Selenium-binding protein 1 (SELENBP1) has been associated with several cancers, although its exact role is unknown. We show that SELENBP1 is a methanethiol oxidase (MTO), related to the MTO in methylotrophic bacteria, that converts methanethiol to H2O2, formaldehyde, and H2S, an activity not previously known to exist in humans. We identified mutations in SELENBP1 in five patients with cabbage-like breath odor. The malodor was attributable to high levels of methanethiol and dimethylsulfide, the main odorous compounds in their breath. Elevated urinary excretion of dimethylsulfoxide was associated with MTO deficiency. Patient fibroblasts had low SELENBP1 protein levels and were deficient in MTO enzymatic activity; these effects were reversed by lentivirus-mediated expression of wild-type SELENBP1. Selenbp1-knockout mice showed biochemical characteristics similar to those in humans. Our data reveal a potentially frequent inborn error of metabolism that results from MTO deficiency and leads to a malodor syndrome.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The authors gratefully thank the patients and families for their help in this study. The contributions of W. Lehnert, I. Goldschmidt, and C. von Schnakenburg to early investigations on family A are gratefully acknowledged. M. Antoine and T. van Alen are kindly acknowledged for their technical expertise, and Ö. Eyice-Broadbent is acknowledged for work on identifying the bacterial MTO. G. Linthorst, F. Wijburg, and H. Blom are acknowledged for help in obtaining good-quality patient samples. The authors thank P. Klaren for statistical consultation. The work was supported by the following grants: a UK Biotechnology and Biological Sciences Research Council grant to H.S. (reference BB/H003851/1), an ERC grant to H.J.M.O.d.C. (ERC 669371-Volcano), and a grant from the E.C. Noyons Foundation to R.A.W.
Integrated Supplementary Information
Breath samples of the extra-oral halitosis patient CII-2 (dotted line), a control person (dashed line), and an intra-oral halitosis patient (solid line) analysed in a portable gas chromatograph (OralChromaTM). Vertical lines indicate the retention times of H2S, MT and DMS. The vertical line indicated with * is a non-sulfur volatile background peak as described in: Hanada, M. et al. Portable oral malodor analyzer using highly sensitive In2O3 gas sensor combined with a simple gas chromatography system. Analytica Chimica Acta 475, 27-35 (2003).
The full-length alignment was performed using the Pairwise Alignment tool of the Protein Information Resource (http://pir.georgetown.edu/pirwww/search/). The Smith-Waterman score was 398. The proteins show 26.0% identity (highlighted in yellow and by *) and 54.2% similarity (highlighted in green) in 461 aa overlap. The signal peptide of the bacterial enzymes is underlined. In red putative TTQ residues (W) and copper ligands (H).
a MT oxidation results in H2S production. The amounts of MT (circles) and H2S (squares) were followed in time. Two standard assay mixtures (in 250 ml serum bottles) containing MT with- (closed symbols) or without (open symbols) Zn (0.2 mM) were started by the addition of 2 µl human control erythrocyte extract as a MTO source at the time point indicated by the arrowhead. Without Zn in the medium the molar amount of MT that has been converted is seemingly higher that the amount of H2S formed. Zn-ions will capture formed H2S. After MT was depleted the reaction mixtures were acidified thus releasing the H2S again (indicated with an arrow). In the incubation mixture with Zn the amount of sulfide formed is equimolar to the amount of MT that has been converted. In the absence of Zn-ions the formed H2S can be further enzymatically metabolized towards thiosulfate (Ref) thus explaining the lower molar recovery of H2S in the absence of Zn-ions. b Kinetic analysis of the MTO assay. Two identical standard assays using 10 µl human control erythrocyte extract were started at t = 0 under anaerobic conditions. The closed arrowhead indicates the addition of 20 ml of oxygen to incubation indicated with ●. The open arrowhead indicates the addition of oxygen to the other incubation (○) twenty minutes later. c MTO activity in colon cancer cell lines HT29 (▲) and SW480 (■). Oxygen dependence of the reaction in 3 ml exetainers is shown with cell line HT29 under anoxic condition (Δ), arrow indicates the addition of 0.25 ml of oxygen. Background disappearance of MT (●). Each data point is the average of duplicate incubations that showed less than 5% difference. d Michaelis-Menten plot of MTO activity measurements done with control erythrocyte extract using different substrate concentrations. The apparent Km is indicated. e-g Kinetic analysis of MTO activity in human erythrocyte extract e, HT29 colon cancer cell extracts f and wild type mouse liver extracts (g, duplicate measurements as open and closed symbols) under low starting concentrations of MT. In each panel the line is fitted on the data using Michaelis-Menten kinetics. This resulted in apparent Km values of 5, 4.7 and 6 nM, respectively. Ref: Szabo, C. et al. Regulation of mitochondrial bioenergetic function by hydrogen sulfide. Part I. Biochemical and physiological mechanisms. Br J Pharmacol 171, 2099-122 (2014).
Stably transduced patient cells were stained for indirect fluorescence analysis with anti-V5 antibodies that detect the C-terminal tag of SELENBP1 (red). Nuclei were stained with Hoechst dye (blue).
Sanger sequencing of the THAP4 mutation in family A shows that the c.1400C>A mutation occurs in homozygous form in the male individual AII-3 with the malodour and the neurological features while in heterozygous form in the female individual AII-2 with the malodour syndrome without neurological features. None of the other 3 sibs in this family has this THAP4 loss of function mutation in homozygous form. Black symbols in the pedigree indicate the malodour in individuals AII-2 and AII-3.
Supplementary Figures 1–5, Supplementary Table 2 and Supplementary Note.
Variants from ExAc and in-house databases for CFTR, PAH and SELENBP1. CADD_PHRED scores and variant frequencies were used to calculate the presumed frequency for SELENBP1 mutations, as described in the Supplementary Note
Full length blots