蛋白質半胱胺酸氧化修飾之偵測與亞磺酸還原酶抗氧化角色之研究

國立臺灣大學生命科學院生化科學研究所 博士論文

Graduate Institute of Biochemical Sciences College of Life Science

National Taiwan University Doctoral Dissertation

蛋白質半胱胺酸氧化修飾之偵測與 亞磺酸還原酶抗氧化角色之研究

On Protein Cysteine Oxidation Detection and the Roles of Sulfiredoxin in Antioxidant Responses

李昱蓉 Yu-Jung Lee

指導教授：張震東 博士 Advisor: Geen-Dong Chang, Ph.D.

中華民國 107 年 7 月

July, 2018

謝誌

不同於以往的學習經驗，博士班除了知識與技術上的追求，更多的是自我的

反思與覺察。感謝過程中曾協助與提點我的每一個人，您們寶貴的分享與建議都

使我獲益良多。感謝張老師多年來的指導，也感謝 R403 實驗室的每一位成員，在

紅館的每一天都使我學習與成長。感謝師長們的教誨，以及同儕們的砥礪，更要

感謝家人的支持與包容，您們都是我學習路上重要的力量。

摘要

氧化壓力與許多疾病與生理功能有關，如癌症、心血管疾病、神經退化疾病。

當細胞處於氧化壓力下，蛋白質中半胱胺酸(cysteine)殘基之巰基(thiol group)易受

攻擊產生氧化相關之轉譯後修飾(oxidative post-translational modification)，進而影

響其功能與結構特性。因此針對蛋白質半胱胺酸殘基之氧化修飾，已有許多功能

之探討及鑑定工具之開發。

半胱胺酸氧化修飾包括可逆(reversible)修飾如形成雙硫鍵(disulfide)或次磺酸

(sulfenation)，在更高的氧化壓力下則可能形成不可逆的亞磺酸和磺酸(sulfination

and sulfonation)。亞磺酸還原酶(sulfiredoxin)為目前被報導唯一可對半胱胺酸亞磺

酸化進行還原的酵素，其已知作用對象為過氧化物還原酶(peroxiredoxin 1-4)。在這

裡我們利用對細胞處理二醯胺 (diamide)，以引發針對巰基攻擊的氧化修飾，進而

探討蛋白質半胱胺酸氧化修飾現象。我們在 HeLa 細胞中對二醯胺引發之氧化攻擊

建立基礎了解，包括蛋白質氧化程度改變以及細胞抗氧化反應的啟動。接著我們

再進一步運用此方法探討亞磺酸還原酶之抗氧化功能，利用SRXN1 敲除(knockout)

之HAP-1 細胞和過表現 SRXN1 之 HEK 細胞兩種模式，我們報導了亞磺酸還原酶

對於二醯胺引發之蛋白質氧化修飾具有保護作用，並探討亞磺酸還原酶在抗氧化

原酶(thioredoxin)，有交互作用。

此外，我們針對改對良蛋白質半胱胺酸之氧化修飾之標定方式，以發展氧化

程度之定量描述方式。目前對於蛋白質氧化修飾之偵測工具，主要是針對特定修

飾的辨識和捕捉。我們優化PEG- maleimide 之反應條件，對蛋白質還原態之半胱

胺酸殘基進行標記，藉由免疫轉漬(immunoblotting)後積分其訊號以計算氧化的比

例，並提出加權的方式將蛋白質的不同氧化程度計分。我們於生物模型驗證此方

式，如胰島素引發之蛋白質酪胺酸磷酸酶1B (PTP1B)氧化，惟此方法仍受如抗體

辨識區域遮蔽之限制。

關鍵字

半胱胺酸氧化修飾、亞磺酸還原酶、抗氧化反應、二醯胺、定量偵測

Abstract

Oxidative stress is relevant to several physiological functions and diseases, including

cancer, cardiovascular disease, and neural degeneration. The reactive oxygen and

nitrogen species (ROS/RNS) such as hydrogen peroxide arise as by-products of

metabolism, while excessive ROS/RNS results in protein oxidation and signaling which

leads to disease.

Cysteine oxidation is the main post-translational modification associated with redox

signaling and oxidative stress. Reversible cysteine oxidations include disulfide and

sulfenic acid, which will be further oxidized to irreversible sulfinic and sulfonic acid

while facing extreme oxidative stress. Sulfiredoxin is the only reported enzyme that

reduces sulfinic acid on hyperoxidized peroxiredoxin 1 - 4. Herein we treated cells with

diamide, which caused thiol-specific attacks, to investigate the antioxidant role of

sulfiredoxin. Utilizing SRXN1 knockout HAP-1 cells and FLAG-SRXN1

over-expressing HEK cells, we reported a protective role of sulfiredoxin in diamide-

induced protein thiolation, and investigated its interactions with other antioxidant

systems such as thioredoxin and glutathione.

Maleimide-polyethylene glycol (m-PEG) has been used to detect reversibly oxidized

proteins by reacting to the reduced cysteine residues leading to mobility shift in

immunoblots; a method called PEG-switch. Following PEG-switch, both reduced and

oxidized proteins can be observed on the same immunoblot simultaneously, providing a

simple quantitative measurement for protein thiol modifications. We optimized the

assay conditions and exploited the applications of PEG-switch in quantitation of the

extent of protein thiol oxidation in cells in response to H2O2 and insulin. In addition, we

have proposed a redox scoring system for measuring the redox status of any given

protein from the m-PEG immunoblot. Some restrictions of the method are also

indicated.

Keywords

Cysteine oxidative modification, sulfiredoxin, antioxidant response, diamide,

quantitative detection

Index

口試委員會審定書 ...

謝誌 ... i

摘要 ... ii

Abstract ... iv

Index ... vi

Figure List...viii

Table List ...viii

Appendix Index ... ix

Abbreviation List ... x

Chapter I. Investigation of the Antioxidant Role of Sulfiredoxin ... 1

1. Introduction...1

1.1 Oxidative stresses and antioxidant systems in mammalian cells...1

1.2 Protein cysteine residues as the victims of oxidative attacks ...3

1.3 Sulfiredoxin reduces peroxiredoxin sulfinylations ...5

1.4 Specific aims and significances ...7

2. Materials and methods ...9

2.1. Chemicals and antibodies ...9

2.2. Cell culture...9

2.3. SRXN1 knockout, genomic DNA extraction and sequencing ...10

2.4. SRXN1 over-expression ...10

2.5 Cell lysate preparation ... 11

2.6. Fractionation ... 11

2.7. SDS-PAGE, Coomassie blue staining, and immnunoblotting ...12

2.8. Immunofluorescence...12

2.9. RNA extraction and RT-PCR...13

2.10. Immunoprecipitation...13

2.11. Diamide treatment and resin-assisted capture...14

2.12. Hydrogen peroxide-induced oxidation ...15

2.13. MTT assay ...15

2.14. Anti-glutathione antiserum preparation ...15

2.15. Thioredoxin reductase activity assay ...16

2.16. In-gel digestion and protein identification by mass spectrometry ...16

3. Results...20

3.1. Diamide induced reversible thiolations and antioxidant responses ...20

3.2. Sulfiredoxin knockout increased diamide-induced protein thiolations ...22

3.3. Sulfiredoxin knockout affected functions of antioxidant enzymes ...23

3.3.1. Glutathione system ...24

3.3.2. Peroxiredoxin and thioredoxin system ...25

3.4. Diamide treatment induced cytosolic sulfiredoxin dimerization ...26

3.5 Hypothetic mechanism of sulfiredoxin under thiol-specific oxidative attacks ...28

4. Discussion ...29

Chapter II. Development of Quantitative Display of Protein Redox Status... 32

1. Introduction...32

1.1. Detection tools of protein cysteine oxidations...32

1.2. PEG-maleimide as a cysteine-specific probe ...34

1.3. Specific aims and significances ...35

2. Materials and methods ...36

2.1. Chemicals and antibodies ...36

2.2. Cell culture and lysis condition ...36

2.3. PEG-maleimide labeling...37

2.4. SDS-PAGE, immunoblotting and reduced/oxidized ratio calculation...37

2.5. Hydrogen peroxide-induced oxidation ...38

2.6. Insulin-induced PTP1B oxidation in HeLa cell ...38

3. Results...39

3.1. Protocol of PEG tagging ...39

3.2 Optimizing tagging efficiency ...39

3.3. m-PEG tagging in other redox-sensitive proteins...41

3.4. Quantitating the redox status of proteins in cells...42

3.4.1. H2O2-induced peroxiredoxin and Hsp27 oxidations in HEK cell ...44

3.4.2. Insulin-induced PTP1B oxidation in HeLa cell ...45

4. Discussion ...47

Chapter III. Conclusions and Perspectives ... 50

Figures ...52

Tables ...84

References...89

Appendix...96

Figure List

Figure 1: Diamide induced reversible thiolations in HeLa cell ...52

Figure 2: Thiolated targets of diamide treatment...54

Figure 3: Diamide affected redox status of glutathione, NADH and NADPH...55

Figure 4: Diamide triggered Nrf2 translocation and downstream gene expressions ...56

Figure 5: Knockout of SRXN1 led to truncation of sulfiredoxin ...58

Figure 6: Diamide-induced thiolations increased in SRXN1 KO HAP-1 cell ...60

Figure 7: Thiolation of antioxidant enzymes increased in SRXN1 KO HAP-1 cell ...62

Figure 8: Diamide-induced protein glutathionylation and glutathione oxidation increased in SRXN1 KO cell ...63

Figure 9: Activity of thioredoxin reductase increased in SRXN1 KO cell...64

Figure 10: Diamide induced Srx dimerization in SRXN1 over-expressed HEK cell...66

Figure 11: Sulfiredoxin dimerization was found in mitochondria...68

Figure 12: Hypothetic mechanism of sulfiredoxin in diamide-induced oxidation ...69

Figure 13: Reaction of PEG-maleimide with protein cysteine residues causing mobility shift on SDS-PAGE ...70

Figure 14: Protein labeling efficiency with PEG-maleimide, PEG-vinyl sulfone...72

Figure 15: Targets of PEG-maleimide labeling. ...74

Figure 16: Detection and scoring of in vitro oxidized proteins ...76

Figure 17: Quantitation of in cellulo H2O2-induced oxidation of Prx in HEK cell ...78

Figure 18: Quantitation of in cellulo H2O2-induced oxidation of Hsp27 in HEK cell ..80

Figure 19: Quantitation of insulin-induced oxidized PTP1B in HeLa cells ...82

Table List

Table 2: Diamide-induced sulfiredoxin disulfide-forming targets ...86

Appendix Index

Figure S1: Antioxidant responses in mammalian cell ...96

Figure S2: Protein cysteine oxidations in mammalian cell ...97

Figure S3: Mechanism of Srx-catalyzed Prx-SO2 reduction ...98

Figure S4: Metabolomics survey of glutathione...99

Figure S5: Design and test of anti-GSH antiserum ...100

Figure S6: Structure and tagging mechanism of polyethylene glycol maleimide ...101

Table S1: Diamide-induced oxidation proteome in HAP-1 cell ...102

Table S2: Identification of sulfiredoxin disulfide-forming targets ... 110

Abbreviation List

ADP Adenosine diphosphate

ARE Antioxidant responsive element ATP Adenosine triphosphate

DMSO Dimethyl sulfone DTT Dithiothreitol

GAPDH Glyceraldehyde-3-phosphate dehydrogenase GPx Glutathione peroxidase

GR Glutathione reductase GSH Glutathione

GSSG Oxidzed glutathione disulfide IAA Iodoacetamide

KO Knockout

LC-MS Liquid-phase chromatography mass spectrometry mRNA Messenger ribonucleic acid

MTT (3-(4, 5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide NADH Nicotinamide adenine dinucleotide

NADPH Nicotinamide adenine dinucleotide phosphate NEM N-ethylmaleimide

Nrf2 NF-E2 related factor 2 PBS Phosphate buffered saline PCR Polymerase chain reaction PEG Polyethylene glycerol Prx Peroxiredoxin

PTP1B Protein tyrosine phosphatase 1 B RAC Resin-assisted capturing

RNS Reactive nitrogen species ROS Reactive oxygen species

RT-PCR Reverse transcription polymerase chain reaction

SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis Srx Sulfiredoxin

TRR /TrxR Thioredoxin reductase Trx Thioredoxin

Chapter I.

Investigation of the Antioxidant Role of Sulfiredoxin

1. Introduction

1.1. Oxidative stresses and antioxidant systems in mammalian cells

Reactive oxygen and nitrogen species (ROS/RNS), such as superoxide (O2^!-),

hydrogen peroxide (H2O2), and nitric oxide (NO^!), are produced as by-products of

mitochondrial metabolic processes including oxidative phosphorylation (OXPHOS),

and they may serve as signaling molecules [1,2]. However, for its high reactivity,

excessive ROS/RNS causes oxidative stress, which is associated with several

pathological and physiological pathways, including infectious diseases, cardiovascular

diseases, cancers, and diabetes [3,4,5].

To maintain the homeostasis of the reducing environment in cell, several antioxidant

enzymes respond to oxidative stress. For example, superoxide dismutase (SOD) and

catalase (CAT) detoxifies superoxide and hydrogen peroxide [6]. Additionally, thiol-

dependent enzymes peroxiredoxin (Prx), thioredoxin (Trx) and glutathione (GSH), play

important roles in protecting cell from oxidative stresses [7,8] (Appendix Fig. S1A).

milli-molar concentrations in mammalian cytoplasm [7,9]. As a tri-peptide containing

one cysteine residue, GSH scavenges ROS/RNS by forming oxidized glutathione

(GSSG) or protein glutathionylation. GSH is required in the reaction of a glutaredoxin

(Grx) and glutathione peroxidase (GPx), and it is recycled by glutathione reductase (GR)

in a NADPH-dependent manner [7]. Meanwhile, peroxidase activity is found in

peroxiredoxin (Prx) family, dependent on conserved cysteine residues in the active site.

Prx reacts with H2O2 at a very high rate, but also reduces peroxynitrite and organic

hydroperoxides [10]. Prx1, 2 and 4 are primarily cytosolic, while Prx3 locates in

mitochondria [11]. Prx2, due to its high reaction rate and abundance, traps almost all

H2O2 in vivo [12]. Prx with hyperoxidation on the catalytic cysteines is catalytically

inactivated [10], whereas still participating in protective signaling [13]. The

hyperoxidized Prx can be recovered by sulfiredoxin (Srx) [14].

Oxidized Prx can be specifically reduced by thioredoxins (Trx), and the oxidized Trx

can be reversed by thioredoxin reductase (TrxR) in a NADPH-dependent manner [15].

Trx1 and TrxR1 present in both cytosol and nucleus, whereas Trx2 and TrxR2 locate in

mitochondria [16]. Trx/TrxR system also has a protective function against oxidative

stress, such as supporting the activity of ribonucleotide reductase and inhibiting

apoptosis signal-regulated kinase-1 (ASK1) [17].

At the transcription level, nuclear factor erythroid-2 related factor (Nrf2) activates

downstream gene expression in response to oxidative stress [7] (Appendix Fig. S1B).

Normally, Nrf2 associates with Kelch-like ECH-associated protein1 (iNrf2, Keap1) by

intermolecular disulfide bridge. ROS/RNS results in Keap1 intramolecular disulfide

formation, releasing Nrf2 and allowing it to translocate into the nucleus and binding to

genes containing antioxidant response elements (ARE) in the promoter encoding several

antioxidant enzymes including Cu-Zn superoxide dismutase (SOD), NAD(P)H quinone

dehydrogenase 1 (NQO1), catalase, and sulfiredoxin (SRXN1) [18].

1.2. Protein cysteine residues as the victims of oxidative attacks

Thiolate (-S^-) is a potent nucleophile and more prone to oxidation than thiols, so that

proteins with cysteine residues of lower ionization constants (pKa) are more likely to be

attacked by oxidants [19,20]. The pKa values for the low-molecular-weight thiols,

cysteine (Cys) and glutathione (GSH), are 8.3 and 8.8, respectively, whereas cysteine

residues in proteins can be strongly influenced by the local environment.

Cysteine oxidation is the main post-translational modification associated with redox

signaling and oxidative stress [21,22]. Reversible oxidation modifications, such as

S-thiolation (S-S), S-nitrosylation (-SNO), disulfides, and Cys sulfenic acid (-SOH), can

be recovered by antioxidant enzyme systems such as the thioredoxin system [7]. The

others, including Cys sulfinic (-SO2) and sulfonic acids (-SO3), are considered

irreversible [2] (Appendix Fig. S2). However, there is a reported exception that

Cys-SO2 on Prx1-4 can be reduced by sulfiredoxin in an ATP-dependent manner [23].

Although cysteine residues consist of only 2.26% of all encoded amino acids in

human proteome [24], cysteine oxidations are important for enzyme activity and

stability of protein structure. Given cysteine residues are found in the active site of

several enzymes, Cys oxidations affect enzyme activity in several cases such as

nitrosylation of protein tyrosine phosphatase 1B (PTP1B) [25], while irreversible Cys

oxidations lead to protein degradation [26]. Disulfide bonds formed in the endoplasmic

reticulum (ER) during protein production are related to structural stability [27], while

some disulfide formations are induced by ROS and relevant to signaling, for instance,

intramolecular disulfide of Keap1 causes its dissociation from Nrf2 and facilitating Nrf2

signaling [28].

There are several detection tools of thiol oxidation. One strategy is reducing the Cys

modifications after alkylating unmodified thiols first, and then reducing the

modifications with proper reducing agents [29]. The nascent thiol groups can be labeled

with specific probes containing biotin (biotin-switch) [30], or directly pulled down with

thiol-specific probe in the immobilized phase (resin- assisted capture) [31]. In addition,

chemicals that can directly interact with oxidized modification are also developed, such

as dimedone, which targets at sulfenic acid [32]. Further, several antibodies of

Cys-oxidative modifications are developed.

1.3. Sulfiredoxin reduces peroxiredoxin sulfinylations

Sulfiredoxin (Srx) was first found in yeast with reactivating function to

hyper-oxidized peroxiredoxin [14]. Peroxiredoxins 1-4 (Prx1-4), which called 2-Cys

peroxiredoxins, form homo-dimer with inter-chain disulfide between peroxidatic

cysteine (CP) and resolving cysteine (CR) when stimulated by oxidative stress [33].

Hyperoxidized Prx1-4 forming sulfinic or sulfonic acid would be catalytically

inactivated [14,34]. Srx is thought to specifically reduce sulfinic acid on 2-Cys Prx in an

ATP-dependent manner [35]. It is the only enzyme reported to reduce protein Cys

sulfinic acid. For reactivating mitochondrial Prx3, Srx disassociates from the complex

consisting of Hsp90 and FKBP and translocates from cytosol to mitochondria induced

by hydrogen peroxide signaling [36,37]. As one of the downstream genes of Nrf2/ARE

signaling, the translation of Srx is induced by the AP-1 promoter [38].

Srx is a small protein with predicted molecular size of 14 kDa. The nucleotide and

amino acid sequences are conserved among species, containing a GCHR region where

ATP binding site and catalytic cysteine residue locates [39]. For mammalian, plant, and

cyanobacteria, there is only one cysteine residue in the amino acid sequence of Srx,

while in prokaryotes and yeast it contains two cysteine residues [40], thus the oxidized

1-Cys Srx cannot be recycled by intramolecular disulfide but relies on the participation

of glutathione [40].

Srx catalyzes the reduction of hyperoxized Prx in an ATP-dependent manner.

Phosphate on ATP transfers to the sulfinic acid group of Prx-SO2, activating the

formation of a thiolsulfinate complex between Prx-SO2 and the catalytic Cys of Srx. In

yeast, 2-Cys Srx is recycled by intramolecular disulfide, releasing Prx-SOH that is

further reduced by intramolecular disulfide and antioxidant systems such as Trx. On the

other hand, in mammals, 1-Cys Srx recycling might depend on other thiol reducer such

as GSH [23,40,41] (Appendix Fig. S3).

Srx is relevant to diseases with a protective function. In assays of human A549 lung

cancer cells, H2O2-induced ROS was decreased in SRXN1-transfected A549 cell [42],

and Srx null cell has increased ROS levels, and significantly slower migration ability

[43]. Srx is also reported to deglutathionylate Prx, actin, and PTP1B [44], and the

deglutathionylation function to Prx is reported to be specific [45]. Srx was also reported

to denitrosylate Prx2 in ATP-consuming manner in vitro [46]. Further, Srx/Prx axis is

relevant to cancer development [23], and it is recently found that Srx may promote

cervical cancer metastasis via Wnt/beta-Catenin signaling [47].

1.4. Specific aims and significances

Herein we investigated the regulatory roles of Srx in antioxidative responses,

focusing on the thiol-specific oxidative stress. First we established a diamide treatment

model in HeLa cells where diamide induced thiol-specific attacks on proteins, and then

clarified the protein thiolations and antioxidant responses triggered by diamide

treatment. Based on this model, we utilized SRXN1 knockout HAP-1 cell and SRXN1

over-expressing HEK 293 cells to elucidate the function of Srx under diamide-induced

thiol-specific attacks. The specific aims of this study were listed as below:

(1) How Srx regulates the antioxidation responses to diamide treatment;

(2) How Srx interacts with antioxidative systems;

(3) How Srx scavengers the thiol-specific oxidative stress.

Since Srx is well recognized regarding its function on reducing sulfinic acid

modification of Prx1–4, here we pointed at an unreported phenomenon, implicating that

there was an additional antioxidant function of Srx.

2. Materials and Methods 2.1. Chemicals and antibodies

Diamide ((CH3)2NCON=NCON(CH3)2) was bought from Sigma (D3648). Antibodies

of sulfiredoxin, peroxiredoxin, and thioredoxin 2 were bought from Santa Cruz, Inc.,

anti-thioredoxin 1 was bought from R&D, Inc., and anti-thioredoxin reductase was

bought from Sigma, Inc. Protein structural prediction was performed by ExPasy online

tool (https://swissmodel.expasy.org/).

2.2. Cell culture

HeLa and HEK cells were obtained originally from American Type Culture

Collection (ATCC). SRXN1 knockout cell line (product ID: HZGHC006356c003) and

its parental HAP-1 cell line were bought from Horizon, Inc. HeLa cells were cultured in

high-glucose Dulbecco's Modified Eagle Medium (DMEM) with 10% fetal bovine

serum and 1% antibiotic/antimycotic solution (HyClone), while HEK cells were

cultured with 10% fetal calf serum. Parental and SRXN1 KO HAP-1 cells were cultured

in Iscove’s modified Dulbecco;s medium (IMDM) with 10% fetal calf serum and 1%

antibiotic/antimycotic solution. All were cultured at 37°C within 5% CO2 atmosphere.

2.3. SRXN1 knockout, genomic DNA extraction and sequencing

SRXN1 knockout HAP-1 cell line (Horizon) was edited by CRISPR/ Cas9 system.

The mutation was 1 bp insertion in exon 2 of SRXN1 (target transcript NM_080725).

Genomic DNA was extracted from both parental and SRXN1 KO cells by EasyPure

Genomic DNA mini kit (Bioman) and utilized as the PCR template. For sequencing, a

200-bp DNA segment containing the mutation site was amplified. The primer sequences

were 5’TTCTCCCCAGGAGGACCCA3’ and 5’GTGTGGATGCTCCCAGGTAC3’.

PCR was done with Solg^TM2X h-Taq PCR Smart mix 1, and the conditions were listed

as below: initial denaturation at 95°C for 15 min, amplification for 30 cycles including

denaturation (95 °C for 20 sec), annealing (60, 62, 64 °C gradient PCR for 40 sec), and

extension (72 °C for 30 sec), and final extension at 72 °C for 5 min, and then holding at

4 °C.

2.4. SRXN1 over-expression

FLAG-SRXN1 clone (target transcript: NM_080725) was bought from Vigene

Biosciences Inc. Bacterial stock was amplified with LB broth containing 1:500 diluted

Ketamycin. Plasmid was extracted with Easy Pure® plasmid DNA miniprep kit

(Bioman), and examined by 1% agarose gel electrophoresis. Extracted FLAG-SRXN1

plasmid was transfected into HEK293 cells with Lipofectamine 3000 (Invitrogen)

followed the suggested concentration and procedure.

2.5 Cell lysate preparation

For lysate preparation, cells were lysed with lysis buffer (50 mM MOPS, 1% CHAPS,

and 8 M urea, pH 7.0) or with Radio immunoprecipitation assay buffer (RIPA) [48],

and harvested the supernatant after centrifugation.

2.6. Fractionation

Cell fractionation was done with Mitoscience® cell fraction kit (ab109719). Cultured

cells were washed twice with ice-cold PBS, and lysed with buffer A containing 1:50

diluted detergent I followed by centrifugation (5,000 rpm for 1 min) to collect the

supernatant as the cytosolic fraction. The pellet was washed in buffer A twice, then

lysed with buffer A containing 1:2000 diluted detergent II, followed by centrifugation

(5,000 rpm for 1 min) to collect the supernatant as the mitochondrial fraction. The final

pellet was washed twice and dissolved as the nuclear fraction.

2.7. SDS-PAGE, Coomassie blue staining, and immnunoblotting

For SDS-PAGE samples preparation, cell lysate or protein extract were mixed with

2x sample buffer with or without 2% β-mercaptoethanol, and boiled for 10 min. The

electrophoresis was carried out with the bis-tricine buffer system, utilizing 10%

acrylamide gel. For Coomassie blue staining, acrylamide gel was washed twice by H2O,

stained with CBR solution overnight and de-stained by H2O. For immunoblotting,

proteins were transferred from gel to PVDF membrane. Blotted membrane was blocked

with 5% skim milk for 30 min, than probed overnight by antibodies (1:1,000 dilution) in

PBST buffer with 0.03% BSA, and followed by secondary antibody probing and

detected by ECL illumination.

2.8 Immunofluorescence

For immunofluorescence, cells were cultured on 20 mm² coverslips (Marienfeld).

After washed in PBS, specimens were obtained by fixing with 4% performaldehyde in

PBS for 15 min, permeabilized with 0.3% Tritone X-100 in PBS for 10 min, and then

blocked with 5% skim milk in PBST for 1 h. For probing the target proteins, specimens

were incubated with 1:200 diluted antibodies in PBST with 0.03% BSA at 4 °C

overnight, and then adding fluorescence-labeled secondary antibodies in dark at room

temperature for 1 h. After washed in PBS, 1:20,000 diluted DAPI was added and

incubated for 5 min to stain cell nucleus. Specimens were mounted by antifade

mounting reagent and analyzed by Leica DM6000B Upright Microscope System.

2.9. RNA extraction and RT-PCR

mRNA was extracted from cultured cells by Oligotex mRNA Mini Kit (Qiagen).

Extracted mRNA was used as template and subjected to reverse transcription by

OneStep RT-PCR kit (Qiagen). The PCR products were analyzed by electrophoresis

with 1.5% agarose gel. PCR condition and primer sequences of SOD2, NQO1, GSTP1,

and actin were referred to the article of Paupe V et al. [49].

2.10. Immunoprecipitation

For immunoprecipitation, cultured cell was lysed with lysis buffer (50 mM MOPS,

pH 7.2, 1 mM EDTA, 1% NP-40, and 1:100 diluted protease inhibitor cocktails) to get

the supernatant, followed by antibody probing at 4°C overnight. Target proteins were

pulled down by Pierce® Protein A Plus Agarose, washed in lysis buffer trice, and then

eluted by sample buffer which containing β-mercaptoethanol.

2.11. Diamide treatment and resin-assisted capture

Diamide at 5 mM was added into cultured medium and incubated for 5 minutes and

washed in PBS. For recovering, removed diamide and incubated with fresh culture

medium for 30 min. For resin-assisted capture, cultured cell was lysed with Tris-urea

lysis buffer (Tris 10 mM, 8 M urea, 1% Triton, pH 8.0) containing 100 mM

iodoacetamide (IAA), incubated for 1 h, and precipitated with saturated ammonium

sulfate solution. The pellet was resolved with the Tris-urea lysis buffer containing 100

mM DTT, incubated for 1 h and then precipitated. After wash twice for completely

eliminating DTT, pellet was resolved with Tris-urea lysis buffer and incubated with

thio-propyl sepharose® 6B (Pharmacia) for 1 h at room temperature, washed in

Tris-urea lysis buffer containing 2% SDS, eluted with protein sample buffer containing

2% β-mercaptoethanol and boiled for 10 min. Harvested samples were objected to

SDS-PAGE or followed by immunoblotting

2.12. Hydrogen peroxide-induced oxidation

Cell was treated with 0 – 10 mM H2O2 in cultured medium for 10 min, washed in

PBS, and lysed with lysis buffer. The lysate was subjected to immunoblotting.

2.13. MTT assay

Diamide-treated cells were washed with PBS, covered with 5 mg/mL MTT in PBS

and allowed to reaction at 37 °C for 2 h. After removing MTT solution, the formed

purple formazan was dissolved by DMSO, and subjected to measuring the absorbance at

570 nm by Molecular Devices M2 Microplate Reader.

2.14. Anti-glutathione antiserum preparation

For preparing the antigen, ovalbumin was utilized as carrier protein, coupled with

glutathione by crosslinker sulfo-succinimidyl (4-iodoacetyl)aminobenzoate (SIAB)

bought from Thermo Fisher, Inc. Antigen was subjected to routinely subcutaneous

immunizations in rabbits, and the blood was collected from the anesthetized animals

after 8 bi-weekly injections and 10 days from the final injection.

2.15. Thioredoxin reductase activity assay

Thioredoxin reductase (TrxR) activity assay was done with Thioredoxin Reductase

Assay Kit (Sigma Aldrich). Harvested cell pellets was lysed with lysis buffer (50 mM

MOPS, 1% Triton X-100) and allowed to reacting with 5,5’-Dithiobis(2-nitrobenzoic

acid) (DTNB) 3 mM in working buffer (10 mM potassium phosphate, 10 mM EDTA,

0.24 mM NADPH), with or without TrxR inhibitor solution (1:1,000 dilution).

Absorbance of 412 nm was measured kinetically for 10 min (interval: 30 sec) with

Molecular Devices M2 Microplate Reader.

2.16. In-gel digestion and protein identification by mass spectrometry

Protein samples were loaded to 10% SDS-PAGE and separated for a distance of 1 cm.

Each lane of gel was cut into 2 mm slices. These gel slices were soaked in 100 mM

dithiothreitol/25 mM ammonium bicarbonate, followed by alkylation with 65 mM

iodoacetamide. The gel slices were washed in 50% acetonitrile in 25 mM ammonium

bicarbonate. The dehydrated gels were moisturized with trypsin solution (10 ng/µL,

Promega, Madison, WI) and incubated at 37 °C for 16 h. The tryptic peptides were

extracted with 50% and 100% acetonitrile containing 0.1% trifluoroacetic acid

sequentially. The peptide mixtures were desalted by C18 Zip-tip and subjected to

proteomics analysis using an LTQ-Orbitrap Velos hybrid mass spectrometer (Thermo

Fisher Scientific).

Peptide mixtures were loaded onto a 75 µm × 250 mm nanoACQUITY UPLC

BEH130 column packed with C18 resin (Waters, Milford, CT) and were separated at a

flow rate of 300 nL/min using a linear gradient of 5 to 40% solvent B (95% acetonitrile

with 0.1% formic acid) in 30 min, followed by a sharp increase to 85% B in 1 min and

held at 85% B for another 10 min. Solvent A was 0.1% formic acid in water. The

effluent from the HPLC column was directly electrosprayed into the mass spectrometer.

The LTQ Orbitrap Velos instrument was operated in data-dependent mode to

automatically switch between full-scan MS and MS/MS acquisition. Instrument control

was through Tune 2.6.0 and Xcalibur 2.1. For the CIDMS/MS top20 method, full-scan

MS spectra (m/z 350−1600) were acquired in the Orbitrap analyzer after accumulation

to a target value of 106 ions in the linear ion trap. Resolution in the Orbitrap system was

set to R = 60, 000 (all Orbitrap system resolution values are given at m/z 400). The

20 most intense peptide ions with charge states ≥2 were sequentially isolated to a target

value of 5000 and fragmented in the high-pressure linear ion trap by low-energy CID

with normalized collision energy of 35%. The resulting fragment ions were scanned out

in the low-pressure ion trap at the normal scan rate and recorded with the secondary

electron multipliers. Ion selection threshold was 500 counts for MS/MS, and the

maximum allowed ion accumulation times were 500 ms for full scans and 100 ms for

CID-MS/MS measurements in the LTQ. An activation q = 0.25 and activation time of

10 ms were used.

The peptides were identified from the MS/MS data searched against the SwissProt

database using the Mascot search engine 2.3.02 (Matrix Science, Boston, MA). Search

criteria used were as follows: trypsin digestion; variable modifications set as acetyl

(protein N-terminal), carbamidomethyl (Cys) and oxidation (Met); up to two missed

cleavages allowed; and mass accuracy of 10 ppm for the parent ion and 0.60 Da for the

fragment ions.

2.17. Metabolomics analysis by mass spectrometry

Cultured cells lysed with H2O and precipitated protein with 66% acetonitrile.

Supernatants were directly applied to the LC-ESI-MS analyses consisted of an

ultra-performance liquid chromatography (UPLC) system (Ultimate 3000 RSLC,

Dionex, Sunnyvale, CA, USA) and an electrospray ionization (ESI) source of

quadrupole time-of-flight (TOF) mass spectrometer (maXis UHR QToF system, Bruker

Daltonics, Bremen, Germany). The samples were kept in an autosampler at 4 ◦C.

Separation was performed with reversed-phase liquid chromatography (RPLC) on a

BEH C18 column (2.1 x 100 mm, Walters). The elution started from 99% mobile phase

A (0.1% formic acid in ultrapure water) and 1% mobile phase B (0.1% formic acid in

ACN), held at 1% B for 0.5 min, raised to 60% B in 6 min, further raised to 90% B in

0.5 min, held at 90% B for 1.5 min, and then lowered to 1% B in 0.5 min. The column

was equilibrated by pumping 1% B for 4 min. The flow rate was set 0.4 ml/min with

injection volume 2 µl. LC-ESI-MS chromatogram were acquired under following

conditions: capillary voltage of 4500 V in positive ion mode, dry temperature at 200℃

dry gas flow maintained at 9.0 l/min, nebulizer gas at 2.0 bar, and acquisition range of

m/z 100-1000. Data were acquired by HyStar and micrOTOF control software (Bruker

Daltonics, Bremen, Germany) and processed by DataAnalysis software (Bruker

Daltonics, Bremen, Germany). Metabolite was identified by the theoretical m/z value

and isotope pattern derived from the chemical formula, and was integrated the area of

signal peaks in the extracted ion chromatogram.

3. Results

3.1. Diamide induced reversible thiolations and antioxidant responses

Diamide (Fig. 1A) reacts with free thiolate anions (RS^-) such as glutathione, changing

the redox equilibrium of thiols in cells, leading to thiol-specific attacks of proteins [50].

According to the mechanism, it would not result in irreversible cysteine oxidations

including sulfinic and sulfonic acids.

Herein we used diamide treatment to focus on protein thiol oxidations, and verified

the oxidized protein targets by resin-assisted capture (RAC) [31]. The oxidized proteins

were first alkylated with iodoacetamide (IAA), and then reduced the oxidative

modifications by dithioeryreitol (DTE). The nascent cysteine thiols were captured by

thiopropyl sepharose thus pulled down the oxidized proteins (Fig. 1B).

To establish a diamide-induced thiolation model, HeLa cells were treated with 0 – 5

mM diamide for 5 min, and the lysate was subjected to RAC (Fig. 1C). When the dose

of diamide was higher than 1 mM, protein thiolations increased in a dose-dependent

manner. Moreover, when diamide was removed and cells were incubated with fresh

culture medium for 5 – 60 min, protein thiolation level recovered in 30 – 60 mins (Fig.

1D). The data showed that treating high-dose diamide for short incubation time induced

reversible protein thiolations.

The thiolated targets were further identified by LC-MS analysis. The potential targets

were in various functions such as glycolysis, pentose phosphate pathway, citric acid

cycle, antioxidant responses, and chaperones (data not shown). Several potential targets,

including pyruvate kinase, aldolase, Hsp90, GAPDH, and peroxiredoxin, were

confirmed by immunoblotting (Fig. 2).

To figure out whether diamide treatment affected metabolism in cells, metabolites

were extracted from both control and diamide-treated cells by 66% acetonitrile and

subjected to LC-MS analysis. Reduced glutathione (GSH) decreased in diamide-treated

cells; in contrary, oxidized glutathione disulfide (GSSG), NADH and NADPH

increased significantly (Fig. 3). After recovering for 30 min, GSH increased, while

GSSG recovered to the level similar to the control group (Appendix Fig. S4). The data

indicate that glutathione is the main victim of the diamide attack.

To elucidate whether diamide triggered antioxidant responses at the transcriptional

level, Nrf2 translocation was examined by subcellular fractionation and

immunofluorescence staining. Both data indicated that Nrf2 translocated into cell

nucleus after 5-min diamide treatment (Fig. 4A, B). The mRNA of Nrf2 downstream

genes, such as superoxide dismutase 2 (SOD2), NAD(P)H quinone dehydrogenase 1

(NQO1), and glutathione S-transferase 1(GSTP1), increased after diamide treatment and

remained high during recovery (Fig. 4C). Therefore, diamide treatment induced

Nrf2-related antioxidant responses, especially in the process of recovery.

3.2. Sulfiredoxin knockout increased diamide-induced protein thiolations

For examining the role of sulfiredoxin under diamide treatment, SRXN1 knockout

cells were used. CRISPR/Cas9-edited SRXN1 (NM_080725) knockout HAP-1 cell line,

which is hyploid and suitable for gene editing [51], was bought from Horizon®. The

mutation was 1 bp adenosine insertion in exon 2 of SRXN1. For confirming, genomic

DNA was extracted as the template in PCR and PCR product was sequenced (Fig. 5A).

This insertion resulted in a truncation of sulfiredoxin at the conserved GCHR

(G98-R101) region [39], which included the active cysteine residue C99 [35] (Fig. 5B).

The metabolic activity of HAP-1 cells under diamide treatment was examined.

According to the results of MTT assay, 5 mM diamide treatment significantly decreased

cell metabolic activity in both parental and KO cells, whereas there was no significant

difference between them (Fig. 6A). Thus we applied diamide-treated model to HAP-1

cell to elucidate the effect of sulfiredoxin on protein thiolations.

The diamide treatment induced protein thiolations in both parental and KO cells,

which recovered after 30 min (Fig. 6B). Interestingly, the thiolation level was obviously

higher in KO cells than in parental cells, implicating that sulfiredoxin might play some

roles cells facing thiol-specific assault. The thiolated targets were also identified by

LC-MS analysis (Appedix Table S1), and several targets also appeared in the HeLa cell

model, such as pyruvate kinase, aldolase, Hsp90, and GAPDHl (Fig. 6C). Some targets,

such as actin and tubulin, were found only in KO group, could be potential interacting

partner with sulfiredoxin (Appedix Table S1).

3.3. Sulfiredoxin knockout affected functions of antioxidant enzymes

There were several antioxidant enzymes identified in RAC data (Table 1). Confirmed

by immunoblotting, the thiolation levels of peroxiredoxin (Prx), thioredoxin 2 (Trx2),

and glutathione reductase increased while treated with diamide, and recovered after 30

min (Fig. 7). All these enzymes exhibited a higher thiolation level in SRXN1 KO cells.

Meanwhile, compared with parental cells, peroxiredoxin and thioredoxin 2 were

thiolated in the sulfiredoxin knocked out cells. Since thiolation in these thiol-dependent

enzymes may be relevant to their antioxidant functions, sulfiredoxin knockout has

impact on these antioxidant systems.

3.3.1. Glutathione system

As the main reductant in mammalian cells, glutathione (GSH) releases cell from

oxidative stress by forming intermolecular disulfide bridge, either protein

glutathionylations or oxidized glutathione (GSSG) [7].

To examine whether sulfiredoxin knockout affects protein glutathionylation, cell

lysates of parental and KO cells were probed with anti-glutathione immunoblotting (Fig.

8A). Although the antiserum recognizes glutathionylation (Appendix Fig. S5), its

specificity was not satisfactory. However, the data indicated that basal levels of protein

glutathionylation were higher in KO cells than in parental cells.

Oxidized glutathione disulfide (GSSG) in parental and KO cells was examined by

LC-MS analysis. Diamide induced significant increase in GSSG in KO cells (Fig. 8B).

The data indicated that sulfiredoxin knockout increased the diamide-induced glutathione

oxidations, both protein glutathionylation and glutathione disulfide.

3.3.2. Peroxiredoxin and thioredoxin system

As the known substrate of sulfiredoxin [35], we studied how peroxiredoxin was

regulated in SRXN1 KO cells under diamide treatment. 2-Cys peroxiredoxins form

homo-dimer with inter-chain disulfide between peroxidatic cysteine (CP) and resolving

cysteine (CR) while facing to oxidative stress [33]. In HAP-1 cells, diamide treatment

induced peroxiredoxin dimerization, and the dimerization increased more in

sulfiredoxin knockout cells (Fig. 9A). Peroxiredoxin thiolations were already present in

knockout cells without diamide treatment (Fig. 7), suggesing that peroxiredoxin

glutathionylation increased while sulfiredoxin was knocked out. Therefore, sulfiredoxin

might relevant to the regulation of peroxiredoxin dimerization and glutathionylation.

According to the immunoblotting, the expressions of peroxiredoxin, thioredoxin 1,

thioredoxin 2, and thioredoxin reductase were not significantly different between

parental and KO cells (Fig. 9B), so that the thiolation increase was not due to the

change of protein amounts.

In general, thioredoxin forms intramolecular disulfide to recycle oxidized

peroxiredoxin dimer, and the oxidized thioredoxin is reversed by thioredoxin reductase

[52]. Examining by Ellman’s assay combined with thioredoxin reductase inhibitor, the

activity of thioredoxin reductase was significantly higher in KO cells than in parental

cells (Fig. 9C). Due to the unchanged expression level of thioredoxin reductase (Fig.

9B), the higher activity might be due to higher reducing status of the enzyme. Besides,

according to the change of the redox status, I suggest that while sulfiredoxin was

knocked out, the responses of thioredoxin system to the diamide-induced attacks might

have a compensatory increase.

3.4. Diamide treatment induced cytosolic sulfiredoxin dimerization

To elucidate how diamide treatment affected sulfiredoxin, diamide-treated cells were

examined with anti-sulfiredoxin antibody. However, given the sensitivity of anti-Srx

antibodies were not enough to recognize endogenous sulfiredoxin, we expressed

FLAG-tagged sulfiredoxin in HEK293 cells.

In immunoblots with both anti-sulfiredoxin and anti-FLAG, diamide treatment led to

signal of sulfiredoxin of higher molecular mass, which could be reduced by

β-mercaptoethanol (Fig. 10A, B). Given there is only one cysteine in mammalian

sufliredoxin, the high molecular-weight sulfiredoxin complex was possibly protein

dimer by an intermolecular disulfide at the only cysteine residue C99. In order to verify

the binding partner of sulfiredoxin, we immunoprecipitated sulfiredoxin complex,

disassociating the complex with reducing agent, and allowing the sample to LC-MS

analysis (Table 2) and immunoblotting (Fig. 10C). I found no potential binding

candidates, thus the complex might be a homo-dimer. Additionally, some thiolated

proteins were found only in diamide-treated group, such as S100A8/9 complex,

suggesting that they also play some roles in antioxidant responses. The

immunoprecipitated sulfiredoxin was also allowed to anti-GSH immunoblotting (Fig.

10D) and found no evidence of being glutathionylated.

Sulfiredoxin was reported to translocate into mitochondria while induced by H2O2

signaling [53]. To figure out whether the diamide-induced sulfiredoxin dimer

translocated into mitochondra, SRXN1 over-expressed cell was objected to subcellular

fractionation, and anti-Srx immunoblotting. After diamide treatment, only cytosolic

sulfiredoxin dimer increased, and the dimer could be reduced by β-mercaptoethanol

(Fig. 11A), thus diamide attack only resulted in cytosolic sulfiredoxin dimerization and

no mitochondrial translocation. Unexpectedly, there was mitochondrial sulfiredoxin

dimer found even without diamide treatment. Meanwhile, 10 mM H2O2 treatment for 10

min also resulted in reversible sulfiredoxin dimerization, implicating the possibility that

sulfiredoxin could release oxidative stress by forming a dimer with disulfide (Fig. 11B).

3.5 Hypothetic mechanism of sulfiredoxin under thiol-specific oxidative attacks

Based on the phenomena observed in knockout and over-expression models, herein a

hypothetic mechanism of sulfiredoxin was proposed (Fig. 11). Besides the reducing

ability of hyperoxidized peroxiredoxin sulfinic acid (Prx-SO2), sulfiredoxin might

scavenge oxidative stress by forming intermolecular disulfide at its only cysteine

residue C99. Its antioxidant function might involve peroxiredoxin dimerization and

glutathionylation. Sulfiredoxin knockout leads to less resistance to diamide-induced

attacks, resulting in more glutathione oxidations and more protein thiolations (including

glutathionylation). However, the thiolations would be recoverable as well, for other

antioxidant systems, such as thioredoxin/ thioredoxin reductase, might be activated in

compensation.

4. Discussion

Diamide treatment induced general protein thiolations, GSSG increase, and Nrf2

translocation in HeLa cell. Although 0.5 – 2 mM diamide treatment for 15 – 60 min is

enough to oxidize membrane and cytoskeleton proteins [50], the thiolations induced by

5 mM diamide treatment for 5 min, which was higher dose and shorter duration, were

still recoverable. The identified thiolation targets included several enzymes that

participated in antioxidant responses and energy metabolic pathways. Metabolomics

survey showed that diamide treatment increased GSSG level as reported [54,55],

indicating GSH might cushion the diamide attacks mainly. Additionally, Nrf2

translocation was induced by diamide treatment, leading to downstream gene expression

of antioxidant enzymes.

Herein we reported a protective function of sulfiredoxin in diamide-induced protein

thiolations, which implicated a more essentially antioxidant role of sulfiredoxin. Since

the mechanism of diamide was attacking thiolate and interrupting the equilibrium of

reducing glutathione pool [50], we utilized diamide treatment to rule out the effect from

ROS production. According to previous literatures, sulfiredoxin reduced peroxiredoxin

S-nitrosylation in 60 min but more than 45 min in vitro [46]. Considering that SRXN1

knockout increased protein thiolations and glutathione oxidations within 5 min

treatment (Fig. 6), here it might release oxidative stress immediately, similar to

glutathione or thioredoxin. Since sulfiredoxin interrupts in vitro glutathionylation

induced by PABA/NO [44], and negatively relates with ROS productions [42,43], it

might scavenge the ROS thus protect proteins from oxidations. The function might be

relevant to its conserved region (Fig. 5), and the over-expression model implicated that

it might release the oxidative stress by intermolecular disulfide on its only cysteine

residue (Fig. 10).

Diamide triggers mainly S-thiolations in the cytoplasmic proteomes of Bacillus

subtilis and Staphylococcus aureus [56]. In our fractionation data of HAP-1 cells, it was

also stated that diamide dominantly induced thiolations in cytosolic, so that the

diamide-induced sulfiredoxin dimer was only found in cytosol and without translocation

to mitochondria (Fig. 11). On the other hand, consistent observation of its dimerization

in mitochondria indicated another possible antioxidant role of sulfiredoxin (Fig. 11).

Due to mitochondrial electron transferation [1,2], high H2O2 concentration in

mitochondria might trigger the antioxidant response of sulfiredoxin, so that increased its

dimerization. However, due to the lack of an efficient anti-Srx antibody, we could only

detect sulfiredoxin in overexpressing system. For further clarification, a better tool is

required.

Althought sulfiredoxin is activated by Nrf2 and AP-1 as downstream signaling [38],

due to the delay of Nrf2 signal transduction observed in the HeLa cell model (Fig. 4C),

it might mainly function in the recovery state but not in the immediate rising of

thiolation level. Meanwhile, although sulfiredoxin knockout, protein thiolations was

still recovered in 30 minutes, implicated the possibility of a compensatory induction of

other antioxidant systems such as thioredoxin and glutathione. Both of them increase

thiolations while sulfiredoxin knocked out (Fig. 7-9), and thioredoxin reductase was

more dominant in knockout cells (Fig. 9C).

Chapter II.

Development of Quantitative Display of Protein Redox Status

1. Introduction

1.1. Detection tools of protein cysteine oxidations

Cysteine oxidative modifications, such as S-nitrosylation, disulfide, and S-sulfenic

acid, participate in several biological functions including protein structuring, signaling,

and enzyme activity [21,22,57]. For detecting thiol modifications, differential alkylation

method was first introduced in 2001 for analysis of S-nitrosocysteine and remains the

most used one [30,58]. The initial step is blocking of all reduced SH groups by

alkylation; after removal of excess alkylating reagent, specific reductant such as

ascorbate is added, thus generated nascent thiols upon the modification-specific

reduction. Another alkylating agent containing a specific tag, such as biotin achieves

final alkylation step, rendering identification and enrichment feasible. Strong reducing

agents like dithioerythritol reduce all reversibly oxidized cysteines, glutaredoxins and

sodium arsenite reducing S-glutathione and SOH, respectively [59,60].

The original differential alkylation method is also called biotin-switch technique

(BST) or tag-switch technique. Meanwhile, the resin-assisted capture (RAC) strategy

was developed in 2009 to globally profile protein S-nitrosylation dynamics in

S-nitrosylcysteine-treated E. coli and HEK293 cells [31]. In this method, redox

modified thiols are tag-switched with a thiol-reactive resin and enrichment is achieved

following extensive wash procedures and elution with β-mercaptoethanol or

dithiothreitol. One major drawback associated with the BST is that some endogenous

proteins containing the biotin prosthetic group will be co-purified and co-identified with

the genuine signals.

Commonly used alkylating reagents include iodoacetamide (IAM), N-ethylmaleimide

(NEM), and bromobimane (BBM) [29]. Cysteine oxidations can also be distinguished

by mass spectrometry, which coupled with alkylation method and amplifying signals by

isotope-coded affinity tag (ICAT), identifying specific modification sites [61,62].

Antibody is also utilized to detect cysteine oxidations; for example, reversed PTP1B

S-nitrosylation could be recognized by anti-iodoacetamide (IAM) antibody [63].

Dimedone, which specific reacting to sulfenic acid, was applied to label cysteine

sulfenic acid and recognized by its antibody [32]. Antibodies directly probing to PTP1B

sulfonic acid and SOD1 sulfonic acid are also reported [64,65].

1.2. PEG-maleimide as a cysteine-specific probe

Maleimide derivatives are widely used for the alkylation of thiols in which thiolates

attack the electrophilic double bond of the maleimide [66] (Appendix Fig. S6).

Maleimide-polyethylene glycol (m-PEG) labeling is commonly used as an easy

approach of reversible protein cysteine oxidation [67]. It is designed for analysis of

disulfide bonds in proteins with and without prior reduction [68], for site-specific

PEGylation of therapeutic proteins to avoid rapid degradation and excretion from

patients [69], for topology determination of membrane proteins by scanning cysteine

mutagenesis and accessibility [70,71]. In addition, m-PEG has been used to detect

reversibly oxidized creatine kinase, protein kinase A and protein kinase G by reacting to

the reduced cysteine residues leading to mobility shift in immunoblots; a method called

PEG-switch [67,72]. M-PEG labeling of proteins leads to a shift to high-molecular-

weight end, thus can be distinguished by SDS-PAGE or immunoblotting. The

ameliorated method, which combining with click-chemistry, is also reported [73].

Compared with other detection strategies, this method is easy, quick and not so

expensive, whereas it might be interrupted due to the conformational change of

antibody-recognized epitopes. Additionally, it has been implicated that quantify the

labeling ratio of proteins reflects the oxidation status [73].

1.3. Specific aims and significances

Herein we optimized the labeling conditions and exploited the applications of

PEG-switch in quantitation of the extent of protein cysteine oxidation in cells which in

response to H2O2 and insulin. Based on the PEG labeling strategy, we proposed a

scoring system to reflect the redox status of target proteins. Since recent detection tools

are most likely to be qualitative rather than quantitative, the assay method we provided

was easy, fast, and promised in quantitatively comparing various protein redox statuses.

2. Materials and methods 2.1. Chemicals and antibodies

Methoxypolyethylene glycol maleimide 5000 (PEG-maleimide 5K) was bought from

Sigma, and PEG-vinyl sulfone 5K was bought from Tours. Antibodies of peroxiredoxin,

protein tyrosine phosphatase 1 B (PTP1B), prohibitin, Heat shock protein 27 (Hsp27),

nm23, histone deacetylase 6 (HDAC 6), glutathione reductase, and deoxythymidylate

kinase (DTYMK) were bought from Santa Cruz, Inc., and which of B-cell lymphoma 2

(Bcl-2), fatty acid synthase, nuclear factor kappa-light-chain-enhancer of activated B

cells (NFκB), calcium/calmodulin-dependent protein kinase II (CamKII), calpain,

caspase 3, and caspase 9 were bought from Cell Signaling Inc. Horse insulin was

bought from Sigma.

2.2. Cell culture and lysis condition

HeLa, HepG2, HEK, and 3T3 cells were obtained originally from American Type

Culture Collection. HeLa and HepG2 cell were cultured in high-glucose Dulbecco's

Modified Eagle Medium with 10% fetal bovine serum and 1% antibiotic/antimycotic

solution (HyClone), while HEK cell and 3T3 cells were cultured in 10% fetal calf serum.

All were cultured at 37°C within 5% CO2 atmosphere. For lysates preparation, cells

were lysed with the lysis buffer (50 mM MOPS, 1% CHAPS, and 8 M urea, pH 7.0),

and harvested the supernatant after centrifugation.

2.3. PEG-maleimide labeling

PEG-maleimide (m-PEG) was dissolved in H2O for 5 mg/mL as 5x stock solution,

and diluted by lysis buffer to 1 mg/mL as final concentration. Dealing with in vitro

oxidized samples, cell lysates were precipitated with ammonium sulfate and dissolved

with lysis buffter containing m-PEG. On in cellulo oxidation samples, cells were lysed

directly with lysis buffer with m-PEG. For labeling protein thiol residue, solution

containing cell lysate and m-PEG was incubated at 37°C for 30 minutes, followed by

adding sample buffer with 2% β-mercaptoethanol immediately to quench free m-PEG.

Labeled samples were applied to SDS-PAGE and immunoblotting. Protein quantitation

was done with Pierce® BCA protein assay kit (Thermo Scientific).

2.4. SDS-PAGE, immunoblotting and reduced/oxidized ratio calculation

previous publication. [74] Signals of immunoblotting were integrated by Quanti-Scan

(BioSoft) for quantification, and the ratio was examined by t test.

2.5. Hydrogen peroxide-induced oxidation

For in vitro oxidation, HeLa cell lysate was treated with 0 – 10 mM H2O2 for 1 h at

room temperature, and precipitated with saturated ammonium sulfate in 4x volume.

Dissolved the pellet with lysis buffer to appropriate concentration of protein (not below

0.5 µg/mL) for applying to labeling. For in cellulo oxidation, cultured HEK cell was

treated with 0 – 10 mM H2O2 in serum-free DMEM for 10 minutes, washed in PBS

trice, and lysed with lysis buffer (with or without PEG-maleimide) immediately for

labeling.

2.6. Insulin-induced PTP1B oxidation in HeLa cell

Confluent HeLa cell cultured in 6-well plate was treated with insulin 10 ng/mL for 0

– 30 minutes, washed by PBS quickly, and lyzed with lysis buffer (50 mM MOPS, 1%

CHAPS, and 8 M urea, pH 7.0) including 1 mg/mL PEG-maleimide immediately at

37°C for 30 min for labeling.

3. Results

3.1. Protocol of PEG tagging

The flowchart of this m-PEG tagging method was shown as Figure 13. Cultured cells

were treated with oxidants, washed and then lysed with a lysis buffer containing 1

mg/ml m-PEG. Alternatively, cell lysate was harvested, treated with oxidants, and

precipitated with ammonium sulfate. The resulting protein extract was dissolved in the

lysis buffer containing 1 mg/ml m-PEG. The m-PEG tagging reaction was carried out at

37°C for 30 min [31], and then quenched with an equal volume of 2x SDS sample

buffer containing 4% β-mercaptoethanol [32] to stop the reaction. The resulting samples

were subjected to SDS-PAGE and immunoblotting analysis. Successful tagging of the

reduced cysteine residues in proteins would result in mobility shift on electrophoresis

due to the large PEG tag, while oxidized cysteine residues are inert to the m-PEG

tagging. However, this method does not differentiate the various forms of cysteine

oxidation.

3.2. Optimizing tagging efficiency

Although m-PEG has been used in labeling protein cysteine residues [28, 29], the

detail of optimal labeling condition in crude protein extract are still uncovered. To

optimize the tagging efficiency, we used peroxiredoxin as a target of m-PEG due to its

well-known sensitivity to oxidative stress [75].

To elucidate the optimal reaction concentration, m-PEG was added to the HeLa cell

lysate to a final concentration of 0-1 mg/ml. The reaction was allowed for 30 min at

37°C. At 1 mg/mL of m-PEG, most peroxiredoxin molecules were tagged and shifted to

higher molecular forms (Fig. 14A). The protein concentration of the cell lysate was 0.54

mg/mL according to BCA quantitation, thus the ratio of m-PEG (at 1 mg/ml) to protein

was about 2:1 (w/w). Assuming average molecular size of human protein is 50 kD, the

molar ratio of m-PEG to protein would be near 20:1. Taking the average protein length

of 400 amino acids [76] and cysteine composition of 2.26% [24], the molar ratio of

m-PEG to cysteine residues is about 2.2:1. Therefore, m-PEG at 1 mg/ml is adequate

for tagging most cysteine residues in the protein extract. Besides HeLa cells, HEK,

HepG2 and 3T3-L1 cells were also examined for m-PEG tagging of peroxiredoxin (Fig.

14B). However, there were different patterns of m-PEG tagging among these cell lines,

possibly due to different redox status within different cells [23].

Small-molecule thiol compounds, such as hydrogen sulfide, cysteine, homocysteine,

and glutathione, are present in animal tissues [77,78], thus may consume m-PEG and

decrease the tagging efficiency. To lessen the effect of endogenous reductants, we

extracted proteins from cell lysate by ammonium sulfate precipitation (Fig. 14C). The

m-PEG tagging efficiency was moderately improved at both 0.1 and 1 mg/ml

concentrations. Accordingly, ammonium sulfate precipitation in advance is suggested

for m-PEG tagging. We also compared the tagging efficiency of m-PEG with another

cysteine-specific probe, PEG-vinyl sulfone (v-PEG), whereas the tagging of v-PEG was

less specific and efficient then m-PEG under the same reaction conditions (Fig. 14D).

The results are in accordance with the previous data [79,80]. Therefore, m-PEG is

recommended for protein cysteine tagging of cell lysate after ammonium precipitation.

3.3. M-PEG tagging in other redox-sensitive proteins

One of the advantages of this m-PEG tagging method is that many proteins can be

monitored using the same cell lysate treated with m-PEG. Here we presented this

application by surveying several redox-sensitive proteins. Procaspase 3, procaspase 9

(Fig 15A), glutathione reductase, B-cell lymphoma 2 (Bcl-2), histone deacetylases 6

(HDAC-6), nucleoside diphosphate kinase (nm23), and fatty acid synthases were shifted

to higher molecular forms caused by m-PEG treatment (Fig. 15B). On the other hand,

the signal of immunoblotting for calpain, glutathione reductase, calcium/calmodulin

dependent protein kinase II (CamkII), nuclear factor kappa B (NFκB),

deoxythymidylate kinase (DTYMK) vanished on treatment of m-PEG (Fig. 15C),

possibly because that the recognition of these proteins by the corresponding antibody

was hindered by m-PEG labeling.

3.4. Quantitating the redox status of proteins in cells

To examine if m-PEG labeling can accurately reflect the redox status of a given

protein in cell lysate, we applied this method to assay cells treated with oxidative stress.

Cysteine thiol oxidation would lead to a differential tagging efficiency of target protein,

and for m-PEG causing a mobility change of tagged protein, it could be detected in

immunoblotting, so that able to be quantitated by integrating the immunoblotting

signals.

First, we utilized an extreme in vitro oxidation example to confirm whether oxidation

cause a difference in m-PEG tagging. We examined the redox status of peroxiredoxin

and Hsp27 in HeLa cell lysate that treated with 10 mM H2O2 in vitro. Most molecules

of peroxiredoxin and Hsp27 were tagged by m-PEG without oxidant treatment, while

few molecules were attacked by m-PEG after oxidant treatment (Fig. 16A, B). It

suggested that most molecules were oxidized by H2O2.

To describe the estimation of redox status, herein we proposed a scoring system,

which could be applied to any particular protein from the m-PEG tagging results. The

equation is as follows:

Redox!score! = !f0!×!2⁰! + !f1!×!2¹!+!. . . +!fn!×!2ⁿ

Formula 1!

where f0 is the fraction of nonshifted signal over total, fn is the fraction of n-tagged

signal over total. Take PHB for example, while without H2O2 treatment, nearly all PHB

was tagged, which could be presented as score of 1, whereas with H2O2 treatment led to

almost completely labeled and presented as 0 (Fig. 16C). Higher redox scores indicate

more reduced status of proteins.

3.4.1. H

O

-induced peroxiredoxin and Hsp27 oxidations in HEK cell

Following, we applied this method to two examples of in cellulo oxidations. Cultured

HEK cells were treated with 0-10 mM H2O2 for 10 min and lyzed the cells with the cell

lysis buffer containing 1 mg/ml PEG (Fig. 17B) or treated the cell lysate with m-PEG

after ammonium sulfate precipitation (Fig. 17A). The m-PEG tagging reactions were

allowed at 37 °C for 30 min. m-PEG tagging was not efficient without prior ammonium

sulfate precipitation.

Peroxiredoxin in HEK cells exhibited four different redox status as suggested by the

non-shifted and three shifted forms (Fig. 17A). Treatment of H2O2 in HEK cells

resulted in changes in the redox status of peroxiredoxin, especially marked increase in

doubly tagged (shifted2) and decrease in triply tagged (shifted3) one.

For Hsp27, which containing only one cysteine residue, singly tagged Hsp27

decreased and non-tagged Hsp27 increased in the m-PEG tagging pattern (Fig. 18A). In

addition, the ratio of shifted/nonshifted signal of Hsp27 was significantly decreased

following 10 mM H2O2 treatment (Fig. 18B).

For peroxiredoxin, treatment of 1 mM and 10 mM H2O2 displayed the redox score of

3, 3.04 and 2.5 (Fig. 17A), and for Hsp27, the redox score was decreased from 1.46 to

1.39 and 1.14 (Fig. 18A).