Halobenzoquinone-Induced Alteration of Gene Expression Associated

3 hours ago - Halobenzoquinones (HBQs) are emerging disinfection byproducts (DBPs) that effectively induce reactive oxygen species and oxidative ...
0 downloads 4 Views 747KB Size
Subscriber access provided by Kaohsiung Medical University

Ecotoxicology and Human Environmental Health

Halobenzoquinone-Induced Alteration of Gene Expression Associated with Oxidative Stress Signaling Pathways Jinhua Li, Birget Moe, Yanming Liu, and Xing-Fang Li Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b06428 • Publication Date (Web): 08 May 2018 Downloaded from http://pubs.acs.org on May 8, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

Environmental Science & Technology

1

Halobenzoquinone-Induced Alteration of Gene Expression Associated with Oxidative

2

Stress Signaling Pathways

3 4

Jinhua Li,1,2* Birget Moe,2,3 Yanming Liu,2 and Xing-Fang Li2*

5 6 7 8 9 10 11 12

1. Department of Health Toxicology, School of Public Health, Jilin University, Changchun, Jilin, China 130021 2. Division of Analytical and Environmental Toxicology, Department of Laboratory Medicine and Pathology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G3 3. Alberta Centre for Toxicology, Department of Physiology and Pharmacology, Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada T2N 4N1

13 14 15 16 17 18 19 20 21

* Corresponding authors: Xing-Fang Li: [email protected], 1-780-492-5094; Jinhua Li:

22

[email protected]

23

ACS Paragon Plus Environment

1

Environmental Science & Technology

Page 2 of 27

24

Abstract

25

Halobenzoquinones (HBQs) are emerging disinfection byproducts (DBPs) that effectively induce

26

reactive oxygen species and oxidative damage in vitro. However, impacts of HBQs on oxidative

27

stress related gene expression have not been investigated. In this study, we examined alteration

28

in the expression of 44 genes, related to oxidative stress-induced signaling pathways, in human

29

uroepithelial cells (SV-HUC-1) upon exposure to six HBQs. The results show structure-

30

dependent effects of HBQs on the studied gene expression. After 2 h exposure, the expression

31

levels of 9 to 28 genes were altered, while after 8 h exposure, the expression levels of 29 to 31

32

genes were altered. Four genes, HMOX1, NQO1, PTGS2, and TXNRD1, were significantly

33

upregulated by all six HBQs at both exposure time points. Ingenuity pathway analysis revealed

34

that the Nrf2 pathway was significantly responsive to HBQ exposure. Other canonical pathways

35

responsive to HBQ exposure included GSH redox reductions, superoxide radical degradation,

36

and xenobiotic metabolism signaling. This study has demonstrated that HBQs significantly alter

37

gene expression of oxidative stress related signaling pathways and contributes to the

38

understanding of HBQ-DBP-associated toxicity.

39

ACS Paragon Plus Environment

2

Page 3 of 27

Environmental Science & Technology

40

41 42 43

TOC graphic

44

ACS Paragon Plus Environment

Environmental Science & Technology

45 46

Page 4 of 27

Introduction Disinfection of drinking water is necessary to inactivate pathogenic microorganisms;

47

however, it leads to the inadvertent formation of drinking water disinfection byproducts (DBPs)

48

from the reaction(s) of disinfectants (e.g. chlorine, chloramine, chlorine dioxide, or ozone) with

49

natural organic matter in source water.1 Human exposure to DBPs in drinking water is a concern

50

driven by epidemiological findings that suggest a potential association between consumption of

51

chlorinated drinking water and an increased risk of developing bladder cancer.2 Although eleven

52

DBPs, including the trihalomethanes (THMs) and haloacetic acids (HAAs), are under regulation

53

by the United States Environmental Protection Agency (US EPA), these regulated DBPs do not

54

appear to account for the observed increased bladder cancer risk based on animal carcinogenicity

55

studies.3

56

Halobenzoquinones (HBQs) are an emerging class of DBPs that have been found to

57

occur widely in treated tap water and recreational waters at levels up to 300 ng/L.4-6 HBQs have

58

also been predicted as potential bladder carcinogens by quantitative structure–toxicity

59

relationship (QSTR) analysis,7 though limited experimental data is available concerning the

60

toxicological effects of HBQs. HBQs have shown greater cytotoxicity than regulated DBPs in

61

Chinese hamster ovary (CHO) cells,8, 9 while also inducing cytotoxic effects in T24 human

62

bladder carcinoma cells with half-inhibitory concentrations (IC50) at micromolar levels.

63

Oxidative stress has been identified as one of the main mechanisms of HBQ-induced cytotoxicity

64

due to production of intracellular reactive oxygen species (ROS), depletion of the cellular

65

antioxidant glutathione (GSH), inhibition of cellular antioxidant enzymes, and oxidative damage

66

to DNA and proteins.10-12 A comparison of HBQ structure and its biological effects identified

67

isomeric structure and the number and identity of halogen substitutions as significant structural

68

features of HBQs resulting in the observed differences in cytotoxicity, ROS generation, and

69

genotoxicity.13 The results of these in vitro studies suggest that HBQs are likely to be genotoxic

70

through oxidative stress-induced DNA damage, the extent of which can be influenced by HBQ

71

structure; however, the molecular basis for this is unclear.

72

Oxidative stress is one of the major contributors to the development of cancer.14-15 In

73

response to increased levels of ROS and oxidative stress, several cell signaling pathways are

74

involved to induce adaptive stress responses through modulation of stress–response genes.16 Of

75

these, the nuclear factor (erythroid-derived 2)-like 2 (Nrf2) signaling pathway is one of the most

ACS Paragon Plus Environment

4

Page 5 of 27

Environmental Science & Technology

76

important pathways to protect cells against oxidative stress.33-34 As a transcription factor, Nrf2

77

regulates the expression of a series of antioxidant genes through its interaction with the

78

antioxidant response element (ARE), which is located on the regulatory region of these genes.17

79

Animal studies have shown that susceptibility to chemical-induced urinary bladder cancer

80

significantly increased in Nrf2-deficient mice in comparison to wild-type mice.18 It has also been

81

reported that the Nrf2 pathway was activated in human cells and in rats by other DBPs, such as

82

HAAs.19-20 However, the effects of HBQ exposure on the expression of Nrf2 pathway-related

83

genes have not yet been studied. The participation of the Nrf2 signaling pathway and the

84

response of Nrf2-regulated genes to HBQ exposure remain unclear.

85

Microarrays are a useful tool to identify multiple pathways or monitor whole gene

86

expression simultaneously after exposure to environmental toxicants.21 In addition, TaqMan

87

probe-based real-time PCR arrays are useful to provide quantitative data on gene expression.22

88

The TaqMan PCR array method has been successfully demonstrated to study the potential

89

toxicity pathways of some DBPs, such as HAAs,19, 23-26 bromate,27-28 and 3-chloro-4-

90

(dichloromethyl)-5-hydroxy-5H-furan-2-one (MX).29 Therefore, we propose to use this method

91

to examine the impact of HBQs on gene expression.

92

The objective of this study was to demonstrate the impact of HBQs on gene expression,

93

particularly on the genes associated with oxidative stress signaling pathways. To achieve this, we

94

used a customized TaqMan Array to investigate the changes in expression of genes related to

95

oxidative-stress induced signaling pathways in human uroepithelial cells exposed to HBQs. A

96

panel of 44 functional genes were selected because they are associated with identified HBQ-

97

induced cellular effects. These genes cover nine functional gene groups, including transcription

98

factors, the GSH production/utilization pathway, the thioredoxin system pathway, stress

99

responsive genes, antioxidant genes, membrane transporters, cell cycle regulation, and oxidative

100

DNA damage repair. We selected an immortalized, non-tumorigenic human urinary tract

101

epithelial cell line, SV-HUC-1, for analysis, as it is a commonly used bladder tissue model.30 The

102

use of this cell line will provide more relevant information on gene expression comparison to

103

data obtained from human cancer cell lines. We will also compare the difference in gene

104

response between HBQ isomers to identify the molecular effects that underlie the different

105

biological responses observed in vitro.13 Thus, this study will provide insights into the molecular

106

basis of antioxidant defense against HBQ-induced oxidative stress in human uroepithelial cells.

ACS Paragon Plus Environment

5

Environmental Science & Technology

Page 6 of 27

107 108

Materials and Methods

109

Reagents. Standards of 2,6-dichloro-1,4-benzoquinone (2,6-DCBQ), 2,5-dichloro-1,4-

110

benzoquinone (2,5-DCBQ), and 2,5-dibromo-1,4-benzoquinone (2,5-DBBQ) were purchased

111

from Sigma-Aldrich (St. Louis, MO, USA). 2,6-dichloro-3-methyl-1,4-benzoquinone (DCMBQ)

112

and 2,3,6-trichloro-1,4-benzoquinone (TriCBQ) were purchased from Shanghai Acana

113

Pharmtech (Shanghai, China), while 2,6-dibromo-1,4-benzoquinone (2,6-DBBQ) was purchased

114

from Indofine Chemical Company (Hillsborough, NJ, USA). The chemical structure, molecular

115

weight, and purity of these HBQs are included in Table S1 of the Supporting Information. HBQ

116

stock solutions were prepared by dissolving each HBQ in 100% methanol (HPLC grade; Fisher

117

Scientific, Ottawa, ON, Canada) and were stored at -20°C in sterile amber glass vials.

118

Cell culture. The human uroepithelial cell line, SV-HUC-1, was obtained from American Type

119

Culture Collection (ATCC® CRL9520™; Manassas, VA, USA). The cells were maintained in

120

Kaighn's modification of Ham's F-12 (F-12K) medium (ATCC) containing 10% fetal bovine

121

serum (FBS; Sigma-Aldrich) and 1% penicillin/streptomycin (100 U/100 µg/ml; Invitrogen,

122

Burlington, ON, Canada) at 37°C in a 5% CO2 incubator.

123

HBQ-induced cytotoxicity and detection of reactive oxygen species (ROS). The cytotoxicity

124

of each HBQ was examined using real-time cell analysis (RTCA; ACEA Biosciences, San

125

Diego, CA, USA). ROS formation was determined using 2’, 7’-dichlorofluorescein diacetate

126

(DCFH-DA) (Sigma-Aldrich), a commonly used fluorogenic dye for the quantification of

127

intracellular ROS generation.31 The detailed procedures for these experiments are described in

128

the Methods section of the Supporting Information.

129

Real-time reverse transcription (RT)-PCR for time-course gene expression analysis. SV-

130

HUC-1 cells were seeded into 6-well plates (Corning Costar, Fisher Scientific) at a density of 1 ×

131

106 cells per well. After 24 h growth, cells were treated with each HBQ (2,6-DCBQ, 2,5-DCBQ,

132

2,6-DBBQ, 2,5-DBBQ, TriCBQ, DCMBQ) at concentrations of equivalent biological response

133

(IC20). At each indicated exposure time point (0.5, 1, 2, 4, 6, 8, and 24 h), cells were harvested

134

for RNA extraction. Total RNA was extracted using TRIzol (Invitrogen) according to the

135

manufacturer’s instructions. The resultant DNA-free RNA samples were quantified with a

136

NanoVue Plus™ spectrophotometer (Biochrom, Cambridge, UK) and stored at -80°C. RNA

137

integrity number (RIN) values were determined using an Agilent 2100 Bioanalyzer (Santa Clara,

ACS Paragon Plus Environment

6

Page 7 of 27

Environmental Science & Technology

138

CA, USA), and RNA purity was determined by the A260/A280 ratio. Only RNA with a ratio within

139

the range of 1.8-2.0 was reverse transcribed into cDNA using SuperScript™ III Reverse

140

Transcriptase (Invitrogen) with random primers (Invitrogen) according to the manufacturer’s

141

instructions. The collected cDNA samples were stored in sterile microcentrifuge tubes at -20°C.

142

The cDNA was used as a template for real-time quantitative PCR using the Fast SYBR®

143

Green PCR kit (Applied Biosystems, Carlsbad, CA, USA). The primers were synthesized by

144

Integrated DNA Technologies (Coralville, IA, USA), and the sequences are presented in Table

145

S2. Real-time fluorescence detection was performed using a StepOnePlus™ Real-Time PCR

146

system (Applied Biosystems). The detailed protocols for these experiments are described in the

147

Methods section of the Supporting Information.

148

Two representative genes, NFE2L2, which encodes the transcription factor Nrf2, and

149

NADPH:quinone oxidoreductase 1 (NQO1), a target gene of Nrf2, were used to select the most

150

informative exposure time points for the array experiments. Gene expression was determined

151

using cycle threshold (Ct) values. A comparative Ct method was used for relative quantification.

152

The Ct values were first normalized to the housekeeping gene glyceraldehyde 3-phosphate

153

dehydrogenase (GAPDH) present in the same samples (∆Ct) and were then expressed as the

154

fold-change over concurrent negative control groups (no HBQ treatment; culture medium)

155

prepared at each treatment time point (2-∆∆Ct). Duplicate samples were used in individual

156

experiments, and three independent experiments were performed (n = 3).

157

TaqMan Array for pathway gene expression analysis. After 2 h and 8 h HBQ exposure at

158

concentrations equal to their respective IC20 value (2,6-DCBQ, 2,5-DCBQ, 2,6-DBBQ, 2,5-

159

DBBQ, DCMBQ, TriCBQ), total RNA was isolated and reverse transcribed to cDNA as

160

described previously. A customized TaqMan Array in a 96-well fast plate was synthesized by the

161

manufacturer (Thermo Scientific, Waltham, MA) using the plate format shown in Figure S1.

162

Table S3 describes the 44 genes selected for the array, categorized into nine functional gene

163

groups, including transcription factors, the GSH production/utilization pathway, the thioredoxin

164

system pathway, stress responsive genes, antioxidant genes, membrane transporters, cell cycle

165

regulation, and oxidative DNA damage repair. Four reference control genes [18S ribosomal

166

RNA (18S rRNA), GAPDH, hypoxanthine phosphoribosyltransferase 1 (HPRT1), β-

167

glucuronidase (GUSB)] and 44 target genes were included in duplicate on each 96-well plate.

168

Real-time PCR analysis was conducted using a StepOnePlus™ Real-Time PCR system.

ACS Paragon Plus Environment

7

Environmental Science & Technology

Page 8 of 27

169

Additional details regarding the protocol for these experiments are described in the Methods

170

section of the Supporting Information. The comparative Ct method was used for relative

171

quantification. The Ct values were first normalized to the four reference control genes in the

172

same samples (∆Ct) and expressed as the fold-change over the concurrent negative control

173

groups (no HBQ treatment; culture medium) prepared at each treatment time point (2-∆∆Ct)..

174

Duplicate samples were analyzed in each replicate experiment. Two independent experiments of

175

the gene expression array were completed (n = 2).

176

Ingenuity pathway analysis (IPA). Briefly, the genes with significantly altered expression were

177

uploaded into the IPA software (www.ingenuity.com) for canonical pathway analysis, which was

178

based on the IPA library of canonical pathways. The significance of the association between our

179

gene list and a canonical pathway was measured by Fisher's exact test.

180

Data analysis. Data are expressed as the mean ± SEM or the mean ± SD. CI values were

181

normalized by the RTCA software (ACEA Biosciences) and were used to fit dose–response

182

curves to obtain the half inhibitory concentrations, IC50, and 20% inhibitory concentrations, IC20,

183

for each HBQ at specific time points. Student’s t-test was performed to determine the statistical

184

significance of the gene expression between negative control groups and HBQ treatment groups.

185

Fitting of dose-response curves and statistical analysis were conducted using GraphPad Prism

186

5.0 software (San Diego, CA). Associations were considered statistically significant at P < 0.05.

187 188

Results and Discussion

189

Optimization of experimental conditions for the TaqMan Array. To obtain reliable gene

190

expression profiles, we first optimized the conditions for several key experimental parameters,

191

including the treatment concentrations of the HBQs and the exposure time. An optimized

192

treatment concentration was identified as one that exhibits low cytotoxicity, but generates

193

sufficient amounts of intracellular ROS to elicit a measurable biological response, specifically

194

the induction of oxidative stress and the subsequent activation of the Nrf2 pathway. To select

195

low cytotoxic concentrations of each HBQ, we used RTCA to continuously monitor the response

196

of SV-HUC-1 cells to different concentrations of HBQs for 80 h. Figure S2 shows the

197

cytotoxicity response profiles for SV-HUC-1 cells exposed to each HBQ. The response profiles

198

clearly show that HBQ-induced cytotoxicity in SV-HUC-1 cells is concentration-dependent.

199

From the concentration–response data, we determined IC20 values, which are concentrations

ACS Paragon Plus Environment

8

Page 9 of 27

Environmental Science & Technology

200

eliciting a 20% decrease in normalized cell index values in comparison to the untreated control

201

cells (no HBQ treatment; culture medium). The IC20 values, therefore, exhibit low cytotoxicity

202

and allow for a comparison between HBQ-induced effects based on equivalent concentrations of

203

biological response. The 24-h IC20 values (mean ± SD) were determined to be 13 ± 2.3 µM (2,6-

204

DCBQ), 17 ± 1.1 µM (2,5-DCBQ), 10 ± 1.4 µM (2,6-DBBQ), 12 ± 0.9 µM (2,5-DBBQ), 17 ±

205

2.9 µM (DCMBQ), and 24 ± 3.3 µM (TriCBQ). When SV-HUC-1 cells were pre-treated with 1

206

mM N-acetyl-L-cysteine (NAC) prior to HBQ treatment, higher concentrations of HBQs were

207

needed to induce a full concentration–response (Figure S3) and resulted in 24, 48, and 72 h IC50

208

values that were significantly higher than the IC50 values of HBQs in cells without NAC

209

pretreatment (Figure S4). Because NAC can function as either a source of thiol in cells or as an

210

antioxidant to scavenge ROS directly,32 these results demonstrate HBQ-induced oxidative stress

211

in SV-HUC-1 cells, which is consistent with results from T24 human bladder carcinoma cells.10

212

To confirm that a treatment concentration equal to the IC20 value for each HBQ can

213

generate sufficient ROS to elicit a measurable biological response in the TaqMan Array, we

214

measured intracellular ROS levels after HBQ treatment from 0.5 h to 72 h. Figure S5

215

summarizes the time-course of cellular ROS generation after each HBQ treatment at their

216

respective IC20 values. All six HBQs share a similar temporal trend of ROS production: ROS

217

production increases from 0.5h, peaks at 8h, and then slowly decreases over the remainder of the

218

exposure period. For all time points examined, ROS generation by each HBQ was statistically

219

greater than the concurrent negative control group (one-way ANOVA, P < 0.05). This is

220

consistent with a study of the quinone compound menadione, where ROS production increased

221

over time until peaking at 8 h after treatment in HepG2 cells.33 Based on our RTCA and ROS

222

results, the IC20 values are an optimal testing concentration for the TaqMan Array.

223

To select the optimal exposure time points to represent both early and late responses

224

within the Nrf2 pathway, we examined the time-dependent effects of HBQ treatment on the

225

expression of two representative genes. The two genes examined were NFE2L2, representing the

226

transcription factor Nrf2, and NQO1, representing a target gene of Nrf2. RNA integrity was

227

assessed to ensure its high quality prior to RT-PCR. Therefore, only RNA with high RIN values

228

were used (Figure S6; RIN > 9, Figure S7). Exposure to 24-h IC20 concentrations of each HBQ

229

resulted in an initial increase in NFE2L2 expression during the first 2 h of exposure, with a

230

subsequent decrease in expression after 4 h exposure (Figure S8A). Because 2 h exposure to each

ACS Paragon Plus Environment

9

Environmental Science & Technology

Page 10 of 27

231

HBQ significantly increased NFE2L2 expression (P < 0.05), it was chosen as the early response

232

time point for our array experiments. Each HBQ also increased the Nrf2 downstream target gene,

233

NQO1, in a time-dependent manner (Figure S8B). The expression of NQO1 increased gradually

234

from 2 h, peaked at 8 h, and subsequently decreased at 24 h, guiding our selection of 8 h as the

235

late response time for our array experiments. Thus, in the following experiments, we analyzed

236

the gene expression changes in SV-HUC-1 cells under the optimized conditions of 2 h and 8 h

237

HBQ treatment at concentrations equal to their respective 24-h IC20 values.

238

Modulation of oxidative-stress induced pathways by HBQs. The 44 genes selected for our

239

customized TaqMan Array focus on the Nrf2 signaling pathway due to its significant impact on

240

the ability of cells to protect against oxidative stress, the main mechanism of HBQ-induced

241

cytotoxicity. Additional genes were also selected that underlie known HBQ-induced effects,

242

including oxidative DNA damage10, 12 and p53 upregulation13, 34. Figure 1 summarizes the

243

relationships between the selected genes and known HBQ-induced effects alongside key

244

oxidative stress-induced pathways. Hence, the modulation of these genes after HBQ exposure

245

will reveal the molecular basis of HBQ-induced responses observed in vitro. Under the

246

optimized experimental conditions, each HBQ altered the transcription of multiple genes at the

247

early and late exposure time points, including significant alterations in genes from all nine

248

functional gene groups. Table S4 lists the genes that exhibited significant changes in expression

249

(P < 0.05) after HBQ exposure. After 2 h exposure, each HBQ altered expression levels in 9 to

250

28 genes, while after 8 h exposure, each HBQ altered expression levels in 29 to 31 genes. As

251

predicted, all six HBQs upregulated NFE2L2, the gene which encodes Nrf2, indicating that

252

HBQs are able to activate the Nrf2 signaling pathway. This has also been shown previously in an

253

ARE-dependent β-lactamase-based reporter gene assay, which indicated that both 2,6-DCBQ and

254

2,6-DBBQ activated Nrf2 with a similar level of potency.34 Interestingly, in this study, we

255

observed that the modulation of NFE2L2 was time-dependent. While 2,5-DCBQ and 2,5-DBBQ

256

upregulated NFE2L2 at the late exposure time point, the remaining four HBQs exhibited

257

upregulation at the early exposure time point. The ability of compounds to activate the Nrf2

258

pathway has been shown to correlate with their reactivity toward thiol groups of Kelch-like

259

ECH-associated protein-1 (Keap1) cysteine residues, which inhibit the ubiquitination of Nrf2.

260

This indicates that the structure of HBQs may influence their ability to activate Nrf2, which is

261

supported by evidence from tert-butylhydroquinone (tBHQ) and 1,2-naphthoquinone.35

ACS Paragon Plus Environment

10

Page 11 of 27

Environmental Science & Technology

262

Furthermore, because HBQs readily bind to thiols in GSH with differing affinity,36 it is likely

263

they could exhibit differing reactivity toward the thiol groups of Keap1, affecting their ability to

264

activate Nrf2.

265

The main targeted genes of the Nrf2 signaling pathway are antioxidant genes and stress

266

responsive genes. While previous studies have indicated activation of Nrf2 by HBQs, no

267

evidence has been presented that demonstrates downstream activation of genes regulated by Nrf2.

268

The selected genes in these functional groups may indicate how SV-HUC-1 cells detoxify HBQs

269

and/or its induced ROS, or mount an inflammatory response. NQO1 is a detoxification enzyme

270

specific to quinones, catalyzing a two-electron reduction of quinones to their less toxic

271

hydroquinone form, a major detoxification pathway of quinones.37 As expected, NQO1 was

272

upregulated in SV-HUC-1 cells exposed to any of the six HBQs. Alternatively, N-

273

acetyltransferase 1 (NAT1) and NAT2 encode N-acetyl-transferases that promote xenobiotic

274

metabolism via transfer of an acetyl group.38 No consistent pattern in expression of NAT1 or

275

NAT2 was observed across the six HBQs, but overall, both genes were downregulated, indicating

276

that this pathway is not likely a defense against HBQs as is the activation of NQO1.

277

HBQs are potent inducers of intracellular ROS.9 Superoxide dismutase (SOD) catalyzes

278

the conversion of superoxide to hydrogen peroxide (H2O2) or O2. At 8 h exposure, SOD1 and

279

SOD2 were significantly upregulated by 2,5-DCBQ and 2,5-DBBQ, but downregulated by 2,6-

280

DBBQ and TriCBQ. Thus, it is possible that the 2,5-HBQs induce higher amounts of superoxide

281

than the other HBQs. Also, the downregulation of SOD could possibly be in response to the

282

upregulation of other superoxide antioxidants in 2,6-DBBQ- or TriCBQ-treated cells, rendering

283

SOD obsolete. The clearest evidence for the production of a specific type of ROS by all six

284

HBQs relates to the expression of antioxidant genes related to the detoxification of H2O2. The

285

genes regulating catalase (CAT), cytoglobin (CYGB), glutathione peroxidase (GPX2), and

286

prostaglandin-endoperoxide synthase 2 (PTGS2) were all significantly upregulated by each HBQ

287

at 8 h exposure, suggesting these four genes may coordinate to eliminate H2O2. The genes

288

associated with the heme peroxidase superfamily, lactoperoxidase (LPO) and myeloperoxidase

289

(MPO), were not modulated by HBQ exposure.

290

The oxidant/antioxidant balance within cells is tightly controlled, with oxidants playing a

291

key role in cellular processes such as metabolism and cell signaling. The induction of oxidative

292

stress can affect the modulation of oxidants in addition to the more commonly assayed

ACS Paragon Plus Environment

11

Environmental Science & Technology

Page 12 of 27

293

antioxidants. NADPH oxidase (NOX) proteins are the main non-mitochondrial source of ROS

294

within cells, generating superoxide.39 Both NOX5, encoding NOX5, and cytochrome b-245 beta

295

chain (CYBB), encoding NOX2, were modulated by HBQ exposure. Hence, it is possible that

296

superoxide radicals produced by HBQ-induced upregulation of CYBB could result in the

297

concomitant downregulation of NOX5, balancing the production and removal of superoxide

298

radicals after exposure to low doses of HBQs. It is clear, however, that other genes are involved

299

to maintain this balance, as the four HBQs most strongly upregulating CYBB, 2,5-DCBQ (17-

300

fold increase), 2,6-DBBQ (14-fold increase), 2,5-DBBQ (20-fold increase), and TriCBQ (14-fold

301

increase), also exhibited differing modulation of the SOD genes (SOD1, SOD2). A more

302

comprehensive screen of genes associated with the oxidative balance of superoxide is necessary

303

to understand these findings.

304

Inflammation is a complex cellular response that is closely linked to xenobiotic-induced

305

oxidative stress.40 Hemeoxygenase-1 (HMOX1) encodes the hemeoxygenase-1 enzyme (HO-1),

306

which is able to cleave heme to produce carbon monoxide (CO), biliverdin, and Fe(II).41-42

307

Among the 44 genes screened in our panel, HMOX1 exhibited the highest upregulation, with

308

expression levels 7- to 50-fold greater after 8 h HBQ exposure and 5- to 10-fold higher after 2 h

309

HBQ exposure. The significant increase of HMOX1 expression suggests that iron regulation may

310

play a role in the detoxication of HBQs by controlling intracellular levels of superoxide with

311

SOD. In addition, the protective role of HMOX1 against oxidative stress is believed to be

312

through its product CO, which is an important signaling molecule in anti-inflammatory

313

pathways.43-44 The PTGS2 gene is also involved in the anti-inflammatory process, although it is

314

not modulated by Nrf2. PTGS2 encodes cyclooxygenase-2 (COX-2), which catalyzes the

315

conversion of arachidonic acid to prostaglandin H2 (PGH2) and further to prostaglandins (PGs),

316

a primary component of the anti-inflammatory process.45 The upregulation of both HMOX1 and

317

PTGS2 suggests a possible pro-inflammatory response in SV-HUC-1 cells after HBQ exposure.

318

This is significant, as inflammation has been linked to the development of every stage of cancer40

319

and to the development of different types of cancer, including bladder cancer.46 Furthermore, the

320

induction of COX-2 itself has been linked to the development of bladder cancer,47-48 which is the

321

target organ of DBP exposure identified in several epidemiological studies.

322

The most prominent thiol-dependent antioxidant systems present in cells are the GSH and

323

thioredoxin systems,49 both of which contain targeted genes in the Nrf2 signaling pathway. After

ACS Paragon Plus Environment

12

Page 13 of 27

Environmental Science & Technology

324

8 h exposure, all six HBQs induced a two-fold upregulation of glutamate–cysteine ligase

325

modifier subunit (GCLM), glutamate–cysteine ligase catalytic subunit (GCLC), and glutathione

326

reductase (GSR). GCLM and GCLC are composed of glutamate cysteine ligase (GCL), which is

327

the rate-limiting enzyme for GSH synthesis.50-51 GSR is responsible for the reduction of oxidized

328

GSH to reduced GSH.51 Thus, it is clear from the upregulation of these genes that Nrf2 increased

329

the synthesis of GSH in response to HBQ-induced GSH depletion, as the ability of HBQs to

330

deplete and/or conjugate GSH has been demonstrated in HBQ-treated cells.11, 36 The conjugation

331

of GSH to xenobiotics is catalyzed by glutathione transferase (GST). In our study, GSTT1,

332

GSTZ1, and GSTP1 were mostly downregulated, although GSTP1 was slightly upregulated in

333

four HBQs after 2 h exposure. Because HBQ-GSH conjugation occurs rapidly in vitro, the

334

overall downregulation of GST isoforms implies that HBQ-GSH conjugation in cells may be

335

mostly non-enzymatic.

336

Although HBQ-GSH interactions have been well-characterized, the ability of HBQs to

337

react with other thiol groups in cells is unclear. The thioredoxin system is composed of

338

thioredoxin (TXNRD), peroxiredoxin (PRDX), and sulfiredoxin (SRXN). After 2 h HBQ

339

exposure, the gene expression levels of thioredoxin reductatse 1 (TXNRD1) increased 2- to 3-fold,

340

and increased 2- to 5-fold after 8 h exposure to each HBQ. TXNRDs are a class of enzymes that

341

reduce oxidized thioredoxin to reduced thioredox.52 Because the altered transcription of TXNRD1

342

is affected by all six HBQs at both exposure time points, thioredoxin and TXNRD1 may be

343

sensitive targets for HBQs in uroepithelial cells, providing the first evidence that demonstrates

344

that HBQs can affect the thioredoxin system. SRXN1 was also upregulated at 2 h (2,6-DBBQ,

345

TriCBQ and DCMBQ) and 8 h exposure (2,6-DCBQ, 2,5-DCBQ, 2,5-DBBQ, and DCMBQ).

346

Benzoquinone-induced induction of SRXN1 was also observed in rodent and human cells

347

exposed to tBHQ.53, 54 Because SRXN proteins re-activate PRDX proteins,55 an upregulation in

348

PRDX genes would be expected. However, we found that most of the PRDX genes were

349

unaffected or downregulated after HBQ exposure.

350

The multi-drug resistance proteins (MRPs) have been shown to protect cells from HBQ-

351

induced cytotoxicity and oxidative stress in cell models overexpressing MRP.56 The genes that

352

encode the MRPs, the ABCC genes, are target genes of Nrf2. After 8 h HBQ exposure, ABCC2

353

was significantly upregulated by five of the HBQs (not DCMBQ), indicating that Nrf2 may

354

control the induction of ABCC2 (MRP2) to increase efflux of HBQs or HBQ-GSH conjugates.

ACS Paragon Plus Environment

13

Environmental Science & Technology

Page 14 of 27

355

However, from our results, it is clear that the expression of ABCC genes varied widely amongst

356

the different HBQs, indicating that individual HBQs may affect MRP expression differently.

357

This could have important toxicodynamic implications for HBQs, as the MRPs are known to

358

affect xenobiotic efflux and distribution.57

359

Two genes included in our 44-gene panel that are not regulated by Nrf2, but are

360

associated with HBQ-induced effects, are 8-oxoguanine glycosylase (OGG1) and cyclin-

361

dependent kinase inhibitor 1 (CDKN1A). In response to DNA damage, cells can directly repair

362

the damaged DNA or stop proliferation through cell cycle arrest. Interestingly, the DNA damage

363

repair gene, OGG1, was downregulated by all six tested HBQs at 8 h exposure. HBQs have been

364

shown to increase levels of 8-hydroxydeoxyguanosine (8-OHdG),10, 13 a key biomarker of

365

oxidative DNA damage,58 which is primarily repaired through the DNA base excision repair

366

(BER) pathway by 8-oxoguanine DNA glycosylase, encoded by OGG1.59 Therefore, HBQs

367

could induce oxidative DNA damage (8-OHdG), while also decreasing DNA repair capacity.

368

This phenomena has also been observed with estrogen, with estrogen-mediated oxidative stress

369

suggested to play an important role in estrogen-induced breast carcinogenesis.60 Because the

370

decreased expression of OGG1 has been shown to be associated with tumor development,61 these

371

findings suggest the importance of examining the carcinogenic potential of HBQs.

372

CDKN1A, regulated by p53, encodes the cyclin-dependent kinase inhibitor 1 (p21)

373

protein, which regulates progression of the cell cycle at G1 and S phases.62 CDKN1A was

374

significantly upregulated by all six HBQs at 2 h exposure, and by most of the HBQs at 8 h

375

exposure, indicating that SV-HUC-1 cells may activate the p53 pathway to promote p21-

376

regulated cell cycle arrest to protect against HBQ-induced DNA damage. This is consistent with

377

several recent reports. A significant increase in p53 protein expression was detected in CHO

378

cells treated with HBQs (2,5-DBBQ, 2,6-DCBQ, and 2,5-DCBQ) after 24 h exposure,13 while

379

activation of the p53 pathway after 2,6-DCBQ and 2,6-DBBQ exposure in Caco-2 cells was also

380

detected using a β-lactamase reporter system.34 Furthermore, cell cycle arrest was induced by

381

2,6-DCBQ and 2,6-DBBQ exposure in human neural stem cells.63 It is also possible that p21 can

382

activate the Nrf2 signaling pathway to protect cells from oxidative stress by releasing Nrf2 from

383

its inhibitor Keap1.64-65 Taken together, p53 may be activated by HBQs to induce p21 to regulate

384

cell cycle arrest, while p21 may also simultaneously activate the Nrf2 pathway to protect against

385

HBQ-induced oxidative stress.

ACS Paragon Plus Environment

14

Page 15 of 27

Environmental Science & Technology

386

Differences of transcriptome changes between HBQ isomers. 2,5-HBQs have been shown to

387

induce higher cytotoxicity and greater ROS production than their corresponding 2,6-HBQ

388

isomers.13 Although this difference has been associated with HBQ structure, as dipole moment is

389

correlated with HBQ isomer reactivity and resulting toxicity, determining the underlying

390

differences in biological activity between isomer pairs may help to further elucidate the

391

differences in cellular response that have been observed. Figure 2 shows the difference in gene

392

expression between the HBQ isomer pairs in SV-HUC-1 cells. Interestingly, 2,6-DCBQ and 2,6-

393

DBBQ both upregulate NFE2L2, while 2,5-DCBQ and 2,5-DBBQ both downregulate Keap1 at

394

the 2 h exposure time point. This indicates that at the earlier exposure time, both the 2,6- and 2,5-

395

HBQ isomers are able to activate the Nrf2 signaling pathway; however, it appears that this

396

activation is achieved in different ways. Tetrachlorobenzoquinone has also been observed to

397

activate NFE2L2, but not Keap1, in HepG2 cells.66 Another interesting finding among the HBQ

398

isomers is that the 2,6-HBQs significantly increased the expression levels of GPX2 and CAT at 2

399

h, while the 2,5-HBQs did not. Because GPX2 and CAT encode proteins responsible for reducing

400

H2O2, the greater expression of these two genes could decrease the cellular ROS induced by 2,6-

401

HBQs. This could explain the greater toxicity induced by 2,5-HBQs in comparison to their 2,6-

402

HBQ isomers. The third and most important finding was that 2,5-DCBQ and 2,5-DBBQ induced

403

significantly higher expression levels of nine genes (2 h: HMOX1; 8 h: GCLC, GSR, TXNRD1,

404

CYBB, CYGB, OGG1, HMOX1, SOD1, SOD2) than their corresponding 2,6-HBQ isomers (P