Data-Independent Acquisition (DIA)-Based Label-Free Redox Proteomics (DIALRP) Identifies Prominent Cysteine Oxidations in Translation Machinery in Prostate Cancer Cells Under Oxidative Stress

Abstract

Oxidative stress is a key factor in numerous physiological and pathological processes, including aging, cancer, and neurodegenerative diseases. Protein cysteine residues are particularly susceptible to oxidative stress-induced modifications that can alter their structure and function, thereby affecting intracellular signaling pathways. In this study, we performed a data-independent acquisition mass spectrometry (DIA-MS)-based label-free redox proteomics method, termed DIALRP, to comprehensively analyze cysteine oxidative modifications in the prostate cancer cell line DU145 under oxidative stress induced by menadione (MND). Of 10,821 cysteine-containing peptides identified, we successfully quantified the redox changes in 3665 peptides. We also observed that 1407 peptides were significantly oxidized in response to MND treatment. Gene ontology analysis revealed that a group of translation-related molecules was most enriched among highly MND-sensitive cysteine-containing proteins. Notably, our data demonstrated that MND-induced oxidative stress inhibits EIF2, EIF6, and EEF2 complex formation, suggesting that these complex inhibitions become functional factors for a dramatic reduction in translation activity. Our results show that DIALRP is utilized as a robust and cost-effective approach for investigating redox-regulated cellular processes. Moreover, these findings provide significant insights into translation regulation under oxidative stress and provide a valuable framework for future studies on redox-mediated cellular processes.

Introduction

Oxidative stress is a hallmark of various physiological and pathological processes, including aging, cancer, and neurodegenerative diseases. − Reactive oxygen species (ROS), the major mediators of oxidative stress, when produced in excess, disrupt cellular homeostasis by oxidizing lipids, nucleic acids, and proteins, leading to functional changes and cellular damage. In particular, proteins can undergo oxidative modifications, resulting in conformational changes, altered enzyme activity, and perturbation of cellular signaling pathways. One of the major oxidative modifications of proteins occurs at cysteine residues, resulting in reversible and irreversible oxidative post-translational modifications of cysteine. Among various ROS, hydrogen peroxide (H₂O₂) is particularly important, as it is relatively stable, membrane-permeable, and a key oxidant that selectively modifies cysteine thiols to sulfenic acid. The reversible modifications alter the structure and function of proteins ,− and several studies have demonstrated their functional roles in signal transduction, metabolic adaptation, and developmental hematopoiesis. However, modifications of specific cysteines can have diverse consequences depending on the target and type; hence, the effects of ROS-induced protein modifications on cellular signaling are not yet fully understood.

Redox proteomics has emerged as a powerful approach to systematically identify and quantify protein cysteine oxidative modifications, providing insights into the cellular responses to oxidative stress. Particularly in cancer biology, it could be useful for a comprehensive analysis of the oxidative stress response. Cancer cells are highly susceptible to oxidative stress, which significantly affects their survival, proliferation, and metastatic potential. , Prostate cancer is one type of cancer in which the relationship between malignant transformation and ROS has been reported. In particular, DU145 is a well-established human prostate cancer cell line derived from a brain metastasis and an Nrf2-dependent cell line that adapts to stress responses. Therefore, this cell line is frequently used in studies of oxidative stress, and redox-related signaling in prostate cancer models. ,

Redox proteomics methods can be broadly performed with two quantification manners: those that quantify relative changes in the levels of oxidized cysteines between conditions (e.g., fold change-based) and those that aim to estimate the percentage of oxidation (occupancy or stoichiometry) for each cysteine site within a sample. The latter provides a more quantitative view of the redox status and is adopted in the present study. Redox proteomics is also often performed using isobaric tag reagents such as tandem mass tags (TMT). To enhance comprehensiveness, cysteine-containing peptides were subjected to enrichment methods utilizing click reaction-, anti-iodoTMT antibody-, and cysteine-reactive phosphate tag (CPT), which selectively labels reduced cysteine thiols with phosphate groups and allows for enrichment similar to phosphorylated peptides such as immobilized metal affinity chromatography (IMAC), to increase the number of identified peptides. A cysteine-labeling method using stable isotope-labeled iodoacetamide has also been reported. This method does not enrich cysteine peptides but instead improves peptide identification by peptide fractionation. These methods require sophisticated protocols, considerable sample volumes for the enrichment of cysteine-containing peptides, and expensive stable isotope-labeling reagents. Recently, data-independent acquisition mass spectrometry (DIA-MS)-based redox proteomics methods using a label-free approach or stable isotopes have been reported. In particular, a DIA-based label-free method allows analysis without the need for the enrichment of cysteine-containing peptides or expensive stable isotope labeling reagents. However, this approach has so far been limited to profiling the cysteine redox state in conditioned media from H₂O₂-treated cultured cells and requires the construction of a cysteine-containing peptide identification data set because this analysis was performed using a spectral library.

In this study, we performed a label-free redox proteomics method based on DIA-MS with a spectral library-free search and measured global cysteine redox changes in DU145 cells under oxidative stress conditions caused by menadione (MND). MND is a widely used reagent that generates ROS during oxidative metabolism, allowing researchers to investigate cellular responses to intrinsic oxidative stress in a controlled manner. Consequently, we identified several translation-related factors undergoing oxidative modification following MND treatment, providing new insights into the molecular mechanisms regulating protein synthesis in response to oxidative stress.

Experimental Procedures

Cell Line and Culture Conditions

DU145 cells were cultured under 5% CO₂ at 37 °C in a RPMI1640 medium (Fujifilm-Wako, Japan) supplemented with 10% fetal bovine serum (FBS) (NICHIREI BIOSCIENCES, Japan), Penicillin/streptomycin (10,000 U/mL) (Thermo Fisher Scientific, MA, USA), and Sodium pyruvate (100 mM) (Thermo Fisher Scientific, MA, USA). HeLa cells were cultured under 5% CO2 at 37 °C in Dulbecco’s modified Eagle’s medium (Fujifilm-Wako, Japan) supplemented with 10% FBS (NICHIREI BIOSCIENCES, Japan), Penicillin/streptomycin (10,000 U/mL) (Thermo Fisher Scientific, MA, USA), MEM Non-Essential Amino Acids Solution (100×) (Thermo Fisher Scientific, MA, USA), and Sodium pyruvate (100 mM) (Thermo Fisher Scientific, MA, USA).

Cell Lysate Preparation

For a reduced cysteine measurement by 5,5′-dithiobis­(2-nitrobenzoic acid) (DTNB), redox proteomics, and redox Western blotting, the cell culture was started in 6 cm dishes or 6 well plates. After 24 h, when cell culture confluency reached 80%, cells were treated with MND for 1 h, washed with PBS twice, and solubilized in lysis buffer containing 7 M urea, 1% Triton X-100, and 100 mM HEPES-NaOH (pH 7.2) for 15 min at 4 °C on a shaker. For FLAG-IP-MS analysis, cells were washed with PBS twice and solubilized in lysis buffer containing 50 mM Tris–HCl (pH 7.4), 150 mM NaCl, 0.5% Triton X-100, 5 mM MgCl₂, and 1% (v/v) protease inhibitor mixture (Roche, Switzerland) for 15 min at 4 °C on a shaker. Lysates were centrifuged at 20,000g for 15 min at 4 °C, and the protein concentration of the supernatants was determined using the BCA Protein Assay Kit (Takara Bio, Inc., Japan).

ROS Measurement

After treatment with 6.25, 12.5, 25, 50, and 100 μM MND for 1 h, culture media was removed, and the cells were incubated with the culture media containing 5 μM CellROX Green Reagent (Thermo Fisher Scientific, MA, USA) and 0.5 μg/mL Hoechst33342 (Dojindo, Japan) at 37 °C for 30 min. Cells were collected using trypsin, washed with PBS, and gently resuspended in PBS. The fluorescence intensities of CellROX and Hoechst were measured using a Fluoroskan FL microplate reader (Thermo Fisher Scientific).

Measurement of Reduced Cysteines in Cell Lysates

Fifty micrograms of protein were treated with 200 μM DTNB (Dojindo, Japan) in 50 μL of lysis buffer for 10 min at room temperature. The absorbance of 2-nitro-5-thiobenzoic acid (TNB) in the sample was measured at 415 nm using a BioRad iMark microplate reader (Bio-Rad Laboratories, TX, USA).

Sample Preparation for Redox Proteome Analysis

Ten micrograms of protein were treated with 10 mM N-ethylmaleimide (NEM) in 50 μL of lysis buffer. The NEM was eliminated using the single-pot solid-phase enhanced sample pretreatment (SP3) method. For the sample to quantify all of cysteine peptides, lysates containing 10 μg of protein were directly applied to SP3 beads. SP3 beads were suspended with 50 μL of 100 mM HEPES-NaOH (pH 8.0) containing 0.4 μg of trypsin (Richcore, India), and trypsin digestion was performed at 37 °C for 16 h. The supernatant was collected and treated with 2.5 mM tris­(2-carboxyethyl)­phosphine hydrochloride (TCEP) for 30 min at 37 °C, followed by treatment with 10 mM iodoacetamide (IAA) for 30 min at room temperature. TFA was added to the samples at a final concentration of 1%. The sample was desalted using styrenedivinylbenzene (SDB)-StageTip and dissolved in 0.1% TFA in 2% acetonitrile for LC–MS analysis.

Plasmid Construction and Transfection

To obtain cDNA, total RNA was extracted from HeLa cells using a TRIzol reagent (Thermo Fisher Scientific) according to the manufacturer’s protocol. The reverse transcription of first strand cDNA was performed with ReverTra Ace (TOYOBO, Japan) using the total RNA as a template. The PCR was performed with KOD One PCR Master Mix (TOYOBO, Japan). The forward and reverse primers for RPL12, RPS11, MRPL50, TSFM, EIF2S2, EIF6, and EEF2 were designed to ligate these cDNAs with an expression plasmid, as listed in Table S1. The PCR was performed using a Bio-Rad T100 Thermal Cycler (Bio-Rad Laboratories, TX, USA). With the following thermal cycling parameters: 94 °C for 2 min, followed by 35 cycles of 98 °C for 15 s, 56 °C for 30 s, and 72 °C for 1 min, and the final extension was performed at 72 °C for 10 min. The PCR products were cloned into pCMV-(DYKDDDDK)-C vector (Clontech, CA, USA) using Gibson Assembly Master Mix (NEB, MA, USA) according to manufacturer’s protocol and were sequenced. The constructed expression plasmids were used to transfect the cultured cells. HeLa cells were transfected with expression plasmids using a Fugene HD reagent (Promega, Madison, WI, USA) according to the manufacturer’s protocol.

FLAG-Immunoprecipitation and Sample Preparation for MS Analysis

Five microliters of anti-FLAG M2Magnetic Beads (Sigma-Aldrich, MO, USA) were washed twice with 200 μL of TBS (50 mM Tris–HCl, 150 mM NaCl, pH 7.4) buffer. The cell lysate containing 100 μg of protein was added to the washed resin beads. The mixture was incubated for 30 min at 4 °C with gentle shaking. The beads were washed three times with 400 μL of TBS. To prepare the sample for mass spectrometry, 50 μL of 0.5 mg/mL FLAG solution in 50 mM Tris HCl pH 7.4, 150 mM NaCl, and 0.05% Triton X-100 were added to the washed resin beads, and the mixture was incubated at room temperature with gentle shaking for 5 min. The supernatant was obtained and further treated to analyze. 160 microliters of solution containing 8.75 M urea and 50 mM Tris–HCl (pH 8.5) were added to 40 μL of eluted sample. Samples were treated with 2.5 mM TCEP for 30 min at 37 °C, followed by treatment with 10 mM IAA for 30 min at room temperature. Acetone was added to the sample at a final concentration of 80%, mixed, and incubated for 30 min at room temperature. The sample was centrifuged at 20, 000g for 15 min at room temperature to precipitate proteins. The precipitate was washed twice with 1 mL of 90% acetone. After being dried, 50 μL of 100 mM HEPES-NaOH pH 8.0 containing 200 ng trypsin were added and incubated at 37 °C for 16 h. TFA was added to the sample at a final concentration of 1%. The sample was desalted using SDB-StageTip and dissolved in 0.1% TFA and 2% acetonitrile for LC–MS analysis. Triplicate biological samples were prepared from all of the bait samples. Triplicate biological samples from HeLa cells transfected with the empty vector were prepared as controls to extract the binding proteins. To evaluate the change in the expression of the binding proteins, HeLa cells treated with 50 μM MND for 1 h were lysed with the same solution using FLAG-IP-MS analysis. Forty microliters of solution containing 8.75 M urea and 50 mM Tris–HCl (pH 8.5) were added to 10 μL of cell lysate containing 10 μg of protein. Samples were treated with 2.5 mM TCEP for 30 min at 37 °C, followed by treatment with 10 mM IAA for 30 min at room temperature. The sample was applied to SP3 methods to remove TCEP and IAA, suspended with 50 μL of 100 mM HEPES-NaOH (pH 8.0) containing 0.4 μg of trypsin (Richcore, India), incubated at 37 °C for 16 h. TFA was added to the sample at a final concentration of 1%. The sample was desalted using SDB-StageTip and dissolved in 0.1% TFA and 2% acetonitrile for LC–MS analysis.

LC–MS

Peptide samples were injected into a precolumn (L-column2 micro, CERI, Japan) and fractionated on an in-house fabricated 15 cm column packed with 2 μm octadecyl silane particles (CERI, Japan). Elution was performed with a linear gradient of 5%–32% solvent B over 60 min at a flow rate of 200 nL/min (solvent A = 0.1% formic acid; solvent B = 0.1% formic acid in acetonitrile) with the use of a Dionex Ultimate 3000 HPLC System (Thermo Fisher Scientific, MA, USA). The eluted peptides were sprayed with a nanoelectrospray source and a column oven set at 42 °C (AMR, Japan). A Q Exactive Hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) was operated in the DIA mode. All data were acquired in profile mode using positive polarity. MS1 spectra were collected in the m/z range of 430–860 at a resolution of 35,000 using an AGC target value of 1 × 10⁶ with a maximum injection time of 50 ms. MS2 spectra were collected in the range of >200 m/z at a resolution of 17,500 using an AGC target value of 1 × 10⁶ and automatic maximum ion ITs. Twenty-one DIA windows of 20 units ranged from 430 to 850 m/z, with an overlap of one unit. The normalized collision energy was set to 25.

MS Data Processing

All raw DIA data were processed using DIA-NN (ver. 1.8.1) using a library-free search mode against human UniProt/SwissProt sequences (release 2024_02, 20,434 entries) to identify tryptic peptides. The enzyme specificity was set to “trypsin”. Up to one missed trypsin cleavages were allowed. Carbamidomethylation of cysteine was set as a fixed modification, and oxidation of methionine was set as a variable modification. The FDR threshold for peptide identification was set to 1%. A match between the run mode was used for both DIALRP and FLAG-IP MS. For DIALRP and FLAG-IP MS, cross-run normalization was set as “RT-dependent and signal dept.” and “None”, respectively. For DIALRP, the positions of the cysteines in all identified peptides were annotated based on UniProt/Swiss-Prot sequences using a Python script. The oxidation percentage was calculated based on the intensity of carbamidomethylated cysteine-containing signals from NEM-treated (Oxi-Cys) and non-NEM-treated (All-Cys) samples. The oxidation percentages were averaged when redundantly quantified due to the difference in charges. Oxidation percentages exceeding 100% due to the quantification noise were capped at 100% for interpretability. Cysteine residues annotated as forming “disulfide bonds” were retrieved from UniProtKB. The frequency of disulfide annotations was compared among three groups of cysteine sites based on their oxidation stoichiometry: low (0–10%), high (90–100%), and total cysteine sites with an oxidation percentage (0–100%). The percentages calculated under DMSO-treated conditions were used to minimize the oxidation bias.

For the evaluation of the detectability of NEM-modified cysteine peptides and the labeling efficiency of IAA, the database search was performed using MSFragger-DIA in FragPipe platform. Carbamidomethylation (+57.0215) and NEM modifications (UniMod:108 and UniMod:320, +125.0477 and +143.0582) were set as variable modifications on cysteine residues. The search was performed against human UniProt/SwissProt sequences (release 2024_02, 20,434 entries), and the FDR threshold for peptide identification was set to 1%.

The significantly oxidized cysteine peptides, differentially expressed proteins, and altered binding proteins were extracted by statistical tests using Perseus software. For the analysis of oxidized cysteine peptides, missing oxidation percentages of peptides quantified in all four DMSO- or MND-treated replicates were applied to imputations from normal distribution. Multiple testing with Student’s t-test and Benjamini-Hochberg FDR, with a q-value of less than or equal to 0.01 as a cutoff value, was performed to extract peptides that underwent significant oxidative modifications and differentially expressed proteins. For the analysis of FLAG-IP-MS data, the quantitation data of all replicates in the bait samples were normalized according to the bait intensities. The missing quantification values of proteins in all three bait samples treated with DMSO-treated were applied to imputations from normal distribution. To extract binding proteins under DMSO-treated conditions and altered binding proteins under MND-treated conditions, multiple testing with Student’s t-test and Permutation-based FDR was performed with q-value less than or equal to 0.05 as the cutoff value.

Gene Ontology Enrichment and Network Analysis

Gene Ontology (GO) enrichment analysis was performed by functional annotation clustering using DAVID. , The background was set to the whole Homo sapiens genome. GOTERM_BP_DIRECT, GOTERM_CC_DIRECT, and GOTERM_MF_DIRECT were selected for the analysis. All proteins, which were annotated with GO “translation (Accession: GO:0006412)”, were extracted and the network of these proteins was visualized using STRING Ver 12.0 (https://string-db.org/). The parameters were set as follows: network type: full STRING network; meaning of network edges: evidence; active interaction sources: text mining, experiments, databases, coexpression, neighborhood, gene function, and cooccurrence; minimum required interaction score: 0.400; maximum number of interactors to show: first shell = no more than 10 interactions; second shell = none. Functional classification was performed using enriched GO terms in the list of functional enrichments in STRING.

Principal Component Analysis

Principal component analysis (PCA) was performed by using a Python script. Prior to analysis, data were log2-transformed and mean-centered. Score plots and loading plots were generated using matplotlib and seaborn. The percentage of variance explained by each component was used to interpret the data structure.

Solvent-Accessible Surface Area Analysis

To analyze the SASA of proteins, we utilized FreeSASA and AlphaFold , within a Python-based workflow. The PDB-formatted structural models were downloaded from AlphaFold. The SASA values (Ų) were computed for each residue in the structural models using FreeSASA. The FreeSASA parameters were set as follow; algorithm: Lee & Richards, probe-radius: 1.400, slices: 20. The SASA values of a specific set of cysteine residues were selectively extracted.

Nascent Polypeptide Detection

The cell culture was started in 6 well plates. After 24 h, when cell culture confluency reached to 80%, cells were treated with MND for 1 h, washed once with PBS and incubated in a methionine-free medium containing 100 μM l-azidohomoalanine (AHA) reagent (Biosynth, UK) at 37 °C for 2 h. The cells were lysed with 50 mM Tris–HCl (pH 7.4), 150 mm NaCl, 0.5% Triton X-100, and 1% (v/v) protease inhibitor mixture (Roche, Switzerland), for 15 min at 4 °C on a shaker. Lysates were centrifuged at 20,000g for 15 min at 4 °C, and the protein concentration of the supernatants was determined using the BCA Protein Assay Kit (Takara Bio, Inc., Japan). Ten micrograms of protein were treated with 10 μM tetramethylrhodamine (TAMRA)-alkyne (Sigma-Aldrich, MO, USA), 1 mM TCEP, 100 μM tris­[(1-benzyl-1H-1,2,3-triazol-4-yl)­methyl] amine (TBTA) and 1 mM CuSO₄ in lysis buffer for 2 h at room temperature. The samples were precipitated with acetone to a final concentration of 80% (v/v). Samples were centrifuged to precipitate the proteins at 20,400g for 15 min at 4 °C. The obtained precipitants were washed twice with 90% (v/v) acetone and dissolved in SDS-PAGE sample buffer with the incubation at 95 °C for 5 min. All proteins were separated on 12% SDS-polyacrylamide gels. The TAMRA-labeled nascent polypeptides were detected using a Bio-Rad ChemiDoc Touch MP imaging system (Bio-Rad Laboratories, TX, USA). The TAMRA fluorescence intensity was quantified using ImageJ software. To normalize the protein amounts, the gels were stained with Coomassie Brilliant Blue G-250 after TAMRA detection and reimaged using the same system.

Redox Western Blotting

The lysate was treated with 10 mM mPEG-maleimide 1 K (Biopharma PEG Scientific Inc., MA, USA) in lysis buffer for 1 h at room temperature. For the validation of EEF2, mPEG-maleimide 2 K (Biopharma PEG Scientific Inc., MA, USA) was used to analyze. To mimic fully oxidized or reduced forms, the lysate was treated with 2.5 mM TCEP for 30 min at 37 °C, followed by treatment with 10 mM NEM or mPEG-maleimide. After the treatment, the sample was precipitated with acetone at a final concentration of 80% (v/v). The sample was centrifuged to precipitate the proteins at 20,000g for 15 min at 4 °C. The precipitant was washed twice with 90% (v/v) acetone and dissolved in SDS-PAGE sample buffer with the incubation at 95 °C for 5 min. Ten micrograms of the labeled protein were separated onto 8, 12, or 14% SDS-polyacrylamide gels according to their molecular weights, transferred onto a PVDF membrane by electroblotting, and subjected to immunoblotting with the primary antimouse anti-DDDDK-tag antibody (MBL Life science, M185-3L, Japan) or antirabbit anti-eIF6 antibody (Cell Signaling Technology, 3833, MA, USA). After incubation with the primary antibody, the membrane was probed with a horseradish-peroxidase-conjugated mouse secondary antibody (Promega). The image was visualized with SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo fisher Scientific, MA, USA) and was detected using a BioRad ChemiDoc Touch MP imaging system (Bio-Rad Laboratories, TX, USA).

Immunocytochemical Analysis

HeLa cells cultured on glass plates were fixed with 3.7% formaldehyde in PBS for 15 min at room temperature and then permeabilized with 0.1% Triton X-100 in PBS on ice for 15 min. After washing with PBS, cells were incubated in primary Anti-DDDDK antibodies (Cell Signaling Technology, 14,793, MA, USA) diluted in PBS containing 5% bovine serum albumin followed by antirabbit Alexa Fluor 488-conjugated IgG (Invitrogen, CA, USA) for 60 min at room temperature. After washing with PBS, the glass plates were mounted on slide glasses using ProLong Glass Antifade Mountant with NucBlue Stain (Invitrogen, CA, USA). Images were acquired using a DMi8 microscope equipped with a Leica Thunder Imaging System (Leica, Germany). Images were processed using Leica LAS X software (Leica, Germany).

Statistical Analysis

All values are expressed as the mean with the ±S.D., and significant differences between groups were assessed using Dunnett’s test, except for MS data.

Results

Overview of DIALRP and Its Application to the Analysis of Menadione Response in DU145 Cells

Mass spectrometry-based methods for redox proteomic analysis rely primarily on stable isotopes. In this study, we performed a label-free redox proteomic analysis method based on DIA-MS (Figure A). Redox proteomics methods require the almost complete removal of excess cysteine-modifying reagents. We employed the single-pot solid-phase-enhanced sample preparation (SP3) method, a bead-based cleanup protocol that enables efficient and scalable sample processing, as previously reported. Each cell lysate was divided into two portions: one where the reduced cysteines were modified with NEM, and the other was left unmodified. Proteins were captured by using magnetic beads, washed to remove residual reagents, and digested with trypsin. After digestion, both samples were treated with TCEP and IAA to modify the residual free cysteines. The resulting peptides were subjected to DIA-MS to identify and quantify cysteine-containing peptides. The percentage of oxidative modification was calculated by comparing the abundance of carbamidomethylated cysteine-containing peptides in each sample (Figure A). Using this method, we investigated the effect of MND on the proteome redox status in DU145 cells. To identify the optimal concentration of MND for redox proteome analysis, we assessed the response of DU145 cells to various doses of MND by measuring intracellular ROS levels with CellROX, a fluorogenic probe that mainly detects intracellular hydroxyl radicals (^•OH) and superoxide anion (^•O^2–) and reduced cysteine levels with DTNB. At 50 μM MND, we observed a significant increase in intracellular ROS and a decrease in reduced cysteine levels (Figure B,C). Based on these results, we selected 50 μM MND as the concentration for subsequent redox proteome analysis. The samples were prepared from four biological replicates, resulting in a data set of 70,983 peptides (Table S2), which included 10,821 cysteine-containing peptides (Figure D, Table S3). A total of 6803 peptides were quantified in all eight samples, in which all cysteines were modified with IAA. To obtain reproducible data, we selected peptides for which quantitative values were calculated consistently in all four Oxi-Cys samples under either DMSO or MND treatment conditions. Consequently, 3072 and 3127 were consistently quantified in all four Oxi-Cys DMSO- and MND-treated samples, respectively. In total, 3665 peptides were quantified in all four DMSO- or MND-treated replicates, with 2534 peptides overlapping (Figure D, Table S4).

1.

DIALRP application to the analysis of menadione response in DU145 cells. (A) Workflow of DIALRP. (B,C) Quantification of ROS using CellROX (B) and reduced cysteines using DTNB (C) in cells under the treatment of indicated concentrations of MND for 1 h. The data were shown in mean ± SD of 3 biological replicates represented as single points. Statistical significance was calculated using Dunnett’s test versus DMSO-treated (−), and the adjusted p-values are indicated. (D) Schematic illustration of identified and quantified cysteine-containing peptides derived from the samples in DMSO- and 50 μM MND-treated cells.

To evaluate the detectability of NEM-modified cysteine peptides and the labeling efficiency of IAA, we compared peptide identification outcomes between samples treated with NEM prior to IAA labeling and those treated with IAA alone. A total of 5757 cysteine-containing peptides were identified. Of these, 3484 peptides were modified with IAA, 2290 with NEM, and 33 were unmodified (Figure S1A, Table S5). Notably, cysteines detected in approximately 13% (537 out of 4021 sequences) of peptides were modified exclusively by NEM and not by IAA (Figure S1B), indicating that a distinct subset of cysteine residues was preferentially modified by NEM. Nevertheless, the low proportion of unmodified peptides (<1%) across all samples suggests that IAA labeling was largely efficient under the experimental conditions used.

Next, we validated the reliability of the DIALRP quantification data by comparing the oxidation percentages across four replicate samples from DMSO- and MND-treated cells (Table S4). The correlation coefficients for the oxidative modification ratios in the DMSO-treated samples ranged from 0.837 to 0.882 (Figure S2). In the quantitative data for the DMSO and MND treatments, 71.8% and 76.7% of values, respectively, had a coefficient of variation (CV) of less than 0.2, with median CVs of 0.142 and 0.132 (Figure A). The median percentage of oxidative modification in each data set averaged 24.5% for DMSO treatment and 36.8% for MND treatment (Figure B). A comparison of the distribution of the average percentage of oxidative modification for each peptide in the DMSO- and MND-treated samples showed a significant increase in the percentage of oxidative modification after MND treatment (Figure C). To further characterize the oxidation stoichiometry, we analyzed the distribution of oxidation percentages across all cysteine sites under DMSO and MND treatment conditions. Cysteine oxidation levels spanned the full range (0–100%), with a global shift toward higher oxidation upon MND treatment (Figure D,E). These results confirm the robustness and reproducibility of DIALRP in quantifying cysteine oxidative modifications under different conditions. To assess whether highly oxidized cysteines are associated with disulfide bonds, we compared UniProtKB annotations with specific oxidation-level subgroups. Among all cysteine sites with calculable oxidation under DMSO conditions, 14.8% were annotated as a “disulfide bond” in UniProt. In contrast, only 0.8% of cysteines in the 0–10% oxidation group were annotated as disulfide forming, while 59.2% of the 90–100% group carried such annotation (Figure F). These findings suggest that highly oxidized cysteines are significantly enriched for disulfide-forming residues, whereas low-oxidation cysteines are predominantly maintained in the reduced state. This supports the biological relevance and validity of our oxidation stoichiometry measurements.

2.

Validation of DIALRP quantitation results. (A) Boxplots of the coefficient of variation (CV) of cysteine oxidation percent in each peptide between four replicates of DMSO- and 50 μM MND-treated DU145 cells. Boxplots display 25th and 75th percentile (bounds of box), median (center line), and largest and smallest values (whiskers) of the distribution. (B) Distribution of cysteine oxidation percent in each data set of DMSO-treated and MND-treated cells. The dotted horizontal line (Y axis = 0.5) showed the median of each cysteine oxidation percent. (C) Boxplots of 3665 mean cysteine oxidation rate (%) in each peptide of DMSO-treated and MND-treated cells. Boxplots display 25th and 75th percentile (bounds of box), median (center line), and largest and smallest values (whiskers) of the distribution. Statistical significance was calculated using the two-tailed paired t-test between DMSO-and MND-treated samples, and the obtained p-values are indicated. (D,E) Distribution of cysteine oxidation stoichiometry in DMSO- and MND-treated cells. Histograms show the distribution of cysteine oxidation percentages calculated for individual peptides under DMSO (D) and MND (E) treatment conditions. Each value represents the average oxidation percentage across four biological replicates. The y-axis indicates the proportion (%) of peptides within each oxidation range relative to the total number of peptides analyzed. Bin size is 10%. (F) Disulfide bond annotation in oxidized cysteines. Cysteine sites were grouped by oxidation stoichiometry under DMSO treatment into total (0–100%), low (0–10%), and high (90–100%) oxidation categories. The proportion of residues annotated as “disulfide bond” in UniProtKB is shown for each group.

DIALRP Revealed that a Group of Thiols in Translation-Related Proteins Are Most Affected in Response to Menadione in DU145 Cells

To evaluate the effect of MND on the oxidation state of cysteine, 3665 peptides quantified in all four DMSO- or MND-treated replicates were used for statistical comparison of cysteine oxidation percentages. Many cysteine residues exhibited low oxidation percentages, particularly under DMSO-treated conditions (Figure D), reflecting the reducing environment of the intracellular space. In several cases, the oxidized form was not detected, resulting in missing oxidation values, despite robust quantification in the All-Cys samples. These missing values likely reflect a biologically meaningful reduction rather than technical failure. To allow a statistical comparison such as log-transformed fold changes, we applied normal-distribution-based imputation for missing oxidation percentages, using the distribution of detected values within the same sample. Among the 1407 cysteine-containing peptides with a q-value cutoff of less than 0.01, 1002 showed a more than 2-fold change in the oxidation state upon MND treatment (Figure A, Table S6). The same set of eight samples, in which all cysteine residues were modified with IAA, was also analyzed to assess changes in the protein expression independent of cysteine oxidation (Figure B, Table S7). While numerous cysteine oxidation sites exhibited significant changes upon MND treatment, the overall protein expression levels remained relatively modest. This contrast highlights the distinct sensitivity of cysteine oxidation to oxidative stress, which may not always be accompanied by detectable changes in the protein abundance. To further compare redox-based changes with global proteomic alterations, we performed principal component analysis (PCA) using two types of quantitative data obtained from the same set of DMSO- or MND-treated samples (n = 4 each): cysteine oxidation percentages and protein abundance. PCA based on cysteine oxidation percentages revealed a clear separation between DMSO and MND conditions along PC1, which accounted for 53.8% of the total variance (Figure S3A). This suggests that the cysteine oxidation status robustly reflects treatment-specific redox changes. In contrast, PCA based on protein abundance showed a less prominent separation, with PC1 and PC2 explaining 32.7% and 19.2% of the total variance, respectively (Figure S3B). These results imply that the redox response to MND treatment is more sensitively captured at the level of cysteine oxidation than by overall changes in protein abundance. Among the 1407 MND-sensitive cysteine-containing peptides, we defined the top 25% (351 peptides) as highly responsive to MND treatment, corresponding to 278 proteins with strongly MND-sensitive cysteines (Figure C).

3.

Extraction of significantly oxidized cysteines in response to MND and in silico prediction of their functions. (A,B) Volcano plot showing log₂(fold change) in oxidation percentage for 3665 unique cysteine-containing peptides (A) and expression of 6104 proteins (B) of MND-versus DMSO-treated cells plotted against –log₁₀(q-value). (C) Order of 1407 significantly oxidized cysteine peptides with a q-value cutoff of less than 0.01. The top 25% (351 peptides with a fold change more than 3.87) were identified as highly responsive to MND treatment, corresponding to 278 proteins with strongly MND-sensitive cysteines. (D) GO analysis of 278 proteins with strongly MND-sensitive cysteines using DAVID functional annotation clustering. Enrichment scores for the top 5 clusters are shown in the bar graph (left panel), and the 3 GO descriptions with the lowest p-value in each cluster are shown in the table (right panel). (E) Functional molecular network of 35 highly MND-responsive cysteine-containing proteins annotated with GO “translation.” The 6 significantly enriched GO annotation selected in the list of functional enrichments in STRING were used to categorize these proteins. (F) Violin plots of the solvent-accessible surface area (SASA) values (Ų) in each data set of cysteine residues in 3665 peptides with detected oxidative modifications (All), 1407 peptides with significantly elevated oxidative modification (Group 1), 351 peptides with highly elevated oxidative modification (Group 2), and 41 peptides within 35 translation-related factors with highly elevated oxidative modification (Group 3). The dotted lines in plot areas show the medians and quartiles. Statistical significance was calculated using two-tailed unpaired t-test versus All data set and the obtained p-values are indicated.

Gene ontology (GO) enrichment analysis using DAVID revealed that the 278 proteins that were highly responsive to MND treatment were associated with various cellular processes, including translation (cluster 1, cluster 4), extracellular secretion (cluster 2), mRNA splicing (cluster 3), and chaperone activity (cluster 5) (Figure D, Table S8). Among these categories, translation-related proteins were the most significantly enriched (cluster 1), indicating the strong impact of MND on the translational mechanisms. Further analysis of the 35 proteins annotated with the GO term “translation” using STRING classified them into six functional groups, such as cytosolic and mitochondrial ribosomal proteins, translation factors, and aminoacyl-tRNA synthetases (Figure E, Table S9). We also performed GO enrichment analysis for the proteins contributing to the PC1 component that represents the MND-treated group in the PCA of cysteine oxidation data. As a result, these proteins were associated with mRNA splicing (cluster 1) and translation (cluster 2) (Figure S3C). These results suggest that various translation-related factors respond to oxidative stress and regulate their function at multiple stages of the translation process.

Because solvent accessibility may influence the extent of oxidative modifications, we next evaluated the solvent accessible surface area (SASA), which represents the degree to which a residue is exposed to the solvent and thus available for chemical modification, for oxidatively modified cysteine residues upon MND treatment. We successfully obtained 1807 structural models out of 1829 proteins, which included 3665 peptides with detected oxidative modifications, from AlphaFold, due to the limitations in modeling proteins larger than 2700 amino acids. SASA values for each residue in these structural models were calculated using FreeSASA. Then, we compared the SASA values of cysteine residues from the following four groups: all 3665 peptides with detected oxidative modifications (All), 1407 peptides with cysteines significantly oxidized by MND (Group 1), 351 peptides with cysteines highly oxidized (Group 2), and 41 peptides within 35 translation-related factors with cysteines highly oxidized (Group 3). The results showed that the SASA values of cysteine residues in the Group 2 data set were significantly higher than those in the All data set (Figure F, Table S10). This indicates that cysteine residues that undergo extensive oxidative modification upon MND treatment are more likely to be exposed on the molecular surface and may represent sites that are highly sensitive to oxidative stress. The SASA values of cysteine residues in Group 3 data set were also significantly higher than those in the All data set and exhibited a distribution with a median and upper quantile at the same level as that of the Group 2 data set (Figure F, Table S10). This similarity in the SASA distribution suggests that cysteine residues in these translation-related factors are similarly surface-exposed and may be particularly vulnerable to oxidative stress.

Validation of Cysteine Redox Response of Translation-Related Proteins

Because the group of proteins most affected by MND-induced oxidative stress was thought to be translation related, the translation status of MND-treated DU145 cells was evaluated by nascent polypeptide labeling using l-azidohomoalanine (AHA), a methionine analogue that is incorporated into nascent proteins during active translation. We observed a decrease in translational activity with increasing concentrations of MND (Figure A,B). Translational activity was drastically decreased under conditions treated with 50 μM MND, which is consistent with increased intracellular ROS and elevated cysteine oxidation (Figure B,C). This decrease in translational activity was reversed by N-acetylcysteine (NAC) treatment (Figure C). The redox statuses of seven translation-related molecules, RPL12, RPL11, EIF2S2, EIF6, EEF2, TSFM, and MRPL50, with particularly elevated oxidative modifications, were selected for the further validation of the redox proteome analysis data (Figure D). The FLAG-tagged forms of these proteins were overexpressed in HeLa cells and analyzed by Western blotting using the maleimide-PEG reagent. We used HeLa cells for these validations because of the high transfection efficiency to sufficiently express all constructs and confirmed that the cysteines of endogenous EIF6 was oxidized in response to MND treatment in HeLa cells, as well as DU145 cells (Figure S4). The results confirmed that these proteins underwent oxidative modification in response to menadione and that NAC treatment restored the redox status (Figure E). These data support the redox proteomic data of the MND treatment response.

4.

Validation of MND-induced oxidative stress response of a group of translation-related factors. (A,B) Quantification of nascent polypeptides in DU145 cells under the treatment of indicated concentrations of MND for 1 h using l-azidohomoalanine (AHA). (A) Total proteins were separated by SDS-PAGE and AHA-tetramethylrhodamine (TAMRA)-labeled nascent polypeptides were detected by a fluorescent scanner. Total proteins were visualized by CBB staining. Representative images of AHA-TAMRA and CBB were shown. The fluorescent intensities were quantified and normalized as 1 for DMSO-treated samples (B). The data were shown in mean ± SD of 3 biological replicates represented as single points. Statistical significance was calculated using Dunnett’s test versus DMSO-treated (−), and the adjusted p-values are indicated. (C) Detection of nascent polypeptides in DU145 cells treated with DMSO, 50 μM MND, or 50 μM MND plus 5 mM N-acetylcysteine (NAC) for 1 h using AHA. (D) Redox proteomics data of peptides with particularly elevated oxidative modifications in selected 7 translation-related proteins. (E) Validation of MND-oxidative stress response by redox Western blotting. HeLa cells transfected with plasmid expressing each FLAG-tagged protein were treated with DMSO, 50 μM MND, or 50 μM MND plus 5 mM NAC. The cell lysate was treated with mPEG-maleimide and analyzed. The proteins treated with TCEP and NEM (mimicking fully oxidized form, Oxi), and TCEP and mPEG-maleimide (mimicking fully reduced form, red) were detected simultaneously. Black and white arrows show fully oxidized and reduced forms, respectively.

Functional Validation of the Translation Factors under MND-Induced Oxidative Stress Conditions

To assess the functional influence of translation factors under oxidative conditions induced by MND treatment, changes in their subcellular localization and protein binding were analyzed. Of the seven translation-related factors evaluated for an oxidative stress response by Western blotting, the functions of three proteins, EIF2S2, EIF6, and EEF2, were successfully evaluated. FLAG-IP-MS analysis of EIF2S2, EIF6, and EEF2 predominantly identified proteins that were previously reported to interact physically and functionally, such as EIF2S1, EIF2S3, and CDC123 in EIF2S2; BCCIP and RPL23 in EIF6; and HGH1 in EEF2 (Figure A,E,I, and Tables S11–S14). We observed a decrease in the binding of EIF2S3 and CDC123 to EIF2S2 in response to MND treatment (Figure A,B), suggesting that the formation of the eIF2 complex is inhibited by MND-induced oxidative stress. Immunohistochemistry revealed no change in EIF2S2 localization, which remained primarily in the nucleoli despite MND treatment (Figure D). We also observed a decrease in the binding of BCCIP and RPL23 to eIF6 (Figure E,F). Previous studies reported that eIF6, BCCIP, and RPL23 form a tripartite complex, with BCCIP functioning to translocate eIF6 to the nucleus and nucleoli. , Although eIF6 was primarily localized in the nucleus and nucleolus, even in its FLAG-tagged form, MND treatment markedly increased its level of cytoplasmic localization. This shift was reversed by simultaneous treatment with MND and NAC, which restored eIF6 localization in the nucleus and nucleolus (Figure H). Moreover, MND treatment decreased the binding of HGH1 to EEF2 (Figure I,J), an interaction in which HGH1 functions as a chaperone of EEF2. , Immunocytochemistry revealed that EEF2-containing aggregates formed in the cytoplasm in response to MND treatment but dissipated with concurrent NAC treatment (Figure L). Notably, quantitative DIA-MS analysis of cell lysates confirmed that the observed changes in protein interactions and localization after MND treatment occurred independent of any changes in protein expression levels (Figure B,F, Table S18).

5.

Functional validation of translation factors under MND treatment by FLAG-IP-MS and immunocytochemistry. (A,E,I) Scatterplots showing quantitative data of FLAG-IP MS analysis of EIF2S2 (A), EIF6 (E), and EEF2 (I). X-axis and Y-axis represent the log₂ fold change of intensities of proteins in FLAG-IP fraction prepared from HeLa cells transfected with a bait-expressing plasmid compared to that with empty vector (EV) and in proteins in FLAG-IP fractions prepared from bait-expressing cells treated with 50 μM MND for 1 h compared to those treated with DMSO, respectively. Proteins that were significantly more abundant in the bait FLAG-IP fraction compared to EV are indicated by blue dots. Of those proteins, their bindings significantly altered by MND treatment are indicated by red dots. The names of baits, and proteins that have been reported to form complexes, are listed in the graph. (B,F,J) Bar graphs showing the intensities of EIF2S2 (B), EIF6 (F), and EEF2 (J) binding proteins to significantly altered by MND treatment in the FLAG-IP fraction and whole cell lysate (WCL). The data were shown in mean ± SD of 3 biological replicates represented as single points. *q < 0.05, **q < 0.01 versus DMSO-treated (−). (C,G,K) Redox proteomics data of the EIF2S2 (C), EIF6 (G), and EEF2 (K) and their binding proteins (indicated as prey). Cysteines that were oxidized significantly with a q-value less than 0.01 was highlighted in red. (D,H,L) Fluorescent images showing the localization of EIF2S2 (B), EIF6 (F), and EEF2 (J) in HeLa cells treated with DMSO, 50 μM MND or 50 μM MND plus 5 mM NAC. The cells were counterstained with Hoechst33342 to detect their nuclei. Scale bar = 10 μm.

These results indicated that specific cysteine residues within translation-related factors are particularly susceptible to oxidative modifications under MND-induced stress, with potential functional implications. In the EIF2 complex, 9 of the 11 cysteine-containing peptides (covering 12 of 14 cysteines) in EIF2S1, EIF2S2, EIF2S3, and CDC123 were significantly oxidatively modified (Figure C). This suggests that oxidative modifications affect the function of the eIF2 complex, possibly affecting the initiation of translation. In the eIF6-BCCIP-RPL23 ternary complex, all three detected peptides (with four cysteines) were significantly oxidatively modified (Figure G), indicating that oxidative stress disrupted the normal localization and function of eIF6 in translational regulation. In contrast, EEF2 showed oxidative modification in five of its nine cysteine-containing peptides, whereas HGH1, a known chaperone of EEF2, displayed no detectable oxidative modifications (Figure K). This difference suggests a protective role for HGH1 under oxidative conditions, potentially helping stabilize EEF2. The FLAG-IP-MS analysis of two proteins, RPL12 and MPRL50, failed to significantly identify binding proteins, despite the high abundance of identified baits; the analysis of TSFM was able to identify TUFM as a binding protein, but MND treatment had no significant effect on binding (Figure S5, Tables S15–S17). In the case of RPS11, enrichment by FLAG-IP was not successful, and did not reach sufficient level for analysis. Immunocytochemistry was performed to evaluate whether their localization was altered in response to the MND treatment (Figure S6). FLAG-tagged RPL12 is mainly localized to the nucleolus. RPL12 localization to the nucleolus was disrupted by MND treatment, suggesting that normal RPL12-related ribosomal biogenesis may be inhibited by oxidative stress. We also observed that the localization of RPS11, TSFM, and MRPL50 was not altered by MND treatment.

Overall, these findings highlight that MND-induced oxidative stress selectively affects cysteine residues in translation-related factors, potentially modulating translation as part of the adaptive response of cells to oxidative stress.

Discussion

Oxidative modification of cysteine has been reported to regulate various biological processes, such as signal transduction, metabolic adaptation, and developmental hematopoiesis. Hence, methods for quantifying the oxidative modification of cysteine are widely applicable. Current redox proteome analysis techniques require relatively complex sample-handling methods. We have demonstrated a workflow that enables redox proteome analysis using an approach comparable to sample handling in general expression proteome analysis. Translation-related factors made up the most significant group of proteins that underwent oxidative modifications during MND-induced oxidative stress.

As redox proteomics is primarily performed using stable isotopes with data-dependent acquisition mass spectrometry, the enrichment or fractionation of peptides has been used to improve the number of identified peptides. This complicates the sample preparation for the redox proteome analysis. To resolve this problem, label-free redox proteomics was reported in Redox Biology in 2019, which profiles the cysteine redox status based on DIA/SWATH. As noted in this earlier report, the advantage of label-free analysis is that the use of nonisotope-labeled alkylation reagents, which are commonly used in most proteomics laboratories, incurs no cost during sample preparation. Additionally, label-free analysis has the advantage of fewer restrictions on the number of samples. In recent years, the use of DIA-MS for protein identification has increased dramatically with the development of software tools. Using a widely available Q-Exactive mass spectrometer, we successfully quantified redox changes in 3665 cysteine peptides with high reliability through a deep learning-based, library-free search using DIA-NN. Several features of the approach contribute to its accuracy. First, it leverages DIA-MS, which is known for its precise and reliable quantitation. Second, unlike enrichment-based methods, which are prone to handling errors and complicated data correction, our method allows straightforward normalization using total peptide ion intensities, enhancing the reliability and consistency of the quantitative results.

Previous redox proteomic studies have also reported enhanced oxidative modification of translation-related factors in response to oxidative stress, indicating that translation and cysteine redox are closely related. , Indeed, the decrease in the translational activity of DU145 cells during MND treatment was accompanied by a significant increase in ROS and a decrease in intracellular reduced cysteine levels (Figures A,B and A,B). The accompanying oxidative response was particularly pronounced in a group of factors that play a central role in translation initiation and elongation such as EIF2, eIF6, and EEF2 (Figure ). Previous studies reported that the functions of eIF2, eIF6, and eEF2 are suppressed under the stress condition through phosphorylation by kinases: eIF2α by PERK, GCN2, PKR, and HFI; eIF6 by RACK1 and GSK3β; , and eEF2 by eEF2K. , A previous report showed that menadione treatment causes increased phosphorylation of Ser51 of eIF2α and decreased translational activity, whereas the inhibition of eIF2α phosphorylation by PERK inhibitors does not restore the decreased translational activity. These findings suggest the existence of a sophisticated regulatory network in which crosstalk among cysteine oxidation, phosphorylation, and other post-translational modifications orchestrates translational control during oxidative stress. Further studies are required to elucidate the intricate details of these multifaceted regulatory mechanisms.

The SP3 method significantly enhanced the throughput and scalability of the DIALRP workflow, making high-throughput redox proteomic analysis more feasible. Unlike traditional methods that often require complex and time-consuming steps for cysteine peptide enrichment or stable isotope labeling, the SP3 approach streamlines sample preparation and enables efficient parallel processing of multiple samples. This streamlined approach is well suited for large-scale studies such as those investigating oxidative stress across various cell types or conditions. The high-throughput capability of DIALRP with the SP3 method accelerates redox proteomics research, enabling comprehensive cysteine oxidation profiling across biological and clinical contexts and providing a valuable tool for broader redox biology studies. However, label-free acquisition necessitates individual runs for each sample, which can be a bottleneck in studies involving large sample numbers such as cohorts. Promisingly, recent advances in high-throughput mass spectrometry technology may help mitigate this limitation, potentially expanding the applicability of label-free workflows, even in large-scale studies.

One potential limitation of our study is the use of NEM, which is known to exhibit high reactivity and may alter labile Cys redox modifications. Consistent with previous reports on differences in reactivity between NEM and IAA, our results demonstrated that approximately 13% of cysteine-containing peptides were exclusively modified by NEM and not by IAA (Figure S1). This finding suggests that NEM may label cysteine residues that are less accessible or less reactive toward IAA, potentially due to its higher reactivity or different chemical properties. In addition, previous studies have reported potential off-target modifications by NEM on other amino acids such as histidine and lysine. Another potential limitation is that oxidative modification can be underestimated, because overoxidized cysteines are not considered. In particular, the oxidation percentage determined in our workflow may underestimate the actual oxidative status because irreversibly overoxidized cysteine residues (e.g., sulfinic or sulfonic acid forms) are not detected under our experimental conditions and are thus excluded from the calculation. Therefore, while our results provide insights into redox-regulated proteins, caution should be exercised in interpreting the precise redox states of the cysteine residues.

In summary, this study provides a comprehensive framework for the analysis of cysteine oxidative modifications using DIA-MS-based label-free redox proteomics, specifically revealing the oxidative stress response in DU145 prostate cancer cells. By identifying translation-related factors as primary targets of oxidative modification, we highlighted a potential mechanism by which oxidative stress modulates protein synthesis and cellular adaptation. However, this study was limited to the oxidative stress response to MND treatment in DU145 cells, which are highly resistant to oxidative stress. , Although we confirmed the oxidative stress response of selected translation-related factors when they were overexpressed in HeLa cells, this was not confirmed in other cell types or under other oxidative stress conditions. Therefore, a detailed analysis of the differences in an oxidative stress tolerance and oxidative stress conditions induced by DIALRP will lead to a better understanding of the oxidative stress response in cancer cells.

Glossary

Abbreviations

AHA

l-azidohomoalanine

DIA

data-independent acquisition

DIALRP

Data-independent acquisition (DIA)-based label-free redox proteomics

DIA-MS

data-independent acquisition mass spectrometry

DMSO

dimethyl sulfoxide

DTNB

5,5′-dithiobis­(2-nitrobenzoic acid)

FBS

fetal bovine serum

GO

gene ontology

MND

menadione

NAC

N-acetylcysteine

NEM

N-ethylmaleimide

ROS

reactive oxygen species

SASA

solvent-accessible surface area

TAMRA

tetramethylrhodamine

TBTA

tris­[(1-benzyl-1H-1,2,3-triazol-4-yl)­methyl] amine

TMT

tandem mass tags

TNB

2-nitro-5-thiobenzoic acid

SDB

styrenedivinylbenzene.

All raw MS data were stored in jPOSTrepo (https://repository.jpostdb.org/). jPOSTIDs/PXIDs for the projects containing these data were JSPT003469/PXD057921 (DIALRP), JSPT003470/PXD057922 (FLAG-IP-MS), and JSPT003471/PXD0579230 (Expression proteome analysis of MND response in HeLa cells).

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jproteome.5c00339.

• Evaluation of the detectability of NEM-modified cysteine-containing peptides and the labeling efficiency of IAA, correlation coefficients of cysteine oxidation rates between each data set and the four replicates of DMSO-treated samples, principal component (PC) analysis of DIALRP quantitation data, evaluation of the oxidative stress response of endogenous EIF6 cysteine in DU145 and HeLa cells, scatter plots showing quantitative data from the FLAG-IP MS analysis of RPL12, MRPL50, and TSFM, and fluorescent images showing the localization of RPL12, MRPL50, TSFM, and MRPL50 in HeLa cells treated with DMSO, 50 μM MND or 50 μM MND plus 5 mM NAC. (PDF)

• Forward and reverse primers used for the construction of expression plasmids, all identified peptides by DIALRP in DMSO- and MND-treated DU145 cells, quantitation data of 10,821 cysteine peptides, all identified peptide by DIALRP in DMSO-treated DU145 cells using FragPipe (set carbamidomethylation and NEM as variable modifications), oxidation percentage of 3665 cysteine peptides, statistical analysis data of 3665 cysteine peptides using Perseus, quantitation data of protein identified by DIALRP in DMSO- and MND-treated DU145 cells, DIVID functional annotation clustering of 278 highly oxidized proteins, list of 35 proteins annotated with GO translation, SASA values of cysteine residues in all, and Group 1–3, quantitation data of protein identified by FLAG-IP-MS analysis, statistical data analysis of EIF2S2 FLAG-IP-MS using Perseus, statistical data analysis of EIF6 FLAG-IP-MS using Perseus, statistical data analysis of EFF2 FLAG-IP-MS using Perseus, statistical data analysis of RPL12 FLAG-IP-MS using Perseus, statistical data analysis of MRPL50 FLAG-IP-MS using Perseus, statistical data analysis of TSFM FLAG-IP-MS using Perseus, and quantitation data of protein in MND-treated HeLa cells (XLSX)

D.K. and M.M. designed the entire study. D.K. collected data. T.T. contributed to the bioinformatics tools. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

This work was supported by MEXT/JSPS KAKENHI under Grant No. JP 20H05930 (M.M.), 22H02607 (M.M.), 22K07144 (D.K.); AMED-CREST under Grant No. 21gm1410006h0001 (M.M.); and GteX Program Japan Grant Number JPMJGX23B3 (M.M.).

The authors declare no competing financial interest.