Methylselenol Produced In Vivo from Methylseleninic Acid or Dimethyl Diselenide Induces Toxic Protein Aggregation in Saccharomyces cerevisiae

Methylselenol (MeSeH) has been suggested to be a critical metabolite for anticancer activity of selenium, although the mechanisms underlying its activity remain to be fully established. The aim of this study was to identify metabolic pathways of MeSeH in Saccharomyces cerevisiae to decipher the mechanism of its toxicity. We first investigated in vitro the formation of MeSeH from methylseleninic acid (MSeA) or dimethyldiselenide. Determination of the equilibrium and rate constants of the reactions between glutathione (GSH) and these MeSeH precursors indicates that in the conditions that prevail in vivo, GSH can reduce the major part of MSeA or dimethyldiselenide into MeSeH. MeSeH can also be enzymatically produced by glutathione reductase or thioredoxin/thioredoxin reductase. Studies on the toxicity of MeSeH precursors (MSeA, dimethyldiselenide or a mixture of MSeA and GSH) in S. cerevisiae revealed that cytotoxicity and selenomethionine content were severely reduced in a met17 mutant devoid of O-acetylhomoserine sulfhydrylase. This suggests conversion of MeSeH into selenomethionine by this enzyme. Protein aggregation was observed in wild-type but not in met17 cells. Altogether, our findings support the view that MeSeH is toxic in S. cerevisiae because it is metabolized into selenomethionine which, in turn, induces toxic protein aggregation.

The solution of DTT that we used contained 0.5% of oxidized DTT. If c 0 represents the initial concentration of DMDSe, c 1 the initial concentration of total DTT (c 1 = [DTT] + [DTT ox ]), α the proportion of DTT ox in the DTT solution, and ξ the advancement of the reaction (defined as ξ = [MeSeH]/(2.c 0 )), the concentrations of the different compounds in the mixture of DMDSe and DTT are: Introduction of the these equations in the expression of K yields an equation from which c 1 can be expressed as a function of the other parameters: c = ( ) .
Because DTT was added at the same concentration in the reference and sample cuvettes, the absorbance of the sample at 252 nm recorded by the spectrophotometer was equal to: where ε refers to the molar absorption coefficients of the compounds at 252 nm. Using the above equations giving the concentrations of the different species, we can write the relation linking ξ to the absorbance: ξ = . ( The combination of equations (1) and (2) gives an implicit equation relating A 252 to c 1 .The value of K was obtained by fitting this implicit function to the experimental data using OriginPro software.

Calculation of the rate constant for MeSeH aerobic oxidation
MeSeH is readily oxidized by oxygen. To evaluate the rate of this oxidation, we monitored at 22°C the change of absorbance at 252 nm of MeSeH solutions prepared in deoxygenized 100 mM potassium phosphate buffer as described in Materials and Methods. At this wavelength, the absorbance of MeSeH is much higher than that of the oxidation product (DMDSe). Various concentrations of MeSeH were produced by reaction of equimolar concentrations of DMDSe and TCEP (15, 30, 45, 60 µM). The observed rates of decay in the first deduced that the oxidation of MeSeH was pseudo-first-order: with a value of k ox experimentally determined equal to 4.2 10 -4 s -1 . For the sake of comparison, we also measured the rate of MeSeH oxidation in a fully oxygenized phosphate buffer. A value of 1.7 10 -3 s -1 was determined.

Kinetic modelling of the reactions between MeSeSG, MSeH, DMDSe, GSH and GSSG
If we consider the system of the two reactions (5) and (6)  Absorbance of the sample was introduced as a 6th variable (y6). The value of y6 is given by the formula (∑ ε i y i ).pl, where ε i represents the molar absorption coefficient of compound i in µM -1 .cm -1 , y i its concentration in µM and pl represents the path length of the acquisition system expressed in cm.
Differentiation of this formula shows that y6 obeys to the equation: When the fits were realized at two wavelengths simultaneously (252 and 340 nm), another variable corresponding to the absorbances at the second wavelength was created and another differential equation using the molar absorption coefficients corresponding to this wavelength was implemented.
In all the above equations, k 2 was then replaced by K2.k -2 , were K2 designates the equilibrium constant of reaction (6) (K2 = k 2 /k -2 ). The reaction rates for reaction (6) are several orders of magnitude larger than those for reaction (5). Therefore, to analyze experiments aiming at determining the rate constants of reaction (5) (k 1 and k -1 ), we assumed that reaction (6) was always at equilibrium.