Selenium-binding protein 1 (SELENBP1) is a copper-dependent thiol oxidase

Abstract

Selenium-binding protein 1 (SELENBP1) was reported to act as a methanethiol oxidase (MTO) in humans, catalyzing the conversion of methanethiol to hydrogen peroxide, hydrogen sulfide and formaldehyde. Here, we identify copper ions as essential to this novel MTO activity. Site-directed mutagenesis of putative copper-binding sites in human SELENBP1 produced as recombinant protein in E. coli resulted in loss of its enzymatic function. On the other hand, the eponymous binding of selenium (as selenite) was no requirement for MTO activity and only moderately increased SELENBP1-catalyzed oxidation of methanethiol. Furthermore, SEMO-1, the SELENBP1 ortholog recently identified in the nematode C. elegans, also requires copper ions, and MTO activity was enhanced or abrogated, respectively, if worms were grown in the presence of cupric chloride or of a Cu chelator. In addition to methanethiol, we identified novel substrates of SELENBP1 from the group of volatile sulfur compounds, ranging from ethanethiol to 1-pentanethiol as well as 2-propene-1-thiol. Gut microbiome-derived methanethiol as well as food-derived volatile sulfur compounds (VSCs) account for malodors that may contribute to extraoral halitosis in humans, if not metabolized properly. As SELENBP1 is particularly abundant in tissues exposed to VSCs, such as colon, liver, and lung, it appears to contribute to copper-dependent VSC degradation.

Keywords: H(2)O(2); H(2)S; MTO; Methanethiol; VSC; Volatile sulfur compounds.

Graphical abstract

Fig. 1

Selenium binding by SELENBP1: effect on MTO activity and contribution of cysteine residues. a, Recombinant Strep-tagged SELENBP1 was isolated from E. coli lysates and washed with sodium selenite during affinity chromatography at the indicated concentrations. The Se content of SELENBP1 was measured by total reflection X-ray fluorescence spectrometry (TXRF). MTO activity was determined by conversion of methanethiol, produced in situ through recombinant MGL, to H₂S and H₂O₂. Data represent means ± SD from three independent experiments. b, Topology of candidate cysteine residues for Se binding in a 3D-model of SELENBP1 (UniProt-ID: Q13228) that is based on a prediction by the AlphaFold Protein Structure consortium. c, Wildtype or cysteine-deficient SELENBP1 (with the given cysteines replaced by serine) was washed with 10 mM selenite during affinity chromatography. Selenium content as well as methanethiol-derived H₂S and H₂O₂ production were analyzed. Data represent means ± SD from three independent experiments, normalized to wildtype SELENBP1.

Fig. 2

MTO activity of SELENBP1 and SEMO-1 is copper-dependent. a, b, Recombinant Strep-tagged SELENBP1 and SEMO-1 were produced in E. coli and washed (pre-treated) with 10 mM EDTA during affinity chromatography where indicated. Thereafter, the given divalent cations were supplemented as chloride salts dissolved in HBS to a final concentration of 10 μM (1:2 protein/cation). MTO activity was determined with MGL-produced methanethiol as substrate. a, MTO activity of SELENBP1, as assessed using H₂S release (upper panel) and H₂O₂ production (lower panel) as readouts. b, MTO activity of SEMO-1. c, SEMO-1 localization in C. elegans, as visualized by fluorescence microscopy of a worm carrying a SEMO-1::GFP translational reporter. d, Relative methanethiol-derived H₂S production (as detected by lead acetate indicator paper) in lysates from wildtype (N2) and SEMO-1-knockout worms (right panel). Synchronized worms were allowed to grow for 72 h to adulthood, and thereafter, they were treated for 24 h with 0.1 mM CuCl₂, 0.1 mM sodium selenite or 1 mM of the copper chelator BCS. Data are normalized against the H₂S production of non-treated wildtype worms and depicted as means +SD of three independent experiments.

Fig. 3

Identification of Cu-binding amino acids in human SELENBP1, and evaluation of their relevance for its MTO activity. a, 3D-model of SELENBP1 (AlphaFold), highlighting the β-propeller structure of the protein. Right panel: magnified cavity showing the modelled topological location of the amino acids proposed to mediate copper coordination (oxygen: blue; nitrogen: red; carbon: yellow). Putative Cu-binding amino acids and motifs were identified using the COFACTOR program. b, c, SELENBP1 and SEMO-1 mutants were generated through site-directed in vitro-mutagenesis. Recombinant wildtype and mutant proteins were produced in E. coli and treated with 10 μM CuCl₂ during affinity purification; excess Cu²⁺ was removed through subsequent washing with HBS. MTO activity was tested assessing the release of H₂S and H₂O₂ from enzymatically generated methanethiol and copper content of the proteins was determined using TXRF; three independent experiments were performed. d, Copper content of the wildtype and mutant recombinant SELENBP1 proteins. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

SELENBP1 substrates other than methanethiol: SELENBP1 is a thiol oxidase. a, Schematic drawing of the assay developed to test for the capability of SELENBP1 to convert different volatile and non-volatile substrates. The assay was set up in reaction wells in a 384 well-plate. Generation of potential gaseous SELENBP1 substrates through MGL-catalyzed elimination reactions was followed (in adjacent reaction wells) by assessment of the volatile thiols/selenols using Ellman's reagent with spectrophotometric measurement of TNB²⁻ at 412 nm, and by testing for thiol oxidase activity of recombinant SELENBP1 using the volatile substrates (left panel). Liquid substrates were added directly to the SELENBP1 reaction wells (right panel). This scheme was created with BioRender.com. b, MGL-catalyzed generation of volatile thiols/selenols, as detected using Ellman's reagent. c, Detection of SELENBP1-catalyzed H₂S and H₂O₂ production from different substrates. Substrates tested and reaction scheme are given on the left. H₂S release (upper right panel) was assessed through lead (II)-acetate indicator paper; image of the lead sulfide precipitates by three independent protein isolates and densitometric quantitation of pixel density are shown. H₂O₂ production (lower right) was quantitated through a fluorometric HRP-coupled assay. All data are given as relative means ± SD from three independent experiments; the standard SELENBP1 substrate methanethiol was set to 1. (MT: methanethiol, ET: ethanethiol, PT: propanethiol, mPT: 2-methyl-1-propanethiol, PeT: 2-propene-1-thiol, BT: butanethiol, PtT: pentanethiol).

Fig. 5

Availability of dietary micronutrients (Cu and Se) is linked to oxidation of food-derived VSCs through the thiol oxidase activity of SELENBP1, resulting in the generation of redox signaling mediators. See text for details. Created with BioRender.com.