHD1) showed no growth under H₂-evolving conditions. This fact, as well as the impaired H₂ generation activity in MHD1, indicated that Mbh is mainly responsible for H₂ evolution. The copresence of Hyh and Mbh raised the possibility of intraspecies H₂ transfer (i.e., H₂ evolved by Mbh is reoxidized by Hyh) in this archaeon. In contrast, the Mbx-deficient mutant (MXD1) showed a decreased growth rate only under H₂S-evolving conditions and exhibited a lower H₂S generation activity, indicating the involvement of Mbx in the S⁰ reduction process. This study provides important genetic evidence for understanding the physiological roles of hydrogenase orthologs in the Thermococcales.

INTRODUCTION

Hydrogenase catalyzes the conversion of H₂ to or from protons and electrons (H₂ = 2H⁺ + 2e⁻). Hydrogenase enables an organism either to utilize H₂ as a source of reducing power or to use protons as a terminal electron acceptor, evolving H₂. Hydrogenases can be divided into two major classes, the [NiFe]-hydrogenases and the [FeFe]-hydrogenases (49), along with the recently identified [Fe]-hydrogenases found in some methanogens (46). The catalytic core of [NiFe]-hydrogenase is a heterodimeric complex with a large subunit containing a binuclear [NiFe] active site and a small subunit bearing one or more Fe-S clusters. On the other hand, [FeFe]-hydrogenase harbors a catalytic center consisting of a binuclear iron-iron core and a 4Fe-4S cluster (designated the H cluster) (34), while [Fe]-hydrogenase harbors a mononuclear iron center (46).

The euryarchaeal order Thermococcales, composed of two major genera, Thermococcus and Pyrococcus, comprises hyperthermophiles living at temperatures close to 100°C (11, 12). Members of the Thermococcales are anaerobic heterotrophs utilizing proteinaceous substrates with the reduction of elemental sulfur (S⁰), forming H₂S. In the absence of S⁰, some of these organisms can grow when pyruvate or carbohydrates such as starch are added. In the latter metabolism, H₂ is evolved as a result of proton utilization as an electron acceptor, suggesting the presence of an H₂-evolving hydrogenase(s) in these organisms.

The first hydrogenase characterized in the Thermococcales was a cytosolic [NiFe]-hydrogenase of Pyrococcus furiosus (Hyh-I) (2). Hyh-I is an NADP(H)-dependent enzyme composed of four subunits (25). Since Hyh-I also exhibits a sulfur (or polysulfide) reductase activity (26, 27), it is also referred to as a sulfhydrogenase (26). Later, Hyh-II, an enzyme paralog of Hyh-I, was found in the cytoplasm of P. furiosus (25, 27). The two Hyhs are very similar in their molecular and enzymatic properties, except that Hyh-II is approximately an order of magnitude less active than Hyh-I. Because P. furiosus is known to evolve H₂ as a metabolite (6, 29, 42), the physiological role of the two Hyhs was suggested to be the disposal of excess reducing equivalents in the form of H₂ (25).

In 2000, two groups independently identified another [NiFe]-hydrogenase localized in the plasma membrane of P. furiosus (Mbh) (38, 44). Using similar detergent extraction procedures, both groups obtained protein fractions with high H₂ evolution activities, from which two proteins, MbhK and MbhL, were identified. Genes encoding the two polypeptides were found in a single operon composed of 14 open reading frames (ORFs) (mbh operon; mbhABCDEFGHIJKLMN). The products of the first eight genes (mbhABCDEFGH) display similarity to the subunits of multimeric Na⁺/H⁺ antiporters and are probably involved in H⁺ and Na⁺ translocation across the membrane. The latter five ORFs (mbhJKLMN) encode proteins showing high sequence similarity with the membrane-associated subunits of Ech hydrogenase of Methanosarcina barkeri (21, 31) and CO-induced hydrogenase (Coo) of Rhodospirillum rubrum (7). In a later study, in vitro experiments using inverted membrane vesicles of P. furiosus suggested that Mbh functions as a redox-driven ion pump generating a proton motive force by means of reduction of protons, thereby generating H₂, using a low-potential ferredoxin (37). The proton motive force generated is used for ATP synthesis driven by the membrane-bound ATP synthase complex (proton respiration). The results show that Mbh functions for energy conservation in P. furiosus as an energy-converting [NiFe]-hydrogenase. Mbh represents a family of multisubunit membrane-bound hydrogenases, the energy-converting [NiFe]-hydrogenases, which show similarity to the catalytic core of energy-conserving NADH:quinone oxidoreductases (complex I) (9, 49).

In addition to Hyh and Mbh, the P. furiosus genome contains another putative [NiFe]-hydrogenase operon (mbx operon) (44). The mbx operon (mbxABCDFGHH′MJKLN) is structurally similar to the mbh operon: every subunit in the mbx operon has a corresponding counterpart in the mbh operon. However, whether Mbx exhibits hydrogenase activity is uncertain, as (i) four cysteine residues that coordinate Ni and Fe metals in the active center of the [NiFe]-hydrogenase large subunit are not fully conserved in MbxL, a homolog of the large subunit encoded by the mbx operon, and (ii) fractionation of the P. furiosus plasma membrane showing H₂-producing activity failed to identify a fraction containing Mbx subunits (44).

The three types of [NiFe]-hydrogenase orthologs (Hyh, Mbh, and Mbx) found in P. furiosus are commonly distributed in the other completely sequenced genomes of Thermococcales, i.e., Pyrococcus horikoshii (16), Pyrococcus abyssi (4), Thermococcus kodakarensis (8), Thermococcus onnurineus (22), Thermococcus sibiricus (30), and Thermococcus gammatolerans (52). Mbh and Mbx orthologs are present in all species; Hyh orthologs are also found in all species, with the exception of T. gammatolerans. Although other types of [NiFe]-hydrogenase orthologs related to the hydrogenase-4 of Escherichia coli are found in P. abyssi, T. onnurineus, and T. gammatolerans, the common presence of the three [NiFe]-hydrogenase orthologs suggests that they play fundamental roles in the energy metabolism of Thermococcales.

In order to obtain genetic evidence to clarify the physiological roles of the three [NiFe]-hydrogenase orthologs, gene knockout mutants of the [NiFe]-hydrogenase orthologs were constructed in T. kodakarensis (previously reported as Thermococcus kodakaraensis) (1, 32). For T. kodakarensis, a system for chromosomal gene disruption/replacement has been established (39, 40). T. kodakarensis-E. coli shuttle vectors have also been developed which can be used for complementation analyses or heterologous gene expression (36). Our genetic analyses support previously reported biochemical data and indicate that the three hydrogenase orthologs exhibit distinct functions in T. kodakarensis. Along with the wealth of biochemical data on Pyrococcus and Thermococcus hydrogenases, the results allow us to propose the metabolic roles of the three hydrogenase orthologs in the Thermococcales.

MATERIALS AND METHODS

Microorganisms and culture conditions.

E. coli DH5α was used for general DNA manipulation and sequencing. E. coli strains were cultivated in LB medium (10 g liter⁻¹ tryptone, 5 g liter⁻¹ yeast extract, and 10 g liter⁻¹ NaCl) at 37°C. Ampicillin was added to the medium at a concentration of 100 μg ml⁻¹.

T. kodakarensis strains and plasmids used in this study are listed in Table 1. T. kodakarensis strains were routinely cultivated under anaerobic conditions at 85°C in a nutrient-rich medium (MA-YT) based on artificial sea salts, yeast extract, and tryptone (14). When growth with S⁰ was required, sulfur powder (Wako Pure Chemical Industries, Osaka, Japan) was added at a concentration of 5 g liter⁻¹ after the MA-YT medium had been autoclaved. In the case of cultivation with pyruvate or maltodextrin, 5 g liter⁻¹ sodium pyruvate (Nacalai Tesque) or 5 g liter⁻¹ maltodextrin (Amycol 3-L; Nippon Starch Chemical, Osaka, Japan) was added to the MA-YT medium, respectively, before autoclaving.

Table 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristics	Source or reference
T. kodakarensis strains
KOD1	Wild type	1, 32
KU216	KOD1 ΔpyrF	39
PHY1	KU216 ΔhyhBGSL::2μ′	This study
PAT1	KU216 Δaat::2μ′	This study
DPHA1	KU216 ΔhyhBGSL::2μ′ Δaat::2μ′	This study
MHD1	KU216 ΔmbhJKL::2μ′	This study
MXD1	KU216 ΔmbxJKL::2μ′	This study
Plasmids
pUC118	Ampr general cloning vector	Takara Bio (Otsu, Japan)
pUD2	pUC118 derivative; pyrF marker cassette	39
pUCMP	pUC118 derivative; 2μ′-pyrF-2μ′	39
pUDHY1	pUC118 derivative; ΔhyhBGSL::2μ′-pyrF-2μ′	This study
pUDAT1	pUC118 derivative; Δaat::2μ′-pyrF-2μ′	This study
pUDMH1	pUC118 derivative; ΔmbhJKL::2μ′-pyrF-2μ′	This study
pUDMX1	pUC118 derivative; ΔmbxJKL::2μ′-pyrF-2μ′	This study

Construction of T. kodakarensis mutant strains.

Disruption of specific genes by double-crossover homologous recombination in T. kodakarensis was performed using previously described methods (39, 40). Four vectors, for disruption of hyhBGSL (TK2072 to TK2069), aat (TK1094), mbhJKL (TK2089 to TK2091), and mbxJKL (TK1217 to TK1215) in T. kodakarensis (named pUDHY1, pUDAT1, pUDMH1, and pUDMX1, respectively), were constructed as follows. Four DNA fragments containing the respective target genes together with the flanking regions (about 1,000 bp) were amplified from T. kodakarensis KOD1 genomic DNA by using the primer sets PHY-F/PHY-R (for hyhBGSL), PAT-F/PAT-R (for aat), PMH-F/PMH-R (for mbhJKL), and PMX-F/PMX-R (for mbxJKL). The sequences of PCR primers used for this study are available upon request. Each amplified DNA fragment was subcloned into pUC118 at the HincII site. The flanking regions of the target gene and the plasmid backbone, excluding the target gene, were then amplified from the respective plasmids by using the primer sets PDHY-F/PDHY-R (for hyhBGSL), PDAT-F/PDAT-R (for aat), PDMH-F/PDMH-R (for mbhJKL), and PDMX-F/PDMX-R (for mbxJKL), and the resulting DNA fragments were designated L-HyhBGSL, L-Aat, L- MbhJKL, and L-MbxJKL, respectively. A 2μ′-pyrF-2μ′ DNA fragment, a universal marker cassette for elimination of the pyrF marker by single-crossover recombination after its insertion into the T. kodakarensis genome, excised from pUCMP (39) by use of SmaI, was ligated with L-HyhBGSL, L-Aat, L-MbhJKL, and L-MbxJKL to construct pUDHY1, pUDAT1, pUDMH1, and pUDMX1, respectively. The constructed vectors were sequenced to confirm the absence of unintended mutations. A T. kodakarensis uracil auxotroph strain, KU216 (39), was used as a host, and transformation procedures, including selection of the pyrF⁺ strains by uracil prototrophy and positive selection of pyrF-eliminated strains by use of 5-fluoroorotic acid, were described previously (39, 40).

Whether successful recombination had occurred was verified by Southern blot analysis and/or PCR. The genotypes of PHY1, PAT1, and DPHA1 strains were analyzed by Southern blot analysis with 5.0 μg of genomic DNA, and the overall procedures were performed as described previously (41). The genotypes of the MHD1 and MXD1 strains were analyzed by PCR, using primer sets that anneal outside the homologous regions (PMHC-F/PMHC-R for the mbhJKL locus and PMXC-F/PMXC-R for the mbxJKL locus).

Growth measurements in batch culture.

Growth characteristics of host cells (strain KU216) and individual mutant cells (PHY1, PAT1, and DPHA1), shown in Fig. S1 in the supplemental material, were measured as follows. Each strain was precultured in 20 ml of MA-YT medium with 0.5% (wt/vol) S⁰ at 85°C for 10 h. After preculture, cells were inoculated into 20 ml of MA-YT supplemented with either 0.5% (wt/vol) S⁰, 0.5% (wt/vol) sodium pyruvate, or 0.5% (wt/vol) maltodextrin and cultured at 85°C. Cell density was monitored by measuring the turbidity at 660 nm at appropriate intervals with an Ultraspec 3300 Pro UV spectrometer (GE Healthcare, Buckinghamshire, United Kingdom). Growth characteristics of KU216, MHD1, and MXD1 shown in Fig. 2 were measured essentially by the same procedure, except that MA-YT medium supplemented with both 0.5% (wt/vol) S⁰ and 0.5% (wt/vol) sodium pyruvate was used as the preculture medium.

Fig. 2.

Growth properties of T. kodakarensis strains in MA-YT medium supplemented with elemental sulfur (+S⁰) (left), sodium pyruvate (+Pyr) (center), and maltodextrin (+Mdx) (right). Symbols: solid diamonds, KU216; open squares, MHD1; open triangles, MXD1. OD₆₆₀, turbidity at 660 nm.

Determination of H₂ and H₂S production.

Production of gas metabolites (H₂ and H₂S) was determined for KU216, PHY1, MHD1, and MXD1 under batch culture conditions. Cells were cultivated in 15 ml of MA-YT medium supplemented with 0.5% (wt/vol) S⁰. After 10 h of cultivation at 85°C, a portion (100 μl) of the gas phase (total volume, 11.5 ml) was extracted with a gas-tight syringe. The amount of H₂ was quantified by gas chromatography (GC-14A; Shimadzu, Kyoto, Japan) with a thermal conductivity detector. The amount of H₂S was quantified by the methylene blue method as described elsewhere (45). The amounts of gas metabolites generated were normalized by cell mass, which was determined by measuring the turbidity of the culture broth.

Continuous culture experiments.

Continuous culture experiments with host cells (strain KU216) and individual mutant cells (PHY1, PAT1, and DPHA1) were performed based on previously described methods (14), using a gas-lift fermenter designed for cultivation of hyperthermophiles (Taiyo Nippon Sanso Corporation, Tokyo, Japan). In a 1-liter cultivation vessel, 500 ml of MA-YT medium supplemented with 0.5% (wt/vol) sodium pyruvate was introduced, and cultivation was performed at 85°C with continuous agitation, using a rotor at 50 rpm. The evolved gas metabolites were flushed out by nitrogen gas, which was introduced continuously into the vessel at a rate of 100 ml min⁻¹. Fresh medium was supplied to the vessel by use of a peristaltic pump (Cole-Parmer, Vernon Hills, IL), and the volume of the culture was monitored with a water level sensor (B. E. Marubishi, Tokyo, Japan) connected to a pump for culture discharging. Cell density was monitored by measuring the turbidity at 660 nm. Biomass was estimated from the cell density data by calculation using previously determined calibration information. The pH of the culture broth was maintained at 7.4.

The amounts of H₂ gas, CO₂ gas, and N₂ gas in the exhaust gas were determined periodically by gas chromatography as described previously (14). The concentration of ammonia in culture medium was determined enzymatically using an F kit for ammonia (Roche Diagnostics, Basel, Switzerland). The concentration of alanine was determined with an L-8800 amino acid analyzer (Hitachi, Tokyo, Japan) after acid hydrolysis. The concentration of acetate was determined by high-performance liquid chromatography, using a Sim-Pack SPR-H column (Shimadzu) and an electrical conductivity detector (Shimadzu).

Protein analysis.

The protein concentration was measured using a Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA), with bovine serum albumin as the standard. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed on 12.5% gels. Western blot analysis was performed as described previously (5), using rabbit polyclonal antibodies against recombinant HyhL, MbhL, and MbxL proteins.

RESULTS

[NiFe]-hydrogenase orthologs in T. kodakarensis.

T. kodakarensis contains genes for three [NiFe]-hydrogenase orthologs (Hyh, Mbh, and Mbx) in its genome (8). Although P. furiosus contains two Hyh orthologs, T. kodakarensis harbors a single Hyh ortholog (encoded by hyhBGSL [TK2072 to TK2069]) that is phylogenetically related to P. furiosus Hyh-I. Biochemical analyses of the Hyh purified from T. kodakarensis revealed that its enzymatic properties were comparable to those of P. furiosus Hyh-I (15). The mbh operon of T. kodakarensis (encoded by mbhABCDEFGHIJKLMN [TK2080 to TK2093]) contains all of the corresponding ORFs in the P. furiosus mbh operon in the same gene order, with each ORF sharing over 50% amino acid identity with its counterpart in P. furiosus. The same situation applies to the mbx operon (encoded by mbxABCDFGHH′MJKLN [TK1226 to TK1214]) with respect to gene members, gene order, and protein similarity.

Expression profiles of the three [NiFe]-hydrogenase orthologs.

T. kodakarensis grows well in a nutrient-rich MA-YT medium supplemented with S⁰, utilizing proteinaceous substrates while evolving H₂S and CO₂. In the absence of S⁰, T. kodakarensis grows with supplementation of pyruvate or maltodextrin in MA-YT medium, evolving H₂ and CO₂ (14). In order to identify the physiological functions of the three [NiFe]-hydrogenase orthologs in T. kodakarensis, their expression profiles were compared using cells cultivated under both H₂S-forming and H₂-forming conditions. Wild-type T. kodakarensis cells (strain KOD1) were cultivated in MA-YT medium containing either S⁰ (MA-YT-S⁰), sodium pyruvate (MA-YT-Pyr), or maltodextrin (MA-YT-Mdx). Western blot analyses were performed against the crude cell extracts obtained from these cultures, using anti-HyhL, anti-MbhL, and anti-MbxL antibodies. As a result, expression levels of HyhL significantly increased under H₂-forming growth conditions (with a sharp increase in cells grown on MA-YT-Pyr and a moderate increase in cells grown on MA-YT-Mdx) compared to those observed under H₂S-forming conditions (MA-YT-S⁰) (Fig. 1A). Similar to the case for HyhL, expression of MbhL was considerably higher under H₂-forming growth conditions than under H₂S-forming conditions. In contrast, expression levels of MbxL were much higher in cells grown under H₂S-forming conditions.

Fig. 1.

(A) Intracellular levels of Hyh, Mbh, and Mbx in T. kodakarensis KOD1 grown under various conditions. Cells were grown in MA-YT medium with 0.5% (wt/vol) elemental sulfur (+S⁰), 0.5% (wt/vol) sodium pyruvate (+Pyr), or 0.5% (wt/vol) maltodextrin (+Mdx). Cell extracts (3 μg) were subjected to SDS-PAGE, followed by Western blot analysis with anti-HyhL, anti-MbhL, or anti-MbxL antibody. (B) Western blot analysis of cell extracts (3 μg) of T. kodakarensis strains KU216 and PHY1, using anti-HyhL antibodies. (C) Western blot analysis of cell extracts (5 μg) of T. kodakarensis strains KU216, MHD1, and MXD1, using anti-HyhL, anti-MbhL, or anti-MbxL antibody. Cells were cultivated using MA-YT medium supplemented with 0.5% (wt/vol) S⁰.

Construction of hyh gene disruption strain.

In order to clarify the roles of Hyh in detail, a gene deletion mutant was constructed. An hyh disruption vector, pUDHY, which contains a 2μ′-pyrF-2μ′ cassette inserted between the upstream and downstream regions of hyhL, was constructed and used to transform the host strain (KU216; ΔpyrF strain). The 2μ′-pyrF-2μ′ cassette contains the pyrF marker gene flanked on both sides by identical sequences (designated 2μ′) (39). In the first step, pyrF⁺ strains, generated via double-crossover recombination with the upstream and downstream regions of hyhL, were selected in liquid synthetic medium based on uracil prototrophy, followed by colony isolation (ΔhyhBGSL::2μ′-pyrF-2μ′). Strains in which the pyrF marker was eliminated through single-crossover recombination between the two 2μ′ regions were positively selected using 5-fluoroorotic acid, resulting in the isolation of the pyrF mutant strain (ΔhyhBGSL::2μ′).

PCR amplification of the Δhyh locus indicated the absence of the hyh genes in the Δhyh strain (data not shown). The genotype of the Δhyh locus was also confirmed by Southern blot analysis (data not shown), and the mutant was named PHY1. NADPH-dependent H₂ evolution activity was not found in the cell extracts prepared from PHY1 (data not shown), and Western blot analysis (Fig. 1B) confirmed the absence of the HyhL protein. These results clearly indicate that PHY1 is a strain deficient in Hyh.

Physiological characterization of Δhyh mutant.

Growth properties of PHY1 were investigated in batch cultures. Cells were cultivated in either MA-YT-S⁰, MA-YT-Pyr, or MA-YT-Mdx. No growth defect of PHY1 was observed in any of the media (see Fig. S1 in the supplemental material), suggesting that Hyh is not essential, at least under the growth conditions tested. We noted that slightly higher cell densities in MA-YT-Pyr were obtained with PHY1 cells than with host cells.

To further examine the contribution of Hyh to H₂ metabolism, a continuous culture experiment was performed and the evolving gas metabolites were analyzed quantitatively. Under steady-state conditions using MA-YT-Pyr, the H₂ evolution rate (per unit cells) of the host cells (KU216) was slightly lower than that of CO₂, resulting in an H₂/CO₂ ratio of 0.88 (see Table S1 in the supplemental material). Interestingly, in the case of PHY1 cells, not a decrease but a slight increase in H₂ evolution rate was observed, resulting in a higher H₂/CO₂ ratio (0.96). The results provide genetic evidence that Hyh is not responsible for H₂ evolution in T. kodakarensis and raise the possibility that Hyh functions in the direction of H₂ uptake.

It has been reported in many cases that Thermococcales cells accumulate alanine as a reduced end product (14, 17, 18, 35). Alanine is formed from pyruvate by alanine aminotransferase (AlaAT) using glutamate as an amino donor, which is reductively regenerated by glutamate dehydrogenase (GDH) using NADPH and NH₃ (17, 18). If Hyh functions in the direction of H₂ uptake, it could provide the NADPH used for alanine production. To examine the presence of a metabolic link between Hyh and alanine production, fermentation products in the continuous culture medium were analyzed. As a result, a clear decrease in the rate of alanine production (−28.3%), as well as an increase in the production rate of NH₃ (3.5-fold), was observed in the culture medium of PHY1 (Table 2). This strongly supports the hypothesis that Hyh contributes to alanine production by supplying NADPH via H₂ uptake. To further examine this metabolic link, a disruption mutant of the AlaAT gene (PAT1 strain) and a double mutant of the Hyh and AlaAT genes (DPHA1 strain) were constructed. Both PAT1 and DPHA1 grew well, just like PHY1, in the three media tested (see Fig. S1 in the supplemental material). In a continuous culture experiment, PAT1 exhibited an enhanced H₂ production rate (see Table S1), a lower alanine production rate, and an increase in the NH₃ production rate (Table 2). The similar phenotypes observed for PHY1 and PAT1 suggest that Hyh and AlaAT function in a common metabolic pathway to produce alanine by using the reducing power generated from H₂. The differences in the production rates of H₂, alanine, and NH₃ in the double mutant were similar to those observed in the Δhyh strain and much smaller than the sum of the differences observed in the individual single-mutant cells (Table 2; see Table S1), further supporting the presence of a metabolic link between Hyh and AlaAT.

Table 2.

Generation rates of fermentation products under steady-state growth conditions (D = 0.25 h⁻¹)

Strain	Generation rate (mmol g−1 h−1) (% of control level)			Alanine/acetate ratio
	Acetate	Alanine	NH3
KU216	18.8 (100)	5.41 (100)	0.46 (100)	0.29
PHY1	17.7 (94.1)	3.88 (71.7)	1.62 (352)	0.22
PAT1	20.8 (111)	4.26 (78.7)	1.25 (272)	0.20
DPHA1	19.4 (103)	3.62 (66.9)	1.84 (400)	0.19

Construction of ΔmbhJKL and ΔmbxJKL strains.

In order to investigate the in vivo functions of membrane-type [NiFe]-hydrogenase orthologs (Mbh and Mbx), gene disruption mutants were constructed. Protein coding genes corresponding to the large and small subunits of [NiFe]-hydrogenase (mbhJKL and mbxJKL) were deleted from their corresponding operons. It should be noted that the coding region corresponding to the large subunit is split into two genes in both operons (mbhKL and mbxKL). The mutant genotypes were confirmed by PCR amplification. DNA fragments with lengths corresponding to those expected from the ΔmbhJKL and ΔmbxJKL loci were obtained (data not shown), and the mutants were named MHD1 (for the ΔmbhJKL mutant) and MXD1 (for the ΔmbxJKL mutant). Crude cell extracts prepared from MHD1 and MXD1 did not contain MbhL and MbxL proteins, respectively, as determined by Western blot analysis (Fig. 1C).

Physiological characterization of MHD1 and MXD1 strains.

Growth properties of MHD1 and MXD1 were investigated in MA-YT-S⁰, MA-YT-Pyr, and MA-YT-Mdx (Fig. 2). In MA-YT-S⁰, MHD1 cells exhibited comparable growth kinetics to those of the host cells. On the other hand, MHD1 cells exhibited severe growth defects in the pyruvate- and maltodextrin-containing media, both of which are conditions that lead to H₂ evolution. The results suggest that Mbh plays significant roles in the H₂ evolution process of T. kodakarensis. Therefore, we next examined H₂ generation by MHD1 grown in a batch culture with MA-YT-S⁰. As also observed for the wild-type strain (14), host cells produced detectable levels of H₂ even in MA-YT-S⁰ (Fig. 3A). In the case of PHY1 (Δhyh mutant), a higher level of H₂ accumulation (per unit cells) than that in host cells was detected under this growth condition, which agrees well with the proposed function of Hyh in H₂ uptake. In contrast, MHD1 exhibited a notable decrease in H₂ accumulation (11.9% compared to the level in the host cells). Since the protein levels of the other [NiFe]-hydrogenase orthologs (Hyh and Mbx) were similar to, not lower than, those of the host cells (Fig. 1C), the results indicate that Mbh is the major H₂-evolving [NiFe]-hydrogenase in T. kodakarensis.

Fig. 3.

Comparison of gas evolution per unit cells among T. kodakarensis strains. The amount of gas evolved from cells grown on MA-YT-S⁰ medium for 10 h at 85°C was divided by the cell density determined by the turbidity at 660 nm. (A) H₂ production; (B) H₂S production.

In the case of MXD1, the growth kinetics were completely different from those of MHD1. The growth rate of MXD1 in MA-YT-S⁰ was significantly lower than that of the host cells (Fig. 2). In contrast, although a prolonged lag time was observed in the case of maltodextrin-containing medium, growth rates under H₂-evolving conditions were not affected. Moreover, MXD1 in MA-YT-S⁰ exhibited higher H₂ accumulation levels than those in host cells (Fig. 3A). Intriguingly, we observed that the protein levels of Mbh and, to a lesser extent, Hyh increased in the MXD1 strain grown under these conditions (Fig. 1C). To further analyze the in vivo function of Mbx, H₂S generation of MXD1 was examined. Production of H₂S by MXD1 was significantly lower than that by the host cells (Fig. 3B). These results suggested that Mbx is not involved in H₂ evolution but rather that it contributes to the process of H₂S generation in T. kodakarensis.

DISCUSSION

In this study, genetic evidence clearly demonstrates that the three [NiFe]-hydrogenase orthologs (Hyh, Mbh, and Mbx) in T. kodakarensis display distinct physiological functions. Hyh, an NADP(H)-dependent [NiFe]-hydrogenase, was initially proposed to be responsible for H₂ evolution in P. furiosus (26, 28). This reaction is thermodynamically unfavorable, however, because of the more positive redox potential of NADP⁺ (NADP⁺/NADPH; E^0′ = −324 mV) than that of H⁺ (H⁺/H₂; E^0′ = −421 mV) (24). Quantitative metabolite analyses of P. furiosus indicated that the total electron transfer necessary to support H₂ production in starch-grown cells is much larger than the level that Hyh can afford (44). Therefore, after the discovery of Mbh, which shows a preference toward H₂ evolution (38, 44), Hyh was recognized as an H₂ uptake hydrogenase for NADPH regeneration (13, 44). This was also supported by kinetic analyses of P. furiosus Hyh-I: its K_m value for H₂ is <20 μM, similar to those of other H₂ uptake [NiFe]-hydrogenases (47, 48). Our genetic study provides in vivo evidence that confirms the physiological role of Hyh as an H₂ uptake enzyme under the growth conditions applied here. This is supported by the higher H₂ production levels observed in the Δhyh mutant than in the host strain in both pyruvate-supplemented (see Table S1 in the supplemental material) and S⁰-supplemented (Fig. 3A) media.

Formation of alanine as a reduced end product in media containing abundant levels of pyruvate or sugars has been reported for several Thermococcales species, such as P. furiosus (17), T. kodakarensis (14), Thermococcus profundus (18), and Thermococcus litoralis (35), suggesting that alanine is a common metabolic end product in this order. Alanine formation is catalyzed by AlaAT through an amino transfer reaction from glutamate to pyruvate (17, 18). In this study using MA-YT-Pyr medium, disruption of TK1094, encoding AlaAT, resulted in a significant decrease in alanine formation, indicating that this gene product is involved in alanine synthesis in T. kodakarensis. However, the fact that the knockout mutant still produced significant levels of alanine (almost 80% of the host cells) indicates the presence of other aminotransferases that contribute to alanine formation. An open reading frame (TK0186) annotated as a multiple-substrate aminotransferase gene could be one of the candidates responsible for alanine synthesis, as the orthologous aminotransferase in P. horikoshii (encoded by PH0207) was shown to exhibit activity toward a broad range of amino acids, including alanine (19).

An AlaAT ortholog of TK1094 has been shown to utilize glutamate as an amino group donor for the synthesis of alanine (51). Glutamate is supplied/regenerated from 2-oxoglutarate via GDH, using NH₃ and NADPH (17, 18). The mechanisms responsible for NADPH regeneration have been contemplated, and the involvement of ferredoxin:NADP⁺ oxidoreductase, utilizing reduced ferredoxin as an electron donor, has been suggested (17). Our genetic results revealing decreased alanine production and increased NH₃ formation in Hyh-deficient T. kodakarensis cells indicate that significant amounts of NADPH for alanine synthesis are provided through the oxidation of H₂. The metabolic link between H₂ uptake and alanine production seems to exist in other Thermococcales members, as cultivation of cells under an increased H₂ partial pressure resulted in increased alanine production in P. furiosus (3, 17) and T. litoralis (35).

High levels of structural similarity between the two membrane [NiFe]-hydrogenase orthologs, the Mbh and Mbx complexes, indicate that they share an evolutionary origin, but their physiological functions are considered different. Biochemical studies of P. furiosus Mbh suggested that it functions as an energy-converting [NiFe]-hydrogenase, creating a proton motive force across the plasma membrane by means of proton reduction using a low-potential ferredoxin (37, 38). On the other hand, although the role of the Mbx complex is not clarified yet, its structural similarity with Mbh suggests that Mbx functions as another energy-transducing membrane complex (44). Since transcription of the mbx operon is induced in the presence of S⁰, it has been proposed to be involved in the process of electron transfer from reduced ferredoxin to S⁰ (43). The present genetic analyses of the T. kodakarensis ΔmbhJKL mutant are consistent with the proposed roles of Mbh participating in H₂ generation. The total lack of growth of the ΔmbhJKL mutant under H₂-evolving conditions (Fig. 2), as well as the nearly complete loss of H₂ evolution activity (Fig. 3A), indicates that Mbh is the primary [NiFe]-hydrogenase responsible for the disposal of excess reducing equivalents to protons. This result, in turn, means that Mbx cannot replace the function of Mbh and that these enzymes play distinct roles in vivo. Genetic analyses of the T. kodakarensis ΔmbxJKL mutant, which showed a decreased specific growth rate in the presence of S⁰ (Fig. 2) as well as a reduced H₂S generation rate (Fig. 3B), were consistent with the proposed involvement of Mbx in the process of H₂S generation. Because a small amount of H₂ was still generated in the ΔmbhJKL strain grown in MA-YT-S⁰, this might be due to Hyh functioning in the reverse direction. It has previously been demonstrated that Hyh can generate H₂ with NADPH as an electron donor (15).

The results of our genetic analyses of T. kodakarensis, taken together with previously reported biochemical properties of Pyrococcus/Thermococcus hydrogenase orthologs, can be interpreted and summarized as follows (Fig. 4). Under H₂-evolving conditions in media supplemented with pyruvate or maltodextrins, Hyh and Mbh are both upregulated, whereas Mbx is downregulated (Fig. 1A). In P. furiosus, transcription of these operons is regulated in a similar manner under H₂-evolving conditions (43). The electrons of reduced ferredoxin generated through the breakdown of pyruvate/maltodextrins are transferred to protons by the function of Mbh, resulting in H₂ evolution and a membrane proton gradient allowing ATP synthesis. Hyh acts as an H₂ uptake hydrogenase that supplies reducing equivalents in the form of NADPH. The metabolic relevance of this function of Hyh is not yet obvious, as hyh disruption did not lead to any growth impairment in the media adopted in this study (see Fig. S1 in the supplemental material). However, the Hyh reaction may play important roles depending on the cell environment. The generation of NADPH from H₂ can be used to promote the fixation or detoxification of free NH₃, leading to the production of amino acids such as alanine, or may simply be a route to supply the NADPH required for biosynthetic purposes. Regardless of the physiological role, the observed increase in H₂ production levels in Δhyh cells (Fig. 3A; see Table S1) demonstrates that intraspecies H₂ transfer (i.e., H₂ evolved by Mbh is reoxidized by Hyh) is potentially possible in T. kodakarensis. Intraspecies H₂ transfer has been observed in several anaerobic microorganisms, such as nitrogen-fixing bacteria (50), sulfate-reducing bacteria (10, 33), and the methanogen M. barkeri (20).

Fig. 4.

Proposed carbon and energy metabolism in T. kodakarensis under H₂-evolving (left) and H₂S-evolving (right) growth conditions. The major substrates for growth, pyruvate (left) and amino acids/S⁰ (right), are indicated with shaded boxes. Abbreviations: Hyh, cytosolic [NiFe]-hydrogenase; Mbh, membrane-bound [NiFe]-hydrogenase (Mbh complex); Mbx, putative membrane-bound [NiFe]-hydrogenase (Mbx complex); AT, amino acid aminotransferase; GDH, glutamate dehydrogenase; POR, pyruvate:ferredoxin oxidoreductase; Fd_ox, oxidized ferredoxin; Fd_red, reduced ferredoxin; NSR, NAD(P)H-elemental sulfur oxidoreductase; XOR, 2-oxoacid:ferredoxin oxidoreductase.

Under H₂S-evolving conditions, Mbh and Hyh are downregulated, and instead, a dramatic elevation in the level of Mbx is observed (Fig. 1A). This is consistent with the results of transcriptome analysis of P. furiosus (43). Amino acids are converted to their corresponding 2-oxoacids via amino acid aminotransferases, with this process sustained by the regeneration of 2-oxoglutarate via the oxidative deamination of glutamate by GDH. S⁰ is essential in this mode of growth to regenerate the NADP⁺ that accepts the electrons from glutamate in the GDH reaction. The 2-oxoacids are converted to their corresponding acyl-coenzyme A (acyl-CoA) compounds via 2-oxoacid:ferredoxin oxidoreductases and subsequently hydrolyzed in a reaction that allows substrate-level phosphorylation. We previously reported that T. kodakarensis cells grown in MA-YT-S⁰ generate predominantly H₂S, along with only low levels of H₂ (14). The reducing equivalents for the reduction of S⁰ derive from the NADPH generated from GDH and the reduced ferredoxin produced by the 2-oxoacid:ferredoxin oxidoreductases. Since the disruption of mbx led to significant decreases in H₂S production rates (Fig. 3B), it is clear that Mbx participates in S⁰ reduction. On the other hand, since approximately half of the H₂S generation was still retained in the ΔmbxJKL strain, this indicates that there is another route, independent of Mbx, that leads to H₂S generation. One feasible scenario, taking into account the model proposed by Schut et al. (43), is the presence of a cytosolic NAD(P)H:S⁰ oxidoreductase (NSR) that is directly responsible for S⁰ reduction by NADPH, along with Mbx, whose function is to transfer electrons from reduced ferredoxin to NADP⁺. It is supposed, based on the similar structures of Mbx and Mbh, that the electron transfer from ferredoxin to NADP⁺ via Mbx is coupled to proton translocation (44). In the ΔmbxJKL strain, the conversion from amino acids to 2-oxoacids, as well as the H₂S generated from this conversion, is not affected, as the reducing equivalents transferred by GDH to NADPH are directly accessible by NSR. The reduced ferredoxin produced by the 2-oxoacid:ferredoxin oxidoreductases, however, cannot release its electrons to Mbx, and thus it must rely on Mbh. Our results revealed that under these conditions, intracellular levels of Mbh increase in the ΔmbxJKL strain (Fig. 1C). This may be one reason for the increase in H₂ production rates (4.4-fold) observed in the ΔmbxJKL strain compared to the host strain (Fig. 3A). Although protein levels are enhanced, the levels of Mbh may still be rate-limiting, as the growth rate of the ΔmbxJKL strain is lower than that of the host strain (Fig. 2).

In order to confirm the present scenarios and to obtain a better understanding of energy metabolism in the Thermococcales, the enzymatic properties of Mbx, particularly the presence of the electron transfer activity from ferredoxin:NADP⁺ and its coupling to proton translocation, must be demonstrated biochemically.

ADDENDUM

During the review process for this article, Lipscomb et al. reported the first example of gene disruption in P. furiosus (23). In addition to the importance of developing a gene disruption system for P. furiosus, the study reports the phenotypes of mutants of two Hyh operons (Hyh-I and Hyh-II). P. furiosus strains with disruption of either or both operons still grew well under H₂-evolving conditions, similar to the results for the Hyh-deficient T. kodakarensis strain (PHY1) observed in this study.