Selective cysteine-to-selenocysteine changes in a [NiFe]-hydrogenase confirm a special position for catalysis and oxygen tolerance

Abstract

In [NiFe]-hydrogenases, the active-site Ni is coordinated by four cysteine-S ligands (Cys; C), two of which are bridging to the Fe(CO)(CN)₂ fragment. Substitution of a single Cys residue by selenocysteine (Sec; U) occurs occasionally in nature. Using a recent method for site-specific Sec incorporation into proteins, each of the four Ni-coordinating cysteine residues in the oxygen-tolerant Escherichia coli [NiFe]-hydrogenase-1 (Hyd-1) has been replaced by U to identify its importance for enzyme function. Steady-state solution activity of each Sec-substituted enzyme (on a per-milligram basis) is lowered, although this may reflect the unquantified presence of recalcitrant inactive/immature/misfolded forms. Protein film electrochemistry, however, reveals detailed kinetic data that are independent of absolute activities. Like native Hyd-1, the variants have low apparent K_MH₂ values, do not produce H₂ at pH 6, and display the same onset overpotential for H₂ oxidation. Mechanistically important differences were identified for the C576U variant bearing the equivalent replacement found in native [NiFeSe]-hydrogenases, its extreme O₂ tolerance (apparent K_MH₂ and V_max [solution] values relative to native Hyd-1 of 0.13 and 0.04, respectively) implying the importance of a selenium atom in the position cis to the site where exogenous ligands (H^-, H₂, O₂) bind. Observation of the same unusual electrocatalytic signature seen earlier for the proton transfer-defective E28Q variant highlights the direct role of the chalcogen atom (S/Se) at position 576 close to E28, with the caveat that Se is less effective than S in facilitating proton transfer away from the Ni during H₂ oxidation by this enzyme.

Keywords: hydrogen activation; hydrogenase; oxygen tolerance; selenocysteine.

Fig. 1.

(A) Amino acid alignment of selected hydrogenases (Hyd-1 numbering) highlighting key residues (Cys/Sec, red; E28, green; D118, pink; R509, yellow) and differences (cyan). See also SI Appendix, Fig. S1 and Table S1. (B and D) The extended active site of Hyd-1 (Protein Data Bank [PDB] ID code 5A4M) (B) and Desulfomicrobium baculatum NiFeSe (PDB ID code 1CC1) (D) hydrogenases. (C) Representation of the active site, where “X” denotes the atom in a bridging position between the Ni and Fe atoms, the identity of which depends on the (in)active state of the enzyme (SI Appendix, Fig. S3).

Fig. 2.

(A) Chemical structures of Cys and Sec show the selenol moiety (red) to be the only difference. (B) EF-Tu–driven site-specific incorporation of Sec at a UAG codon. mRNA, messenger RNA. (C) Coomassie blue-stained denaturing sodium dodecyl sulfate-polyacrylamide gel of Hyd-1 variants (C76U, C79U, C576U, and C579U) shows high purity in each case, comprising HyaA (37 kDa) and HyaB (66 kDa) only. (D) Tandem mass spectra of C76U and C576U show Sec incorporation at the desired position in the designated peptide. Red lines correlate with the cleavage products depicted in the peptide sequence with an accuracy of 5 ppm. See also SI Appendix, Figs. S4–S8.

Fig. 3.

Electrochemical profiles and response to O₂. (A) CVs were scanned from −0.46 to +0.24 V and back (black arrows) at 0.5 mV⋅s⁻¹. Other conditions: 100% H₂, 1,000 standard cubic centimeters (scc) per minute, ω = 3,000 rpm, 37 °C, pH 6.0. (B) Series of CV scans from −0.46 to + 0.24 V and back at 0.5 mV⋅s⁻¹ under 100% H₂ (scan 1), then after injecting O₂ (154 μM) at +0.03 V (red arrows; scan 2), and finally under 100% H₂ to assess post–O₂-exposure effects (scan 3). Other conditions: 100% H₂, 1,000 scc per minute, ω = 3,000 rpm, 37 °C, pH 6.0. (C) The current at 0 V was first measured under 100% H₂ (730 μM), then 10% (blue arrows). Increasing [O₂] levels were introduced into the headspace for 600 s per increment. At 5,000 s, 100% H₂ was restored and spontaneous recovery was monitored. At 7,000 s, potential steps to −0.659 V were performed for 60 or 600 s, and total recovery was monitored at 0 V. Other conditions: 37 °C, pH 6.0, Ar carrier gas, flow 1,000 scc per minute, ω = 3,000 rpm.

Fig. 4.

(A) Apparent K_MH₂ values at each potential were determined by measuring CVs between −0.659 and +0.241 V at 5 mV⋅s⁻¹ in 2 to 730 μM H₂. Total gas flow rate (Ar carrier gas) 1,000 scc per minute, 37 °C, pH 6.0, ω = 3,000 to 4,000 rpm. (B) Activation enthalpies ΔH^‡ at different potentials were determined by measuring CVs under 100% H₂ from −0.659 to +0.241 V at 5 mV⋅s⁻¹ over the temperature range 2 to 45 °C. Gas flow 1,000 scc per minute, pH 6.0, ω = 1,000 rpm. (C) Steady-state H₂ oxidation rates and ΔH^‡ at 0 V (Table 1). Turnover frequencies (TOF; triangles) for native Hyd-1 [the asterisk indicates a previously published result (19)], Sec variants, and R509K (4, 17). (D) Summary: R, response to prolonged exposure to 104 μM O₂; S, ability to recover H₂ oxidation activity spontaneously; T, total recovery level after applying −0.659 V. All error bars represent the standard error of at least three repeats.

Fig. 5.

Overlaid electrocatalytic oxidation profiles of C576U (red) and E28Q (black). Cyclic voltammograms were scanned between −0.6 and +0.241 V at 5 mV⋅s⁻¹ and back (arrows). Conditions: 100% H₂ at 1,000 scc per minute, pH 6.0, and 37 °C (C576U) or 30 °C (E28Q), ω = 1,000 rpm. A modified PGE-multiwalled carbon nanotube electrode was used for E28Q (19).