Selenium in Thioredoxin Reductase: A Mechanistic Perspective

The thioredoxin system, comprised of thioredoxin reductase (TR), thioredoxin (Trx), and NADPH, is essential for maintaining redox balance in eukaryotic cells (1-2). Most eukaryotes contain a high Mr TR that is very similar in structure and mechanism to glutathione reductase (GR) (3). Both enzymes use NADPH as a redox cofactor to reduce a conserved disulfide redox center (…CVNVGC…). TR can be distinguished from GR by the presence of a C-terminal extension of 16 amino acids that contains an additional redox center that is functionally equivalent to oxidized glutathione (GSSG), the disulfide substrate for GR. In the case of GR, the reduced product (GSH) acts as a small molecule shuttle to carry reducing equivalents to many targets, especially glutaredoxin. TR however, utilizes the C-terminal redox center as an internal substrate shuttle (as part of the polypeptide chain) to convey reducing equivalents directly to the target substrate (Trx). The C-terminal redox center of many high Mr TRs are unique in that they contain a conserved tetrapeptide motif of Xaa-Cys1-Cys2-Xaa (where Xaa is usually Gly or Ser) where Cys1 and Cys2 form a rare vicinal disulfide bond during the catalytic cycle (4). This type of disulfide bond has only been found in ~50 protein structures deposited in the Protein Data Bank (PDB) (5). The [Cys-Cys]ox dyad is part of a two residue turn that can fit the molecular constraints of types I, II, VIa, VIb, and VIII beta turns (5). In a few instances, it is thought to be part of a regulatory conformational switch as this motif can undergo a cis/trans isomerization of the peptide bond between Cys1 and Cys2 that would dramatically alter the protein topology where it is found (6). Mammalian and other higher eukaryotic TRs have the interesting and unique feature of replacing Cys2 with a selenocysteine (Sec, U) residue (7). This Sec residue is essential to the function of the mammalian enzyme since mutation of Sec to Cys causes a large drop in kcat (8), however the presence of a selenium atom is not chemically required to catalyze the reduction of the disulfide bond of Trx (9). This point is further illustrated by the data in Table 1, which summarizes kinetic data for Sec-containing enzymes from mammals contrasted with Cys-containing TRs from D. melanogaster (DmTR), C. elegans (CeTR2), and A. gambiae (AgTR). These three Cys-TRs have a conventional Cys residue in the Cys2 position of the tetrapeptide motif and DmTR and AgTR catalyze the reduction of their respective cognate substrates with comparable catalytic efficiencies to the mammalian enzyme with only slightly reduced kcat values. The value of kcat for CeTR is substantial, although E. coli Trx was used as the substrate. In contrast, mutation of Sec to Cys in the mammalian enzyme results in losses of 175-fold and 540-fold in kcat for the cytosolic rat enzyme (rTR) and mitochondrial mouse enzyme (mTR3), respectively2 (8,10). Table 1 Kinetic Constants of Various TRs for the Reduction of Thioredoxin. Why is the activity of the mammalian enzyme so greatly affected by the loss of a selenium atom when the data in Table 1 shows that a sulfur atom can replace it in other eukaryotes without a dramatic loss in activity? As shown in Figure 1A, the presumptive nucleophile for the reduction of the disulfide bond of Trx is the residue that occupies the Cys2 position (X represents either S or Se in Figure 1) of the Cys1-Cys2 dyad. A reasonable assumption (though not explicitly stated in the literature) is that the loss of nucleophilicity that results from Sec to Cys mutation is responsible for the decrease in rate. The nucleophilicity of the Cys2 residue could be increased by lowering its pKa and low pKa thiolates in enzymes are certainly known (13). A mechanism for lowering the pKa of Cys2 (the presumptive nucleophile – Cys490’) in DmTR has been proposed (14-15). Figure 1 Comparison of the catalytic mechanisms of full-length TR (A), truncated TR utilizing a peptide substrate (B), GR utilizing oxidized GSSG as a substrate (C), and truncated TR utilizing the highly reactive disulfide DTNB as a substrate (D). In each case, ... While attenuating the nucleophilicity of the thiolate in a Cys-TR is one possible mechanistic explanation for the high catalytic efficiencies of Cys-TRs relative to Sec-TRs, other possibilities do exist. Figure 1 shows the mechanistic similarity of TR and GR, with TR containing an extra thiol-disulfide exchange step before electrons are transferred to the substrate. We refer to the additional thiol-disulfide exchange step of TR as the “ring opening” step because it involves reduction and opening of the eight-membered ring formed by the Cys1-Cys2/Sec2 dyad (16). We have previously investigated this step in the mechanism because we noted that the truncated enzyme, although mechanistically similar to GR, could only utilize a highly reactive disulfide such as 5,5′dithio-bis(2-nitrobenzoic acid) (DTNB) as a substrate, but not other simple disulfides such as cystine or GSSG. This observation led us to investigate the requirement for Sec in the mammalian enzyme by using the truncated enzyme and peptide disulfide substrates to study the ring opening reaction (16-17). Here we test the hypothesis that the difference between Cys-containing TRs and Sec-containing TRs is the dependence each type of enzyme has on the unique eight-membered ring structure in the ring opening step of the catalytic cycle. Our approach was to synthesize peptide substrates with an eight-membered ring structure formed by a vicinal disulfide/selenylsulfide (cyclic peptide), or peptides substrates containing an interchain disulfide/selenylsulfide linking two peptides together (acyclic peptide). The acyclic peptide had the same sequence as the cyclic peptide except that the peptide bond of the Cys1-Cys2/Sec2 dyad is “broken” in comparison to the cyclic peptide (Figure 2). A second type of acyclic peptide was also synthesized that replaced the Cys2/Sec2 position with a 5-thio-2-nitrobenzoate group. Lastly, in order to differentiate the function of each position of the dyad, we also synthesized peptides in which Sec was placed in the Cys1 position. Our results using these unique substrates provide a mechanistic explanation for the presence of Sec in the mammalian enzyme. Figure 2 Peptide substrates constructed for this study: A) Cyclic, oxidized octamer; B) Acyclic, oxidized octamer; C) Acyclic hexamer in which Cys1 forms a mixed disulfide with the thiol of TNB, referred to as Pep-TNB in the text; and D) Cyclic octamer “switch” ...