Investigation of the C-terminal redox center of high-Mr thioredoxin reductase by protein engineering and semisynthesis

Abstract

High-molecular weight thioredoxin reductases (TRs) catalyze the reduction of the redox-active disulfide bond of thioredoxin, but an important difference in the TR family is the sequence of the C-terminal redox-active tetrapeptide that interacts directly with thioredoxin, especially the presence or absence of a selenocysteine (Sec) residue in this tetrapeptide. In this study, we have employed protein engineering techniques to investigate the C-terminal redox-active tetrapeptides of three different TRs: mouse mitochondrial TR (mTR3), Drosophila melanogaster TR (DmTR), and the mitochondrial TR from Caenorhabditis elegans (CeTR2), which have C-terminal tetrapeptide sequences of Gly-Cys-Sec-Gly, Ser-Cys-Cys-Ser, and Gly-Cys-Cys-Gly, respectively. Three different types of mutations and chemical modifications were performed in this study: insertion of alanine residues between the cysteine residues of the Cys-Cys or Cys-Sec dyads, modification of the charge at the C-terminus, and altering the position of the Sec residue in the mammalian enzyme. The results show that mTR3 is quite accommodating to insertion of alanine residues into the Cys-Sec dyad, with only a 4-6-fold drop in catalytic activity. In contrast, the activity of both DmTR and CeTR2 was reduced 100-300-fold when alanine residues were inserted into the Cys-Cys dyad. We have tested the importance of a salt bridge between the C-terminus and a basic residue that was proposed for orienting the Cys-Sec dyad of mTR3 for proper catalytic position by changing the C-terminal carboxylate to a carboxamide. The result is an enzyme with twice the activity as the wild-type mammalian enzyme. A similar result was achieved when the C-terminal carboxylate of DmTR was converted to a hydroxamic acid or a thiocarboxylate. Last, reversing the positions of the Cys and Sec residues in the catalytic dyad resulted in a 100-fold loss of catalytic activity. Taken together, the results support our previous model of Sec as the leaving group during reduction of the C-terminus during the catalytic cycle.

Figure 1

(A) Proposed pathway for transfer of electrons to Trx by TR. The Cys residue that interacts with the flavin cofactor is labeled as “CT” (charge-transfer), while the Cys residue that acts as the interchange residue is labeled “IC”. Once the interchange Cys becomes reduced by the flavin cofactor, it initiates attack on the 8-membered ring formed by either adjacent Cys residues (DmTR and CeTR2) or adjacent Cys and Sec residues (mTR3) on the C-terminus on the opposing subunit. We refer to this step in the reaction mechanism as the “ring opening” step and is highlighted in red in the diagram. Once this ring is reduced, the attacking nucleophile (either a sulfur atom or a selenium atom – labeled as X), initiates attack on the disulfide bond of Trx. The N-terminal Cys of the dyad would then “resolve” the mixed disulfide formed between TR and Trx and is thus labeled as “Res”. The “prime” designation indicates residues that are on the adjacent subunit. (B) Diagram of the TR/Trx complex formed when the attacking nucleophile (either S or Se) attacks the disulfide bond of Trx. The adjacent Cys residue then attacks to resolve this complex, releasing product and forming the oxidized, 8-membered ring. (C) The 8-membered ring is then reduced by Cys_IC. Cys_IC can attack the N-terminal sulfur atom (pathway 1) or the C-terminal atom (S or Se, pathway 2). We have argued here and previously (17) for pathway 1). Term_aa is the C-terminal amino acid (either Gly or Ser).

Figure 2

The use of “intein engineering” to incorporate changes at the C-terminus of DmTR. DmTR is produced as a fusion protein with the VMA1 intein. The fusion protein exists in equilibrium between amide and thioester forms and the thioester form can be liberated from the intein by the addition of small molecules added to the column buffer. Addition of a thiol to the buffer results in a thioester-tagged protein. This thioester will hydrolyze to form the free carboxylic acid. Addition of hydroxylamine (NH₂OH) liberates DmTR with the C-terminus being functionalized to a hydroxamic acid, while addition of ammonium sulfide ((NH₄)₂S) results in the formation of a thiocarboxylate at the C-terminus.

Figure 3

Activity towards Trx as a function of pH for semisynthetic mTR3. Panel A shows the WT enzyme with the naturally occurring carboxylic acid (closed circle), the WT amino acid sequence produced as a C-terminal carboxamide (open square), the Sec489Cys mutant mutant (closed triangle), and the GUUG mutant (open diamond). Panel B shows the WT-carboxylate enzyme (closed circle) for reference, the GUCG mutant (checkered box), the GCAUG mutant (closed diamond), and the GCAAUG mutant (inverted, open triangle).

Figure 4

Michaelis-Menten plot of the hydrogen peroxidase activity of semisynthetic mouse enzymes. Shown are WT mTR3 (closed circles), the GCAUG mutant (open squares), and the GCAAUG mutant (closed triangles). The dramatic shift in K_m is apparent.