Personal fine particulate matter constituents ... - ACS Publications

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Ecotoxicology and Human Environmental Health

Personal fine particulate matter constituents, increased systemic inflammation and the role of DNA hypomethylation Xiaoning Lei, Renjie Chen, Cuicui Wang, Jingjin Shi, Zhuohui Zhao, Weihua Li, Beizhan Yan, Steve Chillrud, Jing Cai, and Haidong Kan Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.9b02305 • Publication Date (Web): 20 Jul 2019 Downloaded from pubs.acs.org on July 23, 2019

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 29

Environmental Science & Technology

1

Personal fine particulate matter constituents, increased systemic inflammation

2

and the role of DNA hypomethylation

3

Xiaoning Lei1 ，Renjie Chen1, Cuicui Wang1 ，Jingjin Shi1 ，Zhuohui Zhao1 ，

4

Weihua Li2, Beizhan Yan3, Steve Chillrud3, Jing Cai 1,4*，Haidong Kan 1,2*

5

1. School of Public Health, Key Lab of Public Health Safety of the Ministry of

6

Education and NHC Key Laboratory of Health Technology Assessment, Fudan

7

University, Shanghai, China

8

2. Key Laboratory of Reproduction Regulation of National Population and Family

9

Planning Commission, Shanghai Institute of Planned Research, Institute of

10 11 12

Reproduction and Development, Fudan University, Shanghai, China 3. Division of Geochemistry, Lamont-Doherty Earth Observatory of Columbia University, Palisades, New York, USA

13

4. Shanghai Key Laboratory of Meteorology and Health, Shanghai 200030, China

14

*Correspondence: Dr. Haidong Kan, Department of Environmental Health,

15

School of Public Health, Fudan University, P.O. Box 249, 130 Dong-An Road,

16

Shanghai 200032, China. E-mail: [email protected]; or Dr. Jing Cai,

17

Department of Environmental Health, School of Public Health, Fudan University,

18

P.O. Box 249, 130 Dong-An Road, Shanghai 200032, China. E-mail:

19

[email protected]

ACS Paragon Plus Environment


20

ABSTRACT Limited evidence is available on the effects of various fine particulate

21

matter (PM2.5) components on inflammatory cytokines and DNA methylation. We

22

examined whether 16 PM2.5 components are associated with changes in four blood

23

biomarkers, i.e. Tumor Necrosis Factor-α (TNF-α), Soluble Cluster of Differentiation

24

40 Ligand (sCD40L), soluble Intercellular Adhesion Molecule-1 (sICAM-1) and

25

Fibrinogen, as well as their corresponding DNA methylation levels in a panel of 36

26

healthy college students in Shanghai, China. We used linear mixed-effect models to

27

evaluate the associations, with controls of potential confounders. We further conducted

28

mediation analysis to evaluate the potential mediation effects of components on

29

inflammatory markers through change in DNA methylation. We observed that several

30

components were consistently associated with TNF-α and Fibrinogen as well as their

31

DNA hypomethylation. For example, an interquartile range increase in personal

32

exposure to PM2.5-lead (Pb) was associated with 65.20% (95% CI: 37.07, 99.10)

33

increase in TNF-α and 2.66 (95% CI: 37.07, 99.10) decrease in TNF-α methylation,

34

30.51% (95% CI: 0.72, 69.11) increase in Fibrinogen and 1.25 (95% CI: 0.67, 1.83)

35

decrease in F3 methylation. PM2.5 components were significantly associated with

36

sICAM-1 methylation but not with sICAM-1 protein. DNA methylation mediated 19.89%

37

- 41.75% of the elevation in TNF-α expression by various PM2.5 constituents. Our

38

findings provide clues on personal PM2.5 constituents exposure may contribute to

39

increased systemic inflammation through DNA hypomethylation.

40


Page 2 of 29

Page 3 of 29


41



INTRODUCTION

42

Extensive epidemiological studies have linked PM2.5 exposure with the

43

hospitalization and mortality of various cardiovascular diseases (CVDs) 1, 2. Systemic

44

inflammation has been considered as one of common underlying pathological pathways

45

by which PM2.5 affects cardiovascular health3, 4. Previous studies found that short-term

46

exposure to total PM2.5 mass may lead to the upregulation of inflammatory cytokines5.

47

However, the involved biological mechanism remains a matter of speculation.

48

Epigenetic alternations can influence the expression of genes without modifying the

49

genetic code6, and typically DNA hypomethylation may be involved in the

50

development of a wide range of CVDs 7. Our previous panel study hinted that DNA

51

methylation plays an important role in mediating the effect of total PM2.5 mass exposure

52

on inflammatory pathways8.

53

The cardiovascular effect of integrated PM2.5 have been largely evaluated9, 10, but

54

which constituents are most responsible for these effects remain unidentified. It is

55

partially due to the lack of methodology for accurate exposure assessment.

56

Traditionally, PM2.5 exposure assessment relied on fixed monitoring network sites or

57

stationary indoor monitors which usually generate temporally aggregated and averaged

58

data for a certain spatial unit11. However, these methods ignore spatial heterogeneity

59

and difference of activity-patterns among individuals, which could induce bias to the

60

results of health study. The bias thus could be magnified for health effect evaluation of

61

specific PM2.5 constituents, given that PM2.5 composition normally vary with individual

62

and regions12.



63

Therefore, we conducted personal monitoring in a panel of healthy college students

64

in Shanghai, China. We aim to identify the key PM2.5 chemical components responsible

65

for the effects of total PM2.5 exposure on circulating cytokines relevant to systemic

66

inflammation. Moreover, we further explored the mediation effects of specific-gene

67

methylation in the association between inflammatory protein expression and personal

68

PM2.5 constituent exposure.

69



MATERIALS AND METHODS

70

Study population and design. This longitudinal panel study was carried out during

71

the period from December 17, 2014 to July 11, 2015. We excluded 1) participants with

72

any diagnosed cardiopulmonary diseases (including asthma, rhinitis et al. n=2); 2) past

73

or current smoker or alcohol addict (n=0); 3) those who are taking any

74

medication/supplements (n=1) and participants experienced infection within 1 month

75

of baseline physical exam (n=1). Eventually, 36 healthy students from two universities

76

located at downtown area of Shanghai were recruited in this study. To control for the

77

aforementioned potential confounders, all participants were asked to record the

78

information on any use of medications, health conditions, alcohol drinking and physical

79

activity during the follow ups. During the study period, we scheduled 4 rounds repeated

80

measurements for personal exposure and health endpoints, with at least 2-week interval.

81

The study protocol was approved (IRB No. 2014-TYSQ-09-1) by the Institutional

82

Review Board of the School of Public Health of Fudan university, and written informed

83

consent form was obtained from each subject. Detail information about the study design

84

have been described in our previous publication 13.


Page 4 of 29

Page 5 of 29


85

Exposure measurements. For each visit, all subjects were instructed to wear a

86

designed sampling vest, which equipped with HOBO data loggers (Onset Computer

87

Corporation, Pocasset, Massachusetts) and MicroPEM (RTI International, Research

88

Triangle Park, NC, USA). HOBO data loggers were applied to record personal

89

temperature and relative humidity every half-hour. MicroPEM with a 0.50 ( ± 0.05)

90

L/min flow rate was used to log real-time PM2.5 concentrations every 10 seconds based

91

on an on-board micro-nephelometer and collect a Teflon filter (FPTPMP325, Zefon,

92

USA) sample through a one-stage PM2.5 impactor. Under such a low flow rate, we

93

collected a 72-h integrated filter samples for each personal measurement to ensure

94

sufficient amount of loading mass for chemical constitutes analysis.

95

Fourteen elements were measured via polarized-energy dispersive x-ray fluorescence

96

as described by Ngo14, including arsenic (As), barium (Ba), calcium (Ca), chromium

97

(Cr), copper (Cu), iron (Fe), potassium(K), manganese (Mn), phosphorus (P), lead (Pb),

98

silicon (Si), strontium (Sr), titanium (Ti), zinc (Zn). Elemental carbon (EC)

99

concentrations were measured at Columbia University using a multi-wavelength optical

100

measurements15. In the absence of measurements of personal organic carbon (OC) due

101

to the limitations of our instruments and technology, we obtained hourly concentrations

102

of OC measured at a fixed-site of Environmental Science Research Institute (ESRI) of

103

Shanghai which was 0.6 km away from the residence of our participants. The ambient

104

OC was then used as a surrogate of OC exposure. Ambient OC concentrations were

105

measured by a semi-continuous OC/EC analyzer (Model 4G, Sunset Laboratory, OR,

106

USA). More detailed quality assurance/quality control procedures of component



Page 6 of 29

107

measurements were routinely conducted as described in Supplementary Information (SI)

108

materials and methods and our previous publication 16.

109

DNA methylation and biomarker measurements. For each visit, we collected

110

venous peripheral blood samples after 72-h personal sampling for DNA methylation

111

and biomarkers measurements. According to our previous studies 8, 17, we selected four

112

inflammatory proteins, including Tumor Necrosis Factor-α (TNF-α), Intercellular

113

Adhesion Molecule-1 (ICAM-1) and soluble Cluster of Differentiation 40 Ligand

114

(sCD40L), and Fibrinogen. The serum levels of TNF-α, ICAM-1 and Fibrinogen

115

protein

116

cytokine/chemokine kit (Millipore Corporation, Billerica, MA, USA). sCD40L

117

expression was determined using enzyme-linked immune sorbent assays as our

118

previous paper1.

were

detected

by Millipore

MILLIPLEX™

MAP

human

119

We measured DNA methylation levels of TNF-α, ICAM-1 and sCD40L at two to five

120

CpG positions. For Fibrinogen protein, we measured DNA methylation levels of F3

121

instead, as done in previous studies

122

corresponding methylation loci for examining. In addition, Bind et al

123

that F3 hypomethylation may participate in pathways leading to increased Fibrinogen

124

in response to air pollution. The locations of the amplified region and each candidate

125

CpG loci are listed in SI Table S1. Information about the primer sequences are provided

126

in SI Table S2. Two milliliters plasma samples were used for DNA extraction using the

127

QIAmp DNA Mini Kit (Qiagen, Hilden, Germany), and the concentration of purified

128

DNA was determined using a ND1000 spectrophotometer (NanoDrop Technologies

17, 18,

because Fibrinogen does not have


18

documented

Page 7 of 29


129

Inc., Wilmington, DE, USA). We modified aliquots of 300 ng of DNA with sodium

130

bisulfite using the EpiTect Fast DNA Bisulfite Kit (Qiagen, Hilden, Germany) in a final

131

elution volume of 10 μL M-Elution buffer. DNA methylation level was evaluated using

132

highly quantitative bisulfite-PCR pyrosequencing by the PyroMark system (Qiagen,

133

Hilden, Germany). The methylation degree of each CpG dinucleotide was presented as

134

the percentage of methylated cytosines over the sum of methylated and un-methylated

135

cytosines at position 5 (%5mC). More details of DNA methylation and cytokine

136

measurements have been reported in elsewhere 8, 17 and SI materials and methods.

137

Statistical analysis. We used linear mixed-effect (LME) model to estimate

138

association between constituent-specific PM2.5 concentrations and inflammatory

139

markers as well as DNA methylation. All cytokine concentrations were natural log-

140

transformed before statistical analysis to better approximate a normal distribution. In

141

the basic single-constituent model, PM2.5 constituents entered the model as fixed-effect

142

independent variables one at a time, and several covariates including age, gender, body

143

mass index (BMI), 24-h temperature, 24-h relative humidity, and season were also

144

incorporated as fixed-effect predictors. In addition, a random participant-specific

145

intercept was introduced to explain the within-subject correlations of repeated

146

measurements. Moreover, a “constituent-PM2.5 joint model” with the adjustment of

147

total PM2.5 mass (72-h average) was also built to estimate the effects of constituents

148

independent of PM2.5. The Benjamini-Hochberg false discovery rate (FDR) was used

149

for the correction of simultaneous multiple comparisons in dealing with multiple the

150

constituent-specific associations estimating.



151

All LME models were conducted using the package “lme4” in R (version 3.5.0, R

152

development Core Team). The effect estimation for blood biomarkers and methylation

153

were expressed as mean the percent changes and the absolute changes with their 95%

154

confidence intervals (CIs) associated with an interquartile-range (IQR) increase in

155

PM2.5 constituent concentrations. Statistical tests were two-sided, and P-value ≤ 0.05

156

and ≤ 0.10 were considered “significant” and “marginal significant”, respectively.

157

We also conducted mediation analyses to evaluate the potential mediation effects of

158

DNA methylation on the association between a component and an inflammatory

159

biomarker. For this analysis, we used existing “mediation” package in R. Detail

160

information about the methods and materials are described in SI methods and materials.

161

RESULTS

162

Descriptive statistics. Details of the descriptive characteristics of 36 subjects were

163

shown in SI Table S3. The female accounted for nearly 61% (22/36). The average (±

164

standard deviation, SD) age of our participants was 24 ± 2 years and the average (±

165

SD) BMI was 21 ± 3 kg/m2, respectively (SI Table S3). Results from self-reported

166

questionnaires showed that none of them participated in strenuous physical activity,

167

smoked, drank alcohol, took medication or supplements during the study period.

168

Because one participant missed the last visit, a total of 143 valid measurements on both

169

exposure and health were eventually collected.

170

Table 1 provides the summary statistics of personal PM2.5 constituents, temperature

171

and relative humidity. During the study period, the 72-h mean (±SD) mass

172

concentration of total PM2.5 was 45.73 (±31.63) μg/m3. Concentrations of PM2.5


Page 8 of 29

Page 9 of 29


173

constituents vary considerably. Carbonaceous components (OC and EC) were the most

174

abundant composition with a mean (±SD) of 8.00 (±4.92) μg/m3 and 6.14 (±2.84) μg/m3,

175

respectively. Among elements of PM2.5, Fe was the most major component with a mean

176

(±SD) of 0.85 (±0.45) μg/m3, followed by Si with 0.67 (±0.0.35) μg/m3. Our findings

177

were comparable to previous studies reported in Hong Kong12 and Beijing19. In addition,

178

we observed correlations among constituents varied with constituents but they were all

179

statistically significant at 1% levels (SI Table S4).

180

Table 2 provides descriptive statistics for 4 circulating inflammatory cytokine and

181

average DNA methylation. Mean (±SD) concentrations of TNF-α, sCD40L, Fibrinogen,

182

and ICAM-1 were 4.22 (±5.27) pg/mL, 121.50 (±86.61) pg/mL, 3.03 (±3.58) μg/mL,

183

and 172.80 (±98.31) ng/mL, respectively8. The methylation levels and correlation

184

coefficients of CpG loci pairs for TNF-α, ICAM-1, CD40L genes have been reported in

185

our previous work8, and the corresponding detail information for F3 refers to SI Table

186

S5 and SI Figure. S1. Mean (±SD) level of TNF-α, CD40L, F3, and ICAM-1 genes at

187

different CpG locus were 17.28 (±6.60) %5mC, 48.25 (±12.46) %5mC, 3.67

188

(±2.62) %5mC and 3.49 (±2.29) %5mC, respectively. Although the methylated

189

proportions differed across candidate loci, they were highly correlated with each other

190

(all correlation coefficients ranged from 0.58 to 0.92, P  0.05). They may share most

191

of the same functional complexes and traits, we hence introduced the average

192

methylation levels across sites into the model. The correlation coefficients for the 4

193

proteins versus the DNA methylation levels at corresponding upstream candidate CpG

194

sites are shown in SI Table S6. Significant negative correlations were found between



195

TNF-α methylation and TNF-α protein, between F3 methylation and Fibrinogen protein.

196

Regression results. The effect on 4 circulating inflammatory biomarkers varied with

197

constituents and with biomarkers. The percent changes in four inflammatory proteins

198

per an IQR increase in specific 16 PM2.5 constituents' concentrations are illustrated in

199

Figure 1. We observed significantly positive associations of all PM2.5 chemical

200

components with at least one circulating biomarker. sICAM-1 protein was not

201

associated with any constituent; in contrast, TNF-a protein was significantly correlated

202

with all PM2.5 constituents. Especially, an IQR increase in personal exposure to Pb was

203

associated with increases of 42.43% (95% confidence interval (CI): 4.27, 94.57) (P =

204

0.03) and 65.20% (95% CI: 37.07, 99.10) (P  0.01) in sCD40L and TNF-α,

205

respectively. an IQR elevate in personal exposure to Si was associated with increases

206

of 44.46% (95% CI: 16.12, 79.71) in Fibrinogen.

207

Figure 2 presents the changes in DNA methylation associated with increase in 72-h

208

average concentration of PM2.5 constituents. We observed that elevation in all 16

209

components significantly associated with immediate decrease in at least one circulating

210

inflammatory methylation, except Cr negatively linked associated with sICAM-1

211

methylation with marginal significance (P = 0.10). For example, an IQR increase in

212

personal exposure to K and Pb were associated with decreases of 1.28 (95%CI: 0.45,

213

2.11) (P  0.01) and 1.25 (95%CI: 0.67, 1.84) (P  0.01) in CD40L methylation and F3

214

methylation (%5mC), respectively. Si was associated with both 1.21 (95%CI: 0.65, 1.76)

215

(P  0.01) and 3.25 (95%CI: 1.67, 4.82) (P  0.01) decrease in sICAM-1 and TNF-α

216

methylation (%5mC), respectively.


Page 10 of 29

Page 11 of 29


217

Notably, after correcting for multiple comparisons (Table S7), we found robust

218

associations between PM2.5 constituents and TNF-α cytokines; while, the important

219

associations between PM2.5 constituents and sCD40L cytokines as well as CD40L

220

methylation turned to be borderline significant or insignificant. The magnitude of the

221

associations between PM2.5 constituents and Fibrinogen, ICAM-1 cytokines as well as

222

DNA methylation slightly attenuated, and the 95% CIs became wider. Furthermore,

223

except the sCD40L protein, the significant associations between PM2.5 constituents and

224

other inflammation markers as well as methylation of candidate genes did not

225

substantially change in the constituent-PM2.5 joint model (SI Figures S2 and S3).

226

Average methylation level of TNF-α (Pearson r= -0.46, P < 0.01) and F3 (Pearson

227

r= -0.21, P

236

0.10).

237



238

DISCUSSION Previously, we have reported positive associations between integrated PM2.5



239

exposure and the levels of inflammatory cytokines in this group of participants8. Our

240

current results identified specific constituents of PM2.5 exposure substantially

241

contributed to those association, which has implications for developing air pollution

242

abatement strategies to maximize public health benefits. Moreover, this study linked

243

PM2.5 constituents to inflammation methylation on genes related to inflammation

244

pathways and the mediation analysis provided further insight into potential importance

245

of the association between PM2.5 exposure and gene-specific methylation.

246

Systemic inflammation has been suggested as a common underlying pathway of 20.

247

cardiovascular diseases caused by PM2.5 exposure

248

informative set of circulatory cytokines (TNF-α, sICAM-1, sCD40L and Fibrinogen)

249

which are well-established indicators of systemic inflammation and could predict

250

cardiovascular disease risk

251

concentrations 23; however, the results are not consistent 24. The inconsistency of the

252

observations may be partially due to the outdoor fixed-site measurements that were

253

used to assess exposure in these studies which may cause exposure misclassification25.

254

In our study, personal sampling was conducted to minimize the bias. Our observation

255

that TNF-α and sCD40L were associated with a set of PM2.5 components, further

256

confirmed that our previous findings of acute inflammation following personal

257

exposure to total PM2.58. In contrast, none of components were identified to be

258

significantly associated with sICAM-1. It suggests that other components (such as ion

259

fraction) that weren’t examined in this study may be responsible for sICAM-1

260

expression; i.e. Bind et al26 observed that exposure to sulfate could cause sICAM-1

21, 22.

In this study, we selected an

These biomarkers have been linked with PM2.5


Page 12 of 29

Page 13 of 29

261


protein overexpression.

262

We found consistent positive associations between the following PM2.5 components

263

and selected inflammatory biomarkers in both basic single-constituent model and

264

constituent-PM2.5 joint model. EC was found to be associated with sCD40L, Fibrinogen

265

and TNF-α increase. The effect on TNF-α and Fibrinogen remained robust after

266

accounting for multiple comparisons and adjustment for PM2.5, supporting the finding

267

from our prior study in a panel of 28 urban residents in Shanghai 23. Similar results were

268

reported elsewhere

269

exhaust. Lall et al. found that 2.8 µg/m3 increased in traffic-related PM2.5 exposures

270

were positively associated with same day total cardiovascular admissions (RR = 1.041;

271

95% CI, 1.005–1.077) across disease-specific subcategories: stroke and heart failure 30.

272

Thus, our results suggest that the traffic emission might be partly responsible for the

273

influence of PM2.5 exposure on the cardiovascular system disease through inflammatory

274

response.

27-29.

In the metropolis cities, EC is a strong marker of vehicle

275

Although crustal elements constitute only accounts for a small portion of PM2.5 mass,

276

five of them were observed to have stronger influences on inflammatory biomarkers,

277

including K, Ca, P, Si and Sr. Similar effects were also found in transition metals (Cu

278

and Ti) and heavy metals (As and Pb). Among these elements, K and P could trace the

279

biomass burning, Pb and As are widely recognized as markers of sources industry and

280

mobile combustion31, while Ca, Sr and Si mostly are attributed to emissions from mixed

281

sources of construction dust, soil and industry process

282

with the previous findings. For example, Wu et al. found that a significant increase in

32, 33.


Our analyses are in line


283

Fibrinogen is associated with dust and soil source high loading of Ca, Ba, Sr and Ti in

284

Beijing among a panel of 40 healthy university students34. Toxicological and

285

pathological studies documented significant associations of Cu, Ca, K, Si and Ti with

286

inflammatory responses and these components may possess the ability to induce toxic

287

effects in cardiopulmonary system35,

288

cardiovascular hospitalization rates were associated with increased concentrations of

289

road dust factor, which showed high explained variations and concentrations for Si, Ca,

290

Ti and Cu37, 38. Thus, our findings provide evidence that exposure to the aforementioned

291

trace elements could increase cardiovascular risks. These results also highlight that

292

specific sources pollution (i.e., dust, motor vehicles, biomass combustion, and industry)

293

should be of particular concern.

36.

Rich et al. reported that increased acute

294

It is not clear how the systemic inflammation and the adverse cardiovascular

295

outcomes are triggered by PM2.5 exposure. A recent research has identified that DNA

296

methylation can be a potential biological mechanism to explain adverse health effects

297

from air pollution37. This hypothesis could be supported by our observation that similar

298

set of compositions was associated with hypomethylation of genes coding TNF-α and

299

F3 proteins. Our mediation analysis further confirmed that decreased TNF-α

300

methylation potentially mediated the elevated level of TNF-α protein of several PM2.5

301

constituent exposure, including EC, As, Ca, Cu, K, P, Pb, Si, Sr, Ti and Zn. Especially,

302

the mediation effects for these compositions were more than 30%. However, mediated

303

effect of F3 methylation was not significant, which may be due to the much smaller

304

variation of F3 methylation (2.62 vs. 6.60 for TNF-α methylation). We didn’t observe


Page 14 of 29

Page 15 of 29


305

any mediation effects of sCD40L and sICAM-1, which may be impacted by other

306

biological processes.

307

The current study has several strengths. First, this longitudinal panel study

308

measured PM2.5 chemical components in individual levels, which offers an accurate

309

exposure assessment. Second, this study simultaneously evaluated the variation of

310

circulating gene-specific methylation and the expression of the associated inflammatory

311

cytokine after PM2.5 constituent exposure, which provides a unique opportunity to

312

preliminary assess the potential mediation of methylation epigenetic in the

313

cardiovascular effects of PM2.5. Third, all the participants were healthy nonsmoking

314

young college students in the present study, which minimizes the impacts of several

315

confounders, such as unhealthy lifestyle, medication use, chronic disease, ect.

316

Several limitations still exist in our study. First, because the personal exposure

317

measurement is expensive and burdensome, we only recruited 36 participants, which

318

somewhat limited our statistical power and precision. Second, OC concentrations

319

monitored at fixed-site were used to substitute personal OC exposure levels, which may

320

slight the attenuate the accurate of assessment on the relationships between OC fraction

321

and biomarkers. Third, we only obtained the 72-h concentrations of personal PM2.5

322

constituent exposure, other exposure time windows (e.g. 24-h) thus cannot be tested in

323

this study. Finally, we didn’t evaluate the effects of ions (e.g. SO42-, NO3-) and gaseous

324

air pollutant (SO2, NO2 and O3) due to the limitation of instruments, which might also

325

influence DNA methylation and the protein levels.

326

In summary, this longitudinal panel study highlighted that TNF-α hypomethylation



327

might play a modest mediate role in the effects of personal exposure to carbon and trace

328

element components exposure (EC, As, Ca, Cu, K, P, Pb, Si, Sr, Ti and Zn) on increased

329

TNF-α expression. Our results provide evidence for the speculation that methylation

330

epigenetic regulation may be one of the key physiological mechanism linking PM2.5

331

exposure and adverse cardiovascular outcome. More epidemiological and toxicological

332

studies with comprehensive analysis of the exposure, biomarker variables, organ and

333

tissue responses to specific PM2.5 constituents and mixtures are required to replicate our

334

findings, and to decipher the biological mechanism in the future.

335



336

Supporting Information

337

Table S1. Locations of the amplified region and each CpG position relative to the

338

transcription start site. Table S2. Primers used for DNA methylation analysis. Table S3.

339

Description of the 36 study participants in Shanghai, China, 2014–2015. Table S4.

340

Pearson correlation coefficients of the 72-h average concentrations of personal

341

exposure fine particulate matter (PM2.5) and constituents variables. Table S5.

342

Descriptive statistics on F3 methylations for 5 CpG Loci. Table S6. Pearson correlation

343

coefficients for the 4 proteins versus the DNA methylation levels. Table S7. False

344

discovery rate (FDR) - value in blood biomarkers and DNA methylation associated with

345

an interquartile range increase in PM2.5 constituents after correcting for multiple

346

comparisons. Table S8. The mediated effects of F3 methylation in the associations

347

between personal PM2.5 compositions exposure and Fibrinogen protein. Figure S1.

348

Correlations between F3 DNA methylation at the individual CpG site. Figure S2.

ASSOCIATED CONTENT


Page 16 of 29

Page 17 of 29


349

Percent changes in 4 inflammation proteins associated with an interquartile range

350

increase in personal 72-h average concentrations of PM2.5 constituents in the

351

constituents-PM2.5 adjust model. Figure S3. Changes in DNA methylation (5%mC) of

352

4 inflammation genes associated with an interquartile range increase in personal 72-h

353

average concentrations of PM2.5 constituents in the constituents-PM2.5 adjust model. SI

354

Material and Methods. SI References. (PDF)

355



AUTHOR INFORMATION

356

Corresponding Author

357

*Phone/fax: +86 (21) 54237908; e-mail: [email protected].

358

*Phone/fax: +86 (21) 54237908; e-mail: [email protected]

359

Notes

360

The authors declare no competing financial interest.

361



ACKNOWLEDGMENTS

362

The study was supported by the National Key Research and Development Program

363

of China (2016YFC0206504), National Natural Science Foundation of China

364

(91543114) and the Open Foundation of Shanghai Key Laboratory of

365

Meteorological and Health (QXJK201802).

366



367

(1) Requia, W. J.; Adams, M. D.; Arain, A.; Papatheodorou, S.; Koutrakis, P.; Mahmoud, M.,

368

Global Association of Air Pollution and Cardiorespiratory Diseases: A Systematic Review,

369

Meta-Analysis, and Investigation of Modifier Variables. Am J Public Health 2018, 108, (S2),

370

S123-S130.

REFERENCES



371 372

(2) Schwartz, J.; Bind, M. A.; Koutrakis, P., Estimating Causal Effects of Local Air Pollution on Daily Deaths: Effect of Low Levels. Environ Health Perspect 2017, 125, (1), 23-29.

373

(3) Chuang, K. J.; Chan, C. C.; Su, T. C.; Lee, C. T.; Tang, C. S., The effect of urban air pollution

374

on inflammation, oxidative stress, coagulation, and autonomic dysfunction in young adults. Am

375

J Respir Crit Care Med 2007, 176, (4), 370-6.

376

(4) Aung, N.; Sanghvi, M. M.; Zemrak, F.; Lee, A. M.; Cooper, J. A.; Paiva, J. M.; Thomson, R.

377

J.; Fung, K.; Khanji, M. Y.; Lukaschuk, E.; Carapella, V.; Kim, Y. J.; Munroe, P. B.; Piechnik,

378

S. K.; Neubauer, S.; Petersen, S. E., Association Between Ambient Air Pollution and Cardiac

379

Morpho-Functional Phenotypes: Insights From the UK Biobank Population Imaging Study.

380

Circulation 2018, 138, (20), 2175-2186.

381

(5) Cheng, H.; Davis, D. A.; Hasheminassab, S.; Sioutas, C.; Morgan, T. E.; Finch, C. E., Urban

382

traffic-derived nanoparticulate matter reduces neurite outgrowth via TNFalpha in vitro. J

383

Neuroinflammation 2016, 13, 19.

384

(6) Eckhardt, F.; Lewin, J.; Cortese, R.; Rakyan, V. K.; Attwood, J.; Burger, M.; Burton, J.; Cox,

385

T. V.; Davies, R.; Down, T. A.; Haefliger, C.; Horton, R.; Howe, K.; Jackson, D. K.; Kunde,

386

J.; Koenig, C.; Liddle, J.; Niblett, D.; Otto, T.; Pettett, R.; Seemann, S.; Thompson, C.; West,

387

T.; Rogers, J.; Olek, A.; Berlin, K.; Beck, S., DNA methylation profiling of human

388

chromosomes 6, 20 and 22. Nat Genet 2006, 38, (12), 1378-85.

389 390

(7) Baccarelli, A.; Ghosh, S., Environmental exposures, epigenetics and cardiovascular disease. Curr Opin Clin Nutr Metab Care 2012, 15, (4), 323-9.

391

(8) Wang, C.; Chen, R.; Shi, M.; Cai, J.; Shi, J.; Yang, C.; Li, H.; Lin, Z.; Meng, X.; Liu, C.; Niu,

392

Y.; Xia, Y.; Zhao, Z.; Kan, H.; Weinberg, C. R., Possible Mediation by Methylation in Acute


Page 18 of 29

Page 19 of 29


393

Inflammation Following Personal Exposure to Fine Particulate Air Pollution. Am J Epidemiol

394

2018, 187, (3), 484-493.

395

(9) Bonjour, S.; Adair-Rohani, H.; Wolf, J.; Bruce, N. G.; Mehta, S.; Pruss-Ustun, A.; Lahiff, M.;

396

Rehfuess, E. A.; Mishra, V.; Smith, K. R., Solid fuel use for household cooking: country and

397

regional estimates for 1980-2010. Environ Health Perspect 2013, 121, (7), 784-90.

398

(10) Cohen, A. J.; Brauer, M.; Burnett, R.; Anderson, H. R.; Frostad, J.; Estep, K.; Balakrishnan,

399

K.; Brunekreef, B.; Dandona, L.; Dandona, R.; Feigin, V.; Freedman, G.; Hubbell, B.; Jobling,

400

A.; Kan, H.; Knibbs, L.; Liu, Y.; Martin, R.; Morawska, L.; Pope, C. A.; Shin, H.; Straif, K.;

401

Shaddick, G.; Thomas, M.; van Dingenen, R.; van Donkelaar, A.; Vos, T.; Murray, C. J. L.;

402

Forouzanfar, M. H., Estimates and 25-year trends of the global burden of disease attributable

403

to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015.

404

The Lancet 2017, 389, (10082), 1907-1918.

405

(11) Steinle, S.; Reis, S.; Sabel, C. E., Quantifying human exposure to air pollution--moving from

406

static monitoring to spatio-temporally resolved personal exposure assessment. Sci Total

407

Environ 2013, 443, 184-93.

408

(12) Chen, X. C.; Chow, J. C.; Ward, T. J.; Cao, J. J.; Lee, S. C.; Watson, J. G.; Lau, N. C.; Yim, S. H.

409

L.; Ho, K. F., Estimation of personal exposure to fine particles (PM2.5) of ambient origin for healthy

410

adults in Hong Kong. Sci Total Environ 2019, 654, 514-524.

411

(13) Wang, C.; Cai, J.; Chen, R.; Shi, J.; Yang, C.; Li, H.; Lin, Z.; Meng, X.; Liu, C.; Niu, Y.; Xia,

412

Y.; Zhao, Z.; Li, W.; Kan, H., Personal exposure to fine particulate matter, lung function and

413

serum club cell secretory protein (Clara). Environ Pollut 2017, 225, 450-455.

414

(14) Ngo, N. S.; Gatari, M.; Yan, B.; Chillrud, S. N.; Bouhamam, K.; Kinneym, P. L., Occupational



415

exposure to roadway emissions and inside informal settlements in sub-Saharan Africa: A pilot

416

study in Nairobi, Kenya. Atmos Environ (1994) 2015, 111, 179-184.

417

(15) Yan, B.; Kennedy, D.; Miller, R. L.; Cowin, J. P.; Jung, K. H.; Perzanowski, M.; Balletta, M.;

418

Perera, F. P.; Kinney, P. L.; Chillrud, S. N., Validating a nondestructive optical method for

419

apportioning colored particulate matter into black carbon and additional components. Atmos

420

Environ (1994) 2011, 45, (39), 7478-7486.

421

(16) Chen, C.; Cai, J.; Wang, C.; Shi, J.; Chen, R.; Yang, C.; Li, H.; Lin, Z.; Meng, X.; Zhao, A.;

422

Liu, C.; Niu, Y.; Xia, Y.; Peng, L.; Zhao, Z.; Chillrud, S.; Yan, B.; Kan, H., Estimation of

423

personal PM2.5 and BC exposure by a modeling approach - Results of a panel study in

424

Shanghai, China. Environ Int 2018, 118, 194-202.

425

(17) Chen, R.; Meng, X.; Zhao, A.; Wang, C.; Yang, C.; Li, H.; Cai, J.; Zhao, Z.; Kan, H., DNA

426

hypomethylation and its mediation in the effects of fine particulate air pollution on

427

cardiovascular biomarkers: A randomized crossover trial. Environ Int 2016, 94, 614-619.

428

(18) Bind, M. A.; Baccarelli, A.; Zanobetti, A.; Tarantini, L.; Suh, H.; Vokonas, P.; Schwartz, J.,

429

Air Pollution and Markers of Coagulation, Inflammation, and Endothelial Function.

430

Epidemiology 2012, 23, (2), 332-340.

431

(19) Shang, J.; Khuzestani, R. B.; Tian, J.; Schauer, J. J.; Hua, J.; Zhang, Y.; Cai, T.; Fang, D.; An,

432

J.; Zhang, Y., Chemical characterization and source apportionment of PM2.5 personal

433

exposure of two cohorts living in urban and suburban Beijing. Environ Pollut 2019, 246, 225-

434

236.

435

(20) Panasevich, S.; Leander, K.; Rosenlund, M.; Ljungman, P.; Bellander, T.; de Faire, U.;

436

Pershagen, G.; Nyberg, F., Associations of long- and short-term air pollution exposure with


Page 20 of 29

Page 21 of 29


437

markers of inflammation and coagulation in a population sample. Occup Environ Med 2009,

438

66, (11), 747-53.

439 440

(21) Sarzi-Puttini, P.; Atzeni, F.; Doria, A.; Iaccarino, L.; Turiel, M., Tumor necrosis factor-a, biologic agents and cardiovascular risk. Lupus 2005, 14, (9), 780-84.

441

(22) Wu, S.; Deng, F.; Wei, H.; Huang, J.; Wang, H.; Shima, M.; Wang, X.; Qin, Y.; Zheng, C.;

442

Hao, Y.; Guo, X., Chemical constituents of ambient particulate air pollution and biomarkers of

443

inflammation, coagulation and homocysteine in healthy adults: a prospective panel study.

444

Particle and fibre toxicology 2012, 9, (1), 49.

445

(23) Liu, C.; Cai, J.; Qiao, L.; Wang, H.; Xu, W.; Li, H.; Zhao, Z.; Chen, R.; Kan, H., The Acute

446

Effects of Fine Particulate Matter Constituents on Blood Inflammation and Coagulation.

447

Environ Sci Technol 2017, 51, (14), 8128-8137.

448

(24) Brauner, E. V.; Moller, P.; Barregard, L.; Dragsted, L. O.; Glasius, M.; Wahlin, P.; Vinzents,

449

P.; Raaschou-Nielsen, O.; Loft, S., Exposure to ambient concentrations of particulate air

450

pollution does not influence vascular function or inflammatory pathways in young healthy

451

individuals. Part Fibre Toxicol 2008, 5, 13.

452

(25) Goldman, G. T.; Mulholland, J. A.; Russell, A. G.; Strickland, M. J.; Klein, M.; Waller, L. A.;

453

Tolbert, P. E.;, Impact of exposure measurement error in air pollution epidemiology: effect of

454

error type in time-series studies. Environmental health 2011, 10, (1), 61.

455

(26) Bind, M. A.; Lepeule, J.; Zanobetti, A.; Gasparrini, A.; Baccarelli, A.; Coull, B. A.; Tarantini,

456

L.; Vokonas, P. S.; Koutrakis, P.; Schwartz, J., Air pollution and gene-specific methylation in

457

the Normative Aging Study: association, effect modification, and mediation analysis.

458

Epigenetics 2014, 9, (3), 448-58.



459

(27) Croft, D. P.; Cameron, S. J.; Morrell, C. N.; Lowenstein, C. J.; Ling, F.; Zareba, W.; Hopke,

460

P. K.; Utell, M. J.; Thurston, S. W.; Thevenet-Morrison, K.; Evans, K. A.; Chalupa, D.; Rich,

461

D. Q., Associations between ambient wood smoke and other particulate pollutants and

462

biomarkers of systemic inflammation, coagulation and thrombosis in cardiac patients. Environ

463

Res 2017, 154, 352-361.

464 465

(28) Cakmak, S.; Dales, R. E.; Vida, C. B., Components of particulate air pollution and mortality in Chile. Int J Occup Environ Health 2009, 15, (2), 152-8.

466

(29) Ostro, B.; Feng, W. Y.; Broadwin, R.; Green, S.; Lipsett, M., The effects of components of

467

fine particulate air pollution on mortality in california: results from CALFINE. Environ Health

468

Perspect 2007, 115, (1), 13-9.

469

(30) Lall, R.; Ito, K.; Thurston, G. D., Distributed lag analyses of daily hospital admissions and

470

source-apportioned fine particle air pollution. Environ Health Perspect 2011, 119, (4), 455-60.

471

(31) Sun, Y.; Zhuang, G.; Wang, Y.; Han, L.; Guo, J.; Dan, M.; Zhang, W.; Wang, Z.; Hao, Z., The

472

air-borne particulate pollution in Beijing—concentration, composition, distribution and

473

sources. Atmospheric Environment 2004, 38, (35), 5991-6004.

474 475

(32) Song, Y.; Tang, X.; Xie, S.; Zhang, Y.; Wei, Y.; Zhang, M.; Zeng, L.; Lu, S., Source apportionment of PM2.5 in Beijing in 2004. J Hazard Mater 2007, 146, (1-2), 124-30.

476

(33) Sarnat, J. A.; Marmur, A.; Klein, M.; Kim, E.; Russell, A. G.; Sarnat, S. E.; Mulholland, J. A.;

477

Hopke, P. K.; Tolbert, P. E., Fine particle sources and cardiorespiratory morbidity: an

478

application of chemical mass balance and factor analytical source-apportionment methods.

479

Environ Health Perspect 2008, 116, (4), 459-66.

480

(34) Wu, S.; Deng, F.; Wei, H.; Huang, J.; Wang, X.; Hao, Y.; Zheng, C.; Qin, Y.; Lv, H.; Shima,


Page 22 of 29

Page 23 of 29


481

M.; Guo, X., Association of cardiopulmonary health effects with source-appointed ambient

482

fine particulate in Beijing, China: a combined analysis from the Healthy Volunteer Natural

483

Relocation (HVNR) study. Environ Sci Technol 2014, 48, (6), 3438-48.

484

(35) Gurgueira, S. A.; Lawrence, J.; Coull, B.; Murthy, G. K.; & González-Flecha, B., Rapid

485

increases in the steady-state concentration of reactive oxygen species in the lungs and heart

486

after particulate air pollution inhalation. Environmental health perspectives 2002, 110, (8),

487

749-755.

488

(36) Hetland, R. B.; Cassee, F. R.; Refsnes, M.; Schwarze, P. E.; Låg, M.; Boere, A. J. F.; Dybing,

489

E., Release of inflammatory cytokines, cell toxicity and apoptosis in epithelial lung cells after

490

exposure to ambient air particles of different size fractions. Toxicology in Vitro 2004, 18, (2),

491

203-212.

492 493

(37) Squizzato, S.; Masiol, M.; Rich, D. Q.; Hopke, P. K., A long-term source apportionment of PM2.5 in New York State during 2005–2016. Atmospheric Environment 2018, 192, 35-47.

494

(38) Rich, D. Q.; Zhang, W.; Lin, S.; Squizzato, S.; Thurston, S. W.; van Wijngaarden, E.; Croft,

495

D.; Masiol, M.; Hopke, P. K., Triggering of cardiovascular hospital admissions by source

496

specific fine particle concentrations in urban centers of New York State. Environ Int 2019, 126,

497

387-394.

498 499

(39) Bollati, V.; Baccarelli, A., Environmental epigenetics. Heredity (Edinb) 2010, 105, (1), 10512.



Page 24 of 29

501

Table 1. Descriptive statistics on personal PM2.5 and PM2.5 chemical constituents,

502

and weather variables for the study participants (n=36) in Shanghai, China,

503

2014–2015a

504 Variables

Mean

SD

Min

Median

Max

IQR

PM2.5, μg/m3

45.73

31.63

8.43

36.18

139.95

32.92

OC, μg/m3

8.00

4.92

1.07

6.91

25.22

5.75

EC, μg/m3

6.14

2.84

1.06

5.73

15.33

3.11

Ca, ng/m3

393.13

196.83

78.12

352.53

1145.61

192.99

K, ng/m3

Carbonaceous fractions

Crustal elements 550.21

377.78

51.57

442.86

1856.69

296.93

Sr,

ng/m3

2.63

1.99

0.00

2.29

9.39

1.92

Ba,

ng/m3

31.97

19.02

7.41

26.84

108.79

15.35

Si, ng/m3

666.87

350.94

74.09

622.59

1588.21

489.50

P, ng/m3

17.56

10.61

2.57

15.60

87.86

11.62

Fe, ng/m3

848.07

450.93

116.77

754.74

2492.30

518.56

Zn, ng/m3

159.33

114.85

23.03

126.31

713.79

100.58

Cu, ng/m3

18.74

11.19

3.80

16.56

90.81

10.80

Ti,

ng/m3

21.10

10.07

1.94

20.85

45.44

11.98

Cr,

ng/m3

7.88

5.28

0.71

7.06

30.84

5.10

Mn, ng/m3

64.89

32.93

11.30

60.06

205.23

41.90

As, ng/m3

2.43

2.04

0.00

1.80

10.18

2.37

Pb, ng/m3

38.79

34.11

0.37

28.35

145.08

31.20

Temperature, °C

22.44

4.64

10.24

24.60

28.54

6.81

Relative humidity, %

55.13

11.13

31.04

54.20

75.96

17.58

b Temperature,

°C

22.40

4.37

10.50

24.16

29.22

6.35

humidity, %

55.31

9.80

26.31

56.48

75.66

6.44

Transition metals

Heavy metals

b Relative

505

a Definition of abbreviations: SD, standard deviation; IQR, interquartile range.

506

b Data are presented as the 24-h average of weather conditions before blood sample collection.


Page 25 of 29


507

Table 2. Descriptive statistics on inflammation-related DNA methylations and

508

cytokines for the study participants (n=36) in Shanghai, China, 2014–2015a Variables

Mean

SD

4.22

5.27

ICAM-1, ng/mL

172.80

sCD40L, pg/mL

Min

Median

Max

0.66

2.44

38.88

98.31

24.92

143.80

574.40

121.50

86.61

0.33

111.00

372.00

3.03

3.58

0.88

1.88

26.83

17.28

6.60

6.61

15.14

35.43

ICAM-1 methylation, %5mC

3.49

2.29

0.00

2.91

8.88

CD40L methylation, %5mC

48.25

12.46

25.73

55.71

63.64

F3 methylation, %5mC

3.67

2.62

0.00

2.92

9.40

Cytokines TNF-α, pg/mL

Fibrinogen, μg/mL Gene-specific methylation TNF-α methylation, %5mC

509

a Definition of abbreviations: SD, standard deviation; TNF-α, TNF encoding tumor necrosis factor-

510

α; ICAM-1, ICAM1 encoding intercellular adhesion molecule-1; sCD40L, CD40LG encoding

511

soluble cluster of differentiation 40 (CD40) ligand; F3, F3 encoding tissue factor.



Page 26 of 29

512

Table 3. The mediated effects of TNF-α methylation in the associations between

513

personal PM2.5 compositions exposure and TNF-α protein

514 Pathway

PM2.5 constituents → TNF-α methylation → TNF-α protein Mean（%） 95% CI Lower（%） 95% CI Upper（%）

EC

29.34

-7.90

76.00

P-value 0.10

As

33.50

9.30

69.00

0.02

Ca

41.75

-0.53

139.00

0.05

Cu

19.89

3.89

39.00

0.03

K

27.00

13.10

48.00

0.00

P

26.53

6.84

48.00

0.02

Pb

25.21

9.88

48.00

0.00

Si

34.70

17.50

59.00

0.00

Sr

27.18

6.27

57.00

0.01

Ti

32.40

12.10

74.00

0.00

Zn

20.27

0.43

40.00

0.04

515 516


Page 27 of 29


517 518

Figure 1. Percent changes in 4 inflammation proteins associated with an

519

interquartile range increase in personal 72-h average concentrations of PM2.5

520

constituents in the single-constituent model. Abbreviations: TNF-α, tumor necrosis

521

factor alpha; sICAM-1, soluble intercellular adhesion molecule-1; sCD40L, soluble

522

cluster of differentiation 40 (CD40) ligand; CF, carbonaceous fractions; HM, heavy

523

metals.

524



525 526

Figure 2. Changes in DNA methylation (5%mC) of 4 inflammation genes

527

associated with an interquartile range increase in personal 72-h average

528

concentrations of PM2.5 constituents in the single-constituent model. Abbreviations:

529

TNF-α, tumor necrosis factor alpha; sICAM-1, soluble intercellular adhesion molecule-

530

1; CD40L, cluster of differentiation 40 (CD40) ligand; F3: tissue factor; CF,

531

carbonaceous fractions; HM, heavy metals.

532


Page 28 of 29

Page 29 of 29


84x47mm (300 x 300 DPI)


Personal fine particulate matter constituents ... - ACS Publications

Recommend Documents