Deletion of alternative pathways for reductant recycling in Thermococcus kodakarensis increases hydrogen production

Abstract

Hydrogen (H₂) production by Thermococcus kodakarensis compares very favourably with the levels reported for the most productive algal, fungal and bacterial systems. T. kodakarensis can also consume H₂ and is predicted to use several alternative pathways to recycle reduced cofactors, some of which may compete with H₂ production for reductant disposal. To explore the reductant flux and possible competition for H₂ production in vivo, T. kodakarensis TS517 was mutated to precisely delete each of the alternative pathways of reductant disposal, H₂ production and consumption. The results obtained establish that H₂ is generated predominantly by the membrane-bound hydrogenase complex (Mbh), confirm the essential role of the SurR (TK1086p) regulator in vivo, delineate the roles of sulfur (S°) regulon proteins and demonstrate that preventing H₂ consumption results in a substantial net increase in H₂ production. Constitutive expression of TK1086 (surR) from a replicative plasmid restored the ability of T. kodakarensis TS1101 (ΔTK1086) to grow in the absence of S° and stimulated H₂ production, revealing a second mechanism to increase H₂ production. Transformation of T. kodakarensis TS1101 with plasmids that express SurR variants constructed to direct the constitutive synthesis of the Mbh complex and prevent expression of the S° regulon was only possible in the absence of S° and, under these conditions, the transformants exhibited wild-type growth and H₂ production. With S° present, they grew slower but synthesized more H₂ per unit biomass than T. kodakarensis TS517.

Fig. 1

Pathways proposed for ATP synthesis and reduced cofactor recycling in T. kodakarensis. A modified Embden-Meyerhof glycolytic pathway generates ATP and reduced ferredoxin(s) (Fd_red; red). The alternative pathways proposed to regenerate Fd_ox (brick red) are illustrated with enzyme/complexes identified by name and the encoding T. kodakarensis gene number(s). The same colors are used to report the results obtained with each enzyme/complex in Figs. 4 through 7. Electron transfer to protons by the membrane bound hydrogenase (Mbh) complex (yellow; TK2080–93) generates H₂ (red square). Mbh activity also produces a proton-gradient that can be exploited for ATP-production. A cytosolic hydrogenase (cyan; TK2069–72) can consume H₂ to generate NADPH, and NADPH may also be produced by ferredoxin-NADPH oxidoreductases (blue; TK1684–5, purple; TK1325–6). An NADPH:sulfur reductase (brown; TK1299) and NADPH:polysulfide reductase (pink; TK1481) are predicted to generate H₂S when S° is available. A membrane bound oxidoreductase (Mbx) complex (green; TK1214–26) is predicted to functionally replace the Mbh complex when S° is present, although no efflux function for Mbx has been established. SurR (red; TK1086) is a redox-sensitive transcription regulator. When reduced, SurR activates transcription of the Mbh operon and represses transcription of the S° regulon. Transcription of the genes (TK1260–61) encoding SipA and SipB (orange) is highly induced by the presence of S°. A glutamate-dependent alanine aminotransferase (grey; TK1094) converts pyruvate to alanine, generating 2-oxoglutarate from glutamate, and the regeneration of glutamate by glutamate dehydrogenase (TK1431) oxidizes NADPH to NADP.

Fig. 2

Deletion of TK1094 from the T. kodakarensis TS517 genome. (A) Plasmid pLC120 was constructed, as illustrated, with the PCR-amplified regions of the T. kodakarensis genome flanking the [TK0254–TK0664] expression cassette and used to transform T. kodakarensis TS517 (ΔpyrF; ΔtrpE::pyrF; ΔTK0664). The genome organization in a representative transformant (intermediate strain), selected by growth in the absence of tryptophan, was confirmed by diagnostic PCR. The intermediate strain was plated on a medium containing 6MP, and 6MP^R clones were isolated. A representative clone (designated T. kodakarensis TS1100; Table 1) was confirmed (see below) to have the desired ΔTK1094 mutation. As illustrated, a recombination between sequences directly repeated in the intermediate strain genome deleted both the expression cassette and TK1094. (B) Ethidium-bromide stained agarose gel electrophoresis of the PCR products generated by primers A plus D, and B plus C from T. kodakarensis TS1100 genomic DNA. The sites to which the primers hybridized are indicated above and below the T. kodakarensis TS517 genome organization in (A). Consistent with the presence of the TK1094 deletion, a PCR product was not generated using primers B and C, and the PCR product generated with primers A and D was shorter, by the length of the TK1094 deletion, than the PCR product generated using these primers with the parental T. kodakarensis TS517 genomic DNA. As shown in Fig. S1, the same diagnostic PCR procedure was used to confirm the presence of the desired deletion(s) in the genomes of all the T. kodakarensis strains used in this study (Table 1).

Fig. 3

T. kodakarensis genes deleted for this study. The genes precisely deleted from the T. kodakarensis TS517 genome to generate the T. kodakarensis strain(s) listed above each genome organization are shown in black. The organizations are drawn to scale with the length of a 1 kb region noted as a size reference.

Fig. 4

Growth of the T. kodakarensis strains in ASW-YT media containing S°, S° + Pyr, or Pyr. Increases in the optical density (OD₆₀₀) were measured over time. The results shown for each strain, in the color listed below the graphs, are the moving averages (P=2) of the data collected from at least two independent cultures. T. kodakarensis TS1101, TS1103, and TS1104 exhibited essentially no growth when incubated in Pyr medium. OD₆₀₀ measurements below 0.01 are noted as points on the X-axis.

Fig. 5

Comparison of H₂ production by T. kodakarensis strains. The H₂ present in the headspace of a culture during exponential growth was divided by the culture density (OD₆₀₀) to calculate the Ω value. The solid and hatched columns are Ω values from cultures grown in S°+Pyr and Pyr medium, respectively. The Ω values shown for T. kodakarensis TS1106 and TS1110 were measured near the end of exponential growth (see Fig. 6). As indicated, there was essentially no detectable H₂ in the headspace of cultures of T. kodakarensis TS1101, TS1103 and TS1104.

Fig. 6

Comparison of H₂ production by cultures of T. kodakarensis TS517 and TS1106. The Ω values shown were calculated from the H₂ present in the headspace of T. kodakarensis TS517 (black bars) and TS1106 (blue bars) cultures grown in Pyr medium sampled during early, middle and late exponential growth (hrs 5 to 9), at the end of exponential growth (hr 11) and in stationary phase (hr 13). In Pyr medium, T. kodakarensis TS1106 (ΔTK2069–72) cultures grew to a lower final cell densities than T. kodakarensis TS517 cultures (see Fig. 4), but H₂ production was apparently unlinked from growth and the Ω values increased throughout the culture incubation.

Fig. 7

Growth and H₂ production by T. kodakarensis TS1101 (ΔTK1086) strains containing pTS703 (◇), pTS705 (Δ), pTS709 (○) or pTS710 (□). (A) Growth of each plasmid-containing T. kodakarensis TS1101 strain in (ASW-YT) S°, S°+Pyr and Pyr media supplemented with 20 μM mevinolin. (B) The Ω values measured for T. kodakarensis TS517 and TS1101 cultures are shown as bars adjacent to the Ω values (bars on grey background) measured for these strains containing plasmids pLC70, pTS703, pTS705, pTS709 or pTS710. The solid and hatched bars report the Ω values for cultures grown in S°+Pyr and Pyr medium, respectively. As indicated by the very small bar, T. kodakarensis TS1101 cultures grew in S°+Pyr medium but did not accumulate measurable H₂. The adjacent space notes that T. kodakarensis TS1101 did not grow in Pyr medium.

Fig. 7

Growth and H₂ production by T. kodakarensis TS1101 (ΔTK1086) strains containing pTS703 (◇), pTS705 (Δ), pTS709 (○) or pTS710 (□). (A) Growth of each plasmid-containing T. kodakarensis TS1101 strain in (ASW-YT) S°, S°+Pyr and Pyr media supplemented with 20 μM mevinolin. (B) The Ω values measured for T. kodakarensis TS517 and TS1101 cultures are shown as bars adjacent to the Ω values (bars on grey background) measured for these strains containing plasmids pLC70, pTS703, pTS705, pTS709 or pTS710. The solid and hatched bars report the Ω values for cultures grown in S°+Pyr and Pyr medium, respectively. As indicated by the very small bar, T. kodakarensis TS1101 cultures grew in S°+Pyr medium but did not accumulate measurable H₂. The adjacent space notes that T. kodakarensis TS1101 did not grow in Pyr medium.