Thyroid Disruption by Bisphenol S Analogues via Thyroid Hormone

Subscriber access provided by - Access paid by the | UCSB Libraries

Ecotoxicology and Human Environmental Health

Thyroid Disruption by Bisphenol S Analogues via Thyroid Hormone Receptor #: in Vitro, in Vivo and Molecular Dynamics Simulation Study Liping Lu, Tingjie Zhan, Mei Ma, Chao Xu, Jingpeng Wang, Chunlong Zhang, Weiping Liu, and Shulin Zhuang Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b00776 • Publication Date (Web): 15 May 2018 Downloaded from http://pubs.acs.org on May 15, 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 44

Environmental Science & Technology

1

Thyroid Disruption by Bisphenol S Analogues via Thyroid

2

Hormone Receptor β: in Vitro, in Vivo and Molecular Dynamics

3

Simulation Study

4 5

Liping Lu†, Tingjie Zhan†, Mei Ma‡,||, Chao Xu§, Jingpeng Wang†, Chunlong Zhang⊥,

6

Weiping Liu†, Shulin Zhuang†,*

7 8 9

10 11

12 13

14 15

16 17

†

College of Environmental and Resource Sciences, Zhejiang University, Hangzhou

310058, China. ‡

Key Laboratory of Drinking Water Science and Technology, Research Center for

Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China. ||

College of Resources and Environment, University of Chinese Academy of Sciences,

Beijing 100085, China. §

College of Environment, Zhejiang University of Technology, Hangzhou 310032,

China. ⊥

Department of Biological and Environmental Sciences, University of Houston-Clear

Lake, 2700 Bay Area Blvd., Houston, Texas 77058, USA.

18

19

* Address correspondence to E-mail: [email protected] (S. Zhuang).

20

1

ACS Paragon Plus Environment


21 22

ABSTRACT Bisphenol S (4-hydroxyphenyl sulfone, BPS) is increasingly used as BPA

23

alternatives. The global usage of BPS and its analogues (BPSs) caused the frequent

24

detection of their residues in multiple environmental media. We investigated their

25

potential endocrine disrupting effects toward thyroid hormone receptor (TR) β. The

26

molecular interaction of BPSs toward TRβ ligand binding domain (LBD) was probed

27

by fluorescence spectroscopy and molecular dynamics (MD) simulations. BPSs

28

caused the static fluorescence quenching of TRβ LBD. The 100 ns MD simulations

29

revealed that the binding of BPSs caused significant changes of the distance between

30

residue His435 at helix 11(H11) and residue Phe459 at H12 in comparison with no

31

ligand-bound TRβ LBD, indicating relative repositioning of H12. The recombinant

32

two-hybrid yeast assay showed that tetrabromobisphenol S (TBBPS) and

33

tetrabromobisphenol A (TBBPA) have potent antagonistic activity toward TRβ with

34

IC10 of 10.1 nM, 21.1 nM, respectively. BPS and BPA have the antagonistic activity

35

with IC10 of 312 nM, 884 nM, respectively. BPSs significantly altered the expression

36

level of mRNA of TRβ gene in zebrafish embryos. BPS and TBBPS at

37

environmentally relevant concentrations have antagonistic activity toward TRβ,

38

implying that BPSs are not safe BPA alternatives in many of the BPA-free products.

39

Future health risk assessment of TR disruption and other adverse effects should focus

40

more on the structure-activity relationship in designing environmentally benign BPA

41

alternatives.

2


Page 2 of 44

Page 3 of 44


42 43

Keywords: Bisphenol S; Molecular Interaction; Fluorescence Spectroscopy; Molecular modeling; Endocrine disruption

44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63

3



64 65

1. INTRODUCTION Bisphenol S (4-hydroxyphenyl sulfone, BPS) is one of the commercial

66

alternatives for bisphenol A (BPA) due to its enhanced heat and light resistance. BPS

67

and its analogues (BPSs) such as tetrabromobisphenol S (TBBPS) and TBBPS bis

68

(2,3- dibromopropyl ether) (OBBPS, TBBPS-BDBPE) are used commonly in epoxy

69

resins, thermal receipt papers, foodstuff containers, electronic devices, baby bottles

70

and medical appliance1-3. The increasing global usage of BPSs resulted in frequently

71

detected residues of BPSs in foodstuff, sewage sludge, sediment, surface water, and

72

indoor dust4-7. The relatively high lipophilicity of BPSs also contributes to their

73

bioaccumulation in different fish and wildlife species, mollusks and avian eggs

74

through various food chains and trophic levels3, 8-10. BPSs were detected also in

75

human urine, breast milk, blood and cord serum via oral and other exposure pathways,

76

especially for individuals living near e-waste facilities11-15. BPSs showed a detection

77

frequency of 81% and a mean concentration of 0.65 µg/L in urine samples collected

78

from the United States and Asian countries16.

79

The exposure to BPSs has been reported to cause multiple adverse effect to

80

human health and animals, such as cytotoxicity, genotoxicity, immunotoxicity and

81

teratogenic effects17-21. TBBPA and TBBPS stimulated neural differentiation of

82

mouse embryonic stem cell via different disruption of the positive regulator22. BPSs

83

can cause endocrine disruption such as estrogenic, androgenic, anti-androgenic

84

disruption1, 23, 24. BPS was shown to significantly increase progestagens levels through

85

steroidogenic pathway25. Structural differences in the bridging moiety and benzene

4


Page 4 of 44

Page 5 of 44


86

ring of bisphenol analogues lead to distinct estrogenic potencies in cell proliferation

87

and MVLN (MCF-7-p-Vit-tk-Luc-Neo) cell-based assays26, which may relate to their

88

preferential binding toward estrogen receptor subtype20. BPS was revealed to disrupt

89

the membrane-initiated estradiol-induced cell signaling18 and caused adipocyte

90

differentiation more directly than BPA via the activation of peroxisome

91

proliferator-activated receptor gamma (PPARγ)27.

92

Thyroid hormones (TH) play vital role in metabolism, energy expenditure,

93

growth and development in vertebrates and TRs represent potential target of various

94

endocrine disrupting chemicals (EDCs)28, 29. TRβ is the major TR isoform in the

95

thyroid responsible for the regulation of target gene transcription30. Exogenous

96

ligands may induce conformational changes of TRβ ligand binding domain (LBD),

97

thereby affecting transcriptional activity of TRβ31. Increasing reports focus on the

98

adverse health effect of thyroid hormone disrupting chemicals29, 32, 33. BPA and it

99

analogues could regulate TRβ gene and key genes related to

100

hypothalamic-pituitary-thyroid (HPT) axis, thus influencing thyroid hormone

101

homeostasis and inducing thyroid disruption34-36. They may also bind to transthyretin

102

(TTR) or TRs37 and interfere with transcriptional regulation of target genes38,

103

disrupting relevant signaling pathway39. The investigation on how BPSs affect

104

conformations of TRβ LBD at molecular level and whether structural analogues of

105

BPSs bear different TR disruption is of significance for the evaluation of TRβ

106

disruption by BPSs.

5



107

In the present study, we evaluated the potential risk of BPS analogues toward

108

TRβ (Table 1). We purified the TRβ LBD protein and performed in vitro fluorescence

109

spectroscopy together with molecular dynamics (MD) simulations to decipher the

110

molecular mechanism of TR disruption induced by BPSs. The recombinant human

111

TRβ two-hybrid yeast assay combined with the zebrafish assay was used to evaluate

112

the TRβ disruption of BPSs. Our results provided the first in vitro, in vivo and in

113

silico evidence on the binding between TRβ LBD and BPSs. Such information is of

114

significance for understanding physiological effects and endocrine activities of the

115

BPSs in the consumer products and facilitates future design of environmentally

116

benign BPA and BPS substitutes.

117

2. MATERIALS AND METHODS

118

2.1 Materials.

119

BPSs (Table 1), dimethyl sulfoxide (DMSO) and 4-hydroxytamoxifen (4-OHT)

120

with a purity ≥ 98.0% were purchased from Sigma Aldrich (St. Louis, MO, USA).

121

3,3',5-Triiodo-l-thyronine (T3) with a purity of 98% was purchased from J&K

122

Chemical Ltd. (Shanghai, China). SD/-Leu/-Trp medium (Catalogue: 4823-6) was

123

purchased from Mobitec company (Goettingen, Germany). Power-SF DNA

124

polymerase (Catalogue: HG1001) and In-Fusion Enzyme (Catalogue: HG3001) were

125

supplied by HAOGENE Biotech Co., Ltd (Hangzhou, China). The Restriction enzyme

126

FastDigest EcoRI and FastDigest Xho I were purchased from Fermentas China Co.,

127

Ltd (Shanghai, China). Other chemicals were of analytical grade. All chemicals were

128

dissolved in DMSO and diluted using distilled water (18.2 MΩ, Millipore, Bedford, 6


Page 6 of 44

Page 7 of 44


129

MA) for toxicity evaluation and also were diluted in Tris-HCl buffer (0.2 M Tris, 0.1

130

M NaCl, pH 7.4) for spectroscopic measurement. The prepared solutions were stored

131

in amber glass vials at 4 oC.

132

2.2 Expression of TRβ LBD.

133

The coding DNA sequence of a full length Homo sapiens thyroid hormone

134

receptor β (TRβ) (Genebank ID: GI: 358001055) with 1,386 base pairs was amplified

135

by polymerase chain reaction (PCR) method with Power-SF DNA polymerase and

136

specified primers (Table S1) and the DNA fragment was validated by agarose gel

137

electrophoresis (Figure S1). The fragment of TRβ LBD was further amplified based

138

on the full length TRβ. The vector pET-32a (Hangzhou Newbay Biological

139

Technology Co., Ltd., Hangzhou, China) was digested with the enzyme FastDigest

140

EcoRI and FastDigest XhoI. The DNA fragment of TRβ LBD was recominanted into

141

the digested pET-32a using In-Fusion Enzyme. The recominant plasmid containing

142

TRβ LBD was further amplified by Colony PCR using T7 promoter primer and T7

143

terminator. TRβ LBD was expressed in E. coli BL21(DE3) and purified by

144

Ni2+-affinity chromatography (Catalogue: 30210, Qiagen, Germany) and further

145

eluted in 50 mM Tris HCl buffer (pH 8.0).

146

2.3 Fluorescence Spectroscopy.

147

The steady-state fluorescence spectra were measured with FluoroMax®-4

148

spectrofluorometer (Horiba Jobin Yvon IBH) in a 1-mm quartz cell following a

149

reported protocol40, 41. The emission spectra of TRβ LBD were recorded from 320 nm

150

to 400 nm with an excitation wavelength of 280 nm at 303 K and 310 K. The

7



151

fluorescence of a buffer solution was recorded to correct the background noise and

152

eliminate the inner-filter effect. Each assay was performed in triplicate.

153

2.4 Molecular Dynamic Simulations.

154

The tertiary structure of TRβ LBD in complex with BPSs was constructed on the

155

basis of X-ray crystal structure of TRβ LBD (PDB ID: 2J4A, 2.2 Å) by molecular

156

docking using MVD 4.2 program (Text S2). The conventional MD simulations were

157

further performed by Sander module implemented in Amber 14 with Amber ff12

158

force field42. The atomic partial charges of BPSs were derived by R.E.D. Server

159

Development with restrained electrostatic potentials (RESP) method and were used to

160

build the parameters and libraries compatible with Amber ff12 force field. The

161

complexes were solvated by 10 Å TIP3P waters in a periodic rectangular cubic box

162

and were neutralized by 10 Na+ by AmberTools 1.542. The generated complex

163

systems containing 40,000 atoms were minimized for 30 ps and then heated from 0 K

164

to 300 K for 50 ps. The equilibration run of MD simulations were carried out using

165

isothermal-isobaric (NPT) ensemble with unconstrained MD simulations for 100 ns

166

with a time interval of 2 fs following a reported protocol43, 44. The MD trajectories

167

were saved every 10 ps. The particle mesh Ewald (PME) method with a nonbonded

168

cutoff of 10 Å was used to calculate the long range electrostatic interactions. The

169

hydrogen atoms of the complex system were constrained by the SHAKE algorithm.

170

The cpptraj module implemented in AmberTools 1.5 was used for conformational

171

changes and the molecular mechanics generalized Born/surface area (MMGB/SA)

172

with a single trajectory method was used to evaluate the binding free energies.

8


Page 8 of 44

Page 9 of 44


173 174

2.5 Recombinant Two-hybrid Yeast Assay. The reported recombinant human TRβ two-hybrid yeast assay, which is highly

175

speciﬁc to TRβ ligand without cross-talk to other receptors was used to screen the

176

potential TR disruption of BPSs45, 46. The yeast cells were cultured in SD/-Leu/-Trp

177

medium at 30oC overnight. BPSs solutions (5 µL) at varying concentrations were

178

added to 995 µL medium with the OD600 value of approximately 0.75. For

179

antagonistic activity of BPSs, a series of 5 µL BPSs (0.005 nM to 50 µM) and 5 µL

180

T3 (10-4 M) were co-incubated with 990 µL yeast cells. This test culture (200 µL) was

181

further cultured for 2.5 hr at 30oC. After the lysis of the cultured yeast cells using

182

chloroform, the enzyme reaction was initiated by adding 40 µL

183

o-nitrophenyl-β-D-galactopyranoside (13.3 mM). This reaction continued for 60 min

184

and was terminated by adding 100 µL sodium carbonate (1.0 M). The OD420 was

185

measured by Infinite 200 PRO NanoQuant Multimode Reader (Tecan Group Ltd.,

186

Switzerland) and the β-galactosidase activity (U) was calculated following a reported

187

protocol47-49. All assays were repeated three times and triplicate samples were

188

measured in each assay.

189

2.6 Determination of BPSs Concentration in ZebraFish Exposure Water.

190

To measure actual exposure concentrations of BPSs, water samples were

191

collected and determined in triplicate following the reported protocol20 with some

192

modifications (Text S4). The solid-phase extraction (SPE) was performed after the

193

conditioning of an Oasis HLB cartridge (6 mL, 200 mg, Waters, Massachusetts, USA)

194

with 5 mL methanol and ultrapure water, respectively. Water samples of 100 mL

9



195

containing BPSs analogues and 0.2% DMSO were extracted through the Oasis HLB.

196

The compounds were eluted with 20 mL methanol after washed with 5 mL ultrapure

197

water. Extracts were dried with a stream of nitrogen and was then solvated in 1 mL

198

methanol. The water samples were injected in ultra-high-performance liquid

199

chromatography tandem mass spectrometry (UHPLC-MS/MS) (Acquity UPLC,

200

Waters, USA) after passed through 0.22 centrifuge filters. The detection limits were

201

determined by 3/1 signal-to-noise ratio (S/N) and the method detection limits (MDLs)

202

were determined on concentration of detection limits in water sample (n=3). A five

203

point calibration curve was established between the peak area of compounds and their

204

concentrations. The measured water concentrations and qualitative and quantification

205

of BPSs were provided (Table S7-S9).

206

2.7 Transcription of TRβ Gene in ZebraFish.

207

The transcription profile of TRβ gene was evaluated using zebrafish (Danio rerio)

208

by the quantitative real-time polymerase chain reactions (qPCR). The adult zebrafish

209

were maintained following a recently reported protocol35. After spawning, the live

210

embryos were collected and exposed to BPSs at 0.01, 0.1, and 1.0 µM for 72 h. The

211

total RNA was then extracted from the zebrafish larvae using Trizol reagent (Takara

212

Biochemicals, Dalian, China). The reverse-transcription of cDNA of TRβ was

213

performed by the reverse transcriptase kit (Takara Biotechnology Co. Ltd., Dalian,

214

China) using 500 ng total RNA. The cDNA (GenBank ID: GI: 358001055) was

215

amplified in a 10-µL SYBR reaction mixture by the Mastercycler® ep realplex

216

(Eppendorf, Hamburg, Germany) using designed primers (Table S1). Solvent control

10


Page 10 of 44

Page 11 of 44


217

was constructed by equivalent DMSO concentration in ultrapure water. Triplicate

218

samples were analyzed in each assay.

219

2.8 Statistical Analysis.

220

The data obtained from recombinant two-hybrid yeast assay was fitted by

221

Levenberg-Marquardt algorithm with IBM SPSS Statistics 20. Data were analyzed by

222

one-way analysis of variance (ANOVA), followed by LSD’s multiple comparisons. A

223

p value of < 0.05 was considered as significant difference.

224

3. RESULTS AND DISCUSSION

225

3.1 BPSs Altered Conformational Changes of TRβ LBD.

226

The binding of cognate ligands to TRβ LBD enables TRβ to activate

227

transcription of target genes50, 51. The disturbed conformations of TRβ LBD by the

228

binding of contaminants may consequently affect transcriptional activity of TRβ,

229

leading to TR disruption. Studies on how BPSs with different structural moieties bind

230

to TRβ LBD and induce conformational changes are therefore essential for the

231

evaluation of TR disruption. We expressed TRβ LBD and probed the potentially

232

induced conformational changes of TRβ LBD upon binding of BPSs by steady-state

233

fluorescence spectroscopy. The fluorescence spectroscopic data on TRβ LBD, the first

234

in vitro evidence reported to date, was further supported by MD simulations below to

235

elucidate the intrinsic changes of TRβ LBD at the atomic level.

236

3.1.1 BPSs Caused Fluorescence Quenching of TRβ LBD

237 238

Fluorescence spectroscopy is widely used to study the interaction of various xenobiotic contaminants with biomacromolecules. The intrinsic tryptophan

11



239

fluorescence may change due to the microenvironmental disruption of proteins upond

240

the binding of contaminants52, 53. Our in vitro spectroscopic spectra revealed that the

241

ligand free form of TRβ LBD (apo TRβ LBD) has the maximum fluorescence

242

emission at 350 nm (Figure 1). This maximum emission wavelength of TRβ LBD

243

shifted from 350 nm to 351 nm, 352 nm, 352 nm, 357 nm, 355 nm and 352 nm after

244

titrating with BPA, BPS, BPS-DAE, TBBPA, OBBPS and TBBPS, respectively,

245

indicating a slight red-shift (Figure 1). This red-shift of maximum emission with λmax

246

to longer wavelengths was caused by the affected positively charged residues upon

247

the binding, suggesting the alteration of microenvironment in TRβ LBD.

248

Along with the red-shift, the intensity of fluorescence from TRβ LBD was

249

significantly decreased after the successive addition of TBBPS, BPA and TBBPA

250

with the increasing concentration from 1 to 20 µM, implying the

251

concentration-dependent fluorescence quenching. Further examination indicated that

252

BPS and BPS-DAE has a similar extent of small fluorescence quenching, whereas

253

TBBPA led to a more marked quenching of fluorescence in comparison with BPA,

254

BPS, and BPS-DAE (Figure 1). An increasing concentration of OBBPS caused the

255

increase of fluorescence intensity of TRβ LBD. It is likely that the high

256

hydrophobicity of OBBPS as revealed by the relatively large LogKow value (Table 1)

257

may partly contribute to the increased fluorescence intensity. It was reported that

258

perfluorodecanoic acid (PFDA) with a longer side chain bears higher hydrophobic

259

interaction and enhances the fluorescence quantum yield, which in turn increases the

260

fluorescence intensity upon binding with bovine hemoglobin and myoglobin54, 55.

12


Page 12 of 44

Page 13 of 44


261

Fluorescence spectra at 303 K and 310 K were analyzed following the

262

Stern-Volmer equation (Text S2) to further characterize the interaction of these

263

chemicals with TRβ LBD. As detailed in Text S2, BPSs, BPA and TBBPA caused

264

static quenching for TRβ LBD. Moreover, TBBPS and TBBPA have markedly higher

265

values of quenching constants (Ksv and Kq in Table S2), implying that both

266

brominated bisphenols induced more significant microenvironment change.

267

3.1.2 BPSs Induced Cα RMSD Changes of TRβ LBD

268

Whether BPSs can induce conformational changes of TRβ LBD was studied by

269

performing MD simulations on ligand free (apo) form and ligand-bound form of TRβ

270

LBD. The root-mean-square deviation of α carbon atoms (Cα RMSD) was used as an

271

important indicator to monitor conformational changes and hence the structural

272

stability of the complexes with TRβ LBD39. MD simulations were reportedly

273

performed 50 ns and 10 ns for interactions of TRβ with household dust contaminants

274

and its agonists, respectively29, 56. In our study, each simulation lasted for 100 ns and

275

this timescale enables us to probe local conformational changes. We sampled 10,000

276

conformations from 100 ns MD trajectories with an interval of 10 ps to measure Cα

277

RMSD in reference to the initial minimized structures (Figure S2). The averaged Cα

278

RMSD for apo TRβ LBD is 1.78 Å. Upon binding, the averaged Cα RMSD for the

279

complex of TRβ LBD with BPS, BPS-DAE, TBBPS, OBBPS, BPA and TBBPA was

280

1.96 Å, 1.77 Å, 1.90 Å, 1.98 Å, 1.99 Å and 2.22 Å, respectively, suggesting the

281

induced conformational changes of TRβ LBD. As revealed by the Cα RMSD

282

monitored along 100 ns MD trajectories (Figure S2), the binding of BPSs affects the

13



283

structural stability of TRβ LBD in comparison with apo TRβ LBD, generally in

284

agreement with the in vitro fluorescence quenching of TRβ LBD upon binding with

285

BPSs, further confirming the disturbed conformational changes of TRβ LBD.

286

3.1.3 BPSs Induced Repositioning of H12 of TRβ LBD

287

The crucial C terminal H12 constitutes important part of the binding site and the

288

dynamics of H12 are vital to the regulation of the transcriptional activity of nuclear

289

receptors57, 58. Upon the binding of contaminants, H12 of TRβ LBD may undergo

290

repositioning, hindering the recruitment of co-activators and transcriptional activity of

291

transactivation function 2(AF-2)39. Figure 2 and 3 show the distance between the

292

nitrogen atom NE2 of residue His435 at helix H11 and nitrogen atom N of residue

293

Phe459 at H12 during 100 ns MD simulation in the presence and absence of BPSs.

294

The changing distance means the relative repositioning of H12. This averaged

295

distance for apo TRβ LBD was 4.41Å and increased significantly to 12.62, 7.66,

296

10.53, 13.74, 8.00 and 8.22 Å upon the binding of BPA, TBBPA, BPS, BPS-DAE,

297

TBBPS and OBBPS, respectively. Ligands can be locked into binding site and the

298

helices shield them from escape at the binding pocket59. The altered position of H12

299

induced by the binding of BPSs could be related to their potentional disruption of

300

transcriptional activity of TR. Since TR and retinoid X receptor (RXR) can form a

301

heterodimer through their LBD interface60, the disruption of conformation of TR LBD

302

may also affect the heterodimerization of TR with RXR, altering ligand-dependent

303

signaling pathways between dimer partners and consequently exerting adverse

304

physiological effects.

14


Page 14 of 44

Page 15 of 44


305

3.2 Different Binding Characteristics of BPSs

306

3.2.1 Distinct Binding Mode to TRβ LBD

307

The ligand competitive binding assay using T3 as a fluorescence probe revealed

308

that BPSs and T3 have the same binding site within TRβ LBD (Text S3, Table S4).

309

The complex of TRβ LBD with BPSs was constructed by molecular docking and was

310

optimized by 100 ns MD simulations. The hydrogen bonds between BPSs and TRβ

311

LBD were calculated based on 10,000 conformations (Table S5). BPSs exhibited

312

distinct binding mode to TRβ LBD. BPA can form hydrogen bonds with residues

313

Gly344, Phe272, Ser331, Ile275 and His435, with a corresponding occupancy of

314

32.52%, 26.39%, 55.99%, 22.59% and 6.42%, respectively. In contrast, BPS forms

315

hydrogen bonds with Thr273 and His435, having a corresponding occupancy of 83.44%

316

and 92.60%. BPA also has a distinct hydrogen bond mode with important amino acid

317

residues of TRβ in comparison with BPS.

318

BPS-DAE, TBBPA and TBBPS form hydrogen bonds with Ser331. The

319

occupancy for the hydrogen bonds with Ser331 was 49.18% (BPS-DAE), 73.36%

320

(TBBPA), and 53.41% (TBBPS), respectively. OBBPS forms hydrogen bond mainly

321

with His435 with occupancy of 14.38%. It was reported that Gly344, Thr273, Ile275

322

and His435 were the critical residues for the interactions of contaminants with TRβ

323

LBD29, 31, 61. These key residues of TRβ LBD can accommodate different moieties of

324

xenobiotic chemicals. For example, His435 and Thr273 of TRβ LBD can be shifted to

325

cater the large moiety of chemicals, whereas Ile275 and Gly344 can reposition to

326

hydroxyl group-related chemicals31, 61.

15



327 328

3.2.2 Varying Binding Potencies to TRβ LBD The total binding free energies (∆Gcalc) of BPSs were evaluated by MMGB/SA at

329

300 K (Table 2). The van der Waals interactions (∆Evdw), electrostatic interactions

330

(∆Eele) and the hydrophobic interactions (∆Esurf) are favourable for the binding. ∆Evdw

331

constitutes the major component of the ∆Gcalc, suggesting van der Waals interactions

332

are most important for the binding between TRβ LBD and BPSs. TBBPA and TBBPS

333

have more negative ∆Gcalc, compared with BPA and BPS, in line with the determined

334

association constant ( ) from the competitive binding assay (Table S4). The more

335

potent binding of the brominated TBBPA and TBBPS induced stable conformational

336

changes, resulting in higher antagonistic activity toward TRβ. For BPS and BPA, the

337

less potent binding caused less alternation of H12 reposition than TBBPS and TBBPA,

338

consequently causing less TR disruption.

339

TBBPA and TBBPS with four Br atoms at 3, 5,3’,5’-positions in phenolic rings

340

possess more negative ∆Evdw value and less negative ∆Eele in comparison with the

341

other three chemicals without any bromination, suggesting that bromination may

342

contribute favourable to van der Waals force. The number of bromination was

343

reported to positively correlated with van der Waals force and negatively related to

344

the electrostatic interactions between BPA analogues and PPARγ and ERα LBD43.

345

The sulfone moiety and alkyl moiety linking two aromatic rings contribute to the

346

different binding free energies, however, bromination appears to contribute

347

predominantly to the ∆Gcalc.

348

3.3 BPSs Showed Antagonistic Activity toward TRβ.

16


Page 16 of 44

Page 17 of 44


349

The TR disruption of BPSs was evaluated by the recombinant two-hybrid TRβ

350

gene yeast assay. This established yeast assay is well suited for a rapid evaluation of

351

various thyroid disrupting chemicals45, 46. T3 was chosen as the positive control and

352

the β-galactosidase activity of this endogenous TRβ agonist was determined. The

353

standard dose-response curve of T3 (Figure S3) revealed a maximum activity at 5 µM,

354

in line with values reported previously46. The β-galactosidase activities of BPSs at

355

environmentally relevant concentrations (5×10-7 µM to 50 µM) were measured. No

356

significant changes of β-galactosidase activity were observed in comparison with the

357

solvent control (Figure S4), suggesting that BPSs showed no agonistic activity toward

358

human TRβ.

359

We further determined whether BPSs present antagonistic activity toward TRβ.

360

BPSs were co-incubated with T3 and the ratio of β-galactosidase activity (U) induced

361

by BPSs and T3 to the β-galactosidase activity induced by T3 was calculated. As

362

shown in Figure 4, BPSs significantly suppressed β-galactosidase activity in a

363

dose-dependent manner, suggesting their antagonistic effect toward TRβ. The

364

corresponding IC10 and relative potency (RP) values were obtained for each BPS in

365

comparison with T3 (Table S6). TBBPS showed the strongest antagonistic activity to

366

TRβ with IC10 value at 10.1 nM and the largest RP at 87.5, followed by TBBPA (IC10:

367

21.1 nM, RP: 41.9) and BPS (IC10: 312 nM, RP: 2.8). The brominated TBBPS and

368

TBBPA showed much stronger antagonistic activity than their corresponding BPS and

369

BPA. Some brominated phenol-related chemicals reportedly show thyroid disrupting

370

activities62. BPS-DAE has a similar dose-response curve to that of its close BPS

17



371

analogue. The brominated OBBPS has a very high LogKow of 9.19 and the low

372

aqueous solubility may contribute partially to its non-detected antagonistic activity63,

373

which may account for low correlation with the MD simulation. Despite the

374

previously reported estrogenic and androgenic disruption23, we show that BPS

375

exhibited weak TRβ disruption similar to BPA with a relatively higher RP, implying

376

that BPS may not be safe substitute of BPA64.

377

3.4 BPSs Altered mRNA Expression Level of TRβ in Zebrafish Larvae.

378

TRβ gene is responsible for the resistance to thyroid hormones and decreasing

379

sensitivity of T3 target tissues 65. We further evaluated the effect of BPSs on the

380

expression level of mRNA of TRβ gene in zebrafish larvae. As revealed by qPCR, the

381

exposure to BPSs at concentrations of 0.01, 0.1, and 1.0 µM for 72 h could

382

significantly alter the expression level of mRNA of TRβ gene (Figure 5). TBBPS and

383

OBBPS at 0.01 µM significantly up-regulated the mRNA expression levels of TRβ,

384

approximately a two-fold increase in comparison with the DMSO control (Figure 5);

385

however, BPS and BPS-DAE at 0.01 µM did not significantly change the mRNA

386

expression levels. Compared with the DMSO control, OBBPS and BPS-DAE at 0.1

387

µM significantly upregulated mRNA expression of TRβ, while BPS and TBBPS

388

showed no significant effect. BPS, TBBPS and OBBPS with the concentration up to

389

1.0 µM significantly up-regulated the mRNA expression level.

390

BPA showed less disruption on mRNA expression of TRβ at 0.01 µM in

391

comparison with TBBPS (P=0.018) and OBBPS (P=0.0061) and more disruption at

392

0.1 µM in compared with BPSs (P

Thyroid Disruption by Bisphenol S Analogues via Thyroid Hormone

Recommend Documents