Distinction of thioredoxin transnitrosylation and denitrosylation target proteins by the ICAT quantitative approach

Abstract

S-Nitrosylation is a reversible PTM for regulating protein function. Thioredoxin-1 (Trx1) catalyzes either transnitrosylation or denitrosylation of specific proteins, depending on the redox status of the cysteines within its conserved oxidoreductase CXXC motif. With a disulfide bond formed between the two catalytic cysteines, Trx1 is not only inactive as a denitrosylase, but it may also be nitrosylated at Cys73 and serve as a transnitrosylating agent. Identification of Trx1-mediated transnitrosylation or denitrosylation targets will contribute to a better understanding of Trx1's function. Previous experimental approaches based on the attenuation of CXXC oxidoreductase activity cannot readily distinguish Trx1 transnitrosylation targets from denitrosylation targets. In this study, we used the ICAT method in conjunction with the biotin switch technique to differentiate Trx1 transnitrosylation targets from denitrosylation target proteins from neuroblastoma cells. We demonstrate that the ICAT approach is effective for quantitative identification of putative Trx1 transnitrosylation and denitrosylation target peptides. From these analyses, we confirmed reports that peroxiredoxin 1 is a Trx1 transnitrosylation, but not a denitrosylation target, and we found several other proteins, including cyclophilin A to be modulated in this manner. Unexpectedly, we found that many nitrosylation sites are reversibly regulated by Trx1, suggesting a more prominent role for Trx1 in regulating S-nitrosylation.

Fig. 1. Schematic of Trx1 regulation of protein nitrosylation

Trx1 Cys32 and Cys35 can denitrosylate specific SNO-proteins [10], [11] and [12], resulting in the formation of a disulfide bond between Cys32 and Cys35 [10] and [13]. SNO-Trx1 can be produced in cells after the formation of a Cys32 and Cys35 disulfide bond and act as a transnitrosylase for specific targets [19] and [20]. It should be noted that SNO-Trx1 transnitrosylation targets and rTrx1 denitrosylation targets may not overlap. Adapted from [20].

Fig. 2. Comparison of biotin-HPDP and ICAT labeling for detection of SNO-proteins and peptides

(A) Schematic of ICAT-based quantitative proteomics of SNO-Trx1 target proteins. Following incubation of cellular proteins without (control) or with SNO-Trx1, protein SNO-Cys sites are preserved by BST: (I) free thiols are alkylated with methylmethanethiosulfonate (MMTS). (II) Nitrosylated residues are reduced by ascorbate. (III) Nascent free thiols are labeled by a thiol reactive biotin-containing reagent. In this example, SNO-Trx1 treated sample is labeled with ICAT-H, while untreated sample is labeled with ICAT-L. (IV) Labeled samples are combined, proteolytically digested, and their SNO-Cys sites determined and quantified by LC/MS/MS analysis of ICAT-labeled peptides. (B) Comparison of biotin-HPDP and ICAT detection specificity. SH-SY5Y cell lysate was in vitro transnitrosylated by incubation with SNO-Trx1. Resultant SNO-proteins were modified by BST using biotin-HPDP or ICAT-L or -H reagent, resolved by non-reducing SDS-PAGE and detected by Western blotting with an anti-biotin antibody (top panel). Protein levels were normalized to actin (bottom panel). (C) Comparison of biotin-HPDP and ICAT-H for identification of SNO-Trx1 derivated SNO-peptide nitrosylation sites. Biotinylated proteins from B (SNO-Trx1 treated) were trypsin digested, enriched by avidin column, the ICAT-H biotin tether cleaved and eluted peptides detected by LC/MS/MS. Venn diagrams show the reproducibility of detecting biotin-HPDP labeled (top) or ICAT-H labeled (middle) peptides from 2 whole reaction repeats. Bottom: Venn diagram compares the commonality of total peptides detected from biotin-HPDP labeling experiments with ICAT-H labeling experiments. Total number of labeled peptides detected in each experiment is in parentheses and are listed in Table S1. 353 peptides were enriched after biotin-HPDP labeling, 40 peptides were unlabeled (11.3%), including 39 peptides without Cys. 243 peptides were enriched after ICAT labeling, 74 peptides were unlabeled (30.5%), including 64 peptides without Cys. (D) Evaluation of ICAT SNO-peptide quantification statistics. SH-SY5Y cell lysate was incubated with GSNO, then divided equally and labeled by BST using light or heavy ICAT reagent. After tryptic digest, ICAT samples were combined 1:1 and biotinylated peptides detected by LC/MS/MS. A box plot is shown of the ICAT light/heavy ratio of 22 ICAT-labeled peptide pairs averaged among the 3 whole reaction repeats (listed in Table S2 and reproducibility illustrated in Fig. S3). Solid horizontal line indicates the median, vertical lines the maximum and minimum, and the box represents the middle 50% of the data.

Fig. 3. In vitro Trx1 transnitrosylation and denitrosylation

SH-SY5Y cell lysate was incubated with or without SNO-Trx1. Lysate treated with SNO-Trx1 was then incubated with or without the Trx reductive system. SNO-proteins in extract treated with SNO-Trx1 were labeled with ICAT-H, while untreated or those treated additionally with the thioredoxin reductive system were each labeled with ICAT-L. Proteins were then resolved by reducing SDS-PAGE and biotinylated proteins detected by Western blot with an anti-biotin antibody.

Fig. 4. Identification and quantification of Trx1 target SNO-peptides

Example MS and MS/MS spectra of (A) GAPDH 235-VPTANVSVVDLTC*R-248 (M2+ 855.45, C : ICAT-label site identified in the upper panel MS/MS spectrum ) and (B) ΕΝΟΑ 344-VNQIGSVTESIQAC*K-358 (M2+ 907.01, sequence identified in the upper panel MS/MS spectrum). Cys247 of GAPDH was transnitrosylated by SNO-Trx1 and denitrosylated by rTrx1/TrxR (A, compare ICAT-H and -L peaks in the lower panel MS spectra), while Cys357 of α-enolase was transnitrosylated by SNO-Trx1 but not denitrosylated by Trx1/TrxR (B, compare ICAT-H and -L peaks in the lower panel MS spectra).

Fig. 5. Western blot analysis of protein S-nitrosylation status

Biotinylated proteins were enriched by avidin affinity capture, resolved by SDS-PAGE and Western blotted for formerly nitrosylated α-tubulin, GAPDH, Prx1, cyclophilin A and α-enolase. Values are the mean ± S.E. for experiments performed in triplicate (*, p = 0.0005; **, p = 0.001; †, p = 0.0006; ‡, p = 0.0001; #, p = 0.0003; §, p = 0.002; ¶, p = 0.0006; Student’s t-test). Total proteins were analyzed in parallel to detect proteins in cell lysates regardless of their nitrosylation status.