CDDO-Im

Environmental Pollution

Exposure to tris(1,3-dichloro-2-propyl) phosphate (TDCPP) induces vascular toxicity through Nrf2-VEGF pathway in zebrafish and human umbilical vascular endothelial cells

Xiali Zhong, Jiahuang Qiu, Jianmeng Kang, Xiumei Xing, Xiongjie Shi, Yanhong Wei

To appear in: Environmental Pollution

TDCPP exerts its vascular toxicity by suppressing Nrf2 and the downstream VEGF pathway in vascular endothelial cells, which accounts for the impaired angiogenesis in vascular development.

Exposure to tris(1,3-dichloro-2-propyl) phosphate (TDCPP) induces vascular toxicity through Nrf2-VEGF pathway in zebrafish and human umbilical vascular endothelial cells

Xiali Zhong a, Jiahuang Qiu a, Jianmeng Kang a, Xiumei Xing a, Xiongjie Shi b,

Yanhong Wei a Guangdong Provincial Key Laboratory of Food, Nutrition and Health, Department of Toxicology, School of Public Health, Sun Yat-sen University, Guangzhou 510080,
China

b Hubei Key Laboratory of Cell Homeostasis, College of Life Sciences, the Institute for Advanced Studies, Wuhan University, Wuhan 430072, China

* Corresponding author

Mailing address: Department of Toxicology, School of Public Health, Sun Yat-sen University, No.74 Zhongshan Rd. 2, Guangzhou 510080, China.
Phone: +86 20 87331122

E-mail: [email protected]

1 Abstract

2 The growing production and extensive use of organophosphate flame retardants

3 (OPFRs) have led to an increase in their environmental distribution and human

4 exposure. Developmental toxicity is a major concern of OPFRs’ adverse health effects.

5 However, the impact of OPFRs exposure on vascular development and the toxicity

6 pathway for developmental defects are poorly understood. In this study, we

7 investigated the effects of exposure to tris(1,3-dichloro-2-propyl) phosphate (TDCPP),

8 a frequently detected OPFR, on early vascular development, and the possible role of

9 nuclear factor erythroid 2-related factor (Nrf2)-dependent angiogenic pathway in

10 TDCPP’s vascular toxicity. TDCPP exposure at 300 and 500 µg/L impeded the growth

11 of intersegmental vessels (ISV), a type of microvessels, as early as 30 hpf.

12 Consistently, a similar pattern of decreased extension and remodeling of common

13 cardinal vein (CCV), a typical macrovessel, was observed in zebrafish at 48 hpf and

14 72 hpf. Developing vasculature in zebrafish was more sensitive than general

15 developmental parameters to TDCPP exposure. The expression of genes related to

16 VEGF signaling pathway dose-dependently decreased in TDCPP-treated larvae. In in

17 vitro experiments using human umbilical vascular endothelial cells (HUVECs), the

18 increased cell proliferation induced by VEGF was suppressed by TDCPP exposure in

19 a dose-dependent fashion. In addition, we found a repression of Nrf2 expression and

20 activity in TDCPP-treated larvae and HUVECs. Strikingly, the application of

21 CDDO-Im, a potent Nrf2 activator, enhanced VEGF and protected against defective

22 vascular development in zebrafish. Our results reveal that vascular impairment is a

23 sensitive index for early exposure to TDCPP, which could be considered in the

24 environmental risk assessment of OPFRs. The identification of Nrf2-mediating VEGF

25 pathway provides new insight into the adverse outcome pathway (AOP) of OPFRs.

26

27 Capsule of main finding: TDCPP induces vascular toxicity via the suppression of

28 Nrf2-dependent VEGF pathway

29

30 Keywords: Tris(1,3-dichloro-2-propyl) phosphate (TDCPP); Developmental toxicity;

31 Vascular toxicity; Vascular endothelial growth factor (VEGF); Nuclear factor

32 erythroid 2-related factor (Nrf2); Adverse outcome pathway

33

34 Introduction

35 Flame retardants (FRs) are manufactured chemicals that are applied to various

36 materials to prevent fire or diminish the chance of ignition. Over the past several

37 decades, FRs have been widely used in a wide range of commercial products,

38 including electronics, vehicles, furniture, upholstery, textiles, and toys (Cooper et al.,

39 2016). Most are additive to the base materials (without chemical bonds), which allows

40 for the easy release of FRs from products. Consequently, they are ubiquitously

41 distributed in environmental and biological samples (Deng et al., 2018, LeBel and

42 Williams, 1983). In recent years, the production and application of organophosphate

43 flame retardants dramatically increased to serve as the substitutes for PBDEs (Deng et

44 al., 2018, Dodson et al., 2012, Stapleton et al., 2008). A good amount of OPFRs have

45 been detected in human serum (Li et al., 2017). Of OPFRs’ adverse health effects,

46 impact on development is the primary concern. Evidence from epidemiological

47 studies has revealed an association between exposure to OPFRs and developmental

48 problems including lower birth weights, compromised IQ, anxiety, and hyperactivity

49 (Castorina et al., 2017, Hendriks et al., 2015). In addition, developmental defects

50 caused by OPFRs exposure have also been demonstrated in a variety of animal

51 experiments (Dishaw et al., 2014, Oliveri et al., 2015, Oliveri et al., 2018).

52 Tris(1,3-dichloro-2-propyl) phosphate (TDCPP) is one of the most common

53 OPFRs in baby products (Hoffman et al., 2015, Stapleton et al., 2011). Infants have

54 greater exposure to TDCPP than adults, which is reflected by the high concentration

55 of TDCPP metabolites detected in their urine (Butt et al., 2014, Hoffman et al., 2015) .

56 In light of its presence in house dust, daily OPFRs intake, primarily TDCPP, among

57 toddlers was estimated at ~20 times higher than that of adults (Tan et al., 2017).

58 Experimental and epidemiological studies have revealed that developmental exposure

59 to TDCPP leads to gastrulation defects, aberrant germ-layer formation (Dasgupta et

60 al., 2018), hyperactivity (Moser et al., 2015). However, the mechanisms of

61 developmental defects resulting from TDCPP exposure are poorly understood.

62 The formation of circulation system is one of the earliest events in

63 embryogenesis, and proper blood supply is fundamental to cellular differentiation and

64 proliferation (Gore et al., 2012, Walls et al., 2008). Disruption of vascular

65 development has been directly correlated with miscarriages, birth defects, maternal

66 placental complications, and neurodevelopmental problems (Kleinstreuer et al., 2011).

67 As the absorption, distribution, and excretion of exogenous chemicals (xenobiotics)

68 relies on the circulatory system, blood vessels are prone to contact with xenobiotics

69 upon developmental exposure. Thus, the blood vasculature is probably an important

70 target of their developmental toxicity (Helker et al., 2013). Given role of blood vessel

71 growth in the early stages of embryogenesis, the impacts of xenobiotics on vascular

72 development could be good candidates for the identification of biomarkers for their

73 early exposure and effects. Despite the central role of blood vessels in developmental

74 toxicity, the adverse effects on vascular growth by developmental exposure to TDCPP

75 have been under-emphasized in previous studies.

76 Nuclear factor (erythroid-derived 2)-like 2 (Nrf2) is a key transcription factor

77 responsible for the defense response to environmental toxic insults such as hazardous

78 chemicals. It works by regulating the expression of genes related to redox balance

79 (antioxidants) and detoxification (Zhao et al., 2015, Wang et al., 2012, Wei et al.,

80 2016). Apart from cytoprotection, our recent studies have found that Nrf2 plays a

81 critical role in vascular growth via a context-dependent mechanism. When tissue is

82 injured, Nrf2 is activated in the ischemic tissue and promotes vascular regeneration

83 through the regulation of NADPH oxidase-2 (Wei et al., 2016) and semaphorin

84 6A-mediating pathway (Wei et al., 2015). In addition to injury, Nrf2 has been shown

85 to regulate physiological vascular development in an endothelial autonomous fashion

86 via VEGF-Dll4-Notch signaling (Wei et al., 2015). Taken together, the uncovered

87 Nrf2 functions suggest the potential role of Nrf2 in modulating vascular growth upon

88 chemical exposure.

89 The zebrafish is a powerful model for developmental studies, as its rapid

90 organogenesis is readily observed in the live and transparent embryos outside the

91 body (Childs et al., 2002, Hogan and Schulte-Merker, 2017). Vascular development in

92 zebrafish embryos is accomplished within several days, faster than is the case in

93 rodents (Gore et al., 2012, Walls et al., 2008). In addition, the zebrafish model is a

94 valuable, widely used resource in the study of environmental chemicals’

95 developmental toxicities (Hill et al., 2005, Padilla et al., 2012). In this study, we

96 investigate the effect of exposure to TDCPP on the vascular development of zebrafish,

97 and the possible role of Nrf2-dependent angiogenic pathways in TDCPP’s vascular

98 toxicity. Using both in vivo and in vitro approaches, we found that exposure to

99 TDCPP inhibited developmental vascular growth and suppressed VEGF signaling via

100 Nrf2-dependent pathway. Our findings provide new insight into the adverse outcome

101 pathway (AOP) of developmental exposure to OPFRs. They also help identify

102 potential biomarkers for early exposure and effects.
103

104
Materials and methods

105 2.1 Chemicals and reagents

106 Tirs(1,3-dichloro-2-propyl) phosphate (TDCPP) and dimethyl sulfoxide (DMSO)

107 were obtained from sigma (St. Louis, USA). TDCPP was dissolved in DMSO as stock

108 solution and stored at -20℃. CDDO-imidazolide (CODD-Im) was purchased from

109 MedChemExpress (Monmouth Junction, USA). 1-phenyl-2-thiourea (PTU) and

110 tricaine were obtained from Sigma (St. Louis, USA).
111

112 2.2 Zebrafish and chemical exposure

113 Wild-type AB and Tg (kdrl:eGFP) transgenic zebrafish embryos were obtained

114 from the Core Lab for Medical Science at Sun Yat-sen University. Maintenance and

115 exposure were performed according to previous report (Xing et al., 2018). Embryos

116 were collected and randomly distributed into glass dishes containing 10 mL of

117 TDCPP exposure solution at various concentrations (0, 50, 300 and 500 µg/L) at

118 about 2 hours post-fertilization (hpf). The dose range was designed according to the

119 previous studies (Wang et al., 2013) and our preliminary experiment. For the

120 CDDO-Im experiment, embryos were treated with 25 nM CDDO-Im for 2 hours at 2

121 hpf, prior to the addition of TDCPP. Zebrafish embryos and larvae were maintained at

122 26±1 ℃ with a 14/10 h light/dark cycle, and the exposure solution was changed every

123 day. After 24 hpf, PTU was applied to suppress the growth of melanin.

124 124

125 125

126 2.3 Assessment of general developmental toxicity

127 Cumulative mortality, malformation rate, and hatching rate were recorded as

128 preciously described (Xing et al., 2018). Cumulative mortality was calculated as

129 percentage of the total number of dead individuals from 2 to 30 hpf or 72 hpf

130 following TDCPP exposure. Malformations including pericardial edema, yolk sac

131 edema and tail abnormality were recorded at 30 hpf. Pericardial edema, yolk sac

132 edema, trunk curvature, tail abnormality, and craniofacial malformation were recorded

133 at 72 hpf.

134 134

135 2.4 Intersegmental vessels (ISV) measurement

136 Vasculatures were observed and measured, as previously described (Xing et al.,

137 2018). Tg (kdrl:eGFP) transgenic zebrafish embryos were collected at 30 hpf. After

138 removing the chorion with a needle, the embryos were anesthetized in a freshly

139 prepared solution containing 0.2 mg/ml tricaine. They were then embedded on their

140 sides in a petri dish containing 1.0 % low melting agarose. Images were captured

141 under a confocal microscope with a 10× objective at 488 nm (Leica, Germany). The

142 total number of ISV and completed ISV were counted for each larva to calculate the

143 percentage of completed ISV.

144

145 2.5 Common cardinal vein (CCV) measurement

146 Tg (kdrl:eGFP) transgenic zebrafish embryos were collected at 48 hpf or 72 hpf.

147 The following steps for zebrafish mounting and capture of images were performed as

148 mentioned above for ISV measurement. The CCV of each larva was outlined and the

149 area measured using Adobe Photoshop.
150

151
2.6 Measurement of superoxide dismutase (SOD) activity, catalase (CAT) activity

152 and GSH levels in zebrafish

153 Zebrafish larvae were collected at 72 hpf and homogenized in either 0.9% NaCl

154 (for SOD and GSH assay), or RIPA buffer (CAT assay). SOD activity, CAT activity

155 and GSH levels were examined according the manufacturer’s instructions (SOD

156 activity assay kit A001, GSH assay kit A006, Nanjing Jiancheng Bioengineering

157 Institute, China; catalase activity assay kit, S0051, Beyotime, China). Results were

158 normalized by protein concentrations.
159

160
2.7 Cell culture

161 Primary human venous endothelial cells (HUVECs) were obtained from Cell

162 System (USA). HUVECs were cultured in endothelial growth medium-2 (EGM-2,

163 Lonza, USA) and maintained at 37 ℃ with 5% CO2, as previously described (Ge et al.,

164 2014).

165 2.8 HUVEC viability and proliferation assay

166 HUVECs were seeded and cultured in a 96-well plate. When the cells reached

167 80-90% confluence, 10-200 µM of TDCPP were added. TDCPP doses for cell culture

168 study were applied according to the previous study (Xiang et al., 2017a, Ta et al.,

169 2014). Cell viability was measured with a Calcein-AM assay kit (Trevigen, USA)

170 after 24 hours. In VEGF-induced proliferation experiment, HUVECs were seeded in a

171 96-well plate. When cells were at 60-70% confluence, media were replaced with

172 endothelial basal medium-2 (EBM-2, lonza, USA), EBM-2+25 ng/mL VEGFA, or

173 EBM-2+25 ng/mL VEGFA+TDCPP. After 24 hours, cell proliferation was assessed

174 using a Calcein-AM assay kit (Trevigen, USA) according to the manufacturer’s

175 instructions.

176 176

177 2.9 Immunofluorescence staining

178 HUVECs were plated on the coverslip and maintained in EGM-2 medium until

179 the cells reached 80-90% confluence. The cells were treated with TDCPP for 24 hours,

180 then fixed with 4% paraformaldehyde, washed three times in PBS and blocked for 1

181 hour in blocking solution (10% goat serum, 1% BSA, 0.3% Triton X-100 in PBS) at

182 room temperature. Cells were then incubated with primary antibody NRF2 (1:100,

183 Cell Signaling Technology,USA)overnight at 4 °C. Goat anti-rabbit secondary

184 antibody (Alexa Fluor 594, 1:700, Thermo Fisher, USA) was incubated for 1 hour at

185 room temperature. Nuclei were stained with DAPI for 5 min, and mount on slides

186 with coverslips. Images were obtained with confocal microscope (Zeiss, Germany).

187 187

188 2.10 Quantitative real-time PCR

189 Zebrafish larvae were collected at 72 hpf and HUVEC cells were obtained 24

190 hours after TDCPP exposure. Total RNA was extracted with Trizol reagent

191 (Invitrogen, USA). Total RNA was reverse-transcribed to cDNA using a Quantscript

192 RT Kit (Thermo Fisher, USA). Real-time PCR was performed using an SYBR Green

193 PCR Master Mix reagent kit (TOYOBO, China) and along with zebrafish specific-

194 and human based- primers (designed by Primer-blast). Data were analyzed with the

195 comparative threshold cycle (CT) method as means of relative quantitation, and were

196 normalized to an endogenous reference beta-2 microglobulin (B2m).

197 197

198 2.11 Statistical analysis

199 Data are presented as mean± standard error of mean. Statistical analyses were

200 conducted with one way ANOVA to compare the difference among groups. A

201 comparison between the two groups was performed with a Newman-Keuls Multiple

202 Comparison test. P<0.05 was considered statistically significant.

203 203

204 3. Results and Discussion

205 3.1 The growths of ISV and CCV are sensitive to TDCPP exposure

206 ISV and CCV are typical microvessels and macrovessel in zebrafish, respectively.

207 To investigate the effect of TDCPP on microvasculature, we observed ISV growth in

208 zebrafish embryos (Figure 1A). ISV sprouting starts at the 24 somite stage (about 22

209 hpf) from the dorsal aorta. At ~30 hpf, ISV growth is finished and the dorsal

210 longitudinal anastomotic vessel (DLAV) grows to connect with the adjacent ISVs

211 (Figure 1B) (Tobia et al., 2015). Thus, ISV images were captured at 30 hpf, and the

212 completed ISVs (Helker et al., 2013) were calculated. Compared with the control

213 group, zebrafish embryos exposed to 300 µg/L showed a dramatic decrease in the

214 percentage of completed ISV. A more pronounced reduction was exhibited in the 500

215 µg/L TDCPP treatment group (Figure 1C and D). We also assessed cumulative

216 mortality and malformations including yolk sac edema, pericardial edema, and tail

217 abnormality at 30 hpf. Significant increase only exhibited in zebrafish embryos

218 treated with 500 µg/L TDCPP (Figure 1E and F).

219 CCV, a typical macrovessel in zebrafish, is located in the anterior trunk of the

220 embryo, where it collects venous blood and transports it to the heart (Helker et al.,

221 2013). The endothelial cells of the CCV emerge from the junction of the anterior and

222 posterior cardinal veins at ~24 hpf, and gradually migrate ventrally and reach the

223 ventral margin of the yolk sac at ~56 hpf. After CCV area peaks, the CCV area

224 reduces as it is remodeled into a tube at ~74 hpf (Figure 2A) (Bello et al., 2004). To

225 investigate whether the extension and remodeling of CCVs were affected by TDCPP

226 exposure, CCV images were obtained at 48 hpf and 72 hpf. The CCV area in

227 TDCPP-treated embryos at concentrations of 300 or 500 µg/L was less than that of the

228 control group (Figure 2B and C). A similar pattern was observed in the counting of

229 endothelial cells (Figure S1). At 72 hpf, CCV area in zebrafish larvae exposed to

230 TDCPP increased in a dose-dependent fashion (Figure 2D and E). This indicated a

231 delayed CCV remodeling caused by TDCPP exposure. Similar to the results at 30 hpf,

232 TDCPP exposure at 500 µg/L significantly increased cumulative mortality (Figure 2F)

233 and malformations (including yolk sac edema, pericardial edema, trunk curvature and

234 tail abnormality) (Figure 2G) at 72 hpf. In addition, hatching rate at 72 hpf slightly

235 decreased in TDCPP-exposed group at 500 µg/L (Figure 2H). As compared with

236 control group, no significant difference of cumulative mortality, malformations or

237 hatching rate was observed in other groups.

238 In addition, heart rates (including both atrium and ventricle beat) slightly

239 decreased in 300 µg/L and 500 µg/L groups, whereas the atrium to ventricle beat

240 ratios were 1:1 in all groups, indicating that TDCPP exposure had mild effect on heart

241 function and did not lead to atrioventricular block (Figure S1). The alteration was not

242 as remarkable as the vascular phenotype. Although heart and blood vessels belong to

243 cardiovascular system and share many characteristics in common, blood vessels in

244 zebrafish seem to be more sensitive to TDCPP exposure.

245 Collectively, our results indicate that TDCPP exposure impairs the growths of

246 both microvessels and macrovessels in zebrafish. Developing vasculature is more

247 sensitive to TDCPP exposure than the conventional developmental parameters in the

248 early life stage. Thus, the effect on vascular development in zebrafish could be

249 applied to the assessment of developmental toxicity of OPFRs and the screening of

250 biomarkers for early exposure. Consistent with our study, Xiang et al. have reported

251 that the flame retardant TBPH and its metabolites could inhibit HUVEC growth and

252 induce cell cycle arrest and apoptosis (Xiang et al., 2017b). Similar to TDCPP,

253 BDE-47, a predominant brominated flame retardant, was found to impede the growth

254 of ISV and CCV in zebrafish in our previous study (Xing et al., 2018). However, the

255 range of toxic BDE-47 doses for vascular growth was from 30 to 300 µg/L, lower

256 than that of TDCPP (300-500 µg/L) in this study. These results indicate that TDCPP,

257 as an alternative FR, was less toxic to vascular development than BDE-47. In the

258 present study, we emphasized the morphological changes of vasculature. To better

259 understand the vascular toxicity of TDCPP, the functional alterations of blood vessels

260 should be investigated in the further study.

261 Dishaw et al. showed that TDCPP levels were about 10 pmol/fish on 1 dpf and 1

262 pmol/l fish on 5 dpf in zebrafish exposed to 1 µM (430.9 µg/L) TDCPP (Dishaw et al.,

263 2014). Given the similar levels of TDCPP we used (50-500 µg/L), we may speculate

264 the comparable concentrations of TDCPP in zebrafish (about 1-10 pmol/fish) in our

265 experiment. TDCPP concentrations of 50-500 µg/L are relatively higher than the

266 reported environmentally relevant concentrations. Tan et al reported that the

267 concentrations of TDCPP ranged from 0.42 to10.19 µg/g in South China house dust

268 (Tan et al., 2017), and 61.2 µg/g dust in Brazilian city (Cristale et al., 2018). TDCPP

269 ranging from 520 to 4000 ng/L was found in the wastewater (Greaves and Letcher,

270 2017). In zebrafish, vascular growth starts at about 20 hpf and vascular development

271 proceeds until about 1 month post fertilization (Gore et al., 2012). The objective of

272 the study is to investigate the vascular toxicity of TDCPP and identify the sensitive

273 biomarkers for its early exposure. Thus, the early stage of vascular development (from

274 30 hpf to 72 hpf) was emphasized in our study. Whether long-term exposure to

275 TDCPP at environmental related concentrations during vascular development (for

276 example, exposure for 1 month) may impair vascular growth is necessary to be

277 considered in the further study.

278 278

279 3.2 TDCPP exposure suppressed VEGF pathway in zebrafish larvae and

280 HUVECs

281 VEGF pathway governing vascular sprouting, proliferation and EC migration, is

282 the primary pathway for the regulation of developmental angiogenesis (Li and Harris,

283 2009). VEGFA, a predominant isoform of VEGF, is directly induced by HIF1a and

284 binds to VEGFR1 and VEGFR2. This mediates the extension of vascular networks

285 and promotes angiogenesis (Pitulescu et al., 2017). To investigate the potential

286 mechanism by which TDCPP exposure impairs vascular development, mRNA levels

287 of the essential genes in VEGF pathway were assessed by real-time PCR. The mRNA

288 levels of Vegfa, Vegfr1, Vegfr2 and Vegfa inducer Hifa dose-dependently decreased

289 following exposure to TDCPP in zebrafish larvae (Figure 3A). To further confirm the

290 effect of TDCPP exposure on VEGF pathway, we examined the VEGF signaling in

291 HUVECs following exposure to TDCPP for 24 hours. TDCPP at concentrations of 10

292 µM and 100 µM were applied as the cell survival was not affected by treatment with

293 10-100 µM of TDCPP (Figure S3). In line with in vivo results, TDCPP exposure

294 markedly reduced the mRNA level of VEGF at 100 µM in HUVECs (Figure S4).

295 Moreover, we examined cell proliferation in HUVECs following treatment with both

296 VEGF and TDCPP for 24 hours. The results showed that VEGFA-induced cell

297 proliferation was dose-dependently suppressed by TDCPP exposure (Figure 3B). The

298 results suggest that TDCPP exposure inhibits VEGF pathway. This correlates with the

299 defective vascular development in zebrafish as shown in figure 1 and 2.

300 VEGF pathway is sensitive to xenobiotic toxicants in developmental process.

301 Consistent with our results, administration of O-(chloroacetyl-carbamoyl) fumagillol

302 (TNP-470) significantly decreased VEGF expression and VEGFR2 phosphorylation

303 in the ocular vasculature in rodents at early stages of pregnancy (Joussen et al., 2001).

304 Therefore, the results implicate that TDCPP inhibits vascular growth by suppressing

305 VEGF signaling. Wang et al., reported that exposure to 300-600 µg/L TDCPP

306 disrupted the thyroid endocrine system at 144 hpf (Wang et al., 2013). In our study,

307 vascular toxicity could be detected as early as 30 hpf following exposure to 300 or

308 500 µg/L TDCPP. These results showed that TDCPP-induced vascular toxicity may

309 occur earlier than endocrine disruption. Therefore, the results suggested that the onset

310 of vascular damage is likely independent of VEGF signaling. Whether endocrine

311 system is involved in the progression of vascular toxicity needs to be further

312 investigated.

313 313

314 3.3 TDCPP exposure inhibited Nrf2 function in zebrafish larvae and HUVECs

315 Our previous study demonstrated that Nrf2 plays an important role in the

316 regulation of developmental angiogenesis (Wei et al., 2013). To investigate whether

317 Nrf2 is involved in the effect of TDCPP on vascular growth, we then assessed the

318 Nrf2 expression and activity in zebrafish larvae and HUVECs. The results showed

319 that the mRNA levels of the Nrf2 and its target genes Sod1, Sod2, Gclm and Txn in

320 zebrafish were reduced at 72 hpf following exposure to TDCPP in a dose-dependent

321 fashion (Figure 4A), suggesting that TDCPP exposure suppresses Nrf2 expression and

322 transcriptional activity during vascular development. We also assessed the Nrf2

323 expression and activity in cultured HUVECs. Immunofluorescence staining showed a

324 dose-dependent reduction in NRF2 protein (red) in cells treated with 10 or 100 µM

325 TDCPP for 24 hours (Figure 5A and 5B). Likewise, the mRNA levels of Nrf2, Sod1,

326 Sod2, Gclm, and Txn in HUVECs dose-dependently decreased in TDCPP-treated cells

327 (Figure 5C). Taken together, these results suggest the inhibition of Nrf2 function by

328 TDCPP exposure. This correlates with the reduced VEGF pathway and defective

329 vascular development as shown above.

330 Nrf2 is a stress-response transcription factor frequently activated by chemical

331 exposure (Abolaji et al., 2014). . The activation of Nrf2 induces an adaptive response,

332 acting as a powerful intracellular mechanism for coping with unfavorable

333 environments (Gaisina et al., 2018). However in this study, Nrf2 was activated in

334 neither zebrafish in early life stage nor cultured HUVECs following TDCPP exposure.

335 Conversely, we observed the repression of Nrf2 expression and activity due to TDCPP

336 treatment. Upon different stimuli in specific subcellular compartments, there are three

337 E3 ubiquitin ligase complexes responsible for the ubiquitylation and degradation of

338 NRF2. NRF2 binds to the CUL3-RBX1-KEAP1 complex to respond to

339 electrophilic/oxidative stress in the cytosol. The SCF/β-TrCP complex, which can be

340 either nuclear or cytosolic, is more sensitive to metabolic changes and is regulated by

341 GSK3-β. During ER stress, HRD1 localizes to the endoplasmic reticulum and

342 ubquitylates NRF2 (Dodson et al., 2018, Harder et al., 2015, Liu et al., 2018). . In the

343 present study, we speculate that TDCPP exposure may affect KEAP1 complex activity

344 and that the ubiquitylation of Nrf2 suppresses Nrf2 activity. Further studies are

345 needed to investigate the regulatory mechanism of Nrf2 by TDCPP exposure in our

346 future studies.

347 Nrf2 plays a key role in the maintenance of redox homeostasis through

348 regulation of antioxidants. In our study, TDCPP exposure at 300 or 500 µg/L led to a

349 reduction in GSH level (Figure 4B) and SOD activity (Figure S2) in zebrafish larvae.

350 A decrease in GSH is frequently attributed to the elevated ROS and reduced activities

351 of antioxidants including SOD. Thus, we speculated that a perturbed redox balance

352 accounted for the impairment of vascular growth caused by TDCPP. The

353 down-regulation of a variety of antioxidants was observed in both zebrafish and

354 cultured HUVECs following TDCPP exposure (Figure 4A and 5B). Previous studies

355 have shown that OPFRs interrupt the antioxidant enzyme system to cause oxidative

356 stress both in vitro and in vivo (Chen et al., 2015a, Chen et al., 2015b). The

357 endogenous antioxidant defense system maintains redox homeostasis and is essential

358 for angiogenesis. Disruption of endogenous antioxidants hinders angiogenesis (Jin et

359 al., 2016) and application of antioxidants obliterates pathological angiogenesis

360 (Radomska-Lesniewska et al., 2017) or rescue angiogenesis defects (Jin et al., 2016).

361 Therefore, our results suggest that the impaired vascular growth due to TDCPP

362 exposure is correlated with disturbance of redox homeostasis.

363 363

364 3.4 Activation of Nrf2 by CDDO-Im abrogated the impairment of vascular

365 development due to TDCPP exposure

366 We next sought to confirm that Nrf2-depedent VEGF pathway is crucial for the

367 vascular toxicity of TDCPP upon developmental exposure. Since TDCPP exposure

368 suppressed Nrf2 function in vascular development, we aimed to determine whether

369 boosting Nrf2 activity can counteract TDCPP’s impact on vascular growth. CDDO-Im

370 is a potent Nrf2 activator and has been shown to enhance Nrf2 function and induce its

371 target gene expression in our previous studies (Wei et al., 2015). Zebrafish embryos

372 were pretreated with CDDO-Im 2 hours prior to TDCPP exposure. The mRNA levels

373 of the Nrf2 and its target genes Txn, Sod1, Sod2 soared in CDDO-Im treated zebrafish

374 (> 2 fold, Figure 6A). In addition, CDDO-Im also induced Vegfa expression (> 3 fold,

375 Figure 6A). We further investigated whether CDDO-Im treatment could protect

376 zebrafish from vascular toxicity in development. We found that CDDO-IM treatment

377 alleviated the reduction in the percentage of completed ISV by 300 or 500 µg/L

378 TDCPP (Figure 6B and D). Delayed CCV remodeling in zebrafish exposed to 300 or

379 500 µg/L TDCPP was abrogated by CDDO-IM treatment (Figure 6C and D). These

380 results indicate that the activation of Nrf2 induces VEGF signaling and protects

381 against TDCPP’s vascular toxicity in development. This strongly supports that the

382 defective vascular growth caused by TDCPP exposure is mediated by Nrf2-dependent

383 VEGF pathway.

384 Nrf2 has been shown to regulate angiogenesis dependent on VEGF signaling in

385 previous studies. Kuang et al. have reported that the knockdown of Nrf2 significantly

386 decreases VEGF levels, and activation of Nrf2 up-regulates VEGF expression with a

387 concomitant increase in cell migration and vascular tube formation (Kuang et al.,

388 2013). Our previous studies have shown that Nrf2 regulates physiological vascular

389 development in an endothelial autonomous fashion via VEGF-Dll4-Notch signaling

390 (Wei et al., 2013). In contrast to vascular development, Nrf2 is activated in the

391 ischemic tissue and promotes vascular regeneration through the regulation of NADPH

392 oxidase-2 and VEGF-HIF1a-Sema6A pathway in tissue injury (Wei et al., 2016, Wei

393 et al., 2015). In addition to physiological and pathological condition, the present study

394 has demonstrated that the suppression of Nrf2-dependent VEGF pathway was

395 responsible for the impairment of vascular growth due to TDCPP exposure and thus

396 might be regarded as a crucial toxicity pathway in the adverse effect of TDCPP on

397 development.

398 398

399 4. Conclusion

400 To our best knowledge, this study is the first time to report that TDCPP could

401 impede the growth of vasculature during early developmental stage in zebrafish. We

402 further demonstrated that TDCPP dose-dependently suppressed the Nrf2-mediated

403 VEGF pathway in both zebrafish and cultured vascular endothelial cells. Due to the

404 growing production and increased contents of OPFRs in both the environment and

405 biota, identifying powerful biomarkers for early exposure and delineating the toxicity

406 pathways for the characterization of AOP are urgently needed. In accordance with our

407 results, vascular impairment is a sensitive index for early exposure, which should be

408 considered in the environmental risk assessment of OPFRs. In addition, the

409 identification of Nrf2-dependent VEGF pathway by TDCPP exposure provides further

410 insight into the AOP of OPFRs.

411 411

412 Acknowledgements

413 We thank Drs. Jiayin Dai (Chinese Academy of Sciences) and Wen Chen (Sun

414 Yat-sen University) for helpful suggestions and technical assistance. This work was

415 supported by the National Science Foundation of China (21777199) and National Key

416 R & D program (2017YFC1600205).

417

418 Declarations of interest: none

419

420 Abbreviations

421 Tris(1,3-dichloro-2-propyl) phosphate (TDCPP); intersegmental vessel (ISV);

422 common cardinal vein (CCV); nuclear factor erythroid 2-related factor (Nrf2);

423 1[2-cyano-3,12-dioxooleana-1,9(11)-dien-28-oy] imidazolide (CDDO-IM); vascular

424 endothelial growth factor (VEGF); beta-2 microglobulin (B2m); dimethyl sulfoxide

425 (DMSO); glutamate-cysteine ligase, modifier (Gclm); thioredoxin (Txn); Kelch

426 domain of Kelch-like ECH-associated protein (keap1).

427 427

428 References

429 Abolaji, A. O., Kamdem, J. P., Lugokenski, T. H., Nascimento, T. K., Waczuk, E. P.,

430 Farombi, E. O., Loreto, E. L. & Rocha, J. B. (2014) Involvement of oxidative

431 stress in 4-vinylcyclohexene-induced toxicity in Drosophila melanogaster. Free

432 Radic Biol Med, 71, 99-108.

433 Bello, S. M., Heideman, W. & Peterson, R. E. (2004)

434 2,3,7,8-Tetrachlorodibenzo-p-dioxin inhibits regression of the common cardinal

435 vein in developing zebrafish. Toxicol Sci, 78 (2), 258-266.

436 Butt, C. M., Congleton, J., Hoffman, K., Fang, M. & Stapleton, H. M. (2014)

437 Metabolites of organophosphate flame retardants and 2-ethylhexyl

438 tetrabromobenzoate in urine from paired mothers and toddlers. Environ Sci

439 Technol, 48 (17), 10432-10438.

440 Castorina, R., Bradman, A., Stapleton, H. M., Butt, C., Avery, D., Harley, K. G.,

441 Gunier, R. B., Holland, N. & Eskenazi, B. (2017) Current-use flame retardants:

442 Maternal exposure and neurodevelopment in children of the CHAMACOS cohort.

443 Chemosphere, 189, 574-580.

444 Chen, G., Jin, Y., Wu, Y., Liu, L. & Fu, Z. (2015a) Exposure of male mice to two

445 kinds of organophosphate flame retardants (OPFRs) induced oxidative stress and

446 endocrine disruption. Environ Toxicol Pharmacol, 40 (1), 310-318.

447 Chen, G., Zhang, S., Jin, Y., Wu, Y., Liu, L., Qian, H. & Fu, Z. (2015b) TPP and

448 TCEP induce oxidative stress and alter steroidogenesis in TM3 Leydig cells.

449 Reprod Toxicol, 57, 100-110.

450 Childs, S., Chen, J. N., Garrity, D. M. & Fishman, M. C. (2002) Patterning of

451 angiogenesis in the zebrafish embryo. Development, 129 (4), 973-982.

452 Cooper, E. M., Kroeger, G., Davis, K., Clark, C. R., Ferguson, P. L. & Stapleton, H.

453 M. (2016) Results from Screening Polyurethane Foam Based Consumer Products

454 for Flame Retardant Chemicals: Assessing Impacts on the Change in the

455 Furniture Flammability Standards. Environ Sci Technol, 50 (19), 10653-10660.

456 Cristale, J., Aragao Bele, T. G., Lacorte, S. & Rodrigues de Marchi, M. R. (2018)

457 Occurrence and human exposure to brominated and organophosphorus flame

458 retardants via indoor dust in a Brazilian city. Environ Pollut, 237, 695-703.

459 Dasgupta, S., Cheng, V., Vliet, S. M. F., Mitchell, C. A. & Volz, D. C. (2018)

460 Tris(1,3-dichloro-2-propyl) Phosphate Exposure During the Early-Blastula Stage

461 Alters the Normal Trajectory of Zebrafish Embryogenesis. Environ Sci Technol,

462 52 (18), 10820-10828.

463 Deng, W. J., Li, N., Wu, R., Richard, W. K. S., Wang, Z. & Ho, W. (2018) Phosphorus

464 flame retardants and Bisphenol A in indoor dust and PM2.5 in kindergartens and

465 primary schools in Hong Kong. Environ Pollut, 235, 365-371.

466 Dishaw, L. V., Hunter, D. L., Padnos, B., Padilla, S. & Stapleton, H. M. (2014)

467 Developmental exposure to organophosphate flame retardants elicits overt

468 toxicity and alters behavior in early life stage zebrafish (Danio rerio). Toxicol Sci,

469 142 (2), 445-454.

470 Dodson, M., de la Vega, M. R., Cholanians, A. B., Schmidlin, C. J., Chapman, E. &

471 Zhang, D. D. (2018) Modulating NRF2 in Disease: Timing Is Everything. Annu

472 Rev Pharmacol Toxicol. doi: 10.1146/annurev-pharmtox-010818-021856

473 Dodson, R. E., Perovich, L. J., Covaci, A., Van den Eede, N., Ionas, A. C., Dirtu, A. C.,

474 Brody, J. G. & Rudel, R. A. (2012) After the PBDE phase-out: a broad suite of

475 flame retardants in repeat house dust samples from California. Environ Sci

476 Technol, 46 (24), 13056-13066.

477 Gaisina, I. N., Lee, S. H., Kaidery, N. A., Ben Aissa, M., Ahuja, M., Smirnova, N. N.,

478 Wakade, S., Gaisin, A., Bourassa, M. W., Ratan, R. R., Nikulin, S. V., Poloznikov,

479 A. A., Thomas, B., Thatcher, G. R. J. & Gazaryan, I. G. (2018) Activation of Nrf2

480 and Hypoxic Adaptive Response Contribute to Neuroprotection Elicited by

481 Phenylhydroxamic Acid Selective HDAC6 Inhibitors. ACS Chem Neurosci, 9 (5),

482 894-900.

483 Ge, L., Trujillo, G., Miller, E. J. & Kew, R. R. (2014) Circulating complexes of the

484 vitamin D binding protein with G-actin induce lung inflammation by targeting

485 endothelial cells. Immunobiology, 219 (3), 198-207.

486 Gore, A. V., Monzo, K., Cha, Y. R., Pan, W. & Weinstein, B. M. (2012) Vascular

487 development in the zebrafish. Cold Spring Harb Perspect Med, 2 (5), a006684.

488 Greaves, A. K. & Letcher, R. J. (2017) A Review of Organophosphate Esters in the

489 Environment from Biological Effects to Distribution and Fate. Bull Environ

490 Contam Toxicol, 98 (1), 2-7.

491 Harder, B., Jiang, T., Wu, T., Tao, S., Rojo de la Vega, M., Tian, W., Chapman, E. &

492 Zhang, D. D. (2015) Molecular mechanisms of Nrf2 regulation and how these

493 influence chemical modulation for disease intervention. Biochem Soc Trans, 43

494 (4), 680-686.

495 Helker, C. S., Schuermann, A., Karpanen, T., Zeuschner, D., Belting, H. G., Affolter,

496 M., Schulte-Merker, S. & Herzog, W. (2013) The zebrafish common cardinal

497 veins develop by a novel mechanism: lumen ensheathment. Development, 140

498 (13), 2776-2786.

499 Hendriks, H. S., Koolen, L. A., Dingemans, M. M., Viberg, H., Lee, I., Leonards, P. E.,

500 Ramakers, G. M. & Westerink, R. H. (2015) Effects of neonatal exposure to the

501 flame retardant tetrabromobisphenol-A, aluminum diethylphosphinate or zinc

502 stannate on long-term potentiation and synaptic protein levels in mice. Arch

503 Toxicol, 89 (12), 2345-2354.

504 Hill, A. J., Teraoka, H., Heideman, W. & Peterson, R. E. (2005) Zebrafish as a model

505 vertebrate for investigating chemical toxicity. Toxicol Sci, 86 (1), 6-19.

506 Hoffman, K., Butt, C. M., Chen, A., Limkakeng, A. T., Jr. & Stapleton, H. M. (2015)

507 High Exposure to Organophosphate Flame Retardants in Infants: Associations

508 with Baby Products. Environ Sci Technol, 49 (24), 14554-14559.

509 Hogan, B. M. & Schulte-Merker, S. (2017) How to Plumb a Pisces: Understanding

510 Vascular Development and Disease Using Zebrafish Embryos. Dev Cell, 42 (6),

511 567-583.

512 Jin, P., Li, T., Li, X., Shen, X. & Zhao, Y. (2016) Suppression of oxidative stress in

513 endothelial progenitor cells promotes angiogenesis and improves cardiac function

514 following myocardial infarction in diabetic mice. Exp Ther Med, 11 (6),

515 2163-2170.

516 Joussen, A. M., Beecken, W. D., Moromizato, Y., Schwartz, A., Kirchhof, B. &

517 Poulaki, V. (2001) Inhibition of inflammatory corneal angiogenesis by TNP-470.

518 Invest Ophthalmol Vis Sci, 42 (11), 2510-2516.

519 Kleinstreuer, N. C., Judson, R. S., Reif, D. M., Sipes, N. S., Singh, A. V., Chandler, K.

520 J., Dewoskin, R., Dix, D. J., Kavlock, R. J. & Knudsen, T. B. (2011)

521 Environmental impact on vascular development predicted by high-throughput

522 screening. Environ Health Perspect, 119 (11), 1596-1603.

523 Kuang, L., Feng, J., He, G. & Jing, T. (2013) Knockdown of Nrf2 inhibits the

524 angiogenesis of rat cardiac micro-vascular endothelial cells under hypoxic

525 conditions. Int J Biol Sci, 9 (7), 656-665.

526 LeBel, G. L. & Williams, D. T. (1983) Determination of organic phosphate triesters in

527 human adipose tissue. J Assoc Off Anal Chem, 66 (3), 691-699.

528 Li, J. L. & Harris, A. L. (2009) Crosstalk of VEGF and Notch pathways in tumour

529 angiogenesis: therapeutic implications. Front Biosci (Landmark Ed), 14,

530 3094-3110.

531 Li, P., Jin, J., Wang, Y., Hu, J., Xu, M., Sun, Y. & Ma, Y. (2017) Concentrations of

532 organophosphorus, polybromobenzene, and polybrominated diphenyl ether flame

533 retardants in human serum, and relationships between concentrations and donor

534 ages. Chemosphere, 171, 654-660.

535 Liu, D., Xue, J., Liu, Y., Gu, H., Wei, X., Ma, W., Luo, W., Ma, L., Jia, S., Dong, N.,

536 Huang, J., Wang, Y. & Yuan, Z. (2018) Inhibition of NRF2 signaling and

537 increased reactive oxygen species during embryogenesis in a rat model of retinoic

538 acid-induced neural tube defects. Neurotoxicology, 69, 84-92.

539 Moser, V. C., Phillips, P. M., Hedge, J. M. & McDaniel, K. L. (2015)

540 Neurotoxicological and thyroid evaluations of rats developmentally exposed to

541 tris(1,3-dichloro-2-propyl)phosphate (TDCIPP) and tris(2-chloro-2-ethyl)

542 phosphate (TCEP). Neurotoxicol Teratol, 52 (Pt B), 236-247.

543 Oliveri, A. N., Bailey, J. M. & Levin, E. D. (2015) Developmental exposure to

544 organophosphate flame retardants causes behavioral effects in larval and adult

545 zebrafish. Neurotoxicol Teratol, 52 (Pt B), 220-227.

546 Oliveri, A. N., Ortiz, E. & Levin, E. D. (2018) Developmental exposure to an

547 organophosphate flame retardant alters later behavioral responses to dopamine

548 antagonism in zebrafish larvae. Neurotoxicol Teratol, 67, 25-30.

549 Padilla, S., Corum, D., Padnos, B., Hunter, D. L., Beam, A., Houck, K. A., Sipes, N.,

550 Kleinstreuer, N., Knudsen, T., Dix, D. J. & Reif, D. M. (2012) Zebrafish

551 developmental screening of the ToxCast Phase I chemical library. Reprod Toxicol,

552 33 (2), 174-187.

553 Pitulescu, M. E., Schmidt, I., Giaimo, B. D., Antoine, T., Berkenfeld, F., Ferrante, F.,

554 Park, H., Ehling, M., Biljes, D., Rocha, S. F., Langen, U. H., Stehling, M.,

555 Nagasawa, T., Ferrara, N., Borggrefe, T. & Adams, R. H. (2017) Dll4 and Notch

556 signalling couples sprouting angiogenesis and artery formation. Nat Cell Biol, 19

557 (8), 915-927.

558 Radomska-Lesniewska, D. M., Balan, B. J. & Skopinski, P. (2017) Angiogenesis

559 modulation by exogenous antioxidants. Cent Eur J Immunol, 42 (4), 370-376.

560 Stapleton, H. M., Allen, J. G., Kelly, S. M., Konstantinov, A., Klosterhaus, S., Watkins,

561 D., McClean, M. D. & Webster, T. F. (2008) Alternate and new brominated flame

562 retardants detected in U.S. house dust. Environ Sci Technol, 42 (18), 6910-6916.

563 Stapleton, H. M., Klosterhaus, S., Keller, A., Ferguson, P. L., van Bergen, S., Cooper,

564 E., Webster, T. F. & Blum, A. (2011) Identification of flame retardants in

565 polyurethane foam collected from baby products. Environ Sci Technol, 45 (12),

566 5323-5331.

567 Ta, N., Li, C., Fang, Y., Liu, H., Lin, B., Jin, H., Tian, L., Zhang, H., Zhang, W. & Xi,

568 Z. (2014) Toxicity of TDCPP and TCEP on PC12 cell: changes in CAMKII,

569 GAP43, tubulin and NF-H gene and protein levels. Toxicol Lett, 227 (3),

570 164-171.

571 Tan, H., Peng, C., Guo, Y., Wang, X., Wu, Y. & Chen, D. (2017) Organophosphate

572 Flame Retardants in House Dust from South China and Related Human Exposure

573 Risks. Bull Environ Contam Toxicol, 99 (3), 344-349.

574 Tobia, C., Gariano, G., Guerra, J. & Presta, M. (2015) Zebrafish embryo

575 intersegmental vessels: a tool for investigating sprouting angiogenesis. Methods

576 Mol Biol, 1214, 173-184.

577 Walls, J. R., Coultas, L., Rossant, J. & Henkelman, R. M. (2008) Three-dimensional

578 analysis of vascular development in the mouse embryo. PLoS One, 3 (8), e2853.

579 Wang, L., Kou, M. C., Weng, C. Y., Hu, L. W., Wang, Y. J. & Wu, M. J. (2012)

580 Arsenic modulates heme oxygenase-1, interleukin-6, and vascular endothelial

581 growth factor expression in endothelial cells: roles of ROS, NF-kappaB, and

582 MAPK pathways. Arch Toxicol, 86 (6), 879-896.

583 Wang, Q., Liang, K., Liu, J., Yang, L., Guo, Y., Liu, C. & Zhou, B. (2013) Exposure

584 of zebrafish embryos/larvae to TDCPP alters concentrations of thyroid hormones

585 and transcriptions of genes involved in the hypothalamic-pituitary-thyroid axis.

586 Aquat Toxicol, 126, 207-213

587 Wei, Y., Gong, J., Thimmulappa, R. K., Kosmider, B., Biswal, S. & Duh, E. J. (2013)

588 Nrf2 acts cell-autonomously in endothelium to regulate tip cell formation and

589 vascular branching. Proc Natl Acad Sci U S A, 110 (41), E3910-3918.

590 Wei, Y., Gong, J., Xu, Z. & Duh, E. J. (2016) Nrf2 promotes reparative angiogenesis

591 through regulation of NADPH oxidase-2 in oxygen-induced retinopathy. Free

592 Radic Biol Med, 99, 234-243.

593 Wei, Y., Gong, J., Xu, Z., Thimmulappa, R. K., Mitchell, K. L., Welsbie, D. S., Biswal,

594 S. & Duh, E. J. (2015) Nrf2 in ischemic neurons promotes retinal vascular

595 regeneration through regulation of semaphorin 6A. Proc Natl Acad Sci U S A, 112

596 (50), E6927-6936.

597 Xiang, P., Liu, R. Y., Sun, H. J., Yang, Y. W., Cui, X. Y. & Ma, L. Q. (2017a) Effects

598 of novel brominated flame retardant TBPH and its metabolite TBMEHP on

599 human vascular endothelial cells: Implication for human health risks. Environ Res,

600 156, 834-842.

601 Xiang, P., Liu, R. Y., Li, C., Gao, P., Cui, X. Y. & Ma, L. Q. (2017b) Effects of

602 organophosphorus flame retardant TDCPP on normal human corneal epithelial

603 cells: Implications for human health. Environ Pollut, 230, 22-30.

604 Xing, X., Kang, J., Qiu, J., Zhong, X., Shi, X., Zhou, B. & Wei, Y. (2018) Waterborne

605 exposure to low concentrations of BDE-47 impedes early vascular development

606 in zebrafish embryos/larvae. Aquat Toxicol, 203, 19-27.

607 Zhao, M., Chen, J., Zhu, P., Fujino, M., Takahara, T., Toyama, S., Tomita, A., Zhao, L.,

608 Yang, Z., Hei, M., Zhong, L., Zhuang, J., Kimura, S. & Li, X. (2015)

609 Dihydroquercetin (DHQ) ameliorated concanavalin A-induced mouse

610 experimental fulminant hepatitis and enhanced HO-1 expression through

611 MAPK/Nrf2 antioxidant pathway in RAW cells. Int Immunopharmacol, 28 (2),

612 938-944.

613 Figure legends

614 Figure 1. TDCPP exposure inhibited ISV growth in zebrafish embryos. (A) ISV in

615 Zebrafish (outlined in yellow). (B) Schematic of ISV growth. ISV sprouts from the

616 DA. At ~30 hpf, ISV growth is finished and the DLAV grows dorsally to connect with

617 the adjacent ISVs. (C) Representative images of ISV at 30 hpf following exposure to

618 TDCPP. Completed ISVs are denoted by arrows. (D) The percentage of completed

619 ISV at 30 hpf. (E) Cumulative mortality from ~2 hpf to 30 hpf. (F) Percentage of

620 zebrafish embryos displaying malformations (including yolk sac edema, pericardial

621 edema, and tail abnormality) at 30 hpf. ISV: intersegmental vessel; DA: dorsal aorta;

622 DLAV: dorsal longitudinal anastomotic vessel. Scale bars in A: 200 µm; B: 20 µm; C:

623 100 µm. Data are presented as mean ± SEM. n = 10-15 embryos (in A-D) or 5 dishes

624 (in E and F, each dish contains 40 embryos before treatment with TDCPP). * p<0.05,

625 ** p<0.01 and *** p<0.001 as compared with the control group.

626 626

627 Figure 2. TDCPP exposure impeded the extension and remodeling of CCV in

628 zebrafish embryos and larvae. (A) Schematic of CCV in zebrafish. The vascular

629 endothelial cells (*) of CCV begin to emerge from the junction of the anterior and

630 posterior cardinal veins on both sides of the embryo at ~24 hpf. The CCV then

631 extends ventrally from the junction as a sheet of vascular endothelial cells until

632 reaching the ventral margin of the yolk sac at ~56 hpf. The arrow indicates the

633 direction of CCV extension. (B) Representative images of CCV at 48 hpf following

634 exposure to TDCPP. (C) CCV area at 48 hpf. (D) Representative images of CCV at 72

635 hpf following exposure to TDCPP. CCV growth involves a remodeling process from

636 48 to 74 hpf and within the process CCV transforms from a sheet of endothelial cells

637 into a tube. (E) CCV area at 72 hpf. (F) Cumulative mortality from ~2 hpf to 72 hpf.

638 (G) Percentage of zebrafish embryos displaying malformations (including yolk sac

639 edema, pericardial edema, trunk curvature and tail abnormality) at 72 hpf. (H)

640 Hatching rate at 72 hpf. Scale bars in A: 200 µm, 5µm (insert); B and E: 100 µm.

641 Data are presented as mean ± SEM. n = 15-20 embryos or larvae (in A-E) or 5 dishes

642 (in F-H, each dish contains 40 embryos before treatment with TDCPP). * p<0.05, **

643 p<0.01 and *** p<0.001 as compared with the control group.

644 644

645 Figure 3. TDCPP exposure suppressed VEGF pathway in zebrafish larvae and

646 HUVECs. (A) The mRNA levels of Vegfa, Vegfr1, Vegfr2 and Hif1a in zebrafish

647 larvae at 72 hpf following exposure to TDCPP. (B) VEGFA-induced cell proliferation

648 was dose-dependently suppressed by TDCPP exposure. HUVECs were treated with

649 25 ng/ml VEGFA and 10-200 µM TDCPP for 24 hours. Data are presented as mean ±

650 SEM. n = 8 pooled samples (10 larvae in each sample) or 3 independent experiments

651 in cell culture experiment. * p<0.05, ** p<0.01 and *** p<0.001 a as compared with

652 the control group (A) or the group treated with VEGF only (B).

653 653

654 Figure 4. TDCPP exposure inhibited Nrf2 function and decreased GSH content in

655 zebrafish larvae. (A) The mRNA levels of Nrf2, Sod1, Sod2, Gclm, and Txn at 72 hpf.

656 (B) GSH level at 72 hpf. Data are presented as mean ± SEM. n = 5. * p<0.05 and **

657 p<0.01 as compared with the control group.

658 658

659 Figure 5. TDCPP exposure inhibited Nrf2 expression and activity in HUVECs. (A)

660 Representative images of HUVECs subjected to immunofluorescence staining of

661 NRF2 (red). Nuclei were stained with DAPI. Scale bar: 50 µm. (B) Quantification of

662 NRF2 immunofluorescence in (A). The result shows a dose-dependent reduction in

663 NRF2 protein in HUVECs treated with 10 or 100 µM TDCPP for 24 hours. n=8

664 random fields (40 cells) per group. (C) mRNA levels of Nrf2, Sod1, Sod2, Gclm, and

665 Txn in HUVECs treated with 10 or 100 µM TDCPP for 24 hours. Scale bar: 10 µm. n

666 = 3 independent experiments. Data are presented as mean ± SEM. * p<0.05, **

667 p<0.01 and *** p<0.001 as compared with the control group.

668

669 Figure 6. Activation of Nrf2 by CDDO-Im abrogated the impairment of vascular

670 development due to TDCPP exposure. (A) The mRNA expression level of Nrf2, Nrf2

671 target genes, and Vegfa in zebrafish treated with 25 nM CDDO-IM. n = 8 pooled

672 samples (10 larvae in each sample) (B) The percentage of completed ISV following

673 exposure to CDDO-Im and TDCPP at 30 hpf. (C) CCV area following CDDO-Im

674 treatment at 72 hpf. (D) Upper panels: representative images of ISV at 30 hpf

675 following exposure to CDDO-IM and TDCPP; Botton panels: representative images

676 of CCV at 72 hpf following exposure to CDDO-IM and TDCPP. Scale bar: 100 µm.

677 Data are presented as mean ± SEM. n = 15-20 embryos or larvae. * p<0.05, ** p<0.01

678 and ***p<0.001 as compared with the control group.

Highlights

TDCPP exposure impeded the growth of microvessels (ISV), as early as 30 hpf, and delayed the extension or remodeling of macrovessels (CCV) at 48 hpf and 72 hpf in zebrafish.
Developing vasculature in zebrafish was more sensitive than CDDO-Im general developmental parameters to TDCPP exposure.
TDCPP exposure inhibited Nrf2-depedent VEGF signaling pathway in zebrafish and cultured human umbilical vascular endothelial cells.