Pyruvate:ferredoxin oxidoreductase is coupled to light-independent hydrogen production in Chlamydomonas reinhardtii

Abstract

In anaerobiosis, the green alga Chlamydomonas reinhardtii evolves molecular hydrogen (H(2)) as one of several fermentation products. H(2) is generated mostly by the [Fe-Fe]-hydrogenase HYDA1, which uses plant type ferredoxin PETF/FDX1 (PETF) as an electron donor. Dark fermentation of the alga is mainly of the mixed acid type, because formate, ethanol, and acetate are generated by a pyruvate:formate lyase pathway similar to Escherichia coli. However, C. reinhardtii also possesses the pyruvate:ferredoxin oxidoreductase PFR1, which, like pyruvate:formate lyase and HYDA1, is localized in the chloroplast. PFR1 has long been suggested to be responsible for the low but significant H(2) accumulation in the dark because the catalytic mechanism of pyruvate:ferredoxin oxidoreductase involves the reduction of ferredoxin. With the aim of proving the biochemical feasibility of the postulated reaction, we have heterologously expressed the PFR1 gene in E. coli. Purified recombinant PFR1 is able to transfer electrons from pyruvate to HYDA1, using the ferredoxins PETF and FDX2 as electron carriers. The high reactivity of PFR1 toward oxaloacetate indicates that in vivo, fermentation might also be coupled to an anaerobically active glyoxylate cycle. Our results suggest that C. reinhardtii employs a clostridial type H(2) production pathway in the dark, especially because C. reinhardtii PFR1 was also able to allow H(2) evolution in reaction mixtures containing Clostridium acetobutylicum 2[4Fe-4S]-ferredoxin and [Fe-Fe]-hydrogenase HYDA.

FIGURE 1.

In vivo H₂ evolution rates of C. reinhardtii cultures in the light or in the dark. Concentrated cell suspensions were flushed with nitrogen for 4 h until they had reached an in vitro hydrogenase activity of 109 ± 18 nmol of H₂·μg Chl⁻¹·h⁻¹. Then culture aliquots were withdrawn, transferred to gas tight head space bottles, and incubated in the light (white bars) or the dark (gray bars) until the indicated time points before determining the H₂ concentration of the head space by gas chromatography. The results shown are the mean values from three independent experiments carried out as technical duplicates. The error bars indicate the standard deviation.

FIGURE 2.

Stacked polypeptide alignment of pyruvate:ferredoxin oxidoreductase primary sequences. Eight PFOR sequences were used for an alignment using ClustalW2 and WebLogo 3 (89, 90). These sequences were C. reinhardtii PFR1 derived from the cDNA obtained in this study, Volvox carteri f. nagariensis (Phytozome v8.0, V. carteri Vocar20008508m), Chlorella variabilis (GenBank^TM EFN55341.1), C. acetobutylicum PFO (GenBank^TM NP_348846.1), Clostridium pasteurianum (GenBank^TM AAD55756.1), Desulfovibrio africanus POR (GenBank^TM CAA70873.1), E. coli YdbK (GenBank^TM YP_002999180.1), and Synechococcus sp. PCC.7002 NifJ (GenBank^TM ACA99434.1). The conserved YPITP substrate-binding site (58), as well as three [4Fe-4S]-cluster coordinating motifs and the region homologous to TPP-binding (59, 91) sites, are underlined. Note that the third [4Fe-4S]-cluster-binding site is atypical and consists of the CXXC motif at positions 942–945 and two separated cysteines at positions 970 and 1221 (60, 61). The first amino acid of recombinant PFR1 is marked by an asterisk.

FIGURE 3.

Kinetic parameters of recombinant C. reinhardtii PFR1 heterologously produced in E. coli and purified via His-tag affinity chromatography. Enzymatic activity was determined following the reduction of methyl viologen spectrophotometrically in 100-μl reaction mixtures containing 1.4 μm PFR1, 5 mm TPP, 16 mm dithioerythritol in 100 mm Tris-HCl, pH 8. The K_m values of the individual substrates were determined in the presence of 2 mm CoA and 10 mm pyruvate (A, methyl viologen), 10 mm methyl viologen and 10 mm pyruvate (B, CoA), or 2 mm CoA and 10 mm methyl viologen (C, pyruvate). Each kinetic was analyzed from two independent PFR1 preparations as technical duplicates. The K_m and V_max values were calculated using GraphPad Prism® software. The graphs show the mean values, and the error bars indicate the standard deviation.

FIGURE 4.

H₂ generation in reconstitution assays of recombinant C. reinhardtii PFR1 and [Fe-Fe]-hydrogenases. A, each reaction contained PFR1 (0.7 μm) and 0.01 μm HYDA1 of C. reinhardtii (except one reaction, which contained C. acetobutylicum ferredoxin (CAC0303) and [Fe-Fe]-hydrogenase HYDA, indicated by the label CaHYDA), 10 mm pyruvate, and 2 mm CoA in 100 mm potassium phosphate buffer, pH 6.8. Electron carriers were applied as indicated below the x axis (10 mm methyl viologen (MV), 40 μm of the Chlamydomonas ferredoxins PETF, FDX2, or FDX5, or 40 μm clostridial ferredoxin CAC0303). The reaction mixtures were incubated for 30 min at 37 °C before analyzing the amount of H₂ in the gas phase. As controls, reaction mixtures were analyzed that lacked one of the enzymes or proteins indicated by a dash. B, the dependence of PFR1-coupled H₂ generation on PETF concentration was determined using C. reinhardtii HYDA1 in reaction mixtures as described for A and the indicated concentrations of PETF. The values and standard deviations shown in all of the experiments are from three independent PFR1 preparations, and the H₂ production rate was related to mg of PFR1 enzyme. The error bars indicate the standard deviation.

FIGURE 5.

Substrate specificity of C. reinhardtii PFR1. The capacity of PFR1 to reduce methyl viologen (A) and to drive H₂ evolution by C. reinhardtii HYDA1 (B) using oxaloacetate or α-ketoglutarate was examined. The reaction mixtures contained 10 mm pyruvate, oxaloacetate, or α-ketoglutarate and 2 mm CoA in 100 mm potassium phosphate buffer, pH 6.8, and additionally 1.4 μm PFR1 and 10 mm methyl viologen (A) or 0.7 μm PFR1, 40 μm PETF, and 0.01 μm HYDA1 (B). A, methyl viologen reduction was determined spectrophotometrically. B, H₂ evolution rates were determined by gas chromatography as described in the legend of Fig. 4. The results shown are the means and standard deviations from two independent experiments carried out as technical duplicates.

FIGURE 6.

Model of fermentative pathways involved in dark anaerobic H₂ production in C. reinhardtii. In the wild type, PFL1 is the major fermentative enzyme in short term anaerobiosis and cleaves pyruvate into formate and acetyl-CoA. The latter can be reduced to ethanol via bifunctional acetaldehyde-alcohol dehydrogenase (ADH1) (92) or converted to acetate via phosphotransacetylase and acetate kinase (PAT/ACK). In addition to PFL1, PFR1 is capable of pyruvate oxidation. In a pfl1 mutant or in long term anaerobiosis, pyruvate oxidation might be preferably catalyzed by PFR1. PFR1 transfers electrons to ferredoxins and thereby allows H₂ generation via the [Fe-Fe]-hydrogenase. In addition to pyruvate, PFR1 is able to use oxaloacetate as a substrate. This might link the oxidation of other substrates such as fatty acids or amino acids to fermentative H₂ production, possibly via an anaerobically operating glyoxylate cycle or parts thereof.