Research article Received: 14 July 2014
Revised: 3 September 2014
Accepted: 4 September 2014
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/jms.3503
Development and validation of a highresolution LTQ Orbitrap MS method for the quantification of isoflavones in wastewater effluent Michael G. Cahill,a Serena Logrippo,b Brian A. Dineen,a Kevin J. Jamesa and Giovanni Capriolib* Isoflavones and coumestranes are the most important classes of compounds among phytoestrogens; by binding to estrogen receptors, they mimic or modulate the effect on the endogenous receptors. Little information can be found in literature about the presence of isoflavones and coumestrol in the environment, even if it is known that this may have significance, being these substances classified as endocrine disrupting compounds. In this research, we aim to explore the capabilities of the LTQ Orbitrap Discovery hybrid MS in full-scan acquisition mode, with high resolution, to validate an analytical method for the quantification of nine isoflavones (genistein, genistin, glycitein, daidzein, daidzin, (R,S)-equol, biochanin A, formononetin and coumestrol) in wastewater samples. The correlation coefficients of calibration curves of the nine analyzed compounds were in a range of 0.996–0.999; recoveries at two different levels of concentration (0.05 and 0.5 μg/l) were in the range 73–98%, and the limits of detection ranged between 0.0014 and 0.017 μg/l, proving that this method is sensitive enough in comparison with other methods available in literature. This method has been applied for the analysis of 20 wastewater treatment plants in County Cork, Ireland. Copyright © 2015 John Wiley & Sons, Ltd. Additional supporting information may be found in the online version of this article at the publisher’s web site. Keywords: isoflavones; phytoestrogens; LTQ Orbitrap MS; high-resolution MS
Introduction
112
Isoflavones and coumestranes are the most important compounds among phytoestrogens, and they possess higher relative estrogenic activities than lignans.[1] Isoflavones have a structure similar to that of 17β-estradiol[2] and are involved in regulation of plant growth.[3] By binding to estrogen receptors, they mimic or modulate the effect on the endogenous receptors. Studies have shown that daidzein is the isoflavone with the highest biological activity, followed by genistein.[4,5] They are present not only in high concentrations in soy, clover, beans and peas[6] but also in other legumes such as lentils and chickpeas. Isoflavones have been found to have health benefits in agerelated and hormone-dependent diseases, including cancers, menopausal symptoms, cardiovascular diseases and osteoporosis.[7] However, studies of soy isoflavones in experimental animals suggest possible adverse effects, e.g. enhancement of reproductive organ cancer and modulation of endocrine function and antithyroid effects.[8] A study of North American men showed that high soy intake has been correlated with low sperm counts.[9] Because isoflavones are excreted, not only by plants but also by humans through food consumption, it is important to determine their impact in the environment.[9] In fact, Bacaloni et al were the only persons that analyzed isoflavones and coumestrol in river water and domestic water by using high-performance liquid chromatography (HPLC) electrospray
J. Mass Spectrom. 2015, 50, 112–116
ionization MS/MS with triple quadrupole as mass analyzer. They reported concentration levels of isoflavones ranging from 0.012 to 0.454 μg/l in wastewater sewage treatment plant influent samples from Genzano and Albano in Italy; meanwhile, in effluent water and river water, the analytes were present at low concentration.[10] LC ultraviolet detection has been the method of choice for isoflavone analysis for many years.[11] LC-MS offers improved selectivity and sensitivity and is a robust analytical technique for environmental analysis.[12–14] Almost all published methods for the determination of isoflavones in environmental water samples employ solid-phase extraction (SPE) as the preconcentration technique.[13] The sorbent employed for SPE is usual a copolymer, poly(divinylbenzene-co-N-vinylpyrrolidone).[10] In this research, we aim to explore the capabilities of the LTQ Orbitrap Discovery hybrid MS in full-scan acquisition mode, with high resolution, a technique that has great advantages in
* Correspondence to: Giovanni Caprioli, University of Camerino, School of Pharmacy, via S. Agostino 1, 62032 Camerino, Italy. E-mail: [email protected] a Environmental Research Institute, University College Cork, Lee Road, Cork, Ireland b School of Pharmacy, University of Camerino, via S. Agostino 1, 62032 Camerino, Italy
Copyright © 2015 John Wiley & Sons, Ltd.
Orbitrap method for isoflavones in wastewater comparison to other mass spectrometric technologies[15,16] such as high mass accuracy and resolution,[17] in order to evaluate the performance characteristics required for quantitative analysis of trace levels of isoflavones in wastewater samples. To our best knowledge, this is the first time to use LTQ Orbitrap for the analysis of isoflavones in wastewater effluent samples, proving that this method is sensitive enough in comparison with other methods available in literature.[10,13]
Experimental
LC conditions Chromatographic separations were carried out using an LC Quaternary analytical pump (Accela, Thermo Scientific, Hemel Hempstead, UK), equipped with a reversed-phase analytical column Agilent Eclipse XDB-C18 (2.1× 150 mm internal diameter 5 μm, APEX, Dublin Ireland) maintained at 35 °C. The injected sample volume was 10 μl. The gradient started with acetonitrile/water (10/90 v/v) containing 0.01% TFA and increased acetonitrile/water (90/10 v/v) in 12 min. At 12 min, the initial conditions were replaced, and the column was reequilibrated for 4 min before the next injection. The flow rate during analysis was 200 μl/min.
Materials and standards Purchased chemicals included the standards genistein, genistin, glycitein, daidzein, daidzin and (R,S)-equol, from LC Laboratories (Woburn, MA 01801, USA). Biochanin A, formononetin and coumestrol were purchased from Sigma-Aldrich (Dublin). Deuterated standards diadzein-d6 and genistein-d4 were purchased from Toronto Research Chemicals (Canada). Diadzein-d6 was used as internal standard for biochanin A, coumestrol, daidzein and daidzin; meanwhile, genistein-d4 was used as internal standard for equol, formononetin, glycitein, genistein and genistin. All solvents including HPLC-grade acetonitrile, methanol and water were purchased from Thermo Fisher (Dublin). Trifluoroacetic acid (TFA) LC-spectrograde was purchased from Sigma-Aldrich (Dublin). Collection of samples Wastewater treatment plant (WWTP) #1 represents a wastewater treatment facility in a heavily industrialized region with a high population equivalent. The WWTPs, #2, #3, #5, #6 and #7, represent agricultural regions in County Cork with relatively low population equivalents. WWTP #4 represents a large town near Cork city with an agricultural hinterland and a high population equivalent. Grab samples were taken from all the WWTPs except for WWTP #1. Cork County Council operates a refrigerated composite sampler on the primary discharge outlet from the WWTP #1 outfall to the sea. This sampler operates on a 28-day rolling cycle and collects samples on a 24-h basis per sample. A total of 20 samples collected between September 2013 and May 2014 have been analyzed. Sample preparation
J. Mass Spectrom. 2015, 50, 112–116
The LC system was connected to a hybrid linear ion trap (LIT) MS–Fourier transform (FT) mass spectrometer (LTQ Orbitrap Discovery equipped with higher collisional dissociation cell, Thermo Scientific). A heated atmospheric pressure chemical ionization (H-APCI) source was used, operating in positive ionization mode. Instrument settings were as follows: vaporization temperature 370 °C, sheath gas 35 arbitrary units, auxiliary gas 20 arbitrary units, sweep gas 0, discharge current 4 kV, capillary temperature 300 °C, capillary voltage 8 kV and a tube lens of 80. All ion source tune parameters were optimized using chromatography. The LTQ Orbitrap MS was calibrated using a solution containing caffeine, L-methionyl-arginyl-phenylalanyl-alanine acetate and Ultramark 1621, according to the manufacturer’s instructions. For the determination of isoflavones in wastewater, the method consisted of three scan events: (1) full-scan FTMS (LTQ Orbitrap Discovery), (2) full-scan LIT MS method and (3) LIT MS/MS datadependent scan. Full-scan MS spectra were acquired as profile data using an isolation window of ±2 mmu, and data-dependent MS/MS spectra were acquired with optimized relative collision energy (RCE) of centroid mode (genistein 28% RCE, genistin 20% RCE, glycitein 30% RCE, daidzein 28% RCE, daidzin 22% RCE, (R,S)-equol 30% RCE, biochanin A 30% RCE, formononetin 28% RCE and coumestrol 30% RCE). The mass resolution setting was 30 000 [full width at half maximum (FWHM)] with a mass range of 100–450 m/z in the Orbitrap analyzer. The total cycle time depends upon the resolution, and at a resolution value of 30 000 FWHM, the total cycle time is about 0.4 s. The lock mass option was employed by recalibration of the mass scale using the m/z values of the ion, [M + Na]+ from TFA (158.964030). Operation of the entire LC-MS instrumentation was controlled using Xcalibur software 2.0.7. (Thermo Scientific).
Results and discussions Analytical performance The analytical characteristics of the developed methods were investigated, including their linearity, limits of quantitation (LOQs), limits of detection (LODs), accuracy and precision and reproducibility to evaluate their efficacy for application to the analysis of environmental wastewater samples. Table 1 summarizes the detection limits, accuracy determination and recoveries. The linearity of the method for nine isoflavones and two internal standards (genistein-d4 and daidzein-d6) was studied by analyzing spiked wastewater samples. Linearity for the nine isoflavones was calculated in the linear range of 0.0025–0.75 μg/l (variable according to each different compound) by using the ratio
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms
113
Wastewater samples were collected by Cork County Council in amber brown bottles (2 l) and immediately transported to the laboratory within a 4-h period. All samples were filtered (0.45-μm nylon filters, Whatman) and stored at 4 °C. The internal standards, diadzein-d6 and genistein-d4, were added to the filtered wastewater samples to obtain a target fixed level of 0.5 μg/l for each. The SPE was carried out using an Oasis hydrophobic–lipophilic balance 3-cc extraction cartridge (60 mg, 3 ml; Waters, Dublin). The SPE cartridge was conditioned with methanol (3 ml) followed by water (6 ml). The samples (500 ml) were loaded onto each cartridge using a vacuum pump. The cartridge was washed using water (5 ml) and eluted using methanol (8 ml). The eluent was evaporated to dryness under nitrogen using a TurboVap LV evaporator (Zymark, Caliper Technologies, Russelsheim, Germany) and reconstituted in water/methanol (90/10 v/v; 0.5 ml, thus ensuring that the analytes are dissolved), filtered (0.45-μm nylon membrane) and transferred to an amber glass vial before injection.
Mass spectrometry optimization and calibration
M. G. Cahill et al. Table 1. Limit of detection (LOD) and limit of quantitation (LOQ), accuracy determination (n = 5) and recoveries (n = 3) for isoflavones in wastewater samples Accuracy determination
Recoveries
Compounds
LOQ (μg/l)
LOD (μg/l)
True concentrations (spiked level in μg/l)
Measured concentrations (μg/l)
% relative error
% RSD
Spiked concentration (μg/L)
% recovery
% RSD
Glycitein
0.007
0.002
Biochanin A
0.0045
0.0015
Genistein
0.004
0.0014
Daidzein
0.0046
0.0015
Formononetin
0.0087
0.0029
Equol
0.01
0.003
Genistin
0.051
0.017
Daidzin
0.049
0.016
Coumestrol
0.0088
0.0029
0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05
0.55 0.049 0.51 0.048 0.50 0.05 0.51 0.047 0.51 0.047 0.53 0.047 0.46 0.055 0.49 0.048 0.49 0.049
9.5 1.6 2.0 3.3 0.42 0.13 1.56 4.1 2.8 6.3 6.1 5.0 1.2 9.5 1.0 3.7 1.5 1.3
6.6 5.0 8.0 6.1 8.1 5.4 5.3 5.9 3.0 4.3 8.0 6.5 8.4 1.3 4.3 5.6 2.6 1.2
0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05
81.0 83.5 87.3 84.7 73.0 74.9 95.8 94.3 94.6 98.1 90.1 92.2 79.4 84.3 87.0 92.2 82.0 85.6
5.9 5.2 5.9 8.0 2.7 3.0 3.2 4.4 3.8 6.9 0.21 7.1 5.2 5.9 3.1 7.2 1.4 7.9
Data includes true concentrations and measured concentrations with relative error and relative standard deviation and recoveries with relative standard deviation for nine isoflavones at two different levels of concentration.
114
wileyonlinelibrary.com/journal/jms
3000000 8
2500000
Relative abundance
with the internal standards’ area of diadzein-d6 and genistein-d4 (that were added to the filtered wastewater samples to obtain a target fixed level of 0.5 μg/l for each). Diadzein-d6 was used as internal standard for biochanin A, coumestrol, daidzein and daidzin; meanwhile, genistein-d4 was used as internal standard for equol, formononetin, glycitein, genistein and genistin. Satisfactory linearity was achieved for all analytes, with the correlation coefficient (r2) better than 0.996. The LOQs were in the range measured at 0.004–0.05 μg/l, and the LODs for this method were in the range 0.0015–0.17 μg/l (Table 1). The accuracy of this method was expressed using percentage relative error at two spiked concentration levels, 0.5 and 0.05 μg/l, over 5 days. The percentage relative errors for all isoflavones in wastewater were satisfactory and within the range 0.13–9.5% (Table 1). The interday method precision was also satisfactory as expressed by the % relative standard deviation (RSD) values (1.2–8.1%) for all isoflavones in wastewater (Table 1). Stability of the retention time for each isoflavone in wastewater was examined over a 5-day period (n = 25). The retention times using this method were stable over a 5-day period, with an RSD of ≤0.17%. In Fig. 1 is an example of a mixture standard of the nine isoflavones spiked into a blank wastewater sample at a concentration of 0.1 μg/l. Figure 2 demonstrates the parallel acquisition of data of the LIT MS in the LTQ Orbitrap instrument, and in fact, it shows a dependent scan spectra of formononetin at 28% RCE spiked into a blank wastewater sample at a concentration of 0.1 μg/l. For matrix effect experiments, genistein standard solution at two different concentration levels (0.1 and 1 μg/l) was infused into the chromatographic eluate prior to entry into the H-APCI source of the LIT MS instrument. Two injections (10 μl) were performed and repeated for three times: (1) LC water and (2) blank wastewater
2000000
4 3
1500000
6 5 7
9
1000000 500000 1
2
0 0
2
4
6
8
10
12
Time (min)
Figure 1. HPLC-MS chromatograms for (1) daidzin, (2) genistin, (3) daidzein, (4) glycitein, (5) formononetin, (6) genistein, (7) equol, (8) coumestrol and (9) biochanin A, which were spiked into a ‘blank’ wastewater sample, each at a concentration level of 0.1 μg/l.
Figure 2. Dependent scan spectra of formononetin at 28% RCE spiked into a blank wastewater sample at a concentration of 0.1 μg/l.
Copyright © 2015 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2015, 50, 112–116
Orbitrap method for isoflavones in wastewater
Figure 3. Postcolumn infusion of genistein (1 μg/l) at 10 μl/min in (A) HLPC water and (B) blank wastewater sample.
sample. The genistein standard solution was introduced, using a 1-ml gastight syringe (Hamilton, Birmingham, UK) and a syringe pump (Harvard Apparatus, Holliston, MA, USA), at a flow rate 10 μl/min into the LC eluent, using a T-junction. The ion at m/z 271.1, corresponding to the [M + H]+ ion of genistein, was monitored for 11 min (Fig. 3). Figure 3 shows that the matrix analyzed had no significant effect on the response, either positive (ion enhancement) or negative (ion suppression), of the isoflavone genistein that was infused postcolumn. The experiment was repeated for all the nine isoflavones and gave the same results (data not shown). Recovery experiments for the SPE were determined by spiking known concentrations of all nine isoflavones into a blank wastewater sample at two concentrations, 0.5 and 0.05 μg/l (Table 1). The recovery of the isoflavones from wastewater was within the range 73–98% and the % RSD within the range 0.2–8%.
Accurate mass studies for isoflavones in wastewater samples In addition, mass measurement experiments were carried out by using the LTQ Orbitrap MS and spiked wastewater samples at concentrations of 0.5 and 0.05 μg/l. The accurate mass assigned to the [M + H]+ ions of each of the nine analytes in wastewater samples in each acquired mass spectrum was measured. Fourteen mass spectra of each analyte were selected to provide a set of mass measurement data acquired at Orbitrap mass resolving powers of 30 000 FWHM and scan cycle times 0.25 s. The mean of the measured mass and mass error are shown in Table 2.
Table 2. Accurate mass studies for isoflavones Isoflavones Glycitein Biochanin A Genistein Daidzein Formononetin Equol Genistin Daidzin Coumestrol
Application of the full-scan MS method to real sample The developed high-resolution full-scan MS method was then applied to wastewater samples from seven WWTPs in County Cork, Ireland. Isoflavones were not detected in any of the 20 samples analyzed from seven WWTPs during September 2013–May 2014.
Conclusions The analysis of isoflavones has taken giant steps forward as a result of the application of MS techniques. This is the first LTQ Orbitrap method developed for the analysis of isoflavones, and it was successfully applied to effluent wastewater samples. Using this method, none of the isoflavones was identified in several WWTPs in County Cork, Ireland, at trace levels of 0.0014– 0.017 μg/l; therefore, the presence of isoflavones in wastewater does not pose a risk for the population in County Cork, Ireland. Acknowledgements This research was funded by the Higher Education Authority of Ireland, as part of Ireland’s EU Structural Funds Programmes (2007–2013) and the European Regional Development Fund, and Programme for Research in Third-Level Institutions (PRTLI-4), Environment and Climate Change: Impacts and Responses.
References
Measured mass
Theoretical mass
285.0759 285.0758 271.0602 255.0653 269.0809 243.1017 433.1127 417.1177 269.0445
285.0762 285.0762 271.0606 255.0657 269.0813 243.1021 433.1134 417.1186 269.0449
ppm 1.4 1.75 1.66 1.7 1.8 1.72 1.78 2.06 1.85
[1] Z. Liu, Y. Kanjo, S. Mizutani. A review of phytoestrogens: their occurrence and fate in the environment. Wat. Res. 2010, 22, 567. [2] C. F. Skibola, M. T. Smith. Potential health impacts of excessive flavonoid intake. Free Radical Biol. Med. 2000, 29, 375. [3] G. Kuhnle, C. Dell’Aquila, S. Aspinall, S. Runswick, M. Joosen, A. Mulligan, S. Bingham. Phytoestrogen content of fruits and vegetables commonly consumed in the UK based on LC-MS and 13C-labelled standards. Food Chem. 2009, 116, 542. [4] N. Konar, E. S. Poyrazoglu, K. Demir, N. Artik. Determination of conjugated and free isoflavones in some legumes by LC/MS-MS. J. Food Compos. Anal. 2012, 25, 173. [5] M. M. Delgado-Zamarreño, L. Perez-Martin, M. Bustamante-Rangel, R. Carabias-Martinez. A modified QuErChers method as a sample treatment before the determination of isoflavones in foods by ultra-performance liquid chromatography- triple quadrupole mass spectrometry. Talanta 2012, 100, 320.
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms
115
Measured mass, theoretical mass and error values measured in ppm are reported.
J. Mass Spectrom. 2015, 50, 112–116
Mass error values were
Revised: 3 September 2014
Accepted: 4 September 2014
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/jms.3503
Development and validation of a highresolution LTQ Orbitrap MS method for the quantification of isoflavones in wastewater effluent Michael G. Cahill,a Serena Logrippo,b Brian A. Dineen,a Kevin J. Jamesa and Giovanni Capriolib* Isoflavones and coumestranes are the most important classes of compounds among phytoestrogens; by binding to estrogen receptors, they mimic or modulate the effect on the endogenous receptors. Little information can be found in literature about the presence of isoflavones and coumestrol in the environment, even if it is known that this may have significance, being these substances classified as endocrine disrupting compounds. In this research, we aim to explore the capabilities of the LTQ Orbitrap Discovery hybrid MS in full-scan acquisition mode, with high resolution, to validate an analytical method for the quantification of nine isoflavones (genistein, genistin, glycitein, daidzein, daidzin, (R,S)-equol, biochanin A, formononetin and coumestrol) in wastewater samples. The correlation coefficients of calibration curves of the nine analyzed compounds were in a range of 0.996–0.999; recoveries at two different levels of concentration (0.05 and 0.5 μg/l) were in the range 73–98%, and the limits of detection ranged between 0.0014 and 0.017 μg/l, proving that this method is sensitive enough in comparison with other methods available in literature. This method has been applied for the analysis of 20 wastewater treatment plants in County Cork, Ireland. Copyright © 2015 John Wiley & Sons, Ltd. Additional supporting information may be found in the online version of this article at the publisher’s web site. Keywords: isoflavones; phytoestrogens; LTQ Orbitrap MS; high-resolution MS
Introduction
112
Isoflavones and coumestranes are the most important compounds among phytoestrogens, and they possess higher relative estrogenic activities than lignans.[1] Isoflavones have a structure similar to that of 17β-estradiol[2] and are involved in regulation of plant growth.[3] By binding to estrogen receptors, they mimic or modulate the effect on the endogenous receptors. Studies have shown that daidzein is the isoflavone with the highest biological activity, followed by genistein.[4,5] They are present not only in high concentrations in soy, clover, beans and peas[6] but also in other legumes such as lentils and chickpeas. Isoflavones have been found to have health benefits in agerelated and hormone-dependent diseases, including cancers, menopausal symptoms, cardiovascular diseases and osteoporosis.[7] However, studies of soy isoflavones in experimental animals suggest possible adverse effects, e.g. enhancement of reproductive organ cancer and modulation of endocrine function and antithyroid effects.[8] A study of North American men showed that high soy intake has been correlated with low sperm counts.[9] Because isoflavones are excreted, not only by plants but also by humans through food consumption, it is important to determine their impact in the environment.[9] In fact, Bacaloni et al were the only persons that analyzed isoflavones and coumestrol in river water and domestic water by using high-performance liquid chromatography (HPLC) electrospray
J. Mass Spectrom. 2015, 50, 112–116
ionization MS/MS with triple quadrupole as mass analyzer. They reported concentration levels of isoflavones ranging from 0.012 to 0.454 μg/l in wastewater sewage treatment plant influent samples from Genzano and Albano in Italy; meanwhile, in effluent water and river water, the analytes were present at low concentration.[10] LC ultraviolet detection has been the method of choice for isoflavone analysis for many years.[11] LC-MS offers improved selectivity and sensitivity and is a robust analytical technique for environmental analysis.[12–14] Almost all published methods for the determination of isoflavones in environmental water samples employ solid-phase extraction (SPE) as the preconcentration technique.[13] The sorbent employed for SPE is usual a copolymer, poly(divinylbenzene-co-N-vinylpyrrolidone).[10] In this research, we aim to explore the capabilities of the LTQ Orbitrap Discovery hybrid MS in full-scan acquisition mode, with high resolution, a technique that has great advantages in
* Correspondence to: Giovanni Caprioli, University of Camerino, School of Pharmacy, via S. Agostino 1, 62032 Camerino, Italy. E-mail: [email protected] a Environmental Research Institute, University College Cork, Lee Road, Cork, Ireland b School of Pharmacy, University of Camerino, via S. Agostino 1, 62032 Camerino, Italy
Copyright © 2015 John Wiley & Sons, Ltd.
Orbitrap method for isoflavones in wastewater comparison to other mass spectrometric technologies[15,16] such as high mass accuracy and resolution,[17] in order to evaluate the performance characteristics required for quantitative analysis of trace levels of isoflavones in wastewater samples. To our best knowledge, this is the first time to use LTQ Orbitrap for the analysis of isoflavones in wastewater effluent samples, proving that this method is sensitive enough in comparison with other methods available in literature.[10,13]
Experimental
LC conditions Chromatographic separations were carried out using an LC Quaternary analytical pump (Accela, Thermo Scientific, Hemel Hempstead, UK), equipped with a reversed-phase analytical column Agilent Eclipse XDB-C18 (2.1× 150 mm internal diameter 5 μm, APEX, Dublin Ireland) maintained at 35 °C. The injected sample volume was 10 μl. The gradient started with acetonitrile/water (10/90 v/v) containing 0.01% TFA and increased acetonitrile/water (90/10 v/v) in 12 min. At 12 min, the initial conditions were replaced, and the column was reequilibrated for 4 min before the next injection. The flow rate during analysis was 200 μl/min.
Materials and standards Purchased chemicals included the standards genistein, genistin, glycitein, daidzein, daidzin and (R,S)-equol, from LC Laboratories (Woburn, MA 01801, USA). Biochanin A, formononetin and coumestrol were purchased from Sigma-Aldrich (Dublin). Deuterated standards diadzein-d6 and genistein-d4 were purchased from Toronto Research Chemicals (Canada). Diadzein-d6 was used as internal standard for biochanin A, coumestrol, daidzein and daidzin; meanwhile, genistein-d4 was used as internal standard for equol, formononetin, glycitein, genistein and genistin. All solvents including HPLC-grade acetonitrile, methanol and water were purchased from Thermo Fisher (Dublin). Trifluoroacetic acid (TFA) LC-spectrograde was purchased from Sigma-Aldrich (Dublin). Collection of samples Wastewater treatment plant (WWTP) #1 represents a wastewater treatment facility in a heavily industrialized region with a high population equivalent. The WWTPs, #2, #3, #5, #6 and #7, represent agricultural regions in County Cork with relatively low population equivalents. WWTP #4 represents a large town near Cork city with an agricultural hinterland and a high population equivalent. Grab samples were taken from all the WWTPs except for WWTP #1. Cork County Council operates a refrigerated composite sampler on the primary discharge outlet from the WWTP #1 outfall to the sea. This sampler operates on a 28-day rolling cycle and collects samples on a 24-h basis per sample. A total of 20 samples collected between September 2013 and May 2014 have been analyzed. Sample preparation
J. Mass Spectrom. 2015, 50, 112–116
The LC system was connected to a hybrid linear ion trap (LIT) MS–Fourier transform (FT) mass spectrometer (LTQ Orbitrap Discovery equipped with higher collisional dissociation cell, Thermo Scientific). A heated atmospheric pressure chemical ionization (H-APCI) source was used, operating in positive ionization mode. Instrument settings were as follows: vaporization temperature 370 °C, sheath gas 35 arbitrary units, auxiliary gas 20 arbitrary units, sweep gas 0, discharge current 4 kV, capillary temperature 300 °C, capillary voltage 8 kV and a tube lens of 80. All ion source tune parameters were optimized using chromatography. The LTQ Orbitrap MS was calibrated using a solution containing caffeine, L-methionyl-arginyl-phenylalanyl-alanine acetate and Ultramark 1621, according to the manufacturer’s instructions. For the determination of isoflavones in wastewater, the method consisted of three scan events: (1) full-scan FTMS (LTQ Orbitrap Discovery), (2) full-scan LIT MS method and (3) LIT MS/MS datadependent scan. Full-scan MS spectra were acquired as profile data using an isolation window of ±2 mmu, and data-dependent MS/MS spectra were acquired with optimized relative collision energy (RCE) of centroid mode (genistein 28% RCE, genistin 20% RCE, glycitein 30% RCE, daidzein 28% RCE, daidzin 22% RCE, (R,S)-equol 30% RCE, biochanin A 30% RCE, formononetin 28% RCE and coumestrol 30% RCE). The mass resolution setting was 30 000 [full width at half maximum (FWHM)] with a mass range of 100–450 m/z in the Orbitrap analyzer. The total cycle time depends upon the resolution, and at a resolution value of 30 000 FWHM, the total cycle time is about 0.4 s. The lock mass option was employed by recalibration of the mass scale using the m/z values of the ion, [M + Na]+ from TFA (158.964030). Operation of the entire LC-MS instrumentation was controlled using Xcalibur software 2.0.7. (Thermo Scientific).
Results and discussions Analytical performance The analytical characteristics of the developed methods were investigated, including their linearity, limits of quantitation (LOQs), limits of detection (LODs), accuracy and precision and reproducibility to evaluate their efficacy for application to the analysis of environmental wastewater samples. Table 1 summarizes the detection limits, accuracy determination and recoveries. The linearity of the method for nine isoflavones and two internal standards (genistein-d4 and daidzein-d6) was studied by analyzing spiked wastewater samples. Linearity for the nine isoflavones was calculated in the linear range of 0.0025–0.75 μg/l (variable according to each different compound) by using the ratio
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms
113
Wastewater samples were collected by Cork County Council in amber brown bottles (2 l) and immediately transported to the laboratory within a 4-h period. All samples were filtered (0.45-μm nylon filters, Whatman) and stored at 4 °C. The internal standards, diadzein-d6 and genistein-d4, were added to the filtered wastewater samples to obtain a target fixed level of 0.5 μg/l for each. The SPE was carried out using an Oasis hydrophobic–lipophilic balance 3-cc extraction cartridge (60 mg, 3 ml; Waters, Dublin). The SPE cartridge was conditioned with methanol (3 ml) followed by water (6 ml). The samples (500 ml) were loaded onto each cartridge using a vacuum pump. The cartridge was washed using water (5 ml) and eluted using methanol (8 ml). The eluent was evaporated to dryness under nitrogen using a TurboVap LV evaporator (Zymark, Caliper Technologies, Russelsheim, Germany) and reconstituted in water/methanol (90/10 v/v; 0.5 ml, thus ensuring that the analytes are dissolved), filtered (0.45-μm nylon membrane) and transferred to an amber glass vial before injection.
Mass spectrometry optimization and calibration
M. G. Cahill et al. Table 1. Limit of detection (LOD) and limit of quantitation (LOQ), accuracy determination (n = 5) and recoveries (n = 3) for isoflavones in wastewater samples Accuracy determination
Recoveries
Compounds
LOQ (μg/l)
LOD (μg/l)
True concentrations (spiked level in μg/l)
Measured concentrations (μg/l)
% relative error
% RSD
Spiked concentration (μg/L)
% recovery
% RSD
Glycitein
0.007
0.002
Biochanin A
0.0045
0.0015
Genistein
0.004
0.0014
Daidzein
0.0046
0.0015
Formononetin
0.0087
0.0029
Equol
0.01
0.003
Genistin
0.051
0.017
Daidzin
0.049
0.016
Coumestrol
0.0088
0.0029
0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05
0.55 0.049 0.51 0.048 0.50 0.05 0.51 0.047 0.51 0.047 0.53 0.047 0.46 0.055 0.49 0.048 0.49 0.049
9.5 1.6 2.0 3.3 0.42 0.13 1.56 4.1 2.8 6.3 6.1 5.0 1.2 9.5 1.0 3.7 1.5 1.3
6.6 5.0 8.0 6.1 8.1 5.4 5.3 5.9 3.0 4.3 8.0 6.5 8.4 1.3 4.3 5.6 2.6 1.2
0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5 0.05
81.0 83.5 87.3 84.7 73.0 74.9 95.8 94.3 94.6 98.1 90.1 92.2 79.4 84.3 87.0 92.2 82.0 85.6
5.9 5.2 5.9 8.0 2.7 3.0 3.2 4.4 3.8 6.9 0.21 7.1 5.2 5.9 3.1 7.2 1.4 7.9
Data includes true concentrations and measured concentrations with relative error and relative standard deviation and recoveries with relative standard deviation for nine isoflavones at two different levels of concentration.
114
wileyonlinelibrary.com/journal/jms
3000000 8
2500000
Relative abundance
with the internal standards’ area of diadzein-d6 and genistein-d4 (that were added to the filtered wastewater samples to obtain a target fixed level of 0.5 μg/l for each). Diadzein-d6 was used as internal standard for biochanin A, coumestrol, daidzein and daidzin; meanwhile, genistein-d4 was used as internal standard for equol, formononetin, glycitein, genistein and genistin. Satisfactory linearity was achieved for all analytes, with the correlation coefficient (r2) better than 0.996. The LOQs were in the range measured at 0.004–0.05 μg/l, and the LODs for this method were in the range 0.0015–0.17 μg/l (Table 1). The accuracy of this method was expressed using percentage relative error at two spiked concentration levels, 0.5 and 0.05 μg/l, over 5 days. The percentage relative errors for all isoflavones in wastewater were satisfactory and within the range 0.13–9.5% (Table 1). The interday method precision was also satisfactory as expressed by the % relative standard deviation (RSD) values (1.2–8.1%) for all isoflavones in wastewater (Table 1). Stability of the retention time for each isoflavone in wastewater was examined over a 5-day period (n = 25). The retention times using this method were stable over a 5-day period, with an RSD of ≤0.17%. In Fig. 1 is an example of a mixture standard of the nine isoflavones spiked into a blank wastewater sample at a concentration of 0.1 μg/l. Figure 2 demonstrates the parallel acquisition of data of the LIT MS in the LTQ Orbitrap instrument, and in fact, it shows a dependent scan spectra of formononetin at 28% RCE spiked into a blank wastewater sample at a concentration of 0.1 μg/l. For matrix effect experiments, genistein standard solution at two different concentration levels (0.1 and 1 μg/l) was infused into the chromatographic eluate prior to entry into the H-APCI source of the LIT MS instrument. Two injections (10 μl) were performed and repeated for three times: (1) LC water and (2) blank wastewater
2000000
4 3
1500000
6 5 7
9
1000000 500000 1
2
0 0
2
4
6
8
10
12
Time (min)
Figure 1. HPLC-MS chromatograms for (1) daidzin, (2) genistin, (3) daidzein, (4) glycitein, (5) formononetin, (6) genistein, (7) equol, (8) coumestrol and (9) biochanin A, which were spiked into a ‘blank’ wastewater sample, each at a concentration level of 0.1 μg/l.
Figure 2. Dependent scan spectra of formononetin at 28% RCE spiked into a blank wastewater sample at a concentration of 0.1 μg/l.
Copyright © 2015 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2015, 50, 112–116
Orbitrap method for isoflavones in wastewater
Figure 3. Postcolumn infusion of genistein (1 μg/l) at 10 μl/min in (A) HLPC water and (B) blank wastewater sample.
sample. The genistein standard solution was introduced, using a 1-ml gastight syringe (Hamilton, Birmingham, UK) and a syringe pump (Harvard Apparatus, Holliston, MA, USA), at a flow rate 10 μl/min into the LC eluent, using a T-junction. The ion at m/z 271.1, corresponding to the [M + H]+ ion of genistein, was monitored for 11 min (Fig. 3). Figure 3 shows that the matrix analyzed had no significant effect on the response, either positive (ion enhancement) or negative (ion suppression), of the isoflavone genistein that was infused postcolumn. The experiment was repeated for all the nine isoflavones and gave the same results (data not shown). Recovery experiments for the SPE were determined by spiking known concentrations of all nine isoflavones into a blank wastewater sample at two concentrations, 0.5 and 0.05 μg/l (Table 1). The recovery of the isoflavones from wastewater was within the range 73–98% and the % RSD within the range 0.2–8%.
Accurate mass studies for isoflavones in wastewater samples In addition, mass measurement experiments were carried out by using the LTQ Orbitrap MS and spiked wastewater samples at concentrations of 0.5 and 0.05 μg/l. The accurate mass assigned to the [M + H]+ ions of each of the nine analytes in wastewater samples in each acquired mass spectrum was measured. Fourteen mass spectra of each analyte were selected to provide a set of mass measurement data acquired at Orbitrap mass resolving powers of 30 000 FWHM and scan cycle times 0.25 s. The mean of the measured mass and mass error are shown in Table 2.
Table 2. Accurate mass studies for isoflavones Isoflavones Glycitein Biochanin A Genistein Daidzein Formononetin Equol Genistin Daidzin Coumestrol
Application of the full-scan MS method to real sample The developed high-resolution full-scan MS method was then applied to wastewater samples from seven WWTPs in County Cork, Ireland. Isoflavones were not detected in any of the 20 samples analyzed from seven WWTPs during September 2013–May 2014.
Conclusions The analysis of isoflavones has taken giant steps forward as a result of the application of MS techniques. This is the first LTQ Orbitrap method developed for the analysis of isoflavones, and it was successfully applied to effluent wastewater samples. Using this method, none of the isoflavones was identified in several WWTPs in County Cork, Ireland, at trace levels of 0.0014– 0.017 μg/l; therefore, the presence of isoflavones in wastewater does not pose a risk for the population in County Cork, Ireland. Acknowledgements This research was funded by the Higher Education Authority of Ireland, as part of Ireland’s EU Structural Funds Programmes (2007–2013) and the European Regional Development Fund, and Programme for Research in Third-Level Institutions (PRTLI-4), Environment and Climate Change: Impacts and Responses.
References
Measured mass
Theoretical mass
285.0759 285.0758 271.0602 255.0653 269.0809 243.1017 433.1127 417.1177 269.0445
285.0762 285.0762 271.0606 255.0657 269.0813 243.1021 433.1134 417.1186 269.0449
ppm 1.4 1.75 1.66 1.7 1.8 1.72 1.78 2.06 1.85
[1] Z. Liu, Y. Kanjo, S. Mizutani. A review of phytoestrogens: their occurrence and fate in the environment. Wat. Res. 2010, 22, 567. [2] C. F. Skibola, M. T. Smith. Potential health impacts of excessive flavonoid intake. Free Radical Biol. Med. 2000, 29, 375. [3] G. Kuhnle, C. Dell’Aquila, S. Aspinall, S. Runswick, M. Joosen, A. Mulligan, S. Bingham. Phytoestrogen content of fruits and vegetables commonly consumed in the UK based on LC-MS and 13C-labelled standards. Food Chem. 2009, 116, 542. [4] N. Konar, E. S. Poyrazoglu, K. Demir, N. Artik. Determination of conjugated and free isoflavones in some legumes by LC/MS-MS. J. Food Compos. Anal. 2012, 25, 173. [5] M. M. Delgado-Zamarreño, L. Perez-Martin, M. Bustamante-Rangel, R. Carabias-Martinez. A modified QuErChers method as a sample treatment before the determination of isoflavones in foods by ultra-performance liquid chromatography- triple quadrupole mass spectrometry. Talanta 2012, 100, 320.
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms
115
Measured mass, theoretical mass and error values measured in ppm are reported.
J. Mass Spectrom. 2015, 50, 112–116
Mass error values were
