Methionine sulfoxide reduction in mammals: characterization of methionine-R-sulfoxide reductases

Abstract

Methionine residues in proteins are susceptible to oxidation by reactive oxygen species, but can be repaired via reduction of the resulting methionine sulfoxides by methionine-S-sulfoxide reductase (MsrA) and methionine-R-sulfoxide reductase (MsrB). However, the identity of all methionine sulfoxide reductases involved, their cellular locations and relative contributions to the overall pathway are poorly understood. Here, we describe a methionine-R-sulfoxide reduction system in mammals, in which two MsrB homologues were previously described. We found that human and mouse genomes possess three MsrB genes and characterized their protein products, designated MsrB1, MsrB2, and MsrB3. MsrB1 (Selenoprotein R) was present in the cytosol and nucleus and exhibited the highest methionine-R-sulfoxide reductase activity because of the presence of selenocysteine (Sec) in its active site. Other mammalian MsrBs contained cysteine in place of Sec and were less catalytically efficient. MsrB2 (CBS-1) resided in mitochondria. It had high affinity for methionine-R-sulfoxide, but was inhibited by higher concentrations of the substrate. The human MsrB3 gene gave rise to two protein forms, MsrB3A and MsrB3B. These were generated by alternative splicing that introduced contrasting N-terminal and C-terminal signals, such that MsrB3A was targeted to the endoplasmic reticulum and MsrB3B to mitochondria. We found that only mitochondrial forms of mammalian MsrBs (MsrB2 and MsrB3B) could compensate for MsrA and MsrB deficiency in yeast. All mammalian MsrBs belonged to a group of zinc-containing proteins. The multiplicity of MsrBs contrasted with the presence of a single mammalian MsrA gene as well as with the occurrence of single MsrA and MsrB genes in yeast, fruit flies, and nematodes. The data suggested that different cellular compartments in mammals maintain a system for repair of oxidized methionine residues and that this function is tuned in enzyme- and stereo-specific manner.

Figure 1.

Mammalian methionine-R-sulfoxide reductases. (A) Alignment of mammalian MsrBs. The conserved catalytic Cys and Sec (shown as U) residues are shown in red. The violet shows Ser and Thr residues that replaced the cysteine that forms a disulfide bond with the catalytic Cys in many MsrBs. Zinc-coordinating Cys residues are high-lighted in yellow. Other identical residues are indicated in gray. The predicted signal peptides are indicated as follows: green, N-terminal ER or secretory signal peptide of MsrB3A; bluish green, mitochondrial target signal of MsrB3B or MsrB2; turquoise, ER retention signal at C-terminus of MsrB3. GenBank accession numbers are as follows: human MsrB3A, BI257829; human MsrB3B, BC040053; mouse MsrB3B, NM_177092; human MsrB2, AF122004; mouse MsrB2, BC021619; human MsrB1, NM_016332; mouse MsrB1, NP_038787. (B) Genomic structures of MsrBs. Boxes indicate exons. Mouse MsrB1, MsrB2, and human MsrB3 are located on chromosomes 17, 2, and 12, respectively. MsrB3A consists of six exons, whereas MsrB3B is composed of eight exons. Translation start sites are shown by arrows.

Figure 2.

SECIS element in MsrB1-Sec/SECIS construct and amino acid sequences of the Sec-containing region. (A) Secondary structure of SECIS element for incorporating Sec in MsrB1. The mutated nucleotides are indicated as bold letters. (B) Alignment of regions downstream of Sec in MsrB1 sequences.

Figure 3.

Substrate inhibition of MsrB2. One microgram of purified MsrB2 was used in the enzyme reaction. The assay was carried out at various concentrations of dabsyl-Met-R-SO as described in MATERIALS AND METHODS.

Figure 4.

Free Met-R-SO reductase activity of mammalian MsrBs. The assay was carried out using 1 mM free Met-R-SO as described in MATERIALS AND METHODS. Fifty micrograms of MsrB1-Cys and 10 μg of MsrB2 and MsrB3, respectively, were used in the enzyme reaction. Equal amounts of the reaction mixture were spotted on to TLC plate. Lane 1, reaction mixture without enzymes (control); lane 2, MsrB1-Cys; lane 3, MsrB2; lane 4, MsrB3; lane 5, Met as a standard.

Figure 5.

Expression of mitochondrial targeted mammalian MsrBs compensates for growth of yeast cells lacking MsrA and mitochondrial MsrB genes. The yeast cells expressing various mammalian MsrBs were grown at 30°C in YNB media containing 2% dextrose in the presence of 0.14 mM L-Met (A) or 0.14 mM L-Met-R-SO (B). The constructs used in this experiment were as follows: MsrB1-Cys, full-length mouse MsrB1-Cys mutant; MsrB2, full-length mouse MsrB2; MsrB2ΔS, MsrB2 lacking N-terminal 23 aa signal peptide; MsrB3A, full-length human MsrB3A; MsrB3ΔS, MsrB3 without signal peptide of either A- or B-form; MsrB3B, full-length human MsrB3B.

Figure 6.

Schematic representation of GFP fusion constructs. The GFP fusion constructs used in this study were as follows: MsrB1-GFP, full-length mouse MsrB1-Cys fused to the N-terminus of GFP; MsrB1ΔGFP, only full-length MsrB1-Cys expressing construct without GFP; S2-GFP, N-terminal signal peptide of mouse MsrB2 fused to the N-terminus of GFP; MsrB2-GFP, full-length MsrB2 fused to the N-terminus of GFP; MsrB2ΔS-GFP, MsrB2 without signal peptide fused to the N-terminus of GFP; S3A-GFP, N-terminal signal peptide of human MsrB3A fused to the N-terminus of GFP; S3A-GFP-MsrB3, signal peptide of MsrB3A and MsrB3 without signal peptide fused to the N-terminus and the C-terminus of GFP, respectively; S3A-GFP-MsrB3ΔKAEL, the same construct as S3A-GFP-MsrB3 except for deletion of KAEL sequence at the C-terminus; GFP-MsrB3, MsrB3 without signal peptide fused to the C-terminus of GFP; S3B-GFP, N-terminal signal peptide of MsrB3B fused to the N-terminus of GFP; S3B-GFP-MsrB3, signal peptide of MsrB3B and MsrB3 without signal peptide fused to the N-terminus and the C-terminus of GFP, respectively; S3B-GFP-MsrB3ΔKAEL, the same construct as S3B-GFP-MsrB3 except for deletion of KAEL sequence.

Figure 7.

MsrB1 is located in both cytoplasm and nucleus. (A) Green fluorescence images of CV-1 cells expressing GFP-fused MsrB1 and control proteins at 20 h posttransfection. (B) Confocal images of MsrB1ΔGFP-expressing cells. The transfected cells at 18 h posttransfection were stained with anti-MsrB1 antibodies followed by secondary anti-rabbit Cy5. A set of two images is shown: left panel, immunofluorescence; right panel, phase contrast.

Figure 8.

MsrB2 is targeted to mitochondria. Confocal images of CV-1 cells expressing various GFP-tagged MsrB2 and control proteins at 12 h posttransfection. A set of three images is shown for each construct. Left panels: green fluorescence corresponding to transiently expressed fusion proteins. Center panels; fluorescence of the mitochondrial marker, Mito-Tracker Red. Right panels: images obtained by merging left and center panels. The GFP fusion constructs used in this experiment are shown on the left.

Figure 9.

MsrB3A is located in ER. Images of CV-1 cells expressing various GFP-tagged MsrB3 proteins at 20 h posttransfection. A set of three images is shown for each construct. Left panels: green fluorescence corresponding to transiently expressed fusion proteins. Center panels: fluorescence of the ER marker. ER was stained with Calregulin (C-17), followed by secondary antigoat Cy5-conjugated antibody. Right panels: images obtained by merging left and center panels. The GFP fusion constructs are shown on the left.

Figure 10.

MsrB3B is a mitochondrial protein. Images of CV-1 cells expressing various GFP-tagged MsrB3B proteins at 20 h posttransfection are shown. A set of three images is shown for each construct: left panel, green fluorescence; center panel, mitochondrial marker (MitoTracker Red); right panel, merged image. The GFP fusion constructs are shown on the left.