Novel Fluorinated Surfactants Tentatively Identified ... - ACS Publications

Jan 22, 2015 - ... (Entox), The University of Queensland, Coopers Plains, Queensland 4108, ... Man-Technology-Environment (MTM) Research Centre, Öreb...
0 downloads 0 Views 1MB Size
Subscriber access provided by Columbia Univ Libraries

Article

Novel fluorinated surfactants tentatively identified in firefighters using LC-QTOF-MS/MS and a case-control approach Anna Rotander, Anna Kärrman, Leisa Toms, Margaret Kay, Jochen F. Mueller, and María José Gómez Ramos Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/es503653n • Publication Date (Web): 22 Jan 2015 Downloaded from http://pubs.acs.org on February 4, 2015

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 free 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 accessible to all readers and 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.

Environmental Science & Technology 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 25

Environmental Science & Technology

ACS Paragon Plus Environment

Environmental Science & Technology

Page 2 of 25

1

Novel fluorinated surfactants tentatively identified in

2

firefighters using LC-QTOF-MS/MS and a case-control

3

approach

4

Anna Rotander1, Anna Kärrman2, Leisa-Maree L. Toms3, Margaret Kay4, Jochen Mueller1, Maria Jose

5

Gomez Ramos*1

6

1

National Research Centre for Environmental Toxicology (Entox), The University of Queensland, QLD

7

4108, Australia 2

8

Man-Technology-Environment (MTM) Research Centre, Örebro University, SE-701 82 Örebro,

9

Sweden 3

10

School of Clinical Sciences and Institute of Health and Biomedical Innovation, Faculty of Health,

11 12

Queensland University of Technology, QUT, 4001, Australia 4

The University of Queensland, Discipline of General Practice, Royal Brisbane and Women's Hospital

13

Complex, QLD, 4029, Australia

14 15

*Corresponding author:

16

Maria Jose Gomez Ramos

17

National Research Centre for Environmental Toxicology, Entox

18

The University of Queensland

19

QLD 4108, Australia

20

Phone: +61 0437 511 813

21

Fax: +61 7 3274 9003

22

Email: [email protected] ACS Paragon Plus Environment

1

Page 3 of 25

Environmental Science & Technology

23

RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required

24

according to the journal that you are submitting your paper to)

25 26

Title Running Head: Novel fluorinated surfactants discovered in blood serum from people exposed to

27

AFFF

28 29

Key words: Aqueous film forming foams, AFFF, exposure, firefighters, LC-QTOF-MS/MS, unknown

30

PFAS, PFOS, PFHxS

31

ACS Paragon Plus Environment

2

Environmental Science & Technology

32

Page 4 of 25

Abstract

33

Fluorinated surfactant-based aqueous film–forming foams (AFFF) are made up of per- and

34

polyfluorinated alkyl substances (PFAS), and are used to extinguish fires involving highly flammable

35

liquids. The use of perfluorooctanesulfonic acid (PFOS) and other perfluoroalkyl acids (PFAAs) in

36

some AFFF formulations has been linked to substantial environmental contamination. Recent studies

37

have identified a large number of novel and infrequently reported fluorinated surfactants in different

38

AFFF formulations. In this study, a strategy based on a case-control approach, using quadrupole time of

39

flight tandem mass spectrometry (QTOF-MS/MS) and advanced statistical methods has been used to

40

extract and identify known and unknown PFAS in human serum, associated with AFFF exposed

41

firefighters.

42

(perfluoropentanesulfonic

43

perfluorononanesulfonic acid (PFNS), and four unknown sulfonic acids (Cl-PFOS, ketone-PFOS, ether-

44

PFHxS, and Cl-PFHxS) were exclusively or significantly more frequently detected at higher levels, in

45

firefighters compared to controls.

46 47

Two

target

sulfonic acid

acids

(PFPeS),

(PFOS

and

PFHxS),

perfluoroheptanesulfonic

three acid

non-target

acids

(PFHpS)

and

The application of this strategy has allowed for identification of previously unreported fluorinated chemicals in a timely and cost efficient way.

48

ACS Paragon Plus Environment

3

Page 5 of 25

49

Environmental Science & Technology

Introduction

50

Aqueous film–forming foams (AFFFs) are complex mixtures containing fluorocarbon- and

51

hydrocarbon- based surfactants used to extinguish fires involving highly flammable liquids. Fluorinated

52

surfactants in AFFF are made up of per- and polyfluorinated alkyl substances (PFAS), which are both

53

hydrophobic and oleophobic as opposed to hydrocarbon surfactants that are merely hydrophobic in

54

nature1,2. AFFF have been produced either by electrochemical fluorination (ECF) or telomerization. In

55

the ECF process, as opposed to telomerization that exclusively yields an even number of fluorinated

56

carbon chains, a significant amount of cyclic, cleaved, and branched compounds of different

57

perfluoroalkyl chain lengths are formed in a complex mixture3. The old generation AFFFs, produced by

58

ECF and based on perfluorooctanesulfonic acid (PFOS), were phased out after 2002 when 3M

59

voluntarily discontinued its production of PFOS based products4. The use of PFOS and other

60

perfluoroalkyl acids (PFAAs) in AFFF formulations has been linked to environmental contamination

61

related to handling, storage and usage1,5. Substantially elevated levels of PFOS have been reported in

62

water and biological samples, such as molluscs6, turtles5, wild mink7 and fish8, downstream from

63

airports with a history of firefighting training activities.

64

The new generation AFFF is made up of varieties of PFAS and it has been reported that mixtures of

65

6:2 and 8:2 fluorotelomers predominates9. In environmental samples, for example in seawater10,

66

shrimp11, and human serum12, a significant portion of the total organofluorine can be in the form of

67

unknown fluorinated chemicals. It has been suggested that this may originate from proprietary

68

fluorinated surfactants that may enter the environment through the use of AFFF13. Because many of the

69

chemicals included in these new generation foams are unknown, a number of recent studies have

70

focused on optimizing analytical methods to identify these unknown fluorinated compounds13-17. This

71

has resulted in the identification of a large number of novel and infrequently reported fluorinated

72

surfactants in different AFFF formulations. Previous degradation studies have shown that AFFF

73

components can break down to short chain (C5-C7) perfluorinated carboxylic acids (PFCAs), and more

74

research has been called for to examine these novel surfactants for their potential as PFAA precursors 15, ACS Paragon Plus Environment

4

Environmental Science & Technology

75

18

76

compared to AFFF formulations, suggesting that AFFF precursors were transformed to PFAAs19.

Page 6 of 25

. The fraction of PFAAs was also found to be higher in contaminated groundwater and solid samples

77

Current advances in analytical instrumentation, specifically in accurate mass/high resolution mass

78

spectrometry (HRMS), provide new opportunities for the identification of non-target compounds (i.e.

79

compounds not targeted by the analytical method) and unknown compounds (i.e. compounds not

80

included in any chemical structure database or compounds to date not described in the scientific

81

literature). Quadrupole time-of-flight tandem mass spectrometry (QTOF-MS/MS) is increasingly being

82

used for this purpose20,21,22. The identification/elucidation of unknown compounds, not present in any

83

library, is more complicated than non-target analyses. Non-target analyses usually rely upon the use of

84

personal and commercial libraries. Given the high number of compounds that are generated in a QTOF-

85

MS/MS analysis, filtering strategies are necessary to focus attention on a specific group of masses of

86

interest, enabling their subsequent accurate identification. Filtering strategies using mass defect filtering

87

has been recently employed to identify novel fluoropolymer thermal decomposition products22 and

88

unknown pefluoroalkyl species in fish21. Filtering strategies based on case-control samples have proved

89

to be extremely valuable in the field of clinical chemistry making it possible to identify new biomarkers

90

for the early diagnosis of diseases, such as cancer23 or Alzheimer’s disease24. However, this approach

91

has to our knowledge not yet been applied for identification of environmental contaminants in humans

92

or biota.

93

The purpose of this study was to analyze fluorinated surfactants in blood serum of firefighters who

94

had previously been exposed to AFFF in their work, with the aim of discovering “previously

95

unidentified” persistent PFAS. To carry out this study a case-control approach using LC-QTOF-MS/MS

96

and statistical analysis was developed and applied to detect and identify chemicals associated with

97

AFFF exposure.

98 99

Materials and Methods

ACS Paragon Plus Environment

5

Page 7 of 25

Environmental Science & Technology

100

Study participants. Both firefighters and controls had been recruited for participation in two

101

previous studies. For the current study, blood sera of 20 firefighters collected in 2013 with 19-38 years

102

of exposure to AFFF in their work were selected. Sera from 20 university students and office workers,

103

who had not been exposed to AFFF, had been previously collected in 2011/12. These sera were used as

104

controls. Written informed consent was obtained from all participants at recruitment. Ethics approval for

105

this study was granted by The University of Queensland Medical Research Ethics Committee (Number

106

2014000614).

107 108

Chemicals and Extraction. Details of chemicals and standard compounds used in the extraction are

109

available in the Supporting Information. Mass labeled internal standards were added to 0.2 mL serum.

110

The analytes were then extracted with 1.5 mL 100% acetonitrile using ultrasonication followed by

111

vortex extraction, centrifugation and evaporation to 0.2 mL under a gentle stream of nitrogen.

112

Performance standards and 0.3 mL 5 mM ammonium acetate in water were added prior to analysis.

113 114

LC/ESI-QTOF-MS Analysis. Chromatographic separation of the analytes was carried out using a

115

Shimadzu Nexera X2 UHPLC system equipped with a binary pump and a reverse-phase Gemini-NX

116

C18 column (3µm x 2.0 mm x 50 mm, Phenomenex). The injection volume was 10 µL. The column unit

117

was held at 45 °C and the flow rate was 0.6 mL/min. An isolator column (Phenomenex) was placed

118

directly after the mobile phase mixing chamber to delay the elution of solvent derived background.

119

Mobile phases A and B were, respectively, 1% methanol in water and 10% water in methanol with

120

5mM ammonium acetate in both phases. The initial gradient (10% B) was held for 0.20 min. A gradient

121

ramp followed over 6.5 min to 100% B, which was held for 3 min, then returned to the initial

122

composition in 0.10 min, followed by equilibrium for 2.20 min. The UHPLC was coupled to a hybrid

123

quadrupole time-of-flight mass spectrometer system TripleTOF 5600 System (AB Sciex) with

124

electrospray ionization (ESI) interface working in negative ionization mode. The MS was operated in

125

full scan TOFMS and MS/MS mode with information dependent acquisition (IDA) in a single run ACS Paragon Plus Environment

6

Environmental Science & Technology

Page 8 of 25

126

analysis. The data acquisition and processing was carried out using Analyst® TF 1.6, PeakView®, and

127

MultiQuantTM software (AB Sciex). MarkerViewTM software (AB Sciex) was used for the

128

simultaneous feature finding, alignment, and statistical analysis to highlight compounds of interest.

129

Quantification and quality assurance. Quantification of non-target and unknown PFAS was

130

performed using labeled standards and response factors for PFOS and perfluorohexanesulfonic acid

131

(PFHxS). The method detection limits (MDLs) for PFOS and PFHxS in this study were 0.02 ng/mL and

132

0.07 ng/mL, respectively, and the limit of reporting (LORs) were 0.06 ng/mL for PFOS and 0.35 ng/mL

133

for PFHxS. The reproducibility of the method was calculated from multiple analyses of a pooled serum

134

sample on different days using the internal standard method and mass labeled standards. Other method

135

details and the method validation study (repeatability, reproducibility, and accuracy) are provided in the

136

Supporting Information.

137 138

Statistical analysis. Statistical data analysis was performed using MarkerView software to highlight

139

any groups within the data and identify any trends and unexpected compounds. This software

140

automatically aligns mass and retention time to compensate for minor variations ensuring that identical

141

compounds in different samples are accurately compared. The principal component analysis was

142

performed both without and with normalization of the data using the area of 13C4- PFOS and total area

143

sums. Detailed information about the statistical analysis is provided in the Supporting Information.

144 145

Results and Discussion

146

The relatively rapid and cost efficient identification of both known and unknown PFAS within the

147

sera of workers with occupational exposure to AFFF, was facilitated by an analytical methodology

148

based on a case-control filtering strategy, that has not yet been established in environmental studies.

149

This approach helps to address some of the challenges that researchers face when attempting to identify

150

novel organic chemicals in serum or other biological matrices. To date such research has been

151

challenging because the samples being tested contain many compounds at low concentrations. This ACS Paragon Plus Environment

7

Page 9 of 25

Environmental Science & Technology

152

study provides a data reduction strategy for addressing the large amount of data that is generated using

153

non-target QTOF-MS/MS screening analysis. This analytical strategy combines the use of accurate

154

mass data and statistical evaluation of sample constituents to extract the most important components for

155

further identification. After the multivariate analysis, the compounds of interest were reduced from

156

almost 3000 chemicals to a group of about 50 compounds. Figure 1 illustrates the four components of

157

the ‘case-control’ strategy that were applied; 1) Selection of case and control samples, 2) Sample

158

preparation and analysis, 3) Statistical analysis, and 4) Identification of the extracted compounds. For

159

analytical method and statistical analysis details please see Supporting Information.

160 161

All components of the two groups of samples (firefighters and controls) were compared using

162

multivariate statistical tools including supervised and unsupervised methods in order to facilitate the

163

isolation of the compounds responsible for the discrimination between the groups. The principal

164

component analysis (PCA, unsupervised) showed a clear separation between firefighters and controls

165

(Fig 1, 3A). This separation was obtained both with and without data normalization. The corresponding

166

loading plot indicated that PFOS and PFHxS were strongly associated with the firefighters.

167

Interestingly, after removing PFOS and PFHxS and rerunning the PCA the separation of the two groups

168

remained, indicating that there were other chemicals present that were responsible for this separation.

169

The ions associated with the firefighters were extracted with discriminant analysis (PCA-DA) and T-test

170

(Fig 1, 3B). A volcano plot was used to show the significant entities, which are those at the extremes of

171

the axis (low p-value and a large fold change). The ions with a log fold-change higher than 0.90 and a p-

172

value lower than 0.05 (ions within the red squares in Figure 1, 3B) were selected, which are the ions

173

most strongly associated with the firefighters. All the chemicals identified in this study showed a p-

174

value lower than 0.001, except for PFPeS that presented a p-value of 0.04. Most of the ions selected

175

(ions with fold-change higher than 0.90 and p-value lower than 0.05) presented a mass defect higher

176

than 0.90, which is characteristic of PFAS.

ACS Paragon Plus Environment

8

Environmental Science & Technology

Page 10 of 25

177

A profile plot was used to explore the content of the identified chemicals across all samples (Fig 1,

178

3C). This tool was very useful for confirming that the compounds of interest were almost exclusively

179

present or present at significantly higher levels in the firefighters compared to controls. This tool was

180

also useful for confirming that these compounds were absent in the blanks and standards. In the example

181

from Figure 1 (4A and 4B), 1-chloro-perfluorooctane sulfonic acid (Cl-PFOS; m/z 514.9017,

182

C8O3F16SCl-, 2.4 ppm) is used as an example to show the identification process. Fig 1, 3C shows the

183

profile plot for this mass in all 40 investigated sample, and demonstrates that this compound is uniquely

184

present in the firefighters’ samples.

185

For identification of the discriminating components, the masses extracted with statistical analysis

186

were linked back to the raw mass spectra. The first step in the identification process was to ensure that

187

the signal of interest corresponded to a monoisotopic ion and not to an isotope, adduct or ion product

188

generated during the ionization process. The characterization of compounds not present in commercial

189

libraries was achieved in two steps (Figure 1, 4). i). The determination of the elemental composition

190

from the accurate mass and the use of isotope patterns, mass defect and rings and double bonds (RDB)

191

to reduce the list of candidates (Figure 1, 4A). ii). Public libraries (eg., ChemSpider, PubChem) were

192

used to elucidate possible structures from the elemental formula and MS/MS fragment information

193

(Figure 1, 4B). The non-target compounds (PFPeS, PFHpS and PFNS) were found in ChemSpider.

194

Since these chemical databases do not currently have any mass spectra data for these compounds,

195

proposed molecular structure was crosschecked with the mass spectral data to verify that primary

196

fragments corresponded to realistic chemical compositions. For the unknown compounds (Table 1) the

197

consulted on-line chemical databases did not come up with any results. These compounds were

198

tentatively identified based on the precise interpretation of its accurate mass spectra, MS/MS

199

fragmentation, as well as literature and background information. For confirmation the suspected

200

molecular structure was drawn and linked to the experimental MS/MS data (Fig. 1.4.B).

201

ACS Paragon Plus Environment

9

Page 11 of 25

Environmental Science & Technology

202

Target, non-target, and unknown PFAS exclusively present or present at significantly higher levels in

203

firefighters compared to controls that were either identified (using native standards) or tentatively

204

identified are summarized in Table 1 and shown in the chromatogram in Figure 2. Two target ions (m/z

205

498.9302, PFOS and 398.9633, PFHxS), three non-target ions (m/z 348.9398, perfluoropentanesulfonic

206

acid (PFPeS); m/z 448.9334, perfluoroheptanesulfonic acid (PFHpS) and m/z 548.9270,

207

perfluorononanesulfonic acid (PFNS)), and four unknown ions (m/z 414.9071, m/z 414.9315, m/z

208

514.9007 and m/z 476.9283) were strongly associated with AFFF exposure. It should be noted that

209

method development was performed with strong acidic fluorosurfactants such as PFOS and PFHxS

210

hence several classes of PFAS might not have been extracted and isolated by this analysis. For example,

211

telomer alcohols do not ionize under current conditions and can therefore not be analyzed25. Table 1

212

summarizes the main parameters obtained when applying the LC-QTOF-MS/MS screening. Mass

213

accuracy was