Ke He 1, Anne Timm 2, and Lee Blaney 1. University of Maryland Baltimore County. US Department of Agriculture Forest Service

Transcription

1 Simultaneous determination of UV-filters and estrogens in aquatic invertebrates by modified QuEChERS extraction and liquid chromatography tandem mass spectrometry Ke He 1, Anne Timm 2, and Lee Blaney 1 1 University of Maryland Baltimore County 2 US epartment of Agriculture Forest Service 2017 NEMC Aug 7, 2017 Grand Hyatt Washington, C

2 Estrogens and UV-filters have attracted increased attention as contaminants of emerging concern (CECs) Estrogens: Cause endocrine disruption (e.g., feminization of male fish) 1,2 Bioaccumulate in aquatic organisms 3,4 From Bentham Science Publishers UV-filters: Bioaccumulate in aquatic organisms 2,5 emonstrate estrogenic activity 3 Exhibit toxic impacts on coral reefs 6 From Nature News 1. C. Schmitt et al, 2008; 2. R. Cespedes, 2004; 3. A.M. Al-Ansari et al, 2013; 4. B. Huerta et al, 2015; 5. S. Morrison et al, 2016; 6. C. owns et al,

3 Motivation for simultaneous determination of estrogens and UV-filters in invertebrate tissue LC-MS based methods for simultaneous determination of multiple estrogens and UV-filters are not available; Protocols for effective co-extraction of estrogens and UV-filters from tissue samples are scarce; and, Invertebrates, which have a limited amount of tissue, play important ecological roles have not been rigorously studied. 3

4 Objectives of this talk 1. Analyze estrogens and UV-filters simultaneously with one LC-MS/MS method; 2. evelop efficient extraction (a) and cleanup (b) strategies to extract three estrogens and five UV-filters from tissue samples; and, 3. Examine estrogen and UV-filter concentrations in aquatic and marine invertebrates (i.e., Orconectes virilis and Crassostrea virginica) collected in Maryland. 4

5 Part I: Simultaneous determination of estrogens and UV-filters with LC-MS/MS 5

6 Chemical structures of analytes and internal standards HO H C H 3 H H Estrone (E1, E2-d 3 ) O HO H C H 3 H H OH Estradiol (E2, E2-d 3 ) HO H C H 3 H H OH Ethinyl estradiol (EE2, EE2-d 4 ) CH H3C H 3 C CH 3 O 4-methylbenzylidene camphor (4-MBC, 4-MBC-d 4 ) CH 3 H 3 C O OH O Benzophenone-3 (BP-3, BP-3-d 5 ) H 3 C CH 3 O O OH CH 3 Homosalate (HMS, HMS-d 4 ) N O O H 3 C O O O Octocrylene (OC, OC-d 15 ) Ethylhexylmethoxycinnamate (EHMC, EHMC-d 15 ) 6

7 Positive and Negative ESI-MS/MS fragmentations for OC and HMS NH + H 3 C O CH 3 O [OC + H + ] O O H 3 C CH 3 CH O - [HMS H + ] - 7

8 MS/MS response MS/MS response Wrong-way-round ionization behavior of select UV-filters 1E+9 1E+8 ACN (0.1% HCOOH) ACN (10 mm NH4COOH) ACN-water ACN (0.1% NH4OH) MeOH (0.1% HCOOH) MeOH (10 mm NH4COOH) MeOH-water MeOH (0.1% NH4OH) 1E+7 1E+6 1E+5 1E+4 1E+3 1E+9 E1 E2 EE2 HMS 1E+8 1E+7 1E+6 1E+5 1E+4 1E+3 4-MBC BP-3 EHMC OC Analytes were prepared at 10 µg/l in water; each sample was injected five times. 8

9 Relative response Relative response LC-MS/MS workload and sensitivity were improved with wrong-way-round ionization Negative mode Positive mode Retention time (min) Retention time (min) Analytes were prepared at 10 µg/l. A Waters Xbridge BEH C18 column ( mm, 2.5 µm) was used for separation. The elution gradient employed (A) water with 0.1% NH 4 OH (ph 10.5) and (B) MeOH with 0.1% NH 4 OH at a flow rate of 0.2 ml/min. 9

10 Part II (a): Extraction of estrogens and UV-filters from tissue samples 10

11 Conventional techniques involve one-step extraction Soxhlet extraction Accelerated solvent extraction (ASE) Sonication/ultrasound assisted extraction Ref:

12 The QuEChERS (Quick, Easy, Cheap, Efficient, Rugged, and Safe) strategy provides two-step extraction We employed a modified QuEChERS extraction as indicated below: 50 mg freeze-dried tissue samples 5 ml I + 5 ml acetonitrile 2.5 g MgSO g NaCl Extract with 50 % ACN Extract with 92 % ACN ACN layer Aliquot of 2.5 ml up layer extract Water layer Cleanup with dispersive-spe Instrument analysis 12

13 Mean recovery Recovery of estrogens and UV-filters from tissue with ACN extraction 140% 120% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS 100% 80% 60% 40% 20% 0% 50% ACN 100% ACN 100% ACN + SC 100% ACN + SC + 50 C Red swamp crayfish tissue was used for method development; 100 ng of each analyte was spiked into 50 mg freeze-dried tissue mass overnight before extraction with 5 ml solvent; extraction was conducted in triplicate; SC, sonication. 13

14 Mean recovery Recovery of estrogens and UV-filters in tissue with QuEChERS extraction at different initial ACN content 140% 120% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS 100% 80% 60% 40% 20% 0% 67% ACN 50% ACN 33% ACN 17% ACN 100 ng of each analyte was spiked into 50 mg freeze-dried tissue mass overnight before extraction with 10 ml solvent; extraction was conducted in triplicate. 14

15 Mean recovery Estrogens and UV-filters were effectively recovered using the modified QuEChERS extraction strategy 140% 120% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS 100% 80% 60% 40% 20% 0% 50 mg 150 mg 500 mg 100 ng of each analyte was spiked into 50, 150, and 500 mg freeze-dried tissue mass overnight before extracteionwith 5 ml ACN and 5 ml water; extraction was conducted in triplicate. 15

16 Part II (b): Cleanup with a novel reverse-solid-phase extraction (reverse-spe) process 16

17 The dispersive-spe idea was adopted and further developed as reverse-spe Extract dispersive-spe cleanup protocol reverse-spe strategy 17

18 Mean recovery Mean recovery Recovery of estrogens and UV-filters in 5 ml ACN through different cartridges during cleanup 140% 120% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS 100% 80% 60% 40% 20% 0% 140% 120% 100% 80% 60% 40% 20% 0% HLB (60mg, 3cc) HLB (150mg, 6cc) HLB (500mg, 6cc) Sep-pak C18 (200mg, 3cc) Isolute C18 (100mg, 3cc) Isolute C18 (500mg, 3cc) The concentration was 10 µg/l in 5 ml ACN for each target compound; flow rate was by gravity. 18

19 Mean recovery Isolute C18 (100 mg, 3 cc) cartridge provides acceptable recovery of estrogens and UV-filters 120% 100% 80% Avg. 83% 3rd Eluate (1 ml) 60% 40% 20% 2rd Eluate (1 ml) Extract (2.5 ml) 0% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS The spiked concentration was 10 µg/l in the 2.5 ml extract for each target; flow rate was by gravity. 19

20 Mean recovery HLB (60 mg, 3 cc) cartridge provides better recovery for estrogens and UV-filters 120% 100% 80% Avg. 97% 3rd Eluate (1 ml) 60% 40% 20% 2rd Eluate (1 ml) Extract (2.5 ml) 0% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS The spiked concentration was 10 µg/l in the 2.5 ml extract for each target compound; flow rate was by gravity. 20

21 Matrix effects Reverse-SPE with HLB decreased matrix effects (ME) ME = R SO R O R S 1 R so - response of the spiked analyte in the sample extract; R o - response of the unspiked sample extract; R s - response of the spiked analyte in the mobile phase. 40% 20% EE2-d4 E2-d3 BP-3-d5 4-MBC-d4 OC-d15 EHMC-d15 HMS-d4 Extract after salting out 0% -20% -40% -60% Eluate for N 2 evaporation -80% -100% -120% HLB (60 mg, 3 cc) No cartridge The seven internal standards were spiked at 5 µg/l in the reconstitution solution. 21

22 Part III: Occurrence of estrogens and UV-filters in the aquatic and marine environment in Maryland 22

23 Crayfish and oyster sampling strategies Collected from a watershed Used electric shock Randomly picked 20 crayfish Collected from the Chesapeake bay Used hydraulic dredge Randomly picked 3 oysters 23

24 etection of estrogens and UV-filters in (a) stream water, (b) sediment, and (c) crayfish from the Gwynns Falls watershed (Baltimore, M) (a) (b)* (c) BARN, Baisman Run; R1-5, ead Run Sites 1-5; RKR, ead Run at Franklintown; * HMS was not included for the analysis of sediment samples. 24

25 etection of estrogens and UV-filters in (a) seawater, (b) sediment, and (c) oysters from the Chesapeake Bay Nicholas Point Horn Point Manokin B Holland Straits Manokin A Kitts Creek 25

26 Conclusions Estrogens and UV-filters were simultaneously analyzed in LC- MS/MS using wrong-way-round ionization behavior; Low detection limits (i.e., ng/g) were achieved by processing a small sample size (i.e., 50 mg) with a modified QuEChERS protocol followed by a novel reverse-spe cleanup; All eight target analytes were detected at least once in the tissue samples, with the highest concentration being 399 ng/g homosalate in O. virilis; and, Given the high detection frequency of these CECs, it is important to investigate their potential impacts on invertebrates. 26

27 Acknowledgements CBEE colleagues, UMBC Kiranmayi Mangalgiri, Utsav Shashvatt, John Kemper, and Elvis Andino Nolasco for their help in crayfish sample collection; Ethan Hain for his help in sample preparation and analysis. epartment of Natural Resources, Maryland Mitchell Tarnowski and his colleagues for their help in oyster sample collection. US epartment of Agriculture Forest Service USA Forest Service project #

28 Thank you! 28

29 Supplementary slides 29

30 Mean recovery Mean recovery Impact of salt conditions on recovery of estrogens and UV-filters 140% 120% EE2 E2 E1 BP-3 4-MBC OC EHMC HMS 100% 80% 60% 40% 20% 0% 140% 2 g MgSO g NaCl 2 g MgSO4 + 1 g NaCl 2 g MgSO g NaCl 120% 100% 80% 60% 40% 20% 0% 2.5 g MgSO g NaCl 2.5 g MgSO4 + 1 g NaCl 2.5 g MgSO g NaCl 100 ng of each analyte was spiked into 50 mg freeze-dried tissue mass overnight before extraction with 5 ml ACN and 5 ml water; extraction was conducted in triplicate. 30

31 Mean recovery Recovery of UV-filters and estrogens in water, sediment, and tissue samples. 140% 120% EE2 E2 E1 BP-3 4-MBC OC OCX HMS 100% 80% 60% 40% 20% 0% Water Tissue Sediment 31