For Selected Potent Endocrine Disrupting Steroids: Development and Application to Environmental Studies

Home , Estradiol acetate, Estrogenic substances, Ethylestradiol, Mestranol, Quinestrol

Rapid and Sensitive Enzyme-Linked Immunosorbent Assays (ELISAs) for Selected Potent Endocrine Disrupting Steroids: Development and Application to Environmental Studies

Chatchaporn Uraipong

Thesis submitted in partial fulfillment of the requirement for the Degree of Master of Science (Research)

School of Chemical Engineering The University of New South Wales September, 2010

ABSTRACT

Endocrine disrupting chemicals (EDCs) are chemicals that alter functions of the endocrine system and cause health effects in an intact organism, or progeny, or population, with reproductive, developmental, or carcinogenic consequences. In order to facilitate risk assessment of potential endocrine disrupting steroids that are present in ultra low concentrations in the Australian environment, there is a need to boost the analytical capacity for EDC detection. One strategy is to develop antibody-based techniques that can offer simple, cost-effective and reliable analysis with high throughput capacity and portability for real-time monitoring.

This thesis describes the design and synthesis of hapten molecules, raising of specific antibodies, formatting and characterising of a series of sensitive competitive Enzyme-Linked Immunosorbent Assays (ELISAs) for 17β-estradiol (E2), 17α-ethynylestradiol (EE2), ethylestradiol-3-methyl ether (mestranol) and testosterone (T), including validation of their performance as fast and effective water monitoring tools. Application of the developed assays to investigate the levels of the target EDCs in bodies of water and efficiency of water treatment plants in urban and rural areas in New South Wales, Australia, is also discussed.

17α-Ethynylestradiol and related synthetic estrogens, are active ingredients of contraceptive pills and hormone therapy, and have been identified as potent EDCs (Warner and Jenkins, 2007). In Chapter 3, the development of two ELISAs with varying specificity for the synthetic estrogens is described; 1) a highly sensitive ELISA for 17α-ethynylestradiol and mestranol, and 2) a mestranol specific ELISA. Highly specific antibodies was facilitated by synthesising haptens with the attachment of linkers with varying lengths at the C3 position (17α- ethynylestradiol-acetate hapten (EE2-ACT-KLH) and 17α-ethynylestradiol-butyrate hapten (EE2-BUT-KLH)). The optimised ELISAs in heterologous systems displayed high sensitivity, with the best assay exhibiting a limit of detection (LOD) of 70 ng L-1 EE2 in water (without preconcentration). The LOD of the ELISA covering the preconcentration step was 0.13 ng L-1 EE2 in water samples. This ELISA also correlated well with GC-MS (R2 = 0.934) data that were conducted independently for the same spiked samples.

In Chapter 4, the development and validation of competitive direct ELISAs (cd ELISA) for the detection of 17β-estradiol (E2) in water at sub-parts per trillion levels is described. The specific polyclonal antibodies were raised against KLH conjugates of 17β-estradiol-acetate i hapten (E2-ACT-KLH) and 17β-estradiol-butyrate hapten (E2-BUT-KLH), using a similar hapten synthesis approach as for 17α-ethynylestradiol. The developed ELISA (denoted as E2 ELISA) was highly specific to 17β-estradiol with a LOD of 50 ngL-1 in water (without preconcentration), with high matrix tolerance. The LOD of the E2 ELISA covering the preconcentration step was 0.03 ngL-1in water samples. Validation of ELISA performance against an independently performed GC-MS method indicated good correlation with a R2 value of 0.936.

In Chapter 5, the development of an ELISA specific for testosterone with respect to the desired sensitivity and specificity for environmental analysis is described. The ELISA -1 -1 presented with a LOD and IC50 value at 0.07 ng L and 0.4 ng L , respectively, and an excellent specificity. Unlike previously described assays for the estrogens, this assay was more matrix sensitive. Hence, an alternative or enhanced sample preparation step prior to assay was deemed essential to improve assay performance.

Finally, in Chapter 6, the application of these developed ELISAs for the investigation of the quality of influent and effluents from various Sewage Treatment Plants (STPs) and nearby water resources in New South Wales, Australia, is described. The estrogenic and androgenic activities of the respective water samples were 0.20 to 6.28 ng L-1 estradiol equivalents and 0.01 to 0.64 ngL-1 testosterone equivalents, respectively, as measured by the yeast screening assays. Regular monitoring of EDCs should be conducted to ensure the levels and the contribution of estrogenic potency in water bodies is maintained below the ecotoxicologically significant level of 10 ngL-1 (Fenske et al., 2005, Huschek and Hansen, 2005). In addition, agricultural practices including live stock husbandry should be controlled to a certain distance away from water sources as a precautionary measure to reduce estrogenic potency in our water resources.

In conclusion, a satisfactory agreement in data for E2, EE2 and T was obtained between ELISAs, GC-MS, and the estrogenic/androgenic activities measured by the yeast screening assays, although a slightly higher estimation was observed by ELISA. This suggests that ELISAs developed in this project can be used as fast and cost effective water quality monitoring tools. The high sensitivity and specificity of these ELISAs allow them to be used to monitor with acceptable reliability of the estrogenic and androgenic steroids at low parts per trillion levels after a simple concentration step.

ACKNOWLEDGEMENTS

I would like to express my thanks to my supervisor Dr. Nanju Alice Lee for providing me with an opportunity to conduct research in the School of Chemical Engineering, Food Science and Technology. Her innovative research direction, valuable advice, and enthusiastic supervision throughout the course of this study are much appreciated.

Sincere thanks also to Dr. Victor Wong for his timely help in writing skills as well as scientific advice. I would like to thank Mr. Camillo Taraborrelli and Mr. Eryanto Yanasuwan for their kind technical assistance on my work. Further thanks also to Dr. Chunhua Li of the University of Sydney, for her valuable water samples/extracts and Miss Christine Tan of the University of NSW for her valuable haptens.

The ARC is especially acknowledged for funding of my research at the University of New South Wales (ARC Discovery DP0559179). The Glaxo Wellcome and Dr E. Routledge, for their kind gift of the recombinant yeasts to Dr. Alice Lee are also acknowledged.

My sincere gratitude goes to my family for their ongoing encouragement and beliefs in me. I also would like to specially thank Jay and Molly for their unconditional love and prayer throughout my study.

Thanks also to all of my friends. Their cooperative and friendly manner has created a stress- free working environment.

Finally, to others who were not mentioned, your help and support are much appreciated, for without you, this thesis could not have been completed.

iii

LIST OF PUBLICATIONS IN PREPARATION

C. Uraipong, C. H. Tan , V. Wong , C. Li, I. R. Kennedy, R. D. Allan & N. A. Lee, (2011) Sensitive and specific competitive enzyme immunoassays for monitoring of a potential endocrine disrupting hormone, 17α-ethynylestradiol and mestranol, in aquatic system (Journal of Food and Agricultural Chemistry)

C. Uraipong, C. H. Tan , V. Wong , C. Li, I. R. Kennedy, R. D. Allan & N. A. Lee, (2011) Highly sensitive competitive enzyme immunoassays for monitoring of a potential endocrine disrupting synthetic hormone, mestranol, in aquatic system (Analytical and Bioanalytical Methods)

C. Uraipong, C. H. Tan , V. Wong , C. Li, I. R. Kennedy, R. D. Allan & N. A. Lee (2011) Sensitive and specific competitive enzyme immunoassays for monitoring of a potential endocrine disrupting hormone, 17β-estradiol, in aquatic system (Analytica Chimica Acta)

C. Uraipong, C. H. Tan , V. Wong & N. A. Lee (2011) Development of testosterone specific competitive enzyme immunoassay for monitoring in water quality (Analytical and Bioanalytical Methods)

C. Uraipong, C. Li, I. R. Kennedy, Victor Wong & N. A. Lee (2011) Androgenic Activity in Water Bodies of selected urban and rural areas of New South Wales of Australia (Environmental Toxicology and Chemistry)

LIST OF PRESENTATION

N.A. Lee, C. Uraipong, C.H. Tan, V. Wong, C. Li, I.P. Kennedy, & R.D. Allan, Enzyme- Linked ImmunoSorbent Assay (ELISA) as a water Quality Monitoring Tool: Targeting a Potential Endocrine Disrupting Steroids, 17α-ethynylestradiol (EE2) IUPAC/RACI Conference (Chemistry for a Sustainable World, Melbourne 4th – 8th July 2010)

N.A. Lee, C. Uraipong, C.H. Tan and V. Wong, Antibodies for characterizing and tracking of potential endocrine disrupting chemicals (ECDs) Pacifichem 2010 (International Chemical Congress of Pacific Basin Societies, Honolulu, Hawaii, USA 15th – 20th December 2010)

ABBREVIATIONS

Ab purified antibody APCI atmospheric pressure chemical ionization BSA bovine serum albumin CBT checkerboard titration DCC N,N′-dicyclohexylcarbodiimide DDT dichlorodiphenyltrichloroethane DES diethylstilbestrol DMF dimethylformamide DMSO dimethyl sulfoxide E1 estrone E2 17β-estradiol E2-ACT polyclonal antibody of 17β-estradiol-acetate E2-ACT-BSA conjugate of 17β-estradiol-acetate and bovine serum E2-ACT-HRP conjugate of 17β-estradiol-acetate and horseradish peroxidase E2-ACT-KLH conjugate of 17β-estradiol-acetate keyhole limpet haemocyanin E2-BUT polyclonal antibody of 17β-estradiol-butyrate E2-BUT-BSA conjugate of 17β-estradiol-butyrate and bovine serum E2-BUT-HRP conjugate of 17β-estradiol-butyrate and horseradish peroxidase E2-BUT-KLH conjugate of 17β-estradiol-butyrate keyhole limpet haemocyanin E2-HRP conjugate of 17β-estradiol and horseradish peroxidase E2-HS-HRP conjugate of 17β-estradiol-hemisuccinate and horseradish peroxidase E3 estriol EDC endocrine disrupting chemicals EDSTAC endocrine disruptor screening and testing advisory committee EE2 17α-ethynylestradiol EE2-ACT polyclonal antibody of 17α-ethynylestradio-acetate EE2-ACT-BSA conjugate of 17α-ethynylestradiol-acetate and bovine serum EE2-ACT-HRP conjugate of 17α-ethynylestradiol-acetate and horseradish peroxidase EE2-ACT-KLH conjugate of 17α-ethynylestradiol-acetate keyhole limpet haemocyanin EE2-BUT polyclonal antibody of 17α-ethynylestradio-butyrate EE2-BUT-BSA conjugate of 17α-ethynylestradiol-butyrate and bovine serum EE2-BUT-HRP conjugate of 17α-ethynylestradiol -butyrate and horseradish peroxidase EE2-BUT-KLH conjugate of 17α-ethynylestradiol-butyrate keyhole limpet haemocyanin EEQ an estradiol equivalent ELISA enzyme-linked immunosorbent assay ESI electrospray ionization EtOH ethanol FIA fluorescent immunoassay GC gas chromatography GC/MS gas chromatography-mass spectrometry GC/MS/MS gas chromatography-tandem mass spectrometry GC/NCI-MS gas chromatography with negative chemical ionization mass spectrometric vi

HA humic acid hAR human androgen receptor hER human estrogen receptor HPLC high performance liquid chromatography HRP horseradish peroxide IA immunoassay Ig immunoglobulin KLH kehole limpet homocyanin LC liquid chromatography LC/MS liquid chromatography-mass spectrometry LLE liquid-liquid extraction LOD limit of detection LYES lyticase yeast oestrogen screen assay MeOH methanol NHS N-hydroxysuccinimide OCT 4-octylphenol pAb polyclonal antibody PAH polyaromatic hydrocarbon PBS phosphate buffer saline PCB polychlorinated biphenyl PCDF polychlorinated dibenzofuran PCDD polychlorinated dibenzodioxin PhAC pharmaceutical active chemical RIA radioimmunoassay SPE solid-phase extraction STP sewage treatment plant T testosterone TBT tributyltin TEQ a testosterone equivalent TMB 3,3’,5,5’-tetramethylbenzidine T-OX testosterone3-O-carboxymethyl-oxime conjugate of testosterone3-O-carboxymethyl-oxime and bovine serum T-OX-BSA albumin conjugate of testosterone3-O-carboxymethyl-oxime and horseradish T-OX-HRP peroxidase conjugate of testosterone3-O-carboxymethyl-oxime and keyhole limpet T-OX-KLH haemocyanin USEPA the united states environmental protection agency WWTP wastewater treatment plant YAS yeast androgen screen assay YES yeast estrogen screen assay

vii

TABLE OF CONTENT

CHAPTER 1 INTRODUCTION ...... 2

1.1 BACKGROUND OF RESEARCH...... 2 1.1.1 Selection of estrogens as target compounds ...... 3 1.1.2 Method Development for EDCs ...... 4 1.1.3 Sample locations ...... 5

CHAPTER 2 LITERATURE REVIEW ...... 6

2.1 ENDOCRINE SYSTEM AND ENDOCRINE DISRUPTING COMPOUNDS ...... 6 2.1.1 Endocrine System ...... 6 2.1.2 Main categories of endocrine-disrupting chemicals ...... 7 2.1.2.1 Synthetic chemicals ...... 9 2.1.2.2 Natural Chemicals...... 11 2.1.3 Occurrences and adverse effect of EDCs ...... 11 2.1.3.1 Effect found in Wildlife...... 13 2.1.3.2 Effects found in People ...... 14 2.2 HORMONES AND THEIR ACTIONS...... 15 2.2.1 Occurrences and adverse effects of endocrine disrupting steroidal hormones ...... 21 2.2.2 Estrogenic potency of estrogens that is ecologically meaningful ...... 24 2.3 ANALYTICAL METHODS FOR DETECTING ENDOCRINE DISRUPTING COMPOUNDS ...... 25 2.3.1 Instrument methods ...... 25 2.3.1.1 Gas chromatography/ mass spectrometry (GC/MS) and (GC-MS/MS) ...... 26 2.3.1.2 High Performance liquid chromatography (HPLC) ...... 26 2.3.1.3 Liquid chromatography / mass spectrometry (LC/MS) and (LC-MS/MS) ...... 27 2.3.2 Biological methods ...... 27 2.3.2.1 Immunoassay ...... 28 2.4 ELISAS (ENZYME- LINKED IMMUNOSORBENT ASSAY) ...... 30 2.4.1 Principle ...... 30 2.4.1.1 Antibody binding and the interaction between antibody and antigen ...... 31 2.4.1.2 Hapten design and synthesis ...... 32 2.4.2 ELISA format ...... 32 2.4.3 Sensitivity and specificity of the assay ...... 33 2.4.4 Matrix interference ...... 34 2.5 ESTROGENIC HORMONE DETERMINATION BY ELISAS ...... 35 2.6 RECOMBINANT YEAST SCREENING ASSAY ...... 36 2.6.1 Steroidal hormone determination by a Recombinant Yeast Screening Assay ...... 37

CHAPTER 3 DEVELOPMENT OF A SENSITIVE ELISA SPECIFIC TO 17Α-ETHYNYLESTRADIOL (EE2 ELISA) ...... 39

3.1 INTRODUCTION ...... 39 3.2 MATERIALS AND METHODS ...... 41 3.2.1 Materials and Instrumentation ...... 41 3.2.1.1 Materials ...... 41 3.2.1.2 Instrumentation ...... 42 3.2.2 Hapten Synthesis...... 42 3.2.2.1 Attachment of acetate linker onto C3 of 17α-ethynylestradiol ...... 42 3.2.2.2 Attachment of butyrate linker onto C3 of 17α-ethynylestradiol ...... 44 3.2.2.3 Preparation of conjugates of hapten and carrier proteins or enzyme ...... 46 3.2.3 Antibody Production and Characterisation ...... 47 3.2.3.1 Immunisation and antibody production ...... 47 3.2.3.2 Purification of Rabbit IgG ...... 47 3.2.3.3 Determining antibody concentration ...... 48 3.2.3.4 Determining optimum working concentration by checkerboard titration ...... 48 3.2.3.5 Determining Sensitivity ...... 49 3.2.3.6 Optimisation of EE2 ELISA condition ...... 50 3.2.3.7 Determining specificity ...... 50 viii

3.2.3.8 Study of Matrix Effects ...... 51 3.2.4 Spike and recovery studies...... 51 3.2.5 Validation of ELISA with GC/MS ...... 52 3.2.5.1 Water and water samples ...... 53 3.3 RESULTS AND DISCUSSION ...... 54 3.3.1 Hapten selection...... 54 3.3.1.1 Optimal Concentration of Enzyme Conjugates ...... 55 3.3.2 Assay Sensitivity ...... 59 3.3.2.1 Standard curve parameters and precision (IC80, IC50, IC20 and maximum absorbance) ...... 61 3.3.3 Characteristics of EE2 ELISA ...... 66 3.3.3.1 Assay Specificity ...... 66 3.3.3.2 Assay Optimisation ...... 71 3.3.3.3 Matrix Interferences ...... 80 3.3.5.3 Effect of humic acid ...... 85 3.3.6 Recoveries of EE2 from spiked purified and field water ...... 87 3.3.7 Validation of ELISA with GC/MS ...... 89 3.3.3.7 Determination of 17α-ethynylestradiol in water samples by the developed EE2 ELISA ...... 89 3.4 CONCLUSION ...... 90

CHAPTER 4 DEVELOPMENT OF THE 17Β-ESTRADIOL SPECIFIC ELISA (E2 ELISA)...... 92

4.1 INTRODUCTION ...... 92 4.2 MATERIALS AND METHODS ...... 94 4.2.1 Material and Instrument ...... 94 4.2.1.1 Materials ...... 94 4.2.1.2 Instruments ...... 94 4.2.2 Hapten Synthesis...... 94 4.2.2.1 Attachment of acetate linker onto C 3 of 17β-estradiol ...... 95 4.2.2.2 Attachment of butyrate linker onto C 3 of 17β-estradiol ...... 97 4.2.2.3 Preparation of conjugates of hapten and carrier proteins or enzyme ...... 99 4.2.3 Antibody Production and Characterization ...... 99 4.2.3.1 Immunization and antibody production ...... 99 4.2.3.2 Purification of Rabbit IgG ...... 99 4.2.3.3 Determining antibody concentration ...... 99 4.2.3.4 Determining optimum working concentration by checkerboard titration ...... 99 4.2.3.5 Determining Sensitivity...... 100 4.2.3.6 Optimization of EE2 ELISA condition ...... 100 4.2.3.7 Determining specificity ...... 100 4.2.3.8 Matrix interferences study ...... 100 4.2.4 Spike and recovery study ...... 100 4.2.5 Validation of the ELISA with GC/MS ...... 101 4.3 RESULTS AND DISCUSSION ...... 101 4.3.1 Hapten selection ...... 101 4.3.1.1 Optimal Concentration of Enzyme Conjugates ...... 102 4.3.2 Assay Sensitivity ...... 106 4.3.2.1 Standard curve parameters and precision (IC80, IC50, IC 20, and maximum absorbance) ...... 108 4.3.3 Characteristic of EE2 ELISA ...... 112 4.3.3.1 Assay Specificity ...... 112 4.3.3.2 Assay Optimization ...... 117 3.3.3.3 Matrix Interferences ...... 125 4.3.5.3 Effect of humic acid ...... 128 4.3.6 Recoveries of EE2 from spiked purified and field water ...... 131 4.3.7 Validation of ELISA with GC/MS ...... 133 4.3.3.7 Determination of 17β-estradiol in water samples by the developed E2 ELISA ...... 133 4.4 CONCLUSION ...... 134

CHAPTER 5 DEVELOPEMTNOF TESTOSTERONE ELISA ...... 136 (T ELISA) ...... 136

5.1 INTRODUCTION ...... 136 5.2 MATERIALS AND METHODS ...... 137 5.2.1 Materials and Instrumentation ...... 137 5.2.1.1 Materials ...... 137 5.2.1.2 Instrumentation ...... 137 5.2.2 Preparation of conjugates of haptens and carrier proteins or enzymes ...... 137 5.2.3 Antibody Production and Characterisation ...... 138 5.2.3.1 Immunisation and antibody production ...... 138 5.2.3.2 Purification of rabbit IgG ...... 138 5.2.3.3 Determining antibody concentration ...... 138 5.2.3.4 Determining optimum working concentration (Checkerboard Titration) ...... 138 5.2.3.5 Sensitivity Determination of T-ELISA...... 138 5.2.3.6 Determining specificity ...... 139 5.2.3.7 Matrix interferences study ...... 139 5.2.4 Spike and recovery study ...... 139 5.3.1 Hapten selection and conjugation to carrier proteins and an enzyme ...... 140 5.3.1.1 Optimal concentration of enzyme conjugates ...... 140 5.3.2 Assay Sensitivity ...... 141 5.3.2.1 Standard curve parameters and precision (IC80, IC50, IC20 and maximum absorbance) ...... 142 5.3.3 Assay Specificity ...... 146 5.3.4 Matrix Interferences ...... 146 5.3.4.1 Effect of pH ...... 147 5.3.4.2 Effect of water type and ionic strength ...... 148 5.3.4.3 Effect of humic acid ...... 151 5.3.5 Recoveries of testosterone from spiked, purified and field water samples ...... 153 5.4 CONCLUSION ...... 155

CHAPTER 6 VALIDATION OF YEAST SCREENING ASSAY FOR ENDOCRINE DISRUPTING HORMONES ...... 156

6.1 INTRODUCTION ...... 156 6.2 MATERIALS AND METHODS ...... 157 6.2.1 Material and Instrument ...... 157 6.2.1.1 Materials ...... 157 6.2.1.2 Instrument...... 157 6.2.1.3 The recombinant yeast expressing the human estrogen receptor (hER) ...... 158 6.2.2 Preparation of the medium components ...... 158 6.2.3 Preparation of stock culture ...... 160 6.2.4 Preparation of standard solutions ...... 160 6.2.4.1 Stock standard solution...... 160 6.2.4.2 Intermediat standard solution ...... 160 6.2.5 Assay Procedure ...... 160 6.2.6 Response of the yeast screen bioassay ...... 161 6.2.7 Spike and recovery study ...... 161 6.2.8 Water and water samples ...... 162 6.2.9 Analysis of estrogenic and androgenic hormones in water samples by the YES and YAS assays ... 164 6.3 RESULTS AND DISCUSSION ...... 165 6.3.1 The performance of Yeast Assays ...... 165 6.3.2 Assay Specificity ...... 165 6.3.3 EEQs and TEQs of spiked purified water ...... 167 6.3.4 Estrogenic and androgenic potency in Creek samples ...... 170 6.3.4.1 Analysis of estrogenic hormones in water samples by YES assay ...... 170 6.3.4.2 Analysis of androgenic hormones in water samples by the YAS assay ...... 172 6.3.5 Determiination of estrogenic and androgenic compounds ...... 174 6.3.5.1 Estrogenic compounds determined by ELISA, CALUX and yeast assay ...... 174 6.3.5.2 Androgenic compounds determined by ELISA and yeast assay ...... 177 6.4 CONCLUSION ...... 179

CHAPTER 7 CONCLUSION ...... 180

LIST OF FIGURES

LIST OF FIGURES ...... XII FIGURE 2.1 STRUCTURE OF HORMONE STEROIDS ...... 16 FIGURE 2.2 METABOLIC PATHWAYS FOR THE SYNTHESIS OF PROGESTERONE, ESTRADIOL, TESTOSTERONE, AND DIHYDROTESTOSTERONE ...... 19 FIGURE 2.3 CONVERSION OF PREGNENOLONE TO PROGESTERONE ...... 21 FIGURE 2.4 ANATOMY OF AN ANTIBODY ...... 31 FIGURE 2.5 SCHEMATIC PRESENTATION OF A DIRECT ELISA ...... 33 FIGURE 3.1 GEOGRAPHICAL LOCATIONS WHERE WATER SAMPLES WERE COLLECTED IN NSW, AUSTRALIA...... 53 FIGURE 3.2 CHEMICAL STRUCTURE OF HAPTENS ...... 56 FIGURE 3.6 TITRATION CURVES OF EE2-BUT ANTIBODY AGAINST SIX HAPTEN-ENZYME CONJUGATES (E2-ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT20-HRP, E2-HS-HRP AND E2-OX-HRP) ...... 58 FIGURE 3.8 % CV FOR THE ABSORBANCE ( ) AND % INHIBITION ( ) BY EE2-ACT ANTIBODY (AVERAGE OF 27 ANALYSES)...... 63

FIGURE 3.9 PLOT OF IC80 (), IC50 ( ) AND IC20 ( ) VALUES FOR EE2-ACT ANTIBODY (AVERAGE OF 27 ANALYSES). THE MIDDLE LINE INDICATES AVERAGE VALUE. THE DOTTED LINE SHOWS THE UPPER AND LOWER LIMIT (AVERAGE EE2 CONCENTRATION+STANDARD DEVIATION)...... 63 FIGURE 3.10 CALIBRATION CURVE BY EE2-BUT ANTIBODY (AVERAGE OF 12 ANALYSES). THE ABSORBANCE AGAINST THE EE2 CONCENTRATION ( ) AND THE % INHIBITION AGAINST THE EE2 CONCENTRATION ( )...... 64 FIGURE 3.11 % CV FOR THE ABSORBANCE ( ) AND % INHIBITION ( ) BY EE2-ACT ANTIBODY (AVERAGE OF 12 ANALYSES)...... 65

FIGURE 3.12 PLOT OF IC80 (), IC50 ( ) AND IC20 ( ) VALUES FOR EE2-BUT ANTIBODY (AVERAGE OF 27 ANALYSES). THE MIDDLE LINE INDICATES AVERAGE VALUE. THE DOTTED LINE SHOWS THE UPPER AND LOWER LIMIT (AVERAGE EE2 CONCENTRATION + STANDARD DEVIATION)...... 65 FIGURE 3.13 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-EE2-ACT (MILLIQ WATER, 5% ETOH, 10% ETOH AND 20% ETOH )...... 72 FIGURE 3.14 STANDARD CURVE OF EE2 CONCENTRATION IN DIFFERENT SOLVENTS FOR AB-EE2-ACT (MILLIQ WATER, 5% ETOH, 10% ETOH AND 20% ETOH)...... 72 FIGURE 3.15 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-EE2-ACT (5% MEOH, 10% MEOH AND 20% MEOH)...... 73 FIGURE 3.17 STANDARD CURVE OF EE2 CONCENTRATION IN DIFFERENT SOLVENTS FOR AB-EE2-ACT (10% ACETONE, 20% ACETONE, 10% ACETONITRILE AND 20% ACETONITRILE)...... 74 FIGURE 3.19 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-EE2-BUT (MILLIQ WATER, 5% ETOH, 10% ETOH AND 20% ETOH )...... 75 FIGURE 3.20 STANDARD CURVE OF EE2 CONCENTRATION IN DIFFERENT SOLVENTS FOR AB-EE2-BUT (MILLIQ WATER, 5% ETOH, 10% ETOH AND 20% ETOH ...... 75 FIGURE 3.21 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-EE2-BUT (5% MEOH, 10% MEOH AND 20% MEOH)...... 76 FIGURE 3.23 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-EE2-BUT (10% ACETONE, 20% ACETONE, 10% ACETONITRILE AND 20% ACETONITRILE)...... 77

xii

FIGURE 3.24 STANDARD CURVE OF EE2 CONCENTRATION IN DIFFERENT SOLVENTS FOR AB-EE2-BUT (10% ACETONE, 20% ACETONE, 10% ACETONITRILE AND 20% ACETONITRILE)...... 77 FIGURE 3.25 EFFECT OF ORGANIC SOLVENTS ON THE CALIBRATION CURVE FOR AB-EE2- ACT...... 78 FIGURE 3.26 EFFECT OF ORGANIC SOLVENTS ON THE CALIBRATION CURVE FOR AB-EE2- BUT ...... 78 FIGURE 3.29 STANDARD CURVE OF EE2 CONCENTRATION IN DIFFERENT TYPES OF WATER (PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, LAGOON AT TAHBILK WINERY WETLANDS, AND SEA WATER FROM MAROUBRA BEACH)...... 84 FIGURE 3.30 EFFECT OF HUMIC ACID ON ABSORBANCE FOR EE2 ELISA...... 86 FIGURE 3.31 EFFECT OF HUMIC ACID (HA) ON THE CALIBRATION CURVE FOR 17Α- ETHYNYLESTRADIOL OBTAINED WITH SPIKED WATER SAMPLES ...... 87 FIGURE 3.32 AVERAGE VALUE (ΜG L-1) OF SPIKING AND RECOVERY (%) FROM THREE WATER SOURCES...... 88 FIGURE 3.33 AVERAGE VALUE (ΜGL-1) OF SPIKING AND SPIKING LEVEL (ΜGL-1) FROM FOUR WATER SOURCES: PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, TAHBILK WINERY WETLANDS AND SEA WATER FROM MAROUBRA BEACH BY EE2 ELISA...... 89 FIGURE 3.34 COMPARISON OF ANALYTICAL RESULTS BETWEEN EE2 ELISA AND GC-MS INSTRUMENTAL ANALYSIS FOR DETERMINATION OF EE2 ...... 90 FIGURE 4.1 TITRATION CURVES OF E2-ACT ANTIBODY AGAINST FIVE HAPTEN-ENZYMES CONJUGATES (EE2-ACT12-HRP, EE2-ACT20-HRP, EE2-BUT12-HRP, EE2-BUT2HRP AND E1-HS- HRP) ...... 103 FIGURE 4.2 TITRATION CURVES OF E2-ACT ANTIBODY AGAINST SIX HAPTEN-ENZYMES CONJUGATES (E2-ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT2—HRP, E2-HS-HRP AND E2-OX-HRP) ...... 103 FIGURE 4.3 TITRATION CURVES OF E2-BUT ANTIBODY AGAINST FIVE HAPTEN-ENZYMES CONJUGATES (EE2-ACT12-HRP, EE2-ACT20-HRP, EE2-BUT12-HRP, EE2-BUT20-HRP AND E1- HS-HRP) ...... 104 FIGURE 4.4 TITRATION CURVES OF E2-BUT ANTIBODY AGAINST SIX HAPTEN-ENZYMES CONJUGATES (E2-ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT20-HRP, E2-HS-HRP AND E2-OX-HRP) ...... 104 FIGURE 4.5 CALIBRATION CURVE BY E2-ACT ANTIBODY AVERAGED OF 21 ANALYSES. THE SQUARE INDICATES ABSORBANCE AGAINST THE E2 CONCENTRATION AND THE CIRCLE INDICATES % INHIBITION AGAINST THE E2 CONCENTRATION...... 109 FIGURE 4.6 THE %CV FOR THE ABSORBANCE ( ) AND % INHIBITION ( ) OVER 21 ANALYSES BY E2-ACT ANTIBODY ...... 109

FIGURE 4.7 PLOT OF IC80 ( ) , IC50 ( ) AND IC20 ( ) VALUES OVER 21 ANALYSES FOR E2-ACT ANTIBODY. THE MIDDLE LINE INDICATES AVERAGE VALUE. THE DOTTED LINE SHOWS THE UPPER AND LOWER LIMIT (AVERAGE E2 CONCENTRATION + STANDARD DEVIATION) ...... 110 FIGURE 4.8 CALIBRATION CURVE BY E2-BUT ANTIBODY AVERAGED OF 11 ANALYSES. THE SQUARE INDICATES ABSORBANCE AGAINST THE EE2 CONCENTRATION AND THE CIRCLE INDICATES % INHIBITION AGAINST THE E2 CONCENTRATION ...... 111 FIGURE 4.9 THE %CV FOR THE ABSORBANCE ( ) AND % INHIBITION ( ) OVER 11 ANALYSTS BY EE2-BUT ANTIBODY ...... 111

FIGURE 4.10 PLOT OF IC80 ( ) , IC50 ( ) AND IC20 VALUES OVER 11 ANALYSES FOR E2-BUT ANTIBODY. THE MIDDLE LINE INDICATES AVERAGE VALUE. THE DOTTED LINE SHOWS THE UPPER AND LOWER LIMIT (AVERAGE E2 CONCENTRATION + STANDARD DEVIATION) ...... 112 FIGURE 4.11 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-E2-ACT (MILLIQ WATER, 5% ETOH, 10% ETOH, AND 20% ETOH ) ...... 117 xiii

FIGURE 4.12 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT SOLVENT FOR AB- E2-ACT (MILLIQ WATER, 5%, ETOH, 10% ETOH, AND 20% ETOH) ...... 117 FIGURE 4.13 EFFECT OF SOLVENT ON ABSORBANCE FOR AB-E2-ACT (5%, MEOH, 10% MEOH, AND 20% MEOH) ...... 118 FIGURE 4.14 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT SOLVENT FOR AB- E2-ACT ...... 118 (5%, MEOH, 10% MEOH, AND 20% MEOH) ...... ERROR! BOOKMARK NOT DEFINED. FIGURE 4.16 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT SOLVENT FOR AB- E2-ACT (10% ACETONE, 20% ACETONE, 10% ACETONITRILE, AND 20% ACETONITRILE)... 119 FIGURE 4.18 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT SOLVENT FOR AB- E2-BUT (MILLIQ WATER, 5%, ETOH, 10% ETOH, AND 20% ETOH) ...... 120 FIGURE 4.20 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT SOLVENT FOR AB- E2-BUT (5%, MEOH, 10% MEOH, AND 20% MEOH) ...... 121 FIGURE 4.22 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT SOLVENT FOR AB- E2-BUT (10% ACETONE, 20% ACETONE, 10% ACETONITRILE, AND 20% ACETONITRILE) ... 122 FIGURE 4.23 EFFECT OF ORGANIC SOLVENTS ON THE CALIBRATION CURVE FOR AB-E2- ACT...... 123 FIGURE 4.24 EFFECT OF ORGANIC SOLVENTS ON THE CALIBRATION CURVE FOR AB-E2- BUT ...... 123 FIGURE 4.25 EFFECT OF PH ON E2 ELISA STANDARD CURVE. THE SQUARE INDICATES ABSORBANCE AGAINST THE PH AND THE CIRCLE INDICATES IC50 VALUE AGAINST PH. .. 126 FIGURE 4.26 MATRIX EFFECT ON ABSORBANCE FOR E2 ELISA WITH DIFFERENT TYPE OF WATERS (PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, LAGOON AT TAHBILK WINERY WETLANDS, AND SEA WATER FROM MAROUBRA BEACH)...... 127 FIGURE 4.27 STANDARD CURVE OF E2 CONCENTRATION IN DIFFERENT TYPE OF WATERS (PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, LAGOON AT TAHBILK WINERY WETLANDS, AND SEA WATER FROM ...... 127 FIGURE 4.28 HUMIC ACID EFFECT ON ABSORBANCE FOR E2 ELISA WITH DIFFERENT CONCENTRATION OF HUMIC ACID SPIKED IN PURIFIED WATER...... 130 FIGURE 4.29 EFFECT OF ...... 130 FIGURE 4.30 AVERAGE VALUE (ΜGL-1) OF SPIKING AND RECOVERY (%) FROM THREE WATER SOURCES...... 132 FIGURE 4.31 AVERAGE VALUE (ΜGL-1) OF SPIKING AND SPIKING LEVEL (ΜGL-1) FROM THREE WATER SOURCES: PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, LAGOON AT TAHBILK WINERY WETLANDS, AND SEA WATER FROM MAROUBRA BEACH BY E2 ELISA ...... 132 FIGURE 5.1 STRUCTURE OF HAPTENS: TESTOSTERONE3-O-CARBOXYMETHYL-OXIME (T- CMO) AND PROGESTERONE3-O-CARBOXYMETHYL-OXIME (P-CMO)...... 138 FIGURE 5.2 TITRATION CURVES OF AB-T-CMO AGAINST TWO HAPTEN-ENZYME CONJUGATES. THE CIRCLE INDICATES T-CMO-HRP AND SQUARE INDICATES P-CMO-HRP...... 140 FIGURE 5.3 TESTOSTERONE STANDARD CURVES AGAINST TWO ENZYME CONJUGATES, T- CMO-HRP AND P-CMO-HRP)...... 142 FIGURE 5.5 THE % CV FOR ABSORBANCE ( ) AND % INHIBITION ( ) BY AB-T-CMO (AVERAGE OF EIGHT ANALYSES)...... 144

FIGURE 5.6 PLOT OF IC80 (), IC50 ( ) AND IC20 ( ) VALUES FOR AB-T-CMO (AVERAGE OF EIGHT ANALYSES). THE MIDDLE LINE INDICATES AVERAGE VALUE. THE DOTTED LINE SHOWS THE UPPER AND LOWER LIMIT (AVERAGE T CONCENTRATION+STANDARD DEVIATION)...... 144 FIGURE 5.7 EFFECT OF PH ON T-ELISA STANDARD CURVE. THE SQUARE INDICATES ABSORBANCE AGAINST PH AND THE CIRCLE INDICATES IC50 VALUE AGAINST PH...... 148 xiv

FIGURE 5.8 MATRIX EFFECT ON ABSORBANCE FOR T ELISA WITH DIFFERENT TYPE OF WATERS (PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, LAGOON AT TAHBILK WINERY WETLANDS, AND MAROUBRA BEACH SEA WATER)...... 149 FIGURE 5.9 STANDARD CURVE OF T CONCENTRATION IN DIFFERENT TYPE OF WATERS (PURIFIED WATER, MCWILLIAM’S WINERY RESERVOIR, LAGOON OF TAHBILK WINERY WETLANDS, AND MAROUBRA BEACH: SEA WATER)...... 150 FIGURE 5.10 EFFECT OF HUMIC ACID ON ABSORBANCE FOR T-ELISA...... 152 FIGURE 5.11 EFFECT OF HUMIC ACID ON THE CALIBRATION CURVE FOR TESTOSTERONE IN SPIKED WATER SAMPLES...... 153 FIGURE 5.13 COMPARISON OF THE SPIKING OF TESTOSTERONE IN FOUR WATER SOURCES...... 154 FIGURE 6.1 CHEMICAL REACTION OF CPRG ...... 158 FIGURE 6.2 MAP OF SAMPLING SITES, SOUTH CREEK, SYDNEY BASIN, NSW ...... 164 FIGURE 6.4 THE SPECIFICITY OF HAR YEAST SCREEN; PLOT OF ( ) PROG, (-- --) T, ...... 167 ( ) EE2, ( ) E2, ( -- --) E3 AND (--- --) E1 ...... ERROR! BOOKMARK NOT DEFINED. FIGURE 6.6 THE CORRELATION BETWEEN YAS ASSAY AND SPIKING LEVELS ...... 170 FIGURE 6.7 YAS ASSAY VALIDATIONS OF SPIKED SAMPLE ...... 170 FIGURE 6.8 LEVELS AND DISTRIBUTION OF ESTROGENIC ACTIVITY (N=60) ...... 171 FIGURE 6.9 LEVELS AND DISTRIBUTION OF ANDROGENIC ACTIVITY (N=50) ...... 173 FIGURE 6.11 CORRELATION BETWEEN EEQ BY HER ASSAY AND EEQ BY CALUX ...... 176 FIGURE 6.12 CORRELATION BETWEEN EEQ BY HER ASSAY AND EEQ BY CALUX. THE DISPERSION AT 0 TO 2 NGL-1 ...... 177 FIGURE 6.13 CORRELATION OF TEQS, N = 49. THE TEQ LEVEL PREDICTED WITH THE SUMMATION OF DETECTED INDIVIDUAL ANDROGEN LEVELS WAS CORRELATED WITH THE MEASURED TEQS IN WATER SAMPLES ...... 178 FIGURE 6.14 CORRELATION BETWEEN TEQ BY HAR ASSAY AND ESTIMATED TEQ BY ELISA. THE DISPERSION AT 0 TO 1.5 NGL-1 ...... 178

LIST OF TABLES

TABLE 2.1 EXAMPLE OF ENDOCRINE EFFECTS IN WILDLIFE AND HUMANS ...... 12 TABLE 2.2 PHYSIOCHEMICAL PROPERTIES OF STEROIDS ...... 22 TABLE 2.3 EXAMPLE OF CONCENTRATION OF CONTAMINATED HORMONES IN WATER RESOURCES...... 23 TABLE 3.1 THE RATIO OF 17Α-ETHYNYLESTRADIOL (EE2), 17Β-ESTRADIOL (E2) AND ESTRONE (E1) TO PREPARE THE SPIKED SAMPLES ...... 52 -1 TABLE 3.2 IC50 (μGL ) VALUES FOR 11 ENZYME CONJUGATES FOR 17Α- ETHYNYLESTRADIOL ANTIBODIES ...... 59 TABLE 3.3 STANDARD CURVE PARAMETERS AND PRECISION FOR EE2-ACT ANTIBODY ...... 62 TABLE 3.4 STANDARD CURVE PARAMETER AND PRECISION FOR EE2-BUT ANTIBODY...... 64 -1 TABLE 3.9 THE IC50 (μGL ) AND CROSS REACTIVITY (CR %) FOR SELECTED ESTROGENS WITH TWO ETHYNYLESTRADIOL ANTIBODIES...... 68 TABLE 3.10 EFFECT OF ENZYME CONJUGATE DILUENTS ON EE2 ELISA...... 80 TABLE 3.11 EFFECT OF IONIC STRENGTH ON EE2 ELISAS...... 83 TABLE 4.1 THE RATIO OF 17Α-ETHYNYLESTRADIOL (EE2), 17Β-ESTRADIOL (E2), AND ESTRONE (E1) TO MAKE UP THE SPIKE SAMPLES FOR E2 ELISA ...... 101 -1 TABLE 4.2 SUMMARY IC50 (μGL ) FOR ELEVEN DIFFERENT ENZYME CONJUGATES WITH FOR ESTRADIOL ANTIBODIES ...... 107 TABLE 4.3 STANDARD CURVE PARAMETER AND PRECISION FOR EE2-ACT ANTIBODY ...... 108 TABLE 4.4 STANDARD CURVE PARAMETER AND PRECISION FOR E2-BUT ANTIBODY ...... 110 -1 TABLE 4.5 THE IC50 (μGL ) AND CROSS-REACTIVITY (CR %) FOR SELECTED ESTROGENS WITH TWO ETHYNYLESTRADIOL ANTIBODIES ...... 113 TABLE 4.6 EFFECT OF ENZYME CONJUGATE DILUENTS ON EE2 ELISAS ...... 125 TABLE 4.7 EFFECT OF IONIC STRENGTH ON ESTRADIOL ELISAS...... 128 TABLE 5.1 THE RATIO OF EE2, E2 AND TESTOSTERONE (T) TO PREPARE THE SPIKED SAMPLES FOR T-ELISA...... 139 TABLE 5.2 STANDARD CURVE PARAMETER AND PRECISION FOR AB-T-CMO...... 143 TABLE 5.3 THE CROSS-REACTIVITY OF THE TESTOSTERONE ELISA...... 145 TABLE 5.4 EFFECT OF IONIC STRENGTH ON T-ELISA ...... 151 TABLE 6.1 THE RATIO OF 17Α-ETHYNYLESTRADIOL (EE2), 17Β-ESTRADIOL(E2), ESTRONE(E1), ESTRIOL (E3) AND TESTOSTERONE (T) TO MADE UP THE SPIKE SAMPLES . 162 TABLE 6.2 SAMPLING LOCATIONS IN SOUTH CREEK IN SYDNEY ...... 163 TABLE 6.4 TOTAL ESTROGENIC POTENCIES (BY HER ASSAY) OF WATER SAMPLES COLLECTED FROM CREEKS, EXPRESSED AS THE EQUIVALENT ESTRADIOL...... 173 TABLE 6.5 TOTAL ESTROGENIC POTENCIES (BY HAR ASSAY) OF WATER SAMPLES COLLECTED FROM CREEKS, EXPRESSED AS THE EQUIVALENT 17E-ESTRADIOL (NG T L-1) ...... 174

xvi

CHAPTER 1 INTRODUCTION

1.1 Background of Research

The endocrine system produces hormones that trigger responses and promotes normal biological functions such as growth, development, reproduction and behavior of animals and humans (2001). In 1949, the effect of 1,1,1-trichloro-2,2-di(4-chlorophenyl)ethane (DDT) on the endocrine system was accidentally discovered and thereafter many substances have been identified for their potential interference with the normal endocrine function of humans and wildlife. These compounds are called endocrine disrupting compounds (EDCs) (Jobling et al., 1996a). Some of the sources of EDCs include industrial and municipal effluent and agricultural run-off. The waste water treatment effluents, which are not effective in complete removal of EDCs, may release them into the aquatic ecosystem (Fox, 2001). These compounds are emerging environmental issues, increasing attention and concerns among the scientific community, media and general public (Ying et al., 2004). For this reason they are widely presented as environmental contamination and they have since been found to possess hormone-like or blocking activities when entering the endocrine system. They may interact with hormone receptors and interfere with the normal functions of hormones. Chemicals that have the capacity to interfere with the health of living organisms include pesticides (e.g., the organochlorine insecticides), industrial chemicals (e.g., polychlorinated bipenyls (PCBs) and dioxin), and pharmaceuticals (e.g., synthetic hormones in the contraceptive pills) (Barton and Anderson, 1998). In addition, the natural hormones such as phytoestrogens, occurring in plants, and hormones found in humans and animals such as 17β-estradiol and testosterone can interfere with the endocrine system (Tylor et al., 1998).

There is much research focus on the estrogenic hormones following strong evidence that suggests potential adverse effects on the reproduction of wildlife and humans. Significant evidence has emerged showing natural hormones, such as 17β-estradiol and estrone, and synthetic hormones, such as 17α-ethynylestradiol are the most likely concerns among the large number of potential estrogen-mimicking chemicals in the environment (Desbrow et al., 1998, Routledge et al., 1998a). There is evidence pointing to the impact of these compounds on some aquatic species, for example, the exposure to 17α-ethynylestradiol, estrone and 17β- estradiol can lead to vitellogenin induction and the development of intersexuality for aquatic organism (Routledge et al., 1998a, Schultz et al., 2003). Also, it was suggested that an 2 estrogen-like receptor might exist in mollusks and these receptors may be affected by EDCs in the physiological mechanisms underlying reproductive function (Jobling et al., 2003, Lafont and Mathieu, 2007). Many studies from Europe, Northern America and Japan reported inefficiencies of wastewater treatment plants (WWTPs) to completely remove such compounds, hence releasing them into aquatic environments. 17α-Ethynylestradiol has been released into the environment via its incomplete removal by WWTPs (Schultis et al., 2002).

In Australia, only limited studies reported the estrogenic situation in river, coastal water and lakes. Examples before 2006 include the impact of EDCs such as low fertility of sheep in Western Australia (Adams, 1998), and imposex in mollusks by tributylin (Gibson, 2003). Only a few surveys investigated the concentrations of estrogens and androgens in Australian water resources. For instance, Khan et al. (2004) reported E2 concentration below 15 ngL-1 in effluents from WWTPS in southern Queensland. The concentration of 17β-estradiol in southern Victoria (VIC) was found to be in the range of 2−5 ngL-1 (Mispagel et al., 2009), and the treated effluent from 45 WWTPs in VIC showed estrogenic activity of between 0 – 73 ngL-1 17β-estradiol (Allinson et al., 2010). Mispagel and co-workers (2005) stated that Australia has many factors which influence the level of estrogens/androgens in the environment. These include the anthropogenic sources of these chemicals, the differences in population densities, treatment technologies, and socioeconomic factors. Hence, there is a need to continually assess the levels of estrogens and androgens that may present some risks to viable organisms in the aquatic system.

1.1.1 Selection of estrogens as target compounds

There has been an increase in intensive livestock farming such as dairy, beef cattle, swine and poultry in Australia (Smith, 2003), which led to high levels of excretion of estrogen by animals into the environment via discharge of effluents from farm scale waste treatment plants and runoff, as has also been seen to be occurring in European surveys (Hintemann et al., 2006a, López De Alda and Barceló, 2001, Ying et al., 2002, Matthiessen et al., 2006). Little information is known either about the concentrations of specific hormones such E2, EE2 and T or of the overall estrogenic or androgenic activities in Australian WWTP discharges or surface water.

Therefore, 17β-estradiol, 17α-ethynylestradiol and testosterone were selected as the target compounds for this monitoring study due to their occurrence, contribution to environmental toxicity and the potential risks these chemicals pose to natural living systems.

1.1.2 Method Development for EDCs

Many analytical techniques that quantify EDCs at ultra low levels in environmental matrices were developed in the last decade. These include high performance liquid chromatography (HPLC) (Slikker et al., 1981, Itoh et al., 2001), liquid−chromatography mass spectrometry (LC-MS) (Snyder et al., 2003b, Zuehlke et al., 2005), and liquid-chromatography tandem mass spectrometry (LC-MS/MS) (Heisterkamp et al., 2004), gas-chromatography mass spectrometry (GC-MS) (Desbrow et al., 1998) and gas chromatography-tandem mass spectrometry (GC-MS/MS) (Huang and Sedlak, 2001, Fawell et al., 2001). Extensive sample cleanup and concentration has been inevitable for these techniques. Such sample preparation is even more laborious than the typical trace analysis of environmental contaminants at low parts per billion levels. Typically, a 1−4 L water sample is concentrated to 1−5 mL prior to analaysis in order to detect low parts per trillion levels of EDCs. These methods can be reliable if performed correctly, but they do suffer drawbacks, such as expensive instrumentation and maintenance costs, extensive sample preparation, and the requirement of highly trained operators (Ying et al., 2004). Earlier studies indicated lack of adequate sensitivity for detection at low parts per trillion levels. With the advancement in instrumental techniques in recent years, later studies reported significant increases in analytical sensitivity. Immunoassay is an alternative for quantifying estrogenic hormones in the environment.

Immunoassays have the advantage of providing specific detection via antigen-antibody binding, and high sensitivity due to the high affinity of antibodies. It is fast with high- throughput capacity and field portability, for real-time measurement (Brady et al., 1995, Farr´E et al., 2007, Li et al., 2004, Schneider et al., 2005). The ELISA for estrone developed in our group has proven to be very valuable for detecting estrone in water samples (Li et al., 2004). To increase the analytical capacity for monitoring other endocrine disrupting steroids in Australia, development of immunoassays for selected steroids was therefore required.

Following the success with hapten design and synthesis by Ms. Christine Tan (UNSW), this project aims to develop a series of ELISAs with high sensitivity for the detection of extremely low steroid concentrations, with reduced matrix interferences and low cross−reactivity.

1.1.3 Sample locations

In Australia, New South Wales (NSW) is the most heavily populated state with intensive horticulture, intensive poultry, piggery and grazing of cattle. This project also aims to apply the developed immunoassays to screening of EDCs in urban and rural areas of NSW and to identify potential sources of EDCs in water resources, resulting in important contributions to enhance the knowledge of EDCs in Australia.

The field water samples were collected by Dr. Chuhua Li (University of Sydney) from three different parts of NSW; the details of which are as follows; (1) Sydney Basin, South Creek (NSW) (represents an urban area)

South Creek and its confluent, Eastern Creek, are the main contaminant sources to the Hawkesbury-Nepean River; the main water supply of the Sydney Metropolis, due to three sewage treatment plants discharging effluents into these two creeks.

(2) The Murray River, Albury-Wodonga along the NSW/Victoria (VIC) state border (represents a mix of urban and rural activities)

At the boundary of NSW and VIC water samples were collected from the Wodonga Creek and Wangaratta Creek because these two creeks are confluent with the Murray River at Albury-Wodonga, which was impacted by both agricultural practices and discharge of effluents from STPs. (3) Emigrant Creek and Wilson River, and Ballina and Lismore in North-Eastern NSW (represents a rural area).

The water samples were collected from Emigrant Creek before entering the Dam for the reservoir at Ballina. The water of Emigrant Creek was contaminated from the runoff from the catchments and commercial horticultural enterprises for tropical and subtropical fruit and macadamia nut production, and also from grassland used for cattle grazing. Agriculture practices are also a main pollutant of this location. Water samples from locations (2) and (3) were collected for comparison purposes with the South Creek samples.

CHAPTER 2 LITERATURE REVIEW

2.1 Endocrine System and Endocrine Disrupting Compounds

2.1.1 Endocrine System The endocrine system is instrumental in regulating the body's growth, metabolism, behaviour, and sexual development and function, which is essential in most mammals, non-mammalian vertebrates and invertebrates. Essentially, the system is a set of glands which produce and secrete the extracellular signaling molecules known as hormones. The main glands that make up the human endocrine system include hypothalamus, pituitary, thyroid, parathyroid, adrenals, pineal body, and the reproductive organs (ovaries and testes). Hormones are chemical messengers sending information to coordinate the function of different parts of the body. The blood vessels are mainly used as information channels for the endocrine system by passing many hormones or endocrine signalers through the bloodstream. However each hormone will communicate only with specific target cells (Hardly, 1996).

When the hormone is released, it is rapidly channeled from the endocrine gland into the bloodstream directly where it can be transported to designated cells. The special proteins act as a carrier by binding to some hormones in order to control the quantity of hormones for the cells to use. Each type of hormone has its own appropriate receptor to latch on when the hormones arrive at a target organ. The hormone-receptor combinations then pass on chemical signals to the inner working of the cell. In addition, the specific receptor in the target cells reacts to the small amount of the endocrine messengers (hormones) to regulate body function. Moreover, endocrine systems maintain the hormone levels in the blood when they reach a certain amount. For example, the most important part of the endocrine system is the pituitary gland that controls other endocrine glands. It produces thyrotropin to control thyroid hormones produced from the thyroid in burning and producing energy from food. Another organ, the hypothalamus, manages the endocrine and nervous systems by producing chemicals to either stimulate or suppress hormone secretions from the pituitary. In addition, the pituitary gives off chemical signals to the nervous system called endorphins to reduce the feeling of pain. The pituitary also secretes hormones which act on the reproductive organs to control ovulation and the menstrual cycle in women.

2.1.2 Main categories of endocrine-disrupting chemicals

The accepted definition of endocrine disrupting compounds (EDCs) from the United States Environmental Protection Agency (USEPA) was stated by Kavlock et al. (1996) as “an exogenous agent that interferes with the production, release, transport, metabolism, binding, action or elimination of natural hormones in the body responsible for the maintenance of homeostasis and the regulation of development processes”. Rhind’s, definition (2002) is that “endocrine disruptors are exogenous substances that alter functions of the endocrine system and cause health effects in an intact organism, or progeny, or population”. Another definition of endocrine-disruptor is broadly defined as “an exogenous chemical substance or mixture that alters the structure or function(s) of the endocrine system and causes adverse effects at the level of the organism, its progeny, populations, or subpopulations of organisms, based on scientific principles, data, weight-of-evidence, and the precautionary principle.” (Phillips and Harrison, 1999). Currently, endocrine disruptors are considered to operate on a mode or mechanism of action causing potentially detrimental effects such as reproductive effects, developmental effects, and carcinogenic consequences. Overall, the focus of all definitions is that endocrine disruptors are substances that interfere with endogenous endocrine function (Propper, 2006).

Endocrine disrupting compounds include both natural and synthetic chemicals. The main sources for environmental contamination are manufactured chemicals, industrial, agricultural, and municipal wastes. Agricultural runoff can expose organochlorine insecticide such as DDT, dieldrin, lindane, as well as pesticides currently in use such as atrazine, trifluralin, permethrin. Moreover, natural hormones produced naturally by animals such as 17β-estradiol, estrone, testosterone, and hormonal active growth promoters in livestock production such as progesterone, trenbolone, zeranol have also been released from farm animal wastes. Synthetic steroids found in contraceptives, such as 17D-ethynylestradiol and mestranol, are contaminated environment via domestic/ municipal waste. Manufactured chemicals may release unusually high concentrations of surfactants of detergents as discharges from municipal treatment systems such as the alkylphenolic group - nonylphenol, and phthalates found in plasticizers - dibutylphthalate and butylbenzyl.

There are three main categories of EDCs. First are the synthetic chemicals such as personal care products, agrochemicals, pharmaceuticals and industrial chemicals. Secondly, the natural chemicals such as natural hormones and steroids are secreted by animals. Thirdly, 7 phytoestrogens is the last group of EDCs found in plants such as soybean. They are introduced to the environment mostly via rivers downstream of municipal wastewater effluent. The hydrophobic properties help those EDCs to linger on in soil or silt, air, and water, as well as persisting long enough to be assimilated into the food chain. EDCs easily enter into animals when they are consumed. For example, PCBs, dioxin and the insecticide DDT are persistent and slow to degrade in the environment even when there is no longer production (Smith, 1995, Juberg, 2000, Fox, 2001). Hence, potential sources of exposure are through contaminated food eaten by livestock, which can pass to humans, contaminated air and the particle containing air-sprayed pesticides from agriculture inhaled by organisms, drinking water contaminated by chemicals or sewage effluent, and contaminated recycled water (Wild and Jones, 1992).

Potential mechanisms of endocrine disruptors can be observed through interference of the normal functioning of the endocrine system in several ways. Firstly, the substances act as a chemical messenger or hormone and bind to a target receptor and cause a similar action in cells called the agonist response. For example, studies on rainbow trout treated by alkylphenolic compound, that are weakly estrogenic, have shown inhibition of the testicular growth (Jobling et al., 1996b), and when they were treated with 17α-ethynylestradiol, which mimic the natural female sex hormones, resulted in feminizing effects (Jobling et al., 1996a). Secondly, the antagonists response works by the endocrine disrupting compounds binding to receptors and causing abnormal responses in cells. For instance, the function of estrogen hormone receptors was interrupted by polychlorinated dibenzo-p-dioxins (PCDDs). It interacts with the aromatic hydrocarbon receptor of hormones. The complex undergoes a structural transformation (or activation) followed by translocation to the nucleus, and the activated receptor complex then recognizes and binds to specific regions. Hence, the required function cannot occur as well as a multitude of other biological responses such as altered patterns of growth and differentiation (Ohtake et al., 2003). Thirdly, the substance interferes with natural hormones and controls receptors’ function. For example, the storage or release of hormones were disrupted by EDCs, which causes abnormal levels of hormones (Danzo, 1997).

Routledge and Sumpter (1997) stated that EDCs can be estrogenic or anti-estrogenic (blocking natural estrogen), androgenic or anti-androgenic (blocking natural androgen) (Van Wyk et al., 2003), and thyroidal (compounds with direct or indirect impacts on the thyroid gland) (Snyder et al., 2003b). Furthermore, their estrogenic potency depends highly on the

8 size and degree of branching of the alkyl group, and its position on the phenol ring. However, among different groups of EDCs the estrogenic potency cannot be explained solely by structures, hence a concept titled ‘Relative Estrogen Potency’ was proposed for the practical evaluation of the ecotoxicity of suspect compounds. An 17β-estradiol equivalent (EEQ) expressed the estrogenic potentials of an individual EDC for risk assessment of an ecosystem, especially a water system receiving effluent discharge (Wild and Jones, 1992). For example, organotin compounds such as tributylin, dioxin and dioxin-like compounds as polychlorinated biphenyl (PCBs) can act in an anti-estrogenic way, while alkylphenols, pesticides, synthetic steroids and phthalate esters were categorized as estrogenic compounds (Colborn and Clement, 1992, Van Den Berg et al., 1998).

2.1.2.1 Synthetic chemicals

2.1.2.1. Pesticides

Some pesticides, such as atrazine and trifluralin, are potentially endocrine-disrupting and have been known to cause adverse effects on wild animal such as birds, marine mollusks, and alligators (Gibbs and Bryan, 1994, Guillette, 1995, Crisp et al., 1998). DDT has been used as an insecticide since the 1940s, and was forbidden in most developed countries in 1973 because of the persistency in the environment (Crisp et al., 1998). DDT was a major cause of the decline in the bird population because of eggshell thinning and of embryo mortality in fish-eating birds (Bignert et al., 1995, Walker et al., 1996).

The studies of o,p-dicofol showed eggshell thinning, embryo mortality, and feminization of male embryos in American kestrels (Mclachlan and Arnold, 1996). Also, the exposure to dicofol led to a decline in the number of alligators in Lake Apopka in Florida, USA from abnormality in fertility and the decrease in egg viability in the 1980s (Guillette, 1995).

Tributyltin (TBT), which is used in paint as the active ingredient to prevent fouling on boats and harbour equipment, can alter the reproductive organs of mature female mollusks (Gibbs and Bryan, 1994). A herbicide called atrazine resulted in abnormal gonads in African clawed frogs (Hayes et al., 2002). Pesticides such as metribuzin, methomyl, and aldicarb cause a significant increase in thyroxine levels which leads to hyperactivity, hyperirritability and hypermetabolic rates in laboratory rats that resulted in learning difficulties (Porter et al., 1993).

2.1.2.1.2 Industrial chemicals

Alkylphenols such as nonylphenol and octylphenol are widely used in industrial detergent, domestic detergent, and pesticide formulation. These chemicals are exposed to humans and animals via air pollution, water pipe treatment, and absorption through skin from personal care products such as shampoos and cosmetics. Exposure also results through consumption of contaminated drinking water, inhalation and ingestion of pesticide sprays, and contamination of food from sewage sludge (Warhurst., 1995). Furthermore, nonylphenol and related chemicals can induce vitellogenin production and development of intersex at 1-10 μg L-1 in fish (Gray and Metcalfe, 1997, Gronen et al., 1999), and at >5.4 μg L-1 in sheeps head minnow (Hemmer et al., 2001).

Bisphenol A has been used as a plasticizer and is widely used in many food and beverage packaging products. Therefore, exposure to this chemical may result from the consumption of contaminated canned food such as artichokes, peas, beans, mixed vegetables, corn and mushrooms or polycarbonated bottles of beverages (Brotons et al., 1995). Research on bisphenol A indicated that this chemical can behave like an estrogenic-mimic, and anti- androgenic in human breast cancer cell culture assay (Krishnan et al., 1993, Brotons et al., 1995). As shown in Table 2.1, bisphenol A at 1 μgL-1 also caused feminization in freshwater and marine snails (Oehlmann et al., 2000).

PCBs, PCDDs, and PCDFs are mainly used in the electrical supply and mining industries. Exposure to these chemicals can be via the atmosphere or water from waste incineration, and chemical manufacturing. The chlorinated PCDDs, furan, and PCBs have anti-estrogenic potential, while those that are less chlorinated are estrogenic (Chaloupka et al., 1992, Safe, 1995). The adverse effects on neurological and intellectual function on humans were found in children born from women who consumed PCB contaminated fish (Jacobson et al., 1990). In wildlife, PCB contaminated sediments can be exposed to bottom-dwelling fish, and result in decreasing sperm quality (Nagler and Cry, 1997). Baltic grey seals, Baltic ringed seals, common seals, and Beluga whales also suffered from reproductive impairment when fed with PCB contaminated fish (Bergman and Olsson, 1985, Brouwer et al., 1989, Reijnders, 1986).

2.1.2.1.3 Pharmaceutically active chemical (PhACs)

Pharmaceutically active chemicals (PhACs), used in the medical industries, can enter into the environment easily because of their high water solublility. These PhACs and their metabolites

10 contaminate the environment and food supply because of incomplete removal in wastewater treatment plants (WWTPs) (Richardson and Bowron, 1985, Singer et al., 2002).

Antibiotics, antiphlogistics, antiepileptic, lipid regulators, β-blockers, and synthetic estrogens are some of the pharmaceutically active compounds, which often have contaminated drinking water and source water via sewage outflow (Desbrow et al., 1998, Rhind, 2002).

2.1.2.2 Natural Chemicals. Natural EDCs include the natural hormones, which are excreted from mammals, including humans. These hormones are 17β-estradiol, estrone, estriol and testosterone, which are released into the environment from domestic discharge, agricultural runoff, and sewage treatment plants (STP) (Desbrow et al., 1998, Routledge et al., 1998b). The natural hormones and their actions are described in section 2.2

Phytoestrogens are natural plant chemicals, which present estrogenic activity. These chemicals are found in fibres, whole grains, and soybeans. Exposure to a phytoestrogen, genistein, has been shown to cause the development of uterine cancer in mice, the decrease in the size of the thymus, and impacts immunological indicators (Newbold et al., 2001, Yellayi et al., 2002). However, Barrett (1996) and Zava and co-workers (1997) declared that phytoestrogens are not on the priority lists for management actions. Phytoestrogens, such as genistein, differed from the xenoestrogen DDT in their ability to regulate cell proliferation

2.1.3 Occurrences and adverse effect of EDCs The adverse effects of EDCs on humans and wildlife are a cause of concern as exposure to EDCs cause adverse effects shown from certain observations on wildlife, fish and ecosystems, the large amount of evidence of endocrine-related human diseases, and the adverse effect from experimental animals in laboratories (Table 2.1). The confirmable effects of EDCs on wildlife are impaired reproduction and development, deformities and embryo mortalities, abnormal reproduction, depressed thyroid and immune functions, and feminization (Sharpe, 1993, Routledge et al., 1998a, Rhind, 2002, Young et al., 2002, Roepke et al., 2005).

Table 2.1 Example of Endocrine Effects in Wildlife and Humans

Chemicals Species Dose and Effects References Route Bisphenol A Male fish 0.001 to 1.4 μg inducing the production (Young et L-1 in sewage of vitellogenin al., 2002) effluent 17β-Estradiol Male Fish < 1ng L-1 to inducing the production (Young et 308ng L-1 in of vitellogenin al., 2002) sewage effluent Estradiol Sea Urchin 0.7 to 77.3 ng Mimicking endogenous (Roepke et mL-1 in treated hormones al., 2005) environment 4-Octylphenol Sea Urchin 0.035 to 0.489 Inhibiting normal (Roepke et (OCT) ng mL-1 in hormone activities and al., 2005) treated metabolism environment Tributyltin (TBT) Sea Urchin 0.731 to 1.689 Inhibiting normal (Roepke et ng mL-1 in hormone activities and al., 2005) treated metabolism environment Bisphenol A and freshwater 1 μg L-1 in Feminization such as (Oehlmann octylphenol and marine sediment development of addition et al., 2000) snail female organ as well as gross malformations of the pallial oviduct section Atrazine African 0.1μg L-1 in vitellogenin induction, (Hayes et clawed frogs living changes in sexual al., 2002) (Xenopus environment differentiation, changes laevis) in hormone levels and developmental Diethylstilbestrol Observed - the increasing incidence (Crisp et al., (DES) children of hypospadias and/or 1998) cryptorchidism in children during development 17α- Observed long-term a risk cause of (Palmer et Ethynylestradiol women contraceptive development of breast al., 2002) use cancer Hexachlorobenzene Observed - three or four time to have (Hardell et mother son with testicular cancer al., 2003) p,p’-DDE and Observed - the reduction of the (Sharpe, phthalates men average sperm count by 1993) 50% from 113 million mL-1 to 66 million mL-1

2.1.3.1 Effect found in Wildlife

A variety of endocrine-related effects in fish and wildlife such as impaired reproduction and development, deformities and embryo mortalities, abnormal reproduction, depressed thyroid and immune functions, and feminization, have been researched by scientists in many parts of the world.

2.1.3.1.1 Fish

The studies on water contamination found that fish have been exposed to EDCs via water they live in. For example, the sewage and pulp mill effluent, and pesticide contamination in agricultural runoff, surface water and groundwater can expose fish and destroy a host of aquatic life (Wester, 1991, Routledge et al., 1998a, Fox, 2001). The concentrations of endocrine disrupting chemicals in sewage effluents and river water have been a cause of concern as they are present in the part per trillion (ppt) levels. The levels of 17β-estradiol and estrone range from < 1 ng L-1 to 308 ng L-1, as well as a plasticizer, bisphenol A, was present in sewage effluent ranging from 0.001 to 1.4 μg L-1 (Young et al., 2002). Jobling and co- worker (1996a, 2002b) also indicated that estrogenic chemicals induced the production of vitellogenin in male fish at either the same or higher level with spawning females. Hermaphroditic fish developed with intersex characteristics when receiving sufficiently estrogenic chemicals.

2.1.3.1.2 Invertebrates

The knowledge on the effect of endocrine disruptors for invertebrates is limited because endocrine-disruptors are caused via an endocrine related mechanism. Freshwater and marine snails have been studied extensively because of the decrease or even extinction of local populations of marine snails in coastal areas all over Europe and in the open North Sea (Matthiessen and Gibbs, 1998). Bisphenol A, octylphenol, and tributyltin (TBT) can lead to feminization, such as the development of additional female organs as well as gross malformations of the pallial oviduct section. Imposex and loss of shell development were also observed in marine snails (Matthiessen and Gibbs, 1998, Oehlmann et al., 2000, Gooding et al., 2003).

2.1.3.1.3 Reptiles and Amphibians

Reptiles and amphibians had altered hormone levels, sexual differentiation, as well as vitellogenin induction and developmental difficulties as a result of the effect of EDCs. For example, Guillette (1995) demonstrated that there was distorted sex-organ development and 13 function in alligators and also most of their eggs did not hatch. Additionally, abnormal plasma sex hormone levels and reduced phallus size were found in juvenile alligators and turtles after the spill of the pesticide, dicofol, in a Florida lake. Furthermore, the African clawed frog (Xenopus laevis) exposed to atrazine at 0.1μg L-1 showed gonadal abnormalities, and chlordane, dieldrin, endosulfan, and toxaphene caused a considerable morose of vitellogenin (Hayes et al., 2002)

2.1.3.1.4 Birds

Endocrine disruptors caused a decline in egg hatching, eggshell thinning, skewed sex ratios, abnormal reproductive tract development, abnormal plasma level of sex hormones, abnormal nesting and mating behavior, and poor egg quality (Fox, 2001, Feyk and Giesy, 1998, Bignert et al., 1995). For example, DDT and DDE inhibit the synthesis of prostaglandin, causing a reduction in calcium transport to the eggshell. Eggshell thinning in birds caused a severe population decline in the number of raptor bird species (Bignert et al., 1995). Fox (2001) indicated that birds living near pulp and paper mills showed a significant lack of survival in the young. PCBs and PAHs cause feminization in gulls living around the Great Lakes, USA. (Feyk and Giesy, 1998).

2.1.3.1.5 Mammals

As demonstrated by Crisp and co-workers (1998), an exposure to PCBs, PCDDs, mercury and dieldrin can affect reproduction impairment, immune repression and masculinization in mammals. Both reproduction and immune function of mammals have been impaired by EDCs in the food chain. For instance, the contamination of PCBs, dieldrin, and PCDDs in fish caused abnormalities such as a high prevalence of cancer in beluga whales and a severe decrease in population of Wadden Sea harbour seals. Moreover, the servere infertility in sheep was a result of phytoestrogen in subterranean clover. It caused differentiation of cells in the cervix, resulting in sperm finding it difficult to travel through the cervix and into the reproductive tract, which leads to severe infertility in sheep (Adams, 1998).

2.1.3.2 Effects found in People

Damstra and co-worker (2002) mentioned that there are two possible ways for humans to be exposed to EDCs. The most direct way is via the workplace or via consumer products such as food, plastics, detergents, paints, and cosmetics, as well as indirect exposure to EDCs from the air, water, and soil (the environment). 14

2.1.3.2.1 Effects in children

A decreased proportion of boys to girls has been another cause for concern and is due to the effect of EDCs in children (Hood, 2005) . Damstra and co-workers (2002) indicated that EDCs influence the development and function of the reproductive organs causing a decline in the amount of males born. The result was in good agreement with the PCB and dioxin studies by James (1997) and Vreudgenhil and co-workers (2002). The more feminine and less masculine behaviours in boys were observed as the effects of exposure to dioxin or PCB (James, 1997, Vreudgenhil et al., 2002). In addition, increasing amounts of hypospadias and/or cryptorchidism in children during development resulted from exposure to DES and various pesticides (Crisp et al., 1998).

2.1.3.2.2 Effects in women

Reproductive disorders and breast cancer were found as the common effect of EDCs in women. PCBs and PCDDs were found to cause severe reproductive disorders, which relate to malfunctions in the immune system and the endocrine system (Johnson et al., 1997). Long- term contraceptive use is a potential cause of the development of breast cancer (Palmer et al 2002), although a variety of factors such as age, race, family history, and total lifetime estrogen exposure may increase a risk of breast cancer. From this point of view, through exposure to 17β-estradiol and DES, the development of breast cancer, vaginal cancer, and infertility is possible.

2.1.3.2.3 Effects in men

Estrogenic chemicals including PCBs, dichlorodiphenyldichloroethylene (p,p’-DDE) and phthalates were observed to have negative effects in men, such as reduction in average sperm count, testicular cancer and prostate cancer. The reduction of the average sperm count by 50% from 113 million mL-1 to 66 million mL-1 was influenced by the exposure to PCBs (Sharpe, 1993). Moreover, significant levels of the PCBs, hexachlorobenzene and nonachlordane, in pregnant women increased the chance of having a son with testicular cancer by a factor of four (Hardell et al., 2003).

2.2 Hormones and their actions

The endocrine system is actually a series of specialized glands, which releases hormones that perform as endocrine signalers. Hormones are released into the bloodstream and pass directly

15 through several bodily functions. The lipophillic property supports hormones diffusing through the cells to bind the intended protein called ‘hormone receptors in nucleus’. The hormone chemical messengers interact with the specific hormone receptors in cells, as the key fitting into the lock, to protect against false messages, and consequently result in prompt normal biological functions (Gower, 1975).

The human steroidogenesis generates androgen, testosterone corticoids, cortisol, aldosterone, estrogen and progesterone from cholesterol.

O OH CH CH 22 3 3 21 OH CH CH 12 3 17 11 20 13 H 16 1 9 14 2 10 8 15 H H HO 3 5 18 19 HO 7 HO 23 4 6

17 β-estradiol Estrone 17α-ethynylestradiol

OH OH O CH CH CH 3 3 3 OH CH 3 CH 3 H H CH 3 H H H H H H H O HO O

Estriol Progesterone Testosterone

Figure 2.1 Structure of hormone steroids.

Estradiol

Estradiol is also known as estradiol-17 β; estradiol; estradiol-17 β; beta-estradiol; dihydrofolliculin; dihydroxyestrin; 1,3,5(10)-estratriene-3,17b-diol; 3,17-dihydroxy- delta(1,3,5-10)-estratriene; 3,17-epidihydroxyestratriene; estradiol-17 beta; 17 beta-estradiol; 16 estra-1,3,5(10)-triene-3,17beta-diol (Figure 2.1) (Barton and Andersen, 1998). The molecular formula is C18H24O2 and the molecular weight is 272.39.

Estradiol (17β-estradiol) is the most active form of estrogen in humans. It is present in both males and females as a sex hormone. The production of estradiol starts from a derivative of cholesterol. The androstenedione is converted to testosterone, followed by conversion of testosterone to estradiol by aromatase. Alternatively, androstenedione is aromatized to estrone, which in turn undergoes conversion to estradiol by 17β-hydroxysteroid reductase (Barton and Andersen, 1998).

In women, the granulose cells of the ovaries produce most of the 17β-estradiol during the reproductive years, and smaller amounts are produced by the adrenal cortex. Furthermore, fat cells are active in converting precursor hormones, specifically testosterone, to estradiol, resulting in the production of 17β-estradiol even after menopause. In men, 17β-estradiol is produced by the testes. Additionally, both sexes produce estradiol in both the brain and arterial walls.

In women 17β-estradiol effects reproduction. It acts as an important female hormone which is necessary to maintain oocytes in the ovary for a normal monthly menstrual cycle. Moreover, 17β-estradiol increases during pregnancy to facilitate placental production. As well as being one of the vital components to sexual development and secondary sex characteristics in women, which is monitored most during the reproductive years and during its decrease in peri- and post-menopausal women, 17β-estradiol is responsible for supporting normal function in the physiological process. This includes enhancing breast development, changing body shape affecting bone and joint mineral density and osteoporosis (as an accelerated loss of bone mass related to estrogen deficiency), preventing arteriosclerosis and sustaining heart and blood vessel health. This is due to 17β-estradiol having complex effects on the production of multiple proteins such as lipoproteins, binding proteins, and proteins responsible for blood clotting. In addition, skin elasticity and fat structure are modified by 17β-estradiol, which are reduced without 17β-estradiol resulting in lean skin and wrinkles (Barton and Andersen, 1998).

In males, 17β-estradiol is produced in the sertoli cells of the testes and also plays a significant a role.There are small amounts of 17β-estradiol normally in men’s bodies, but the responsibility of testosterone does not allow 17β-estradiol to have any significant physiological effect on the male’s body. An increasing level of 17β-estradiol in men affects a 17 noticeable weightgain, especially in mid-age men, and the development of enlarged breasts, and the abatement of sex drive. A high level of 17β-estradiol can be found in males with sex chromosome genetic conditions, such as Klinefelters Syndrome (Barton and Andersen, 1998).

Estrone

Estrone is also known as estrol; estrin; estrone; folliculin; ketohydroxyestrin; 1,3,5(10)- estratrien-3-ol-17-one; Oestrone; beta- Estrone; Estra-1,3,5(10)-trien-17-one, 3-hydroxy-; and 3-Hydroxy-1,3,5(10)- estratrien-17-one (Figure 2.1).

Estrone is a hormone, presenting itself in post-menopausal women. It is the least abundant of the three major estrogens, which include estriol (the primary hormone during pregnancy), and 17β-estradiol (the mainly active female hormone in the reproductive female body). The synthesis of estrone occurs in both the ovary and adipose tissue via aromatase from androstenedione (Figure 2.2) (Barton and Andersen, 1998).

The conversion of estrone to estrone sulfate is the general effect to human health because of a long-lived derivative, and also it can be converted to 17β-estradiol by 17β-hydroxysteroid reductase (Figure 2.2) (Barton and Andersen, 1998).

Estriol

Estriol is also known as orstriol. Its molecular formula is C18H24O3 and the molecular weight is 288.38 (Figure 2.1).

Estriol is a metabolite of androgens and estrogens produced by the human body. It is the hormone produced during pregnancy by the placenta. Normally, levels of estriol in non- pregnant women, menopause women and men are not significantly different (Melamed et al., 1997).

H C 1) Cholesterol sidechain 3 CH 3 cleavage cytochrome P-450 CH 3 comples CH 3 CH a. 20- hydroxylase 1b 1a 3 b. 22-hydroxylase HO c. 20,22-lyase 1c Cholesterol 2 )C21 sidechain cleavage CH 3 cytochrome P.450 complex O CH CH 3 a. 17-hydroxylase 3 O 4 CH b. 17,20-lyase 3 CH 3 CH 3 3) 17β-hydroxysteroid

HO dehydrogenase

O 4) 3β-hydroxysteriod 2a 4,5- Pregnolone Progesterone 2a dehydrogenase/ Δ Isomerase complex CH CH 3 4 3 O O CH CH 3 3 5) 5α-reductase, type I and II OH OH CH CH 3 3 6) Aromatse (cytochrome P- 450) O HO

17α-Hydroxyprenenolone 17α-Hydroxyprogesterone

CH O 2b 2b 3 O O CH CH 3 4 3 6 CH CH 3 3

HO O 3

Dehydroepiandrosterone Androstenedione Estrone OH CH 3 3 OH CH CH OH 3 3 CH 4 6 3 CH 3

HO O HO Androstenediol Testosterone 5 17β-estradiol

CH OH 3

CH 3

O H 5α-Diheydrotestosterone

Figure 2.2 Metabolic pathways for the synthesis of progesterone, estradiol, testosterone, and dihydrotestosterone (Barton and Andersen, 1998)

Testosterone

Testosterone is known as testosterone; trans-testosterone; G4-androsten-17b-ol-3-one; 17b- hydroxy-4-androsten-3-one; androlin; android; halotensin; oreton; testex; testoderm; testred; virilon; 17β-Hydroxyandrost-4-en-3-one; Androst-4-en-3-one, 17-hydroxy-, (17beta)-; and 4- androstene-17beta-ol-3-one (Figure 2.1). Its molecular formula is C19H28O2 and the molecular weight is 288.429 (Mooradian et al., 1987).

Testosterone acts primarily as the male hormone, although it is also synthesized in women. It is secreted by the testes in males and the ovaries in females, and is also secreted in small amounts by the adrenal glands in both sexes (Swerdloff et al., 1992).

In females, testosterone is responsible for maintaining the proper sex drive, bone health, mood and blood production. In males, decreasing testosterone levels especially with age, causes decreasing sex drive, decreasing body hair, shrinking of the testes, a gain in body weight and thinning of the skin. Low testosterone has also been linked to fatigue, depression, irritability, infertility and osteoporosis because testosterone is generally associated with libido, which when impaired result in improper testosterone levels (Mooradian et al., 1987, Swerdloff et al., 1992).

Progesterone

Progesterone is the main naturally occurring human progestogen. Progestrogens are characterised by having a basic 21-carbon skelton (Figure 2.1). Progesterone is a steroid hormone primarily produced by the ovaries and adrenal glands in women, which is involved in the female menstrual cycle, pregnancy and embryogenesis of mammals. Progesterone synthesis is from a derivative of cholesterol called pregnenolone (Figure 2.3). In addition, progesterone plays a major role in potential hormone production by being a significant precursor used to synthesize testosterone, estrone and 17β-estradiol.

O O CH CH 3 3 CH CH 3 3

CH CH 3 3

O HO

Progesterone Pregnenolone

Figure 2.3 Conversion of Pregnenolone to Progesterone

The level of progesterone is also a significant influence on human health. Excessive amounts of progesterone potentially affect the reproductive system relating to abnormal menstrual bleeding, increased urinary incontinence and inhibition of lactation. Hence, progesterone is the major component used in contraceptive pills, and it is one of the most widely used medications. Moreover, progesterone can affect the nervous system because it also belongs to a neurosteroid group of dehydroepiandrosterone and pregnolone. Accordingly, excessive progesterone can cause slowed response, depression, impaired memory, tiredness and sleepiness in both sexes.

17α-Ethynylestradiol

17α-Ethynylestradiol is an active synthetic steroidal estrogen, used commonly in pharmaceuticals as the formulation of oral birth control pills. 17α-Ethynylestradiol is the derivative of 17β-estradiol. It is more resistant to degradation, thus it is released into the environment from the urine and feces of people who take it as a xenoestrogen (Keam and Wagstaff, 2003).

2.2.1 Occurrences and adverse effects of endocrine disrupting steroidal hormones

Exposure to steroid hormones in humans and animals is through the discharge of domestic sewage effluents and disposal of animal waste. The excreted hormones from humans can be in active forms or inactive conjugated forms such as sulfate or glucuronide derivatives (Ternes et al., 1999). They can become active via microbial degradation in sewage treatment plants.

Table 2.2 Physiochemical properties of steroids (Ying et al., 2002)

Chemical Name Molecular Water Solubility (mgL-1 Vapour LOG ο 3 Weight at 20 C) Pressure Kow

(mm Hg) Estrone 270.4 13 2.3 × 10 -10 3.43 17β-Estradiol 272.4 13 2.3 × 10 -10 3.94 Estriol 288.4 13 6.7 × 10 -10 2.81 17α-Ethinylestradiol 296.4 4.8 4.5 × 10 -10 4.15 Mestranol 310.4 0.3 7.5 × 10 -10 4.67

All narutral and synthetic estrogen belong to the C18 steriodal group, sharing the same backbone, and tetracyclic molecular framework (Figure 2.1). They contain one phenol, two cyclohexanes and one cyclopentane rings. As demonstrated in Figure 2.1, the difference in E1, E2, EE2, and E3 is in the configuration of the D-ring at position C16 and C17.

Based on the physicochemical properties (Table 2.2) of natural and synthetic hormones, the distribution and partition of estrogenic steroids in the environment can be predictable.

Estrogen was moderately high octanol-water partition coefficient (log Kow) values, indicating the moderate hydrophobic and poor water solubility. The solubility of E1, E2 and E3 are two to three fold higher than that of EE2, so E2 and E1 are easily degraded within the body and also in the environment (Reddy et al., 2004). Chang and Jang’s (2005) study supported the lesser absorbtion of E1 being predominantly in sediments. Also, the synthetic estrogenic steroids such as EE2 and mestranol have lower water solubility, suggesting longer persistence in the environment. They are hydrophobic organic compounds, which are more susceptible to bio-degradation in water, but they avoid biodegradation by being present in lower concentrations in aqueous phase and partitioning in sediments and soil. Interestingly, Kile and Chiou (1989) suggested higher water solubiluity of EE2 due to the presence of surfactant in effluents.

Table 2.3 Concentrations of hormones in water source samples.

Compounds Source of sample Methods Concentration References (ngL -1) Estrone Italy (STP effluent) LC-MS-MS 9.3 (Baronti et al., 2000) Estrone France (Sludge) GC-MS 1-2 (Muller et al., 2008) Estrone France (STP effluent) GC-MS 0.2-5 (Muller et al., 2008) Estrone Japan ( wastewater) LC-MS 5200 (Furuichi et al., 2006) Estrone Denmark (Manure treated GC-MS/MS 68.1 (Kjaer et al., 2007) soils) 17β-Estradiol Australia (STP effluent) GC-MS 4.2 (Ying et al., 2008) 17β-Estradiol New Zealand (STP MCF-7 70 (Gadd et al., 2005) effluent) 17β-Estradiol Italy (STP effluent) LC-MS-MS 1.0 (Baronti et al., 2000) 17β-Estradiol Germany (STP effluent) GC-MS-MS <0.15-5.2 (Ternes et al., 1999a) 17β-Estradiol California (surface water) GC-MS-MS 0.05-0.8 (Huang and Sedlak, 2001) 17β-Estradiol New York (STP effluent) HPLC 0.18 (Ying et al., 2008) 17β-Estradiol Denmark (Manure treated GC-MS/MS 2.5 (Kjaer et al., 2007) soils) Estriol Italy (STP influent) LC-MS-MS 0.6 (Baronti et al., 2000) Estriol Italy (STP effluent) LC-MS-MS 0.2 (Baronti et al., 2000) Estriol Italy (river water) LC-MS-MS 0.02 (Baronti et al., 2000) Ethynylestradiol Italy (STP effluent) LC-MS-MS 0.45 (Baronti et al., 2000) Ethynylestradiol Australia (STP effluent) GC-MS 1.3 (Ying et al., 2008) Ethynylestradiol Germany(STP effluent) GC-MS-MS <0.10- 8.9 (Ternes et al., 1999a) Ethynylestradiol California (surface water) GC-MS-MS <0.05-0.07 (Huang and Sedlak, 2001)

As shown in Table 2.3, the concentrations of the estrogenic hormones in water from many developed countries indicated very low levels; around 0.1 to 10 ng L-1. In addition, the hormones are excreted in the urine of mammals, and enter the aquatic environment via wastewater treatment plant effluent at the ng L-1 level, which are sufficient to initiate negative impacts on aquatic organisms. This is due to the low molecular weight and the hydrophobic property of all hormones that helps them pass through biological membranes, causing adverse effects to humans and wildlife (Pacakova et al., 2009).

As Roepke and co-workers (Roepke et al., 2005) demonstrated, 17β-estradiol and 17α- ethynylestradiol at concentrations as low as 0.02 and 0.03 ng mL-1, respectively, were effective in decreasing the development of sea urchin embryos. The 17α-ethynylestradiol at < 23

0.4 ng L-1 caused vitellogenin induction, and inhibited growth and development in male fish such as rainbow trout (Sumpter and Purdom, 1994, Lange et al., 2001) The study in male wild fish found that 17β-estradiol at 1-10 ngL-1 and 17α-ethynylestradiol at 0.1 ng L-1 lead to feminization (Desbrow et al., 1998). Exposure to 17β-estradiol at 20, 100 and 500 ng L-1 showed a significant decrease in sexual activity in juvenile male mosquitofish (Doyle and Lim, 2002).

These studies suggest that the sewage treatment technologies have limitations in removing endocrine disrupting steroids from sewage. The concentration of estrogenic hormones, which were detected from STP effluents as shown in Table 2.3, can pose significant risks to the environment.

2.2.2 Estrogenic potency of estrogens that is ecologically meaningful

The most potent natural steroidal hormone is E2. It’s IC50 (the concentration of estradiol required to reduce the colour development by 50%) in an immunoassay has been used as the reference for calculating the estrogenic potency for other compounds. Also, 17β-estradiol equivalent (EEQ) is used for comparing estrogenic potency of E2 and other compounds. The

EEQ is calculated by the ratio of the IC50 of E2 to the IC50 of a target compound.

Cargouet and co-worker (2004) used an estrogen responsive receptor cell line (MELN) assay to demonstrate that E1 showed roughly four-fold less estrogenic potency than E2, and E3 300−fold less estrogenic potency to equal potency with E2, which agreed with an estrogenic yeast screening assay of Wood and co-worker (2007). Moreover, EE2 has been shown to have higher estrogen potency than E2, with an EEQ ranging from 1.03 to 1.2 (Cargouet et al., 2004, Woods, 2007)

The levels of E2 and EE2 causing adverse effects have been related to their estrogenic potency. As the most potent estrogen, exposure to E2 by rainbow trout, even at < 10 ng L-1 resulted in a measurable reproductive change (Routledge et al., 1998a). Also, the induction of vitellogenin synthesis in male fish has been induced by E2 at 1 ng L-1 (Dorabawila and Gupta, 2005). In the same way, EE2 can cause a dramatic reduction in reproductivity of fish at 2 to 6 ng L-1 (Routledge et al., 1998a, Snyder et al., 1999, Robinson et al., 2003). Therefore, it is

24 important to assess regularly the quality of water receiving discharges of effluents from wastewater treatment plants.

2.3 Analytical Methods for Detecting Endocrine Disrupting Compounds

During the last two decades, many researchers have reported steroid sex hormones and related synthetic compounds in wastewater effluents in the range of lower ng L-1, and the relative effects that these hormones have on aquatic organisms that receive significant inputs of wastewater effluents. The detecting estrogenic hormones in wastewater effluents or in estaurinewater at extremely low concentrations are an important analytical challenge. As can be seen in Table 2.3, instrument analysis methods such as gas chromatography/ mass spectrometry (GC/MS), or gas chromatography coupled with tandam mass spectrometry (GC/MS/MS) were often employed to analyze these estrogenic hormones in environmental matrices. These methods provide sensitivity, but less accurate quantification because of the interfering organic matter( wastewater contains many compounds that can affect the target analytes) (Hegari and Andersson, 2007). Moreover, the levels of estrogenic hormones in water (< 0.05- 0.1 ng L-1 Huang and Sedlak, 2001) is much lower than the detection limit of many analytical instruments such as GC-MS/MS (0.1-2.4 ng L-1) ( Belfroid et al., 1999), LC- MS/MS (2.7 – 11.7 ng L-1) (Matani et al., 2005), GC-MS (1.3-11 ng L-1) (Morteani et al., 2006). Hence, immunoassay, with a general detection limit of 0.05 - 0.2 ng L-1(Schneider et al., 2005, Ingerslev and Sorensen, 2003), is an alternative choice for quantifying estrogenic hormones in environmental samples.

2.3.1 Instrument methods

Instrumental methods are mostly used for environmental monitoring in order to identify and quantify mixtures of compounds in water. Despite the sensitivity of these techniques, they are often expensive, time-consuming, labor intensive, a technical expert is needed, and thus they are not widely utilized, except in well-established laboratories. Additionally, the extensive cleanup and pre-concentration step required by instrumental methods makes these methods even more challenging. The established sample preparation step for estrogenic hormones is solid phase extraction, which is time-consuming for sample loading, washing, and sample elution.

2.3.1.1 Gas chromatography/ mass spectrometry (GC/MS) and (GC-MS/MS)

GC/MS combines features of gas chromatography and mass spectrometry to identify different substances within a test sample. This tool has been used for tracking organic pollutants in the environment, especially estrogens in sewage samples (Desbow et al. (1998);Ternes et al. (1999a), While GC/MS can directly detect many endocrine disrupting compounds (EDCs), it often requires derivatization to make analytes compatible with it (Snyder et al., 1999). Kelly (2000) stated that estrogenic hormones including estrone, 17β-estradiol, estriol and 17α- ethynyloestradiol can easily form conjugates thus becoming less volatile. Therefore, GC needs to introduce a particular label into those compounds to raise their volatility for sensitivity and selectivity. An unique derivatization procedure for estrogenic hormones was described by Kuch and Ballschmiter (2001) by converting the estrogens to pentrafluorobenzoylate esters and then detecting them by high−resolution gas chromatography with negative chemical ionization mass spectrometry (GC/NCI-MS) in selective ion monitoring (SIM) mode.

Another powerful technique is tandam mass spectrometry (MS/MS), which is the adding of a second phase of mass fragmentation for an improved identification. A primary MS is applied to isolate a precursor and a second MS is applied to analyze the product ions. This technique enables quantitation of low concentrations of target analytes in the presence of a high sample matrix background. The limit of quantification (LOQ) for such a method is typically in the ngL-1 range. Ternes et al., (2002) developed GC/MS/MS to detect estrogens down to 2 ng g-1 in sludge, and down to 0.2 ng g-1 fresh water sediments. Although this analytical method is very powerful and has a number of advantages, the high capital cost in the instrumentation, the complicated operation and maintenance are limitations to the use of these methods in a routine fashion.

2.3.1.2 High Performance liquid chromatography (HPLC)

The HPLC/UV method was employed by Itoh et al (2001) to analyze an estrogen sulphate mixture and unconjugated oestrogen in pharmaceutical formulations, and Penalver et al. (2002) determined estrogen in natural water resources by HPLC/UV and found the limit of detection between 0.3 to 0.4 μg L-1 for 17β-estradiol, estriol and estrone.

2.3.1.3 Liquid chromatography / mass spectrometry (LC/MS) and (LC-MS/MS)

Liquid chromatography coupled with mass spectrometry (LC/MS) is a technique that merges the physical separation technique of liquid chromatography with the mass analysis capabilities of mass spectrometry. It is a powerful technique which was developed to overcome the limitation of GC/MS. Generally, no derivatization is required prior to analysis, which reduces the operating time and sample preparation. Due to the sensitivity and specificity, the application is generally oriented towards the specific detection and potential identification of chemicals. LC/MS is a powerful tool for analysis of compounds with high polarity, thermally labile, and relatively larger molecular weight (Snyder et al., 2003b). Zuehlke et al. (2005) analyzed the estrogens in sewage treatment plants using LC/MS, and found the LOQ of estrone, 17β-estradiol, and 17α-ethynylestradiol <2 ngL-1. However, the method suffered from low selectivity when applied to very complex matrices such as sewage treatment samples.

Another powerful method, which promises higher−sensitivity and greater−selectivity for the quantification of trace levels of analysis in environmental monitoring, is LC/MS/MS. Lagana` et al. (2004) demonstrated the capability of LC/MS/MS to quantify the natural sex hormones such as estrone, estriol, 17β-estradiol, as well as other estrogenic compounds such as phytoestogen, daidezin, and 4-nonylphenol. Applying this method to analyzing estrone, 17β- estradiol, and estriol in STP revealed levels below 10 ng L-1 and other estrogenic compounds such as zeranols, bisphenol A at level below 30 ng L-1.

2.3.2 Biological methods

The biological techniques are gaining more recognition for the analysis of estrogens and progesterone in the environment because they offer fast, cost effective and simple methods of detection, with sensitivity and specificity compared to the classical analytical methods. Several bioassays such as cell proliferation assays, receptor dependant and receptor binding assays, yeast screening assays and immunoassay are used for the recognition of the biological activity of a variety of endocrine disrupting compounds. Moreover, most biologically based methods are usually carried out in combination with one conventional instrument method such as GC, LC or MS.

2.3.2.1 Immunoassay

Immunoassay is a technique combining chemistry and immunology principles to measure analytes normally present at very low concentrations. The basic principle is based on the specificity of the antibody-antigen reaction. Immunoassay is commonly used in the measurement of drugs, hormones, specific proteins, tumor markers, and markers of cardiac injury.

Immunoassay was introduced for environmental monitoring due to the contaminants present at very low levels in certain environments which cannot be determined accurately by other less expensive tests.

Immunoassay is divided into two types, homogenous and heterogenous and the heterogenous immunoassays can be further divided into competitive or non- competitive. A competitive immunoassay is a format whereby an antigen in a sample and a labeled antigen compete to bind to an antibody. The response is measured by the amount of labeled antigen bound to the antibody, so the greater the response, the less antigen in the sample. On the other hand, non- competitive immunoassays are binding of antigen in the unknown to the antibody site, and then the labeled antibody is bound to the antigen. The results of non-competitive methods are directly proportional to the concentration of antigen (Wild, 2000).

Radioimmunoassay (RIA)

Radioimmunoassay is the standard method for quantitative determination of small amounts of clinically significant compounds since Yalow and Berson (1994) developed the first radioisotopic technique to study blood volume and iodine metabolism. This method uses radioactive isotopes to label either the antigen or antibody and measure the gamma rays which are emitted from those isotopes following removal of free radiolabel (Berson and Yalow, 1996). This technique is very sensitive, as low as one trillionth of a gram of material per milliliter of sample, and it also requires only a small sample for measurement. A RIA is a rapid assay with direct signal detection. However, RIA is limited by high investment costs of expensive instrumentation, a need for special disposal of radioactive waste, and associated potential health and safety risks. Alternative methods such as enzyme immunoassays have largely replaced the RIA in routine laboratories (Schuurs and Van Weemen, 1977).

Enzyme (EIA) immunoassay

Because of the drawbacks of RIA, enzyme immunoassay (EIA), particularly the ELISA method has replaced RIA. The ELISA methods use colourimetric signals to measure the antigen-antibody reaction, and the enzyme is labeled either on the antibody or antigen.

Fluorescent immunoassay (FIA)

Fluorescent immunoassay (FIA) is a technique which has greater analytical sensitivity than Enzyme (EIA) immunoassay due to using a fluorescent molecule to label either the antigen or antibody to form a fluorescent product. In simple terms, an unknown amount of antigen is affixed to a surface, and then a specific antibody is washed over the surface so that it can bind to the antigen. This antibody is linked to a fluorescent product. When light of the appropriate wavelength is shone upon the sample, any antigen/antibody complexes will fluoresce so that the amount of antigen in the sample can be inferred through the magnitude of the fluorescence. These methods have inherent advantages such as no radioactivity, a much longer useful lifetime of the assay kit, and less expensive measurement instrumentation. The associated disadvantages are improper instrumentation, and the variation and magnitude of the sample’s own fluorescence which often lead to decreasing sensitivity (Soini and Hemila, 1979).

Chemiluminescent immunoassays

In chemiluminescent immunoassays, the response is measured from the emission of chemiluminescent light. The light that is produced from chemiluminescent molecules are excited by chemical energy and detected by a light detector.

2.3.2.1.1 Advantages and disadvantages of immunoassays.

Immunoassays offer a number of advantages for numerous kinds of contaminants, over the conventional methods, such as HPLC and GC. Immunoassays provide a specific method because of the molecular recognition of antigen-antibody binding, and the remarkable distinguishing capability of antibodies. Immunoassays are sensitive because of the high affinity of antibodies. Immunoassays are fast methods due to their high throughput capability that reduces the time of analysis, which suits routine monitoring. Cost efficiency is another advantage of immunoassays. They are less than ten times lower than the capital cost for

29 setting up the laboratory for instrumental analyses. Hence immunoassays provide a fast, simple, and a cost-effective method with high−sensitivity and specificity.

Although immunoassays provide numerous advantages, the limitations of these methods should be considered. The use of immunoassay as fast screening with further confirmation by conventional methods as quality assurance is still required (Lee and Kennedy, 2007). Meulenberg and co-workers (1995) stated that immunoassays are not widely used in laboratory as routine analysis because an antibody that lacks single compound specificity, cross−reacts with structurally related compounds in samples and therefore, leads to overestimation. The use of broad broad-specificity antibodies that may detect a large range of related compounds, however, can be very useful in selecting positive samples that may be further analyzed by an instrumental method. Such approach can significantly save time and cost of analysis. The time required to develop immunoreagents for new analytes is significantly long (Meulenberg et al., 1995). Moreover, immunoassays are not suitable to analyze for multiresidues without prior knowledge, because two to five related compounds can be detected but with different sensitivity by one antibody, leading to overestimation (Lee and Kennedy, 2007). Finally, potential matrix interference can lead to false positive results.

2.4 ELISAs (Enzyme- Linked Immunosorbent Assay)

ELISAs are a common clinical diagnostic tool, which are mainly used in immunology to determine the presence of an antibody or an antigen. This technique has been developed and applied in a wide variety of environmental samples. ELISAs have gained popularity because of a low limit of detection, easy-handling techniques, and because they are user-friendly and cost effective. Highly selective and sensitive ELISA methods have been developed for determination of steroid hormones and pharmaceuticals, and the ELISAs have demonstrated excellent LODs of less than 10 ng L-1 without cross reaction (Goda et al., 2000, Schneider et al., 2004, Schneider et al., 2005, Hientemann et al., 2006).

2.4.1 Principle

Antigens in the human body or foreign substances are counteracted by antibodies. Antibodies are part of a class of proteins called immunoglobulins (Igs) which are comprised of two pairs of heavy chains and light chains joined to form a “Y” shape molecule (Nahm and Hoffmann,

1990). There are two classes or isotypes of the light chain called kappa (N) and lambda (O). The heavy chains contain five different isotypes which divide the Igs into five different classes (i.e., IgG, IgA, IgD, IgM, and IgE) with different functions. However, the most common immunoglobulin class is IgG. Each chain is divided into constant and variable regions. The constant region dominates 90% of the total mass of the antibody. However, the variable region is the most important with regard to antibody-antigen interaction (Henniona and Barcelo, 1998).

2.4.1.1 Antibody binding and the interaction between antibody and antigen The fundamental principle of an immunoassay is based on the antibody-antigen interaction. Antibodies are special molecules produced by the body to fight off infection. They are protein molecules, as are many of the important, hard working components of the body. The antibodies have prominently versatility to make molecules that bind specifically to particular target molecules or antigens and to no other. The body produces a vast range of different antibodies so that if it has to combat a new bacteria or virus it is certain to find an antibody that is just right for it. The antibodies are produced by special factory cells called B cells, and each B cell produces its own particular form of antibody (Crowther, 2000).

Antigen binding sites; every antibody has

Protein chains; there are four in each antibody

Carbohydrate part of antibody Stem region: The structure of the stem decide the typ of antibody: IgA,IgG,IgE,IgM and IgD

Figure 2.4 Anatomy of an antibody (Jonathan and Linda, 1989)

Antibodies are y-shaped molecules, as illustrated in Figure 2.4. At the tip of each arm is an antigen-binding site where the antibody can bind to the particular feature of the antigen that it recognizes. These antigen-binding sites are the most changeable part of the antibody molecule. The stem of an antibody can also vary, although not as much. There are five basic types of stem, and they produce five different types of antibodies, known as isotypes. The names of these isotypes are: IgG. IgA, IgM, IgD and IgE. Ig stands for immunoglobulin- another name for an antibody (Mian et al., 1991, Roux, 1999) .

2.4.1.2 Hapten design and synthesis

Appropriate hapten design and synthesis is important in manipulating the specificity and sensitivity of an immunoassay (Crowther, 2000, Lee and Kennedy 2007). Since selectivity is dependent on the immunogen used to raise an antibody, structural design of a hapten is critically important, especially when the size of the target molecule is small and the number of determinant groups becomes fewer. There are several criteria for hapten synthesis and design for antibody production (Goodrow and Hammock, 1998). They are as follows: 1. A good hapten should be a near perfect mimic of the target structure in size, shape and electronic properties. 2. The chemistry of the target molecule must be understood, reviewed and evaluated for methods of insertion of the attachment handled by established chemistry. 3. The hapten handle (required for connecting the target-mimic to a carrier protein) should not elicit antibody recognition. Thus, an innocuous methylene handle often is most appropriate, the length of which should be evaluated using one or more homologs. 4. An appropriate functional group for covalent attachment of the handle to the protein must be compatible with the chemistry of the functional groups on the target.

2.4.2 ELISA format An ELISA is a heterogeneous analysis since either the antibody or an antigen is coated on a solid phase. ELISA can be divided into three different formats: competitive ELISA, including direct and competitive indirect ELISA, and non-competitive ELISA, involving a double antibody sandwich ELISA. Both competitive direct and indirect ELISAs (cd ELISA and ci ELISA) are normally used for sensitive analysis of small molecules such as pesticides, mycotoxins and antibiotics. In contrast, a double antibody sandwich ELISA is used for

macromolecule detection such as protein. However, the common ELISA used widely for detecting contaminants is the competitive direct ELISA format.

A competitive direct ELISA relies on the measurement of un-occupied sites by limiting antibody concentration (Crowther, 2000). In Figure 2.5, the antibody is immobilized onto a solid phase such as tubes or microwells. Then, a sample and a known amount of enzyme- labeled antigen are added to compete for restricted antibody sites on the support. After incubation, unbound sample antigens and labeled antigens are removed. The amount of bound enzyme-labeled antigen is detected by a spectrophotometer after adding a suitable substrate/chromogen solution. Since the colour produced is inversely proportional to the amount of analytes in the sample, the more analyte in the sample the lighter the colour in the solution.

Sample and labeled Two molecules Elimination of Enzymatic reaction analyte added to compete for the unbound molecules started by addition of antibody coated binding sites by washing substrate/chromogen plate Sample Labeled analyte

Figure 2.5 Schematic presentation of a direct ELISA.

2.4.3 Sensitivity and specificity of the assay

Sensitivity is the ability of an assay to detect the smallest amount of target analyte under the defined conditions (Crowther, 2000). The sensitivity of an immunoassay is generally defined

as an IC50; which was the concentration of analyte required to reduce the colour development by 50% (Lee and Kennedy, 2007).

Specificity is a crucial factor indicating the accuracy of an assay (Crowther, 2000). Monoclonal antibodies, as the name implies, can offer better specificity than polyclonal antibodies because they react as a single population of antibody with a single affinity (Crowther, 2000). Immunogens were prepared using the hapten conjugate and with the large

33 molecules of protein, antibodies subsequently were raised in the animal. Rabbit is one of the best sources of high−affinity polyclonal antibodies for small molecular analytes (Huang, 1999). Cross−reactivity and the variation in a specific response against a required single site of antigen make the measurement of a specific activity more complex (Crowther, 2000). Determination of assay specificity can be expressed by cross−reactivity of an antibody for similar structural compounds of target analytes. Cross−reactivity is calculated as the ratio between the IC50 values of the antigen and the cross-reactant as follows (Henniona and Barcelo, 1998; Lee and Kennedy, 2007).

-1 % Cross reactivity = IC50 of test compound (mol L ) ×100 -1 IC50 of analyte (mol L )

2.4.4 Matrix interference

Since a competitive ELISA is based on interaction between an antibody, an antigen, and a labeled antigen, it can be affected by pH, ionic strength of a water sample and presence of other interfering substances from the sample matrices (Henniona and Barcelo, 1998; Lee and Kennedy, 2007). In general, structurally related compounds, and high concentration of synthetic or natural compounds such as dissolved or halogen organic carbon in a sample matrix which can lead to false positive responses (Oubiña et al., 2000). Li et al (2004) found that methanol greater than 10% reduced the assay sensitivity. Similarly, Schneider et al (2004) stated that the absorbance declined and there was a clear decrease in sensitivity with increasing amounts of organic solvent in the sample. From the above studies, it can be illustrated that both solid phase extraction (SPE) and liquid-liquid extraction (LLE) were employed for simple extraction and pre-concentration prior to measurement, since the estrogenic hormones in aquatic samples was very low at less than 1 ng L-1. This step can reduce high−matrix interferences (Li et al., 2004, Packov et al., 2009). For example, the ELISA itself has a LOD of 0.1 μg L-1 estrone in water (without preconcentration), which in combination with SPE was decreased down to 1.3 ng L-1 (Li et al., 2004)

The matrix effects can be determined by measuring the recovery of a real sample spiked at known levels using a standard curve, with the standard solutions provided within the kit, or by

34 constructing dose-response curves with real samples spiked within the working range, including a blank run (Henniona and Barcelo, 1998). Any change between the curves (e.g, shift to the left or right) whilst keeping parallelism, corresponds to a loss or an gain in the assay sensitivity. The percent recovery can be expressed as the following:

Measured increase in concentrat ion (ng/mL) % Recovery (R) u 100 Spike level (ng/mL)

2.5 Estrogenic hormone determination by ELISAs

As mentioned above, the ELISA technique is indispensable for rapid and inexpensive screening for steroid compounds in both clinical diagnostic and environmental monitoring. The commercial ELISA test kits are available from several manufacturers. They were used to determine steroidal hormones in water samples in many studies. For example, Huang and Sedlak (2001) assessed the ability of commercially available ELISAs to quantify estrogenic hormones from conventional wastewater treatment plants and compared the results with GC/MS/MS. The study showed that the ELISA kit provided lower detection limits than GC/MS/MS. the ELISA kits that were employed showed higher sensitivity for estrogen than HPLC-MS/MS and GC/MS/MS (Farr´E et al., 2007, Hirobe et al., 2006).

These estrogens ELISA kits are marketed for clinical diagnostic purposes, and mainly for high−concentrations in biological samples. It was demonstrated that these assays were sufficiently sensitive for the environmental application with much greater throughput capacity than other chemical based methods. ELISA test kits have reported an overestimation of results when they were employed to environmental samples. One of the limitations of the commercial ELISA is insufficient specificity for structural similarity of estrogens, and the complexity of wastewater samples which may occasionally cause incorrect results.

Development to improve the precision and accuracy of an ELISA is challenging. Solid phase extraction (SPE) has been used for pre-treatment water samples to reduce water matrices. The results obtained were in good agreement and SPE is a suitable means of pretreating the samples for ELISA (Farr´E et al., 2007, Li et al., 2004). Li et al (2004) have found that the polyclonal antibodies produced by using estrone-3-hemisuccinate, coupled to keyhole limpet heamocyanin (KLH) raised in rabbits, exhibited high specificity to estrone, and low cross- 35 reactivity with other structurally-related compounds, including 17β-estradiol, estriol, and 17α- ethynyloestradiol. It showed good recoveries of spiked river water samples in Penrith (Australia) at 20 and 100 μgL-1 of 90-105 and 99-110%, respectively, indicating that river water following SPE cleaned up did not interfere with the ELISA. Schneider and co-workers (2004) synthesised the highly specific hapten using a biotinylated long chain for the determination of ethynylestradiol with the LOD at 14 ngL−1. The sensitivity of the existing ELISA can be improved by using chemiluminescence detection. A chemiluminescence ELISA was developed for the detection of ethinylestradiol (EE2) with the detection limit found to be 0.2 + 0.1 ngL−1 (Schneider et al., 2005). In another study, a specific polyclonal antibody was produced against a conjugate of estrone-3-hemisuccinate and keyhole limpet hemocyanin (KLH) in order to determine estrone in water samples. The obtained ELISA showed very low cross−reactivity to the major estrogens (17β-estradiol, estriol and 17α- ethynylestradiol) with LOD at 0.14 μgL−1(Li et al., 2004).

2.6 Recombinant Yeast Screening Assay

In general, a yeast recombinant screening assay is a technique, which relies on yeast constructs expressing a human estrogen receptor. The yeast strain Saccharomyces cerevisae was introduced with the mammalian steroid receptors in order to function as steroid- dependent transcriptional activators (Wright et al., 1990). The recombinant yeast characterizes chemical interaction with the estrogen and androgen receptors.. In most cases, it produces enzymes as the response to an estrogen/androgen stimulus, which can be easily spectrophotometrically quantified by using a chromogenic substance. (Wright et al., 1990, Rehmann et al., 1999).

The yeast assay system can detect chemicals whether they are agonists or antagonists. The estrogenic activity of p-nonylphenol, biphenol A, o,p’-DDT, o,p’-TDE and methoxychlor was evaluated by a yeast estrogen screen (YES) assay, (Soto et al., 1994). These synthetic compounds disturb endocrine systems by competing with 17β-estradiol for binding to the estrogen receptor. Several synthetic chemicals such as DDT metabolites, like DHT and p,p’- DDE, were determined for the interaction of, and their activity compared to dihydrotestosterone for androgen receptor assay. The study found that half the maximal -9 6 effective concentration (EC50) of DHT was estimated at 3.5 x 10 M , and p,p’-DDE was 10 - fold less potent than dihydrotestosterone Also p,p’-DDE has currently been identified as an 36 androgen receptor antagonist by using the yeast androgen receptor assay by interacting at a dose approximately 106-fold higher than testosterone which is active at a concentration of 10- 10 M (Kelce et al., 1995, Kevin et al., 1997a).

2.6.1 Steroidal hormone determination by a Recombinant Yeast Screening Assay

The endocrine disrupting steroids have been investigated for their potential to affect reproductive health and development at extremely low concentrations. Those compounds within effluents have been tested for toxicity or hormone disrupting activity, usually in order to determine safe discharge levels (Gaido et al., 1997b, Aneck-Hahn et al., 2005). As Wright and co-workers (1990) introduced a useful technique by applying the mammalian steroid receptors into the yeast strain Saccharomyces cerevisae for producing steroid-dependent transcriptional activators. This technique resulted in the ability to quickly process large sample numbers inexpensively, with rapid attainment of stable transformance, limited metabolic capabilities, and ease of manipulation (Wright et al., 1990, Gaido et al., 1997b, Rehmann et al., 1999, Céspedes et al., 2004). As previously described, steroidal hormones are mainly excreted by mammals and enter into the environment, especially though the aquatic system. These estrogens have been determined to be the major contributors of estrogenic activity in water.

The yeast-based assay has been used to assess chemical interactions with estrogen, androgen and progesterone receptors. The synthetic estrogen diethystilbesterol (DES) was as effective as estradiol, and they were the most potent in the estrogen receptor assay. Testosterone and dihydrotestosterone were most active in the androgen receptor assay (Gaido et al., 1997). The yeast-based assays were developed for greater sensitivity, specificity and were much faster than the conventional YES assay by assisting enzymatic digestion with lyticase (LYES). This technique provided a very high-quality alternative by performing 65 hours faster than the existing estrogenic yeast assay because of a digesting step with lyticase enhanced the sensitivity and speed of the assay (Metzger et al., 1992). Furthermore, the estrogen receptor assay determined a concentration range from 10-12 to 10-8 M of 17β-estradiol, while the androgen receptor assay only responded to the concentration range from 10-10 and 10-8 M testosterone and dihydrotestosterone, respectively. In addition, progesterone and hydrocortisone were inactive on the yeast estrogen receptor assay and cortisol was inactive for the yeast androgen receptor assay (YAS). The YAS assay determined the weakly androgenic

37 response at approximately 80,000 times less potentency than testosterone for diethylstiboestrol (DES), while progesterone and estrone found very slight activity concentrations above 10-8 (Metzger et al., 1992, Gaido et al., 1997a, Sohoni and Sumpter, 1998).

CHAPTER 3 DEVELOPMENT OF SENSITIVE ELISAs SPECIFIC TO SYNTHETIC HORMONES, 17α- ETHYNYLESTRADIOL AND MESTRANOL

3.1 Introduction

Water is a very valuable and limited resource, particularly in drought-prone countries with unpredictable weather patterns. The idea of introducing treated recycled water into the portable water supply system remains attractive for general household and agricultural purposes. However, the use of recycled water is not without issues, namely accumulative residual chemicals that may have adverse effects on natural ecosystems and, subsequently, the health of the entire food chain. In recent years, a great deal of research interest has been directed towards studying chemicals that possess the ability to mimic and/or disrupt the endocrine system by altering the natural balance of hormones in the human body. These compounds have been termed endocrine disrupting compounds (EDCs) as their biological and ecological activities influence the endocrine system of animals (Colborn et al., 1993, Oberdorster and Cheek, 2001). Gutendorf and Westendorf (2001) identified endocrine disruptors as being estrogenic in nature. Estradiol, estrone and estriol have been detected in municipal and industry effluents in Europe, North America and Asia (Belfroid et al., 1999, Ternes et al., 1999, Oh et al., 2000, Sole´ et al., 2000).

The synthetic hormone, 17α-ethynylestradiol (EE2), is one of the most common synthetic hormones prescribed as an oral contraceptive. It has been detected at trace levels in the low ngL−1 range in sewage effluents and surface water (Ternes et al., 1999, Baronti et al., 2000, Huang and Sedlak, 2001, Kuch and Ballschmiter, 2001, Matêjícek and Kubă, 2007). Intersex fish, inhibition of gonadal growth, and delayed progression of spermatogenesis were also discovered in aquatic wildlife exposed to EE2 at the ng L-1 levels. Other examples include the inhibition of gonadal and spermatogenic development in Sydney rock oysters exposed to 5 ng L-1 EE2 (Andrew et al., 2008), the changing in the AhR and ER-signalling gene expression pattern in Atlantic salmon exposed to 5 ngL-1 EE2 (Mortensen and Arukwe, 2007), and the reduction of fecundity and gemetes development in zebrafish exposed to 2 ngL-1 EE2 (Xu et al., 2008).

The extensive research carried out to develop analytical methods in the past few decades unfortunately suffers from one of the following shortcomings: inability to detect ultra low concentration levels such as parts per trillion, extensive sample preparation, potential inaccuracies due to matrix interferences, and cross−reactions with other molecules with similar chemical structures or properties, leading to false results in the case of antibody or receptor-based assays (Huang and Sedlak, 2001, Farr´E et al., 2007, Li et al., 2007). Much of the recent research has been conducted to develop more effective analytical methods to quantify parts per trillion levels of these EDCs, using gas-chromatography mass spectrometry (GC-MS), gas-chromatography tandem mass spectrometry (GC-MS/MS), liquid chromatography mass spectrometry (LC-MS) and liquid chromatography tandem mass spectrometry (LC-MS/MS) (Ternes et al., 1999a, Céspedes et al., 2004, Heisterkamp et al., 2004, Yang et al., 2006, Zhang et al., 2006).

These chemical methods are highly precise and accurate compared to biological methods. However, they do suffer drawbacks of the expensive instrumentation, high running costs and also the requirement for highly trained staff to operate them properly. In addition, because they are limited by their detection level in uncleanup samples, additional sample preparation is required prior to the analysis, often with the use of solid-phase extraction for extensive concentration and cleanup as well as derivatision to enhance volatility for GC.

Immunochemical assays such as the ELISA can be used because it has high−throughput capacity, analysing a greater number of samples at a lower cost, which is beneficial for large scale screening purposes. ELISAs can be used for identifying and quantifying specific hormones even at ultra low concentrations (at low parts per trillion levels) in a mixture of chemicals due to high specificity of antibody to its antigen, and with no or little sample preparation. Several studies have been published using clinical diagnostic ELISA test kits for measuring environmental EE2; however, the applicability of such ELISA test kit was neither considered nor properly assessed.

A radioimmunoassay method (RIA) using a monoclonal antibody (MAb) reagent for detecting EE2 with a sensitivity of the assay at 10 pg mL-1 in a buffer sample, and 43 pg mL-1 in unpurified plasma samples, was clinically useful to assess low levels of EE2 in patients having the estrogenic hormone contained in many oral contraceptives. Fluorescent immunoassay using total internal reflection fluorescence (TIRF) detection and energy transfer immunoassay (ETIA) achieved a limit of detection for EE2 of 0.01- 0.85 μg L-1 (Coille et al., 40

2002). A chemiluminescence enzyme-linked immunosorbent assay was developed for the direct detection of EE2 in water with the limit of quantification (LOQ) at1.4 + 0.8 ng L-1 (Schneider et al., 2005). The technique of enzyme-linked immunosorbent assay (ELISA) was also employed in determining EE2 in environmental and industrial waste samples, and commercial ELISA test kits demonstrated method sensitivity with sample preparation as the lowest quantification limit of 0.05 μg L-1 (Hirobe et al., 2006) and 0.01 ng L-1 (Hintemann et al., 2006). A new long chain biotinylated EE2 derivative was used to develop ELISA for the determination of EE2, and the limit of detection was found to be 14 ng L-1 (Schneider et al., 2004).

In this chapter, the facile synthesis and complete characterisation of the EE2-hapten with both acetate and butyrate linkers through its carbon number 3 is presented. The development and characterisation of the assay for detecting ultra low 17α-ethynylestradiol concentrations with reduced matrix interference and low cross−reactivity are also included. Validation of the sensitive immunoassays by evaluation of assay performance in various water matrices is also demonstrated.

3.2 Materials and Methods

3.2.1 Materials and Instrumentation

3.2.1.1 Materials

17α-ethynylestradiol, 17β-estradiol, estriol, estrone, estradiol dipropionate, progesterone, medroxyprogesterone, ethynylestradiol-3-methyl ether, 17α-ethynylestradiol-3-cyclopentyl ether, bovine serum albumin (BSA), 3,3’5,5’-tetramethylbenzidine (TMB), horseradish peroxidase (HRP) and keyhole limpet homocyanin (KLH), goat anti-rabbit IgG-peroxidase, bovine serum albumin (BSA) and Tween 20 were purchased from Sigma (St. Louis, MO). Analytical grade organic solvents (dimethyl sulfoxide (DMSO), dimethylformamide (DMF) and ethanol (EtOH)) and general chemicals for ELISA reagents were purchased either from Ajax Finechem (NSW, Australia) or Sigma Aldrich (St. Louis, MO, USA). Maxisorp polystyrene 96-well plates were supplied by Nunc (Roskilde, Denmark).

3.2.1.2 Instrumentation

Antibody concentration and immunoassay absorbance were measured by a SpectraMax® M2, multi-detection microplate reader from Molecular Devices (Sunnyvale, California). A pH metre was from TPS Pty. Ltd. (Brisbane, Australia).

3.2.2 Hapten Synthesis

Two haptens of 17α-ethynylestradiol were synthesised by Ms Christine Tan (UNSW). The preparation of 17α-ethynylestradiol-acetate (EE2-ACT) and 17α-ethynylestradiol-butyrate (EE2-BUT) is described in the following sections.

3.2.2.1 Attachment of acetate linker onto C3 of 17α-ethynylestradiol

Overall, 17α-Ethynylestradiol was first alkylated to attach a linker onto a phenolic hydroxyl group at C3 to give the ester compound 1 (Scheme 1) which was further hydrolysed to give the acid 2 (Scheme 2). Compound 2 was purified via crystallisation before being converted into the active ester 3 (Scheme 3).

Procedure for Synthesis of Ester compound (1), Scheme 1

To a solution of 17α-ethynylestradiol (1.0 g, 3.37 mmol) in acetone (50 mL) was added potassium carbonate (931.5 mg, 6.74 mmol) and ethylbromoacetate (450 μL, 4.04 mmol). The reaction mixture was stirred under reflux for 24 h at 70-75ºC with a calcium chloride drying tube. The reaction mixture was evaporated to dryness under vacuum, redissolved in dichloromethane (50 mL) and washed with citric acid (2 x 30 mL, 2 g/100 mL) and saturated brine (30 mL). The organic layer was dried over sodium sulfate and evaporated to dryness under vacuum, giving a yellow oil of crude compound with a yield of 1.436 g, Rf = 0.67 1 (dichloromethane/ethyl acetate, 12:1). H NMR (300 MHz, acetone-d6): δ 0.89 (s, 1H), 1.25 (t, J = 7.1 Hz, 3H), 2.98 (s, 1H), 4.21 (q, J = 7.2 Hz, 2H), 4.64 (s, 2H), 6.62 (d, J = 2.7 Hz, 1H), 6.69 (dd, J = 2.9, 11.4 Hz, 1H), 7.21 (d, J = 8.7 Hz, 1H).

Procedure for the Synthesis of Acid compound (2), Scheme 2

Compound 1 (1.29 g, 3.76 mmol) was dissolved in distilled tetrahydrofuran (THF)/methanol (MeOH) (40 mL, 3:1) and lithium hydroxide (37.6 mL, 1.0 M) was added. The reaction 42 mixture was stirred at room temperature for 2 h and washed with citric acid (2 x 50 mL, 10 g/100 mL) followed by saturated brine (1 x 60 mL) then extracted with ethyl acetate (2 x 100 mL). The organic layer was dried over sodium sulfate and evaporated to dryness under vacuum, giving 1.25 g (94% yield) of a white solid of crude compound 2, Rf = 0.38 (ethyl acetate/methanol, 19:1). Crude compound 2 (1.25 g, 3.53 mmol) was dissolved in ethyl acetate (12 mL) and cyclohexane (24 mL) was added to initiate the crystallisation process. The mixture was left overnight at room temperature before the crystals were filtered under vacuum using a sintered glass filter and rinsed first with the remaining solution mixture followed by cyclohexane. The crystals were left to dry under vacuum for an hour, yielding 1 987 mg (79% yield) of a white solid of compound 2. H NMR (300 MHz, acetone-d6): δ 0.89 (s, 3H), 2.98 (s, 1H), 4.62 (s, 2H [CH2 beside O]), 6.63 (d, J = 2.6 Hz, 1H), 6.70 (dd, J = 2.6, 11.3 Hz, 1H), 7.21 (d, J = 8.5 Hz, 1H).

Procedure for the Synthesis of Active Ester (3), Scheme 3

Compound 2 (500 mg, 1.41 mmol) and N-hydroxysuccinimide (194.7 mg, 1.692 mmol) were dissolved in dry THF (35 mL) under nitrogen with constant agitation. N,N’- dicyclohexylcarbodiimide (349.11 mg, 1.69 mmol) was dissolved in dry THF (10 mL) and added to the above mixture. A white precipitate formed during the initial stage of the reaction and the mixture was left to stir overnight at room temperature. The reaction mixture was filtered through a sintered glass filter under vacuum to remove the urea derivative, then evaporated to dryness under vacuum and re-dissolved in ethyl acetate (50 mL). The mixture was allowed to sit in an ice bath for 1 h before filtering through a sintered glass filter under vacuum to remove as much of the urea derivative as possible. The mixture was then evaporated to dryness under vacuum, yielding 738 mg of a white fluffy solid of crude 1 compound 3. Rf = 0.88 (ethyl acetate). H NMR (300 MHz, acetone-d6): δ 0.89 (s, 3H), 2.91 (s, 4H), 2.98 (s, 1H), 5.13 (s, 2H), 6.63 (d, J = 2.6 Hz, 1H), 6.78 (dd, J = 2.89, 11.5 Hz, 1H), 7.25 (d, J = 8.0 Hz, 1H).

18 17 OH OH 1. EtO CCH Br 2 2 K CO 2. 2 3

Acetone(reflux) OH EtO C 2 O 3 1

Scheme 1. Synthesis of compound 1

OH OH

1. LiOH THF/MeOH

EtO C O HO C 2 2 O 1 2 Scheme 2. Synthesis of compound 2

OH OH

1. NHS O 2. DCC O O N O HO C THF 2 O O 3 2

Scheme 3. Synthesis of compound 3

3.2.2.2 Attachment of butyrate linker onto C3 of 17α-ethynylestradiol The general synthetic approach for the butyrate linker is as follows: 17α-ethynylestradiol was first alkylated to attach a linker arm onto the phenolic hydroxyl group at position 3 of the steroid to give the ester compound 4 (Scheme 4) which was further hydrolysed to give the acid compound 5 (Scheme 5). Compound 5 was purified via crystallisation before being converted into the active ester 6 (Scheme 6).

Procedure for Synthesis of Ester compound (4), Scheme 4

To a solution of 17α-ethynylestradiol (500 mg, 1.69 mmol) in acetonitrile (50 mL) was added potassium carbonate (467.1 mg, 3.38 mmol) and ethyl-4-bromobuytrate (229.3 μL, 1.60 mmol). The reaction mixture was stirred under reflux for 48 h at 80–85 ºC with a calcium chloride drying tube. The reaction mixture was evaporated to dryness under vacuum, redissolved in dichloromethane (25 mL) and washed with citric acid (2 x 15 mL, 2 g/100 mL) and saturated brine (30 mL). The organic layer was dried over sodium sulfate and evaporated

44 to dryness under vacuum, giving a quantitative yield of 782 mg of a yellow oil of crude compound 4, including impurities and residual solvent, Rf = 0.61 (dichloromethane/ethyl 1 acetate, 12:1). H NMR (300 MHz, acetone-d6): δ 0.87 (s, 3H), 1.21 (t, J = 7.2 Hz, 3H), 2.03 (q, J = 2.6 Hz, 6.9 Hz, 2H), 2.47 (t, J = 7.3 Hz, 2H), 2.95 (s, 1H), 3.96 (t, J = 12.3 Hz, 2H), 4.10 (q, J = 7.1, 21.4 Hz, 2H).

Procedure for the Synthesis of Acid compound (5), Scheme 5

Compound 4 (357 mg, 0.87 mmol) was dissolved in THF/MeOH (6 mL, 3:1) and lithium hydroxide (8.97 mL, 1.0 M) was added. The reaction mixture was stirred at room temperature overnight and first washed with ethyl acetate (15 mL) to remove unreacted organic material, then with citric acid (15 mL, 10 g/100 mL) followed by saturated brine (1 x 30 mL) before being extracted with ethyl acetate (2 x 25 mL). The organic layer was dried over sodium sulfate and evaporated to dryness under vacuum, giving 312 mg (93.8% yield) of a white 1 solid of crude compound 5, Rf = 0.42 (ethyl acetate/hexane, 1:1). H NMR (300 MHz, acetone-d6): δ 0.88 (s, 3H), 2.04 (q, J = 3.6, 10.5 Hz, 2H), 2.49 (t, J = 7.4 Hz, 2H), 2.97 (s, 1H), 3.99 (t, J = 12.8 Hz, 2H), 4.05 (q, J = 7.2, 21.4 Hz, 2H), 6.62 (d, J = 2.6, 1H), 6.70 (dd, J = 2.9, 11.3 Hz, 1H), 7.19 (d, J = 8.1, 1H).

Procedure for the Synthesis of Active Ester (6), Scheme 6

Compound 5 (50 mg, 0.131 mmol) and N-hydroxysuccinimide (NHS) (18.07 mg, 0.16 mmol) were dissolved in dry THF (2.5 mL) under nitrogen with stirring. N,N- dicyclohexylcarbodiimide (DCC) (32.39 mg, 0.16 mmol) was dissolved in dry THF (1 mL) and added to the above mixture. A white precipitate was formed during the initial stage of the reaction and the mixture was left to stir overnight at room temperature. The reaction mixture was filtered through a sintered glass filter under vacuum to remove the urea derivative, prior to being evaporated to dryness under vacuum and subsequently re-dissolved in ethyl acetate (5 mL). The mixture was allowed to sit in an ice bath for 1 h before filtering through sintered glass under vacuum to remove as much of the urea derivative as possible. The mixture was then evaporated to dryness under vacuum, yielding 68 mg of a white fluffy solid of crude compound 6, Rf = 0.88 (ethyl acetate). Compound 6 (68 mg, 0.142 mmol) was dissolved in ethyl acetate (2 mL) and hexane (3.5 mL) was added to initiate the crystallisation process. The mixture was left overnight at room temperature before the crystals were filtered under vacuum using a sintered glass filter and rinsed first with the remaining solution followed by hexane. 45

The crystals were left to dry under vacuum for an hour, yielding 40.2 mg (59% yield) of white 1 solid. H NMR (300 MHz, acetone-d6): δ 0.89 (s, 3H), 2.17 (q, J = 6.9, 20.7 Hz, 2H), 2.83 (s, 2H), 2.88 (s, 4H), 2.98 (s, 1H), 4.08 (t, J = 12.4 Hz, 2H), 6.66 (d, J = 2.6 Hz, 1H), 6.73 (dd, J = 2.7, 11.1 Hz, 1H), 7.21 (d, J = 8.4, 1H).

OH OH 1. EtO C(CH ) Br 2 2 3 K CO 2. 2 3

Acetonitrile OH (reflux) EtO C 2 O 4

Scheme 4. Synthesis of compound 4

OH OH

1. LiOH THF/MeOH HOOC O EtO C 2 O 5 4

Scheme 5. Synthesis of compound 5

OH OH

1. NHS 2. DCC O O O HOOC O THF N O

O 5 6

Scheme 6. Synthesis of compound 6

3.2.2.3 Preparation of conjugates of hapten and carrier proteins or enzyme

Immunogens, coating antigen and hapten-enzyme conjugates were synthesised by Ms Christine Tan (UNSW). Immunogens were the conjugates of a carrier protein (KLH) and haptenic analogues of the hormone, 17α-ethynylestardiol-acetate (denoted as EE2-ACT) and 17α-ethynylestardiol-butyrate (denoted as EE2-BUT). Hapten-enzyme conjugates were the conjugates of HRP and haptenic analogues. All conjugations were performed via the NHS/DCC active ester approach.

For EE2-ACT and EE2-BUT haptens, each active ester dissolved in dry DMF was added dropwise to a protein in pre-cooled buffer solution (50 mM K2HPO4, pH 9.1). The reaction solution was gently mixed and let stand at 4qC overnight. The conjugated solution was then dialysed against phosphate buffered saline (PBS, pH 7.0) and stored at 4qC.

The estrone active ester was prepared according to Li et al. (2004). The conjugation of the active ester and a carrier protein/enzyme was performed in a similar manner as above.

For E2-OX and E2-HS (which were commercially obtained), conjugation to a carrier protein and enzyme was performed via the NHS/DCC active ester method without isolation. Briefly, NHS and DCC were added into a flask containing E2-OX in DMF, and the reaction mixture was allowed to stand overnight at 4qC. The precipitate of cyclohexyl urea was filtered through a cotton filter, leaving a clear supernatant for the subsequent conjugation. The active ester solution was added dropwise into a flask containing a carrier protein (e.g., KLH) in pre- cooled 50 mM K2HPO4 (pH 9.1). It was gently mixed and allowed to stand at 4qC overnight. The conjugated solution was then dialysed against phosphate buffered saline (PBS, pH 7.0).

3.2.3 Antibody Production and Characterisation

3.2.3.1 Immunisation and antibody production

New Zealand white rabbits were immunised with the immunogen EE2-ACT-KLH by subcutaneous injections. The rabbits were given the immunogen solution containing EE2- ACT-KLH in 0.9% NaCl (saline) (0.5 mL) emulsified with TiterMax adjuvant or Freund’s incomplete adjuvant (0.5 mL). Subsequent booster injections were given at monthly intervals. The blood was collected from the marginal ear vein and the antiserum was isolated by centrifugation. The same procedure was carried out with the EE2-BUT-KLH immunogen.

3.2.3.2 Purification of Rabbit IgG

Serum was purified by affinity chromatography on a protein A/G Sepharose. Firstly, a serum which was diluted with an equal volume of phosphate buffer was loaded onto the protein A/G column. The serum was eluted with acetate buffer (pH 4) at 1 mL/min. The eluent which contained an antibody fraction was immediately neutralised with 1 M Tris base (pH 11). The 47 purified antibody was extensively dialysed against phosphate buffer saline (PBS, pH 7.8) before storing at 4°C.

3.2.3.3 Determining antibody concentration

The concentration of purified polyclonal antibodies (PAbs) was determined by measuring the absorbance at 280 nm. The absorbance was measured against PBS as a reference. The concentration of antibody was calculated using the following formula:

u dA Antibody concentration (mg/mL) = H Where: A = absorbance at 280 nm d = dilution factor ε = IgG extinction coefficient, which is 1.35 (Nikolayenko et al., 2005)

3.2.3.4 Determining optimum working concentration by checkerboard titration

Checkerboard titration (CBT) was performed to determine the optimum working concentrations of each antibody and enzyme conjugate pair. The CBT was carried out as follows: 1. A microwell plate was initially coated with 10 μg mL-1 purified antibodies in coating buffer (50 mM carbonate buffer, pH 9.6), and left to stand overnight at room temperature. The plate was washed with the washing solution (0.05% Tween 20) and dried on an absorbent paper. 2. The plate was blocked with 1%BSA/PBS at 200 μL per well for 1 h at room temperature. Then, it was washed three times with the washing solution. 3. A HRP conjugate in 1% BSA/PBS was titrated in 3-fold dilutions, followed by incubation for 1h. Then, the plate was washed five times with the washing solution. 4. The colour reaction was initiated by incubating with a substrate/chromogen solution (1.25 mM 3,3’5,5’-tetramethylbenzidine-1.6mM hydrogen peroxide in acetate buffer, pH 5.0), 100 μL/well for 30 min. 5. The reaction was stopped by adding the stop solution (1.25M sulfuric acid, 50 μL/ well). The absorbance was measured at a dual mode at 450 nm and 650 nm.

3.2.3.5 Determining Sensitivity

3.2.3.5.1 Preparation of standard solution

Stock solutions of 100 and 10 mg L-1 17α-ethynylestradiol were prepared in anhydrous ethanol. To construct a calibration curve, a standard solution was freshly prepared in glass tubes by diluting 10 mgL-1 of stock solution, giving a 100 μg L-1 standard solution (in 10% EtOH). Then the 100 μgL-1 standard solution was serially diluted with 10% EtOH to obtain 33.3, 11.1, 3.3, 1.1, 0.4, 0.1, 0.04 and 0.01 μg L-1 standards.

3.2.3.5.2 Preparation of enzyme conjugate

The optimum concentrations of enzyme conjugates were determined by the CBT, as described earlier. The working solutions of enzyme conjugates for the ELISA were prepared by diluting an appropriate amount in 1% BSA/PBS.

3.2.3.5.3 Direct Competitive ELISA protocol

The purified anti-17α-ethynylestradiol antibodies were coated onto the microwell plate as previously described. After that, the plate was washed three times with washing solution (0.05% Tween 20), then dried on an absorbent paper. The unbound sites on the microwells were blocked with 200 μL of 1% fish gelatin in PBS (FG/PBS) per well by incubating for 1 h at room temperature. Then, the plate was washed and dried as previously described.

For the control, the wells were loaded with 100 μL of 10% methanol and 100 μL of enzyme conjugate. For the blank, the wells were loaded with 100 μL of 10% methanol and 1% BSA/PBS (as an enzyme conjugate diluent). Standard solution (100 μL per well) and enzyme conjugate (100 μL per well) were applied to the allocated wells and were incubated for 1 h at room temperature. Following the washing, 100 μL of substrate solution was added to each well. The colour reaction was developed for 30 min and then stopped by adding 50 μL of

1.25M H2SO4. The absorbance was measured using a dual mode at 450 and 650 nm.

3.2.3.5.4 Determination of standard curve parameter

The calibration curves (dose-response) of the analytes were obtained by plotting concentration on the x-axis versus % inhibition on the y-axis. Percent inhibition (%I) for each standard and sample was determined using the following formula:

ª AA blank º % Inhibition «1 » u 100 ¬ control AA blank ¼

Where: A = Average absorbance for each standard and sample

Ablank = Average absorbance for blank

Acontrol = Average absorbance for control Blank = matrix with no analyte, but with enzyme conjugate diluent (no enzyme conjugate added and gives background colour) Control = matrix with no analyte but with enzyme conjugate (no inhibition and maximum absorbance)

The assay sensitivity was determined as assessed by an IC50 and limit of detection (LOD) from the calibration curves. An IC50 is the concentration of analyte required to inhibit the antigen-antibody interaction by 50%. The LOD for the assay was the concentration that reproducibly provides 20% inhibition (IC20) of colour development.

3.2.3.6 Optimisation of EE2 ELISA condition

3.2.3.6.1 The effect of enzyme diluents on assay performance

BSA is commonly used in diluent buffers to reduce non−specific interactions and to stabilise the enzyme conjugate. Tween 20 can reduce non−specific binding of the conjugate. In order to determine their influence on assay performance, four enzyme-conjugate diluents consisting of BSA in PBS were tested. They were 1% BSA/PBS, 1%BSA/PBS+0.05% Tween 20, 1% BSA/PBS + 0.1% Tween 20 and 1% BSA/PBS + 0.25% Tween 20

3.2.3.6.2 The effect of organic solvents on the ELISA

The aqueous solubility of the analyte affects the performance of the immunoassays. In this experiment, several water miscible solvents (acetone, acetonitrile, methanol and ethanol) ranged from 5 to 40% were examined for their compatibility with the ELISA.

3.2.3.7 Determining specificity

Specificity of these ELISAs was tested by evaluating the relative sensitivity of structurally similar natural and synthetic estrogens to that of EE2. The nine test compounds included 17β- estradiol, estriol, estrone, estradiol dipropionate, progesterone, 17α-estradiol,

50 medroxyprogesterone, ethynylestradiol-3-methyl ether and 17α-ethynylestradiol-3- cyclopentyl ether. The % cross−reactivity is calculated as follows:

IC of test compound % Cros s reactivity 50 u 100 IC 50 of anlayte

3.2.3.8 Study of Matrix Effects

Assay stability was evaluated against metal ions, ionic strengths of dissolved salts and humic acid. Humic acid was tested because of its performance as absorbents and ion exchangers in aquatic environments. Different salts ((NH4)2SO4, NaCl, MgSO4, CaCl2, MnSO4, KCl,

NH4Cl, CuSO4 and Fe2(SO4)3) were tested for their compatibility with the ELISA at 0.01 to 1 M. Humic acid was tested at concentrations ranging from 0.1 to 1000 μg L-1.

Matrix effects were evaluated using three water samples collected from a lagoon at Tahbilk winery wetlands, McWilliam’s winery water channels, and sea water from Maroubra Beach; in New South Wales (NSW), Australia.

3.2.4 Spike and recovery studies

The spike and recovery studies were conducted to assess any matrix interference by the environmental water samples on the developed ELISA, as well as to assess the analytical performance. The spiked samples were assayed using the ELISA to quantify 17α- ethynylestradiol, which was calculated by intrapolating the EE2 concentration from % inhibition of the calibration curve. Table 3.1 shows the concentrations EE2, E2 and E1 used to prepare the spiked samples.

Table 3.1 The concetrations of 17α-ethynylestradiol (EE2), 17β-estradiol (E2) and estrone (E1) in the spiked samples. Spike 17α-ethynylestradiol 17β-estradiol Estrone Total No. (μgL-1 ) (μgL-1 ) (μgL-1 ) (μgL-1 ) 1 0.1 0.0 0.5 0.6 2 3.0 0.5 0.5 4.0 3 0.1 0.0 0.0 0.1 4 0.3 0.5 0.5 1.3 5 1.0 0.0 0.0 1.0 6 15.0 0.0 0.0 15.0 7 10.0 0.0 0.0 10.0 8 0.5 0.5 0.0 1.0

3.2.5 Validation of ELISA with GC/MS

Validation of the ELISA with GC/MS for real-world water samples was conducted in collaboration with Dr. Chunhua Li at the University of Sydney, NSW, Australia. All water samples were collected by Dr Li.

As shown in Figure 3.1, three different regions of New South Wales were involved in the sampling. These were South Creek in the Sydney Basin, Emigrant Creek at Ballina, and Wilson River at Lismore, the Northern area, NSW, as well as the Murray River at Albury- Wodonga at the border of NSW and Victoria. The samples were collected upstream of the discharge, at the discharge point and downstream of the discharge point. Effluent from selected sewage treatment plants was also collected.

All water samples were extracted and concentrated using SM2-Biobeads solid phase extraction (performed by C.H. Li). Briefly, 8 L of sample was extracted using the SM2- Biobeads solid phase extraction method to a final 10 mL volume and stored in dichloromethane (DCM) before being transferring into 2 mL LC sample vials, and blown to o dryness under N2 at 60 C, then re-dissolved in 1000 μL of 10% methanol.

An ELISA was performed as described in section 3.2.3.5.3 with 17α-ethynylestradiol as a standard. Each sample was assayed in at least three replicate wells. (All sample preparation for ELISA and ELISAs were performed by the candidate of this thesis).

3.2.5.1 Water and water samples Left: Map of Australia

A Right: Map of New South Wales, Australia A: Emigrant Creek in Ballina and Wilsons River at Lismore, Northern River area B B: South Creek in the Sydney Basin C: Murray River at Albury- Wodonga, border of NSW and C Victoria

Figure 3.1 Geographical locations where water samples were collected in NSW, Australia.

The analysis of estrogen by GC-MS was reported by Li et al. (2007). Briefly, an Agilent 6890 gas chromatography system coupled to a 5973 spectrometer (Agilent, USA) was used. A DB- 1701 capillary column (0.25 mm × 30 m × 0.25μm) (Agilent, USA) was used for separating the sample. The column temperatures were programmed as follows: the initial oven temperature was 80°C, increasing to 280°C with a ramp rate of 20°C per min. The temperature was kept at 280°C for 10 min and then cooled down to 80°C. The total running time was 30 min. The carrier gas was helium at 11.06 psi (on), purge flow 0.5 mL/min, purge time 2.0 min and total flow 8.7 mL/min. The injector temperature was maintained at 230°C and the injector volume was 4 μL in spitless inlet mode. Gas flow was constant with an initial flow of 1.1 mL/min under a nominal initial pressure of 11.07 psi. The software for data collection and analysis was Chemstation.

3.3 Results and Discussion

3.3.1 Hapten selection

The specificity of the antibodies produced towards the targeted steroid is significantly related to the moeity of a hapten molecule that is attached to a carrier protein. Much of the past work has centred on designing and synthesising the 17α-ethynylestradiol hapten by focusing on a linker attached to the 6-position (Kundu et al., 1977, Weber et al., 1989, Schneider et al., 2004). These synthesised derivatives were then used for developing immunoassays (Exley, 1972, Exley and Choo, 1974, Exley and Abuknesha, 1977, Exley and Abuknesha, 1978, Marcus and Durnford, 1988). A hemisuccinate was conjugated at C6 position, led to polyclonal antibodies better recognize the changes in the chemical structure of the steroids (Lindner et al., 1972). Li et al. (2004) also showed that the attachment of a linker to the –OH at the 3-position of the steroid molecule resulted in polyclonal antibodies with good sensitivity and high specificity. All estrogenic hormones share the same backbone structure of the 4 rings, differing only in their substituent groups on carbons 17 and 18. Therefore, the rationale of a hapten with a linker on carbon 3 is to trigger an immunogenic response towards a unique moiety of the 17α-ethynylestradiol. This mode of conjugation was chosen such that it preserves the acetylene substituent group on carbon number 17 which would allow the production of polyclonal antibodies that can specifically recognize this substituent group that is different from those of other estrogenic steroids. Hence, it is predicted to produce polyclonal antibodies that are able to differentiate 17α-ethynylestradiol from other steroids such as 17β-estradiol and estriol. These compounds have one and two hydroxyl groups, respectively, instead of an acetylene group as substituents on carbon number 17, thus reducing the cross-reactivity with these structurally similar steroid molecules.

Ms Christine Tan has successfully synthesized two hapten molecules, one of which has an acetate linker, previously mentioned in a Russian patent (Komarov et al., 2003) and the other with a butyrate linker. The short chain lengths of 2−4 carbons consisting of saturated carbon bonds ensures the flexibility of the hapten molecule after it is conjugated to the carrier molecule. The flexibility of the linker can allow for greater interaction with the immune system upon immunization, and saturated bonds in the linker is less likely to trigger an immunogenic response against the linker. Equally important is the length of the linker as a short linker will decrease the possibility that the hapten will fold back towards the carrier

54 molecule, reducing the chances for specific antibodies to be produced (Lee and Kennedy, 2007).

3.3.1.1 Optimal Concentration of Enzyme Conjugates

Estrogenic haptens used for conjugate preparation included 17α-ethynylestradiol-acetate (EE2-ACT), 17α-ethynylestradiol-butyrate (EE2-BUT), 17β-estradiol-acetate (E2-ACT), 17β- estradiol-butyrate (E2-BUT) and estrone-3-hemisuccinate (E1-HS) ( Li et al., 2004 ) , which were synthesised in our lab, and 17β-estradiol-one6-O-carboxymethyl-oxime (E2-OX) and 17β-estradiol-hemisuccinate (E2-HS), which were purchased commercially (Figure 3.2). These haptens were coupled to HRP in two varying ratios of hapten to enzyme. This gave a total of 11 enzyme conjugates denoted as EE2-ACT 12 – HRP, EE2-ACT 20 – HRP, EE2- BUT 12 – HRP, EE2-BUT 20 – HRP, E2-ACT 12 – HRP, E2-ACT 20 – HRP, E2-BUT 12 – HRP, E2-BUT 20 – HRP, E2-HS-HRP, E2-OX-HRP and E1-HS-HRP. For example EE2- ACT 12 – HRP is EE2-ACT hapten conjugated to HRP at a ratio of 12:1 (hapten to enzyme).

Optimum conditions for each enzyme-conjugate were assessed by CBT against immobilised antibodies and selecting a concentration/dilution that yielded approximately 1–1.5 absorbance units. As shown in Figures 3.3 and 3.4, the optimum dilutions for EE2-ACT 12 – HRP, EE2- ACT 20 – HRP, EE2-BUT 12 – HRP, EE2-BUT 20 – HRP, E2-ACT 12 – HRP, E2-ACT 20 – HRP, E2-BUT 12 – HRP, E2-BUT 20 – HRP, E2-HS-HRP and E1-HS-HRP were 1/81000, 1/243000, 1/729000, 1/ 243000, 1/ 243000, 1/243000, 1/243000, 1/ 81000, 1/1500 and 1/27000, while E2-OX-HRP did not show adequate absorbance when titrated against EE2- ACT antibodies, hence no further evaluation was conducted.

The CBT on immobilised EE2-BUT antibody showed that E2-OX-HRP, E2-HS-HRP and E1- HS-HRP exhibited low absorbance, while the other enzyme conjugates displayed strong responses (Figures 3.4 and 3.6). The optimum dilutions for EE2-ACT 12 – HRP, EE2-ACT 20 – HRP, EE2-BUT 12 – HRP, EE2-BUT 20 – HRP, E2-ACT 12 – HRP, E2-ACT 20 – HRP, E2-BUT 12 – HRP, E2-BUT 20 – HRP, E2-HS-HRP and E2-OX-HRP were 1/2187000, 1/2187000, 1/81000, 1/81000, 1/81000, 1/81000, 1/3000, 1/3000, 1/1000 and 1/1000. Both antibodies titrated against E2-HS-HRP and E2-OX-HRP exhibited low absorbance.

Ab-EE2-ACT generated 4-fold higher titres in the heterologous assay with EE2-BUT12 – HRP than in the homologous assay. The use of bridge heterology with Ab-EE2-BUT also improved colour development by 25-fold compared to its homologous counterpart. Future characterisation of the hetrologous assay was carried out to evaluate assay sensitivity.

OH OH CH CH 3 3 CH CH

O O O O O O N O N O

O O

7α-ethynylestardiol-acetate (EE2-ACT) 17α-ethynylestradiol-butyrate (EE2-BUT)

OH OH CH CH 3 3

O O O O O O N O N O O

17β-estradiol-acetate (E2-ACT) 17β-estradiol-butyrate (E2-BUT)

17β-estradiol-one6-O-carboxymethyl-oxime 17β-estradiol-hemisuccinate (E2-HS) (E2-OX)

O CH 3

CHOOCH2CH2COO

Estrone-3-hemisucciante (E1-HS) (Li et al., 2004)

Figure 3.2 Chemical Structure of haptens

3.5

EE2-ACT12-HRP 3.0 EE2-BUT12-HRP EE2-ACT20-HRP 2.5 EE2-BUT20-HRP E1-HS-HRP 2.0

1.5

1.0

Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000

Dilution of hapten-enzyme conjugates

Figure 3.3 Titration curves of EE2-ACT antibody against five hapten-enzyme conjugates (EE2-ACT12-HRP, EE2-ACT20-HRP, EE2-BUT12-HRP, EE2-BUT20-HRP and E1-HS-HRP).

3.5

E2-ACT12-HRP 3.0 E2-ACT20-HRP E2-BUT12-HRP 2.5 E2-BUT20-HRP E2-HS-HRP 2.0 E2-OX-HRP

1.5

1.0

Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000

Dilution of hapten-enzyme conjugates

Figure 3.4 Titration curves of EE2-ACT antibody against six hapten-enzyme conjugates (E2- ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT2-HRP, E2-HS-HRP and E2-OX-HRP).

3.5

EE2-ACT12 - HRP 3.0 EE2-ACT20 - HRP EE2-BUT12 - HRP 2.5 EE2-BUT20 - HRP E1-HS-HRP 2.0

1.5

1.0

0.5 Absorbance (450 nm) (450 Absorbance

0.0 100 1000 10000 100000 1000000 10000000 10000000

Dilution of hapten-enzyme conjugates

Figure 3.5 Titration curves of EE2-BUT antibody against five hapten-enzyme conjugates (EE2-ACT12-HRP, EE2-ACT20-HRP, EE2-BUT12-HRP, EE2-BUT20-HRP and E1-HS-HRP).

3.5

E2-ACT12 - HRP 3.0 E2-ACT20 - HRP E2-BUT12 - HRP E2-BUT20 - HRP 2.5 E2-HS-HRP E2-OX-HRP 2.0

1.5

1.0

Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000 Dilution of hapten-enzyme conjugates

Figure 3.6 Titration curves of EE2-BUT antibody against six hapten-enzyme conjugates (E2- ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT20-HRP, E2-HS-HRP and E2-OX-HRP).

3.3.2 Assay Sensitivity

Using the established optimised concentrations of the enzyme conjugates, the sensitivity of each antibody-enzyme-conjugate combination was investigated by running standard curves of 17α-ethynylestradiol. That is, four antibodies (two antibodies against each of EE2-ACT and EE2-BUT haptens) x 11 enzyme conjugate combinations (Figures 3.3 to 3.6). As well, the sensitivity of different bleeds of each antibody was investigated (Table 3.2). For EE2-ACT antibodies, the second, third and fourth bleeds were tested, and for the EE2-BUT antibodies only the first and second bleeds were tested. Both EE2-ACT and EE2-BUT antibodies demonstrated improved sensitivity with booster injections, indicating affinity maturation of the antibodies. As expected, the employed immunisation regime for polyclonal antibody production provided low level stimulation of the immune system, and generated a strong response.

-1 Table 3.2 The IC50 (μgL ) values for the calibration curves using 11 enzyme conjugates against two 17α-ethynylestradiol antibodies.

-1 IC50 (μgL ) Enzyme Ab-EE2 ACT Ab-EE2 BUT Conjugates Second Third Fourth First Second bleeding bleeding bleeding bleeding bleeding EE2-ACT 12-HRP 11.7 9.5 0.3 2.4 0.6 EE2-ACT 20-HRP 8.8 2.7 0.3 5.0 0.5 EE2-BUT 12-HRP 8.0 4.5 1.9 5.3 0.8 EE2-BUT 20-HRP 5.5 2.3 0.5 6.0 0.8 E2-ACT 12-HRP 0.7 0.5 0.2 1.1 0.6 E2-ACT 20-HRP 0.5 0.5 0.5 1.6 0.9 E2-BUT 12-HRP 1.0 0.7 0.7 9.0 5.5 E2-BUT 20-HRP 1.5 0.9 0.9 6.5 6.4 E2-OX-HRP * * 4.9 * 4.9 E2-HS- HRP * * NA * * E1-HS-HRP * * 2.9 * 1.9 * The assay exhibited very low colour development, and did not show the correct standard curve. NA – note tested

As shown in Table 3.2, the absolute homologous systems (i.e., both hapten and linkage are the same) evaluated were Ab-EE2-ACT x EE2-ACT12-HRP or EE2-ACT20-HRP, and Ab-EE2- 59

BUTxEE2-BUT12-HRP or EE2-BUT20-HRP. These homologous assays exhibited -1 acceptable IC50 values of 0.3 and 0.8 μg L for Ab-EE2 ACT and Ab-EE2 BUT respectively. The hapten homologous with linker heterologous assays showed slightly lower sensitivity than their respective absolute homologous assays. For example, the IC50 values for Ab-EE2- ACT x EE2-BUT12-HRP or EE2-BUT20-HRP ranged between 0.5–1.9 μgL-1, while Ab-

EE2-BUT x EE2-ACT12-HRP or EE2-ACT20-HRP exhibited better IC50 values, ranging from 0.5–0.6 μgL-1.

As expected, the heterologous assays were able to further improve the sensitivity, but not all heterologous assays are able to improve the sensitivity. The hapten heterologous systems such as Ab-EE2-ACT or Ab-EE2-BUT used with E2-ACT12-HRP, E2-ACT20-HRP, E2-BUT12-

HRP, and E2-BUT20-HRP demonstrated increased sensitivity with IC50 values of 0.2 to 1.5 μgL-1 and 0.6 to 9.0 μgL-1 for Ab-EE2 ACT or Ab-EE2 BUT respectively. The absolute heterologous assays (i.e., both hapten and linkage are different) such as Ab-EE2 ACT x E2- BUT12-HRP or E2-BUT20-HRP, as well as Ab-EE2-BUT x E2-ACT12-HRP or E2-ACT20- -1 HRP demonstrated good sensitivity, exhibiting IC50 values ranging from 0.7 to 1.5 μgL and 0.9 to 1.6 μg L-1 respectively.

The relatively high concentrations of enzyme conjugates (E2-HS-HRP, E2-OX-HRP and E1- HS-HRP) needed in the absolute heterologous assays made them impractical for routine analyses, although it did not increase the background colour, and further characterisation was not conducted.

A hapten to enzyme conjugation ratio of 20:1 demonstrated better sensitivity than that of 12:1 for Ab-EE2-ACT except for the absolute heterologous systems of Ab-EE2-ACT vs E2- BUT12-HRP or E2-BUT20-HRP. This is probably due to better conjugation being achieved at higher molar ratio of hapten to enzyme due to the hydrophobic nature of haptens in aqueous solution. This was more notable in the hapten with a longer linker. Interestingly, only Ab- EE2-BUT worked in an opposite manner.

To sum up, for EE2-ACT antibodies, the assay under hapten heterology with linkage -1 homology demonstrated the best sensitivity (IC50), exhibiting an IC50 value of 0.2 μg L (Ab- EE2-ACT vs E2-ACT 12-HRP), while the assay using a homologous hapten and heterologous -1 linkage system showed the best sensitivity (IC50) value of 0.5 μg L for EE2-BUT antibodies (Ab-EE2-BUT vs EE2-ACT 20-HRP). 60

3.3.2.1 Standard curve parameters and precision (IC80, IC50, IC20 and maximum absorbance)

Nine point calibration curves for Ab-EE2-ACT and Ab-EE2-BUT are shown in Figures 3.7 and 3.10. The results were averaged over 27 analyses with Ab-EE2-ACT and 12 analyses with Ab-EE2-BUT conducted on different days over three seasons (including summer and winter). The % CV of % inhibition decreased as the EE2 concentration increased for both antibodies, ranging from 0.5–32% and 4–66% for Ab-EE2-ACT and Ab-EE2-BUT, respectively (Figure 3.8). This is a typical precision of sigmoidal dose-response relationship of an immunoassay (Lee and Kennedy, 2007). With Ab-EE2-ACT and Ab-EE2-BUT, the -1 -1 IC50 value was 0.29 ±0.12 μgL (% CV = 40%), and 0.44 ±0.10 μg L (% CV = 22%), respectively. The limit of detection (LOD) was determined by using the concentration of standard solution causing 20% inhibition of colour development (IC20). For Ab-EE2-ACT -1 and Ab-EE2-BUT, the IC20 values were 0.07 ±0.01 μgL (% CV = 17%) (Figure 3.9 and Table 3.3), and 0.05 ±0.02 μgL-1 (% CV = 48%), respectively (Figure 3.9 and Table 3.4). The -1 EE2 that could inhibit 80% of the antibody binding (denoted as IC80) was 1.74 ±0.62 μgL (% CV = 36%) for Ab-EE2-ACT (Table 3.3) and 10.0 ±6.63 μg L-1 (% CV = 66%) for Ab-

EE2-BUT (Table 3.4). The IC80 formed the upper limit of the linear range in the sigmoidal calibration curve.

The average maximum absorbance was slightly lower for Ab-EE2-ACT (0.778±0.408 with % CV = 52%) (Table 3.3) than for Ab-EE2-BUT (0.813±0.219 with % CV = 27%), and was more susceptible to various environmental conditions (e.g., temperature) (Table 3.4). The % CV for the maximum absorbance at nine EE2 concentrations ranged between 22 to 42% for Ab-EE2-ACT (Figure 3.8) and 13 to 55% for Ab-EE2-BUT (Figure 3.10 and 3.11). The

%CV of IC50 was lower than those of IC20 and IC80 for Ab-EE2-BUT as expected from a typical signoidal dose-response relationship (Lee and Kennedy, 20067, Figure 3.12 and Table

3.4). Interestingly, the %CV of IC20 was lower than those of IC50 and IC80 for Ab-EE2-ACT, which was unexpected (Table 3.3).

Table 3.3 Standard curve parameters and precision for EE2-ACT antibody. Parameters Concentration S.D.a % CV b -1 IC50 0.29 μgL 0.12 40 -1 IC20 0.07 μgL 0.01 17 -1 IC80 1.74 μgL 0.62 36 Maximum absorbance 0.778 Abs 0.408 52 a standard deviation b percent coefficient variation

100 0.9

90 0.8

80 0.7 70 0.6 60 0.5 50 0.4 40 %inhibition 0.3 30 Absorbance ( 450 nm) 450 ( Absorbance 20 0.2

10 0.1

0 0 0.01 0.1 1 10 EE2 concentration ( μg L-1 )

Figure 3.7 Calibration curve for EE2-ACT antibody (average of 27 analyses).The absorbance against the EE2 concentration ( ) and the % inhibition against the EE2 concentration ( ).

30 %CV 20

0 0.01 0.1 1 10 EE2 concentration (μgL-1)

Figure 3.8 % CV for the absorbance ( ) and % inhibition ( ) by EE2-ACT antibody (average of 27 analyses).

10.00 )

-1 1.00 g L μ

0.10 EE2 concentration (

0.01 0 3 6 9 12 15 18 21 24 27 Number of assay

Figure 3.9 Plot of IC80 ( ), IC50 ( ) and IC20 ( ) values for EE2-ACT antibody (average of 27 analyses). The middle line indicates average value. The dotted line shows the upper and lower limit (average EE2 concentration+standard deviation).

Table 3.4 Standard curve parameter and precision for EE2-BUT antibody Parameters Concentration S.D. a %CV b -1 IC50 0.44 μgL 0.10 22 -1 IC20 0.05 μgL 0.02 48 -1 IC80 10.05 μgL 6.63 66 Maximum absorbance 0.813 Abs 0.219 27 a standard deviation b percent coefficient variation

120 1.2

100 1

80 0.8

60 0.6 % Inhibition % 40 0.4 Absorbance ( 450 nm) 450 ( Absorbance

20 0.2

0 0 0.01 0.1 1 10 EE2 concentration ( μg L-1 )

Figure 3.10 Calibration curve by EE2-BUT antibody (average of 12 analyses). The absorbance against the EE2 concentration ( ) and the % inhibition against the EE2 concentration ( ).

%CV 30

0 0.01 0.1 1 10 EE2 concentration (μgL-1)

Figure 3.11 % CV for the absorbance ( ) and % inhibition ( ) by EE2-BUT antibody (average of 12 analyses).

)

-1 10 gL μ

0.1 EE2 concentration (

0.01 0 1 2 3 4 5 6 7 8 9 10 11 12

Number of assay

Figure 3.12 Plot of IC80 ( ), IC50 ( ) and IC20 ( ) values for EE2-BUT antibody (average of 27 analyses). The middle line indicates average value. The dotted line shows the upper and lower limit (average EE2 concentration + standard deviation).

3.3.3 Characteristics of EE2 ELISA

3.3.3.1 Assay Specificity

Recent papers on the determination of natural and synthetic estrogen and progestogen in wastewater revealed that the levels of E2, E1 and EE2 in influent of many countries such as Italy, Brazil, South Korea and Japan were detected with average concentrations in parts per trillion or nanogram per litre levels (Ternes et al., 1999b, Baronti et al., 2000, Furuichi et al., 2006, Kim et al., 2007, Streck, 2009, Nakamura et al., 2001). 17α-Estradiol has been found in the Tama River in Japan at an extremely low level of 5.6 ngL-1 (Nakamura et al., 2001), and mestranol, estradiol, ethynylestradiol, and estrone have been detected at 0.3 to 0.7 μgL-1 in the Llobregat River in Spain (Farr´E et al., 2007). In addition, Gabet et al. (2007) reported that mestranol and 17α-estradiol were analyzed in river water and effluent, and the LOD was in the range of 0.4 to 1.9 ngL-1. Estradiol dipropionate is a main active ingredient for treatment of menopause symptoms, which has an apparent superiority over other estrogens (Greene, 1941). This estrogen has been found in the environment at μgL-1 levels, although it failed to cause any significant alteration in the glycogen, lipid and water contents of fish (Dasmahapatra and Medda, 1982). Also, medroxyprogesterone, one of the synthetic hormones widely used for contraception and treatment of endometriosis, has been determined by several methods in water samples (Zhang et al., 2008, Vulliet et al., 2008, Xie et al., 2010). Progesterone and medroxyprogesterone have been detected in wastewater and surface water samples from the Koyama River basin in Japan at 0.06 and 0.03 ngL-1, respectively (Chang et al., 2008). Quinestrol is commonly used in the control of wild rodents, and it continues to be released into the environment, predominantly from the rodent’s urine and faeces. Quinestrol levels have been determined in soil and domestic wastewater (Tang et al., 2009).

It is important to assess the ability of the antibodies to interact with substances other than the analyte of interest. The study of cross−reactivity is indicative of the degree of assay specificity. Consequently, the specificity of the EE2 ELISAs was determined by comparing relative affinities for the selected estrogens of structural similarity. These included 17β- estradiol, estriol, estrone, 17α-estradiol, estradiol dipropionate, progesterone, medroxyprogesterone, ethynylestradiol-3-methyl ether, and 17α-ethynylestradiol-3- cyclopentyl ether. The selection of test compounds for cross−reactivity studies made based on the occurrence and contribution to the environment reported by the previous research (Baronti et al., 2000, Chimchirian et al., 2007)

Based on the results of this study, the two antibodies, Ab-EE2-ACT and Ab-EE2-BUT, can be considered to be specific for EE2 and the related compound, mestranol. In fact, the sensitivity for mestranol was two-fold higher than EE2. Specificity for Ab-EE2-ACT are presented in Table 3.9. The IC50 values for estriol, estrone, estradiol dipropionate, progesterone, 17α-estradiol and medroxyprogesterone were >100 μgL-1. 17β-Estradiol, and -1 17α-ethynylestradiol-3-cyclopentyl ether gave IC50 values of 11.1 and 3.1 μg L , respectively (<13%). It was noted that the % CR of the developed EE2 ELISA for E2 was higher than in the previous studies in which <0.3% cross−reactivity was reported (Schneider et al., 2004, Schneider et al., 2005) .

-1 Table 3.9 The IC50 (μgL ) and cross reactivity (CR %) for selected estrogens with two ethynylestradiol antibodies.

Compounds Structure Ab-EE2-ACT Ab-EE2- BUT

IC50 CR (%) IC50 CR (%) IC50 CR IC50 (μgL-1) (mol L-1) (μgL-1) (%) (mol L-1) CR (%)

-9 -9 17α- OH 0.4 100 1.3 x 10 100 0.7 100 2.410 100 CH 3 ethynylestradiol CH

-8 -8 17β-estradiol OH 11.1 3.6 4.1 × 10 3.3 62.1 1.1 2.310 1.0 CH 3

-7 -7 Estriol OH >100 <0.4 3.5 × 10 0.4 >100 < 0.7 3.5× 10 0.7 CH 3 OH

O -7 -7 Estrone CH3 >100 <0.4 3.7 × 10 0.4 >100 < 0.7 3.7× 10 0.6

HO Estradiol O >100 <0.4 2.6 ×10-7 0.5 >100 < 0.7 2.6× 10-7 0.9 O CH 3 CH dipropionate 3

H C 3 O

-7 -7 Progesterone O CH >100 <0.4 3.2 × 10 0.4 >100 < 0.7 3.2× 10 0.7 3 CH 3

CH 3

-7 -7 17 α-estradiol OH >100 <0.4 3.7 × 10 0.4 >100 < 0.7 3.7× 10 0.6 CH 3

-7 -7 Medroxyprogestero O >100 <0.4 2.9 × 10 0.5 >100 < 0.7 2.9× 10 0.8 ne CH CH 3 3

CH 3

Ethynylestradiol-3- 0.20 200.0 6.4 × 10-7 209.5 0.10 700 3.2× 10- 733.2 10 OH CH methyl ether 3 CH (Mestranol)

H C 3 O

17α- 3.10 12.9 8.5×10-9 15.9 9.80 7.1 2.7× 10-9 8.8

OH CH ethynylestradiol 3- 3 CH cyclopentyl ether (Quinestrol) O

The specificity of Ab-EE2-BUT showed < 0.7% of cross−reactivity for estriol, estrone, estradiol dipropionate, estradiol dipropionate, progesterone, 17α-estradiol, and medroxyprogesterone. Only 17β-estradiol and 17α-ethynylestradiol-3-cyclopentyl ether presented 1.1% and 8% cross−reactivity, respectively. Surprisingly, the cross−reactivity of the EE2-BUT ELISA for ethynylestradiol-3-methyl ether (mestranol) is 700%. The substantial cross−reactivity existed for ethynylestradiol-3-methyl ether is due to the C3 position of EE2 hapten (EE2-BUT) and mestranol are very similar. The C3 position of EE2 hapten is –O-

CH3CH3-, while mestranol is –O-CH3. In general, Ab-EE2-BUT exhibited a similar pattern to Ab-EE2-ACT.

This is going to pose a serious analytical impediment if only EE2 is desired in the detection, because mestranol is pharmacokinetically bioequivalent to ethynylestradiol. Additionally, mestranol is a biologically inactive form of ethynylestradiol that is generally used for absorption and to improve oral bioavailability, so this substance is certainly found after ingestion and digestion of the currently available contraceptive pill (Fraser, 1998). There may be a need to monitor this pharmaceutical residue in water in the future.

Therefore, judging from the cross−reactivity studies, the assay based Ab-EE2-ACT is specific to both EE2 and mestranol, both of which are synthetic estrogens used in phamaceutricals. Whereas, the assay based on Ab-EE2-BUT is highly sensitive and specific to mestranol. Mestranol has been found at < 0.01 μg L-1 in the environment (Schneider et al., 2004). Thus using the Ab-EE2-ACT based assay one can theoretically quantify for both EE2 and mestranol residues in environment. Using the Ab-EE2-BUT based assay, metranol residues can be quantified at ultra low concentrations. EE2 in the same samples can, and then be quantified by the subtraction of the Ab-EE2-BUT from the Ab-EE2-ACT.

3.3.3.2 Assay Optimisation

3.3.3.2.1 Effects of Solvents

Based on structure and solubility of 17α-ethynylestradiol, methanol and ethanol are generally used to extract estrogen from various matrices or to elute analyte from solid-phase extraction (SPE) cartridges. For this reason, the effect of organic solvents on the ELISA was tested by running EE2 calibration curves prepared in different concentrations. The effects of solvent on colour development of assay and on the assay sensitivity are presented in Figures 3.13 to 3.24.

1.80

1.60 5% EtOH 10% EtOH 1.40 20% EtOH MilliQ Water 1.20

1.00

0.80

0.60

Absorbance ( 450 nm) 450 ( Absorbance 0.40

0.20

0.00 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.13 Effect of solvents on colour development of the ELISA based on Ab-EE2-ACT (MilliQ Water, 5% EtOH, 10% EtOH and 20% EtOH ) on absorbance using Ab-EE2-ACT.

100

% Inhibition % 5% EtOH 40 10% EtOH 20% EtOH 30 MilliQ Water

0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.14 Standard curves of EE2 in different solvents for Ab-EE2-ACT (MilliQ Water, 5% EtOH, 10% EtOH and 20% EtOH).

2.0

1.5 5% MeOH

10% MeOH

20% MeOH

1.0

0.5 Absorbance (450 nm) (450 Absorbance

0.0 0.01 0.1 1 10 -1 17α-ethynylestradiol (μgL )

Figure 3.15 Effects of solvents (5% MeOH, 10% MeOH and 20% MeOH) on colour development of the ELISA based on Ab-EE2-ACT.

100

% Inhibition % 40

30 5% MeOH

20 10% MeOH

10 20% MeOH

0 0.01 0.1 1 10 -1 17α-ethynylestradiol (μgL ) Figure 3.16 Standard curves of EE2 in different solvents for Ab-EE2-ACT (5% MeOH, 10% MeOH and 20% MeOH).

1.2

1.0 10% acetone

20% acetone

10% acetronitrile 0.8 20% acetronitrile

0.6

0.4 Absorbance (450 nm) (450 Absorbance

0.2

0.0 0.01 0.1 1 10 17α-ethynylestradiol (μgL-1)

Figure 3.17 Standard curves of EE2 in different solvents for Ab-EE2-ACT (10% acetone, 20% acetone, 10% acetonitrile and 20% acetonitrile).

100

% Inhibition % 40 10% acetone

30 20% acetone

20 10% acetronitrile 20% acetronitrile 10

0 0.01 0.1 1 10 17α-ethynylestradiol (μgL-1) Figure 3.18 Standard curves of EE2 in different solvents for Ab-EE2-ACT (10% acetone, 20% acetone, 10% acetonitrile and 20% acetonitrile).

1.2

5% EtOH 10% EtOH 0.9 20% EtOH MiiliQ Water

0.6

0.3 Absorbance ( 450 nm ) nm 450 ( Absorbance

0.0 0.01 0.1 1 10 -1 17α-ethynylestradiol (μgL )

Figure 3.19 Effect of solvents (MilliQ Water, 5% EtOH, 10% EtOH and 20% EtOH ) on colour development of the ELISA based on Ab-EE2-BUT

100

90 5% EtOH

10% EtOH 80 20% EtOH

70 MiiliQ Water

50 % Inhibition % 40

0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1) Figure 3.20 Standard curves of EE2 in different solvents for Ab-EE2-BUT (MilliQ Water, 5% EtOH, 10% EtOH and 20% EtOH).

1.2

0.9 5% MeOH

10% MeOH

20% MeOH

0.6

Absorbacne (450 nm) (450 Absorbacne 0.3

0.0 0.01 0.1 1 10

-1 17α-ethynylestradiol (μgL )

Figure 3.21 Effect of solvents (5% MeOH, 10% MeOH and 20% MeOH) on colour development of the ELISA based on Ab-EE2-BUT

100

60 5% MeOH 50 10% MeOH % Inhibition % 40 20% MeOH 30

0 0.01 0.1 1 10 17α-ethynylestradiol (μgL-1)

Figure 3.22 Standard curves of EE2 in different solvents for Ab-EE2-BUT (5% MeOH, 10% MeOH and 20% MeOH).

0.5

10% acetone

0.4 20% acetone

10% acetronitrile

20% acetronitrile 0.3

0.2 Absorbance ( 540 nm) 540 ( Absorbance

0.1

0.0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.23 Effect of solvenst (10% acetone, 20% acetone, 10% acetonitrile and 20% acetonitrile) on colour development of the ELISA based on Ab-EE2-BUT.

100

10% acetone % Inhibition % 40 10% acetronitrile

30 20% acetone

20% acetronitrile 20

0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.24 Standard curves of EE2 in different solvents for Ab-EE2-BUT (10% acetone, 20% acetone, 10% acetonitrile and 20% acetonitrile).

0.8 Water 5% of the solvents

0.7 10% of the solvents 20% of the solvents

0.6

0.5 ) -1 g L

μ 0.4 ( 50

IC 0.3

0.2

0.1

0 Ethanol Methanol Acetone Acetronitrile Organic Solvent Figure 3.25 Effect of organic solvents on the calibration curve of Ab-EE2-ACT.

1.4

Water 5% of the solvents 1.2

10% of the solvents 20% of the solvents 1 ) -1 g L

μ 0.8 ( 50 IC 0.6

0.4

0.2

0 Ethanol Methanol Acetone Acetronitrile Organic Solvent

Figure 3.26 Effect of organic solvents on the calibration curve of Ab-EE2-BUT

Ethanol at 5–20% in water, methanol at 5–20% in water, acetone at 10–20% in water and acetonitrile at 10–20% in water were used to investigate the effect of solvents on the assay. The standard curves prepared in each of these concentrations were compared against that in water as a control. Results are illustrated in Figures 3.25 and 3.26.

Methanol and ethanol up to 10% have no effect on the sensitivity or colour development of -1 both assays. For Ab-EE2-ACT, 5 and 10% ethanol yielded IC50 values of 0.26 and 0.31 μg L

, respectively (Figure 3.13 and 3.14), and 5 and 10% methanol showed IC50 values of 0.46 and -1 0.36 μg L , respectively (Figure 3.15 and 3.16). For Ab-EE2-BUT, the IC50 of 5 and 10% of -1 ethanol were 0.24 and 0.37 μg L (Figure 3.20), and the IC50 of 5 and 10% methanol were -1 0.29 and 0.23 μg L , respectively (Figure 3.22). For ethanol and methanol, the IC50 value increased rapidly with increasing concentrations to 20% V/V, and the colour development greatly decreased for both assays (Figure 3.13, 3.15, 3.19 and 3.21). Poor colour development and assay sensitivity were observed for Ab-EE2-ACT when using 20% acetone or acetronitrile (Figure 3.17 and 3.18). Moreover, 10% and 20% acetone and acetonitrile showed significantly low colour development with 0.296 to 0.448 Abs (Figure 3.23), and reduced the -1 assay sensitivity with IC50 values between 0.59 to 0.93 μg L (Figure 3.24), for Ab-EE2-BUT. The effect of organic solvent on the immunoassay might be explained by denaturation of the antibody in an organic solvent environment, especially at high concentrations (Goda et al., 2005).

In conclusion, up to 10% methanol or ethanol could be used in the EE2-ELISA without significantly interfering with the assay sensitivity. Keeping the EE2 soluble in the calibration solutions is an important factor for analytical consistency. Based on the above findings, 10% ethanol was chosen for all subsequent assays.

3.3.3.2.2 Effect of enzyme conjugate diluents

Crowther (2000) stated that the way to reduce non−specific interactions is to dilute the enzyme conjugate in diluents containing immunologically inert substances such as PBS and Tween 20, the latter being a non−ionic detergent which is commonly used to prevent non −specific binding, especially in Western blot analysis (Wedge and Svenneby, 1986, Mohammed and Esen, 1989). Therefore, the determination of their influence on the ELISA, four enzyme-conjugate diluents, consisting of BSA in PBS (i.e., 1% BSA/PBS, 1% BSA/PBS + 0.05% Tween 20, 1% BSA/PBS + 0.1% Tween 20, 1% BSA/PBS + 0.25% Tween 20), 79 were tested. It was found that these diluents have considerable effects on assay performance, especially with respect to colour development (Table 3.10).

Table 3.10 Effect of enzyme conjugate diluents on EE2 ELISA Ab-EE2-ACT Ab-EE2-BUT Assay diluents Maximum IC Maximum IC 50 50 -1 -1 Absorbance (μg L ) Absorbance (μg L ) 1% BSA/PBS 0.932 0.4 0.400 0.7 1% BSA/PBS+0.05% Tween 20 0.638 0.6 0.370 0.8 1% BSA/PBS+0.1% Tween 20 0.687 0.7 0.320 0.8 1% BSA/PBS+0.25% Tween 20 0.545 1.0 0.272 1.1

When Tween 20 was added to the enzyme diluents, a significant effect was apparent. The IC50 values increased approximately two-fold for both assays when 1% BSA/PBS with 0.25% Tween 20 was used. However, a number of research papers have previously shown that Tween 20 decreased the sensitivity due to its non−specific hydrophobic interactions. The results from this study indicate that Tween 20 considerably affected the binding between antibody and hapten. Results from these studies are in accordance with Shan and co-workers (1999) who found that the non−specific hydrophobic interactions between Tween 20 and non− polar small analytes in an aqueous environment appeared to have interfered with the specific analyte and antibody interaction. Since Tween 20 did not improve the sensitivity but has a potential to reduce the colour development even at a concentration generally used in an immunoassay, it was not used in further experiments.

3.3.3.3 Matrix Interferences

Antibodies and enzyme-conjugates are generally relatively sensitive to the matrix as well as harsh conditions (Pfortner et al., 1998, Lee and Kennedy, 2007). The study of matrix interference is required as an integral part of assay development validation because of the considerable variability in environmental samples such as surface water and effluent from STP. Matrix effects were evaluated by examining both the colour development as a measure of the effects on the enzyme-conjugate, and the sensitivity (IC50) as a measure of effects on the antibody-antigen interaction.

3.3.3.3.1 Effect of pH

Proteins, such as antibodies are susceptible to pH variations. The pH of a solution can affect the structure and activity of a protein. For example, enzymes generally have a pH optimum, so a change in pH can considerably reduce the rate of the Michelis-Menton kinetic reaction of enzymes (Bergmeyer, 1974, Abe et al., 1987). Assay performance was evaluated at various pH values (3, 5, 8, 9, 10 and 11) – and compared with the control, at pH 7.2.

-1 As illustrated in Figure 3.27, the control showed an IC50 value of 0.35 μg L with colour development at 0.722, while pH 3 and 5 gave interestingly low colour development at below 0.300. There was a noticeable effect when the pH was higher than 7. Colour development declined gradually and the IC50 value increased. The pH at 3, 9 and 11 reduced the assay sensitivity, while it remained stable at pH 5 and 8 when compared to the sensitivity of the control (pH 7.2). The study found that the optimum pH is between 7 and 8. Therefore, the pH of samples must be adjusted prior to analysis.

IC 50 value 4 0.8 Color development 3.5 0.7

3 0.6

) 2.5 0.5 -1 g L

μ 2 0.4 ( 50

IC 1.5 0.3

1 0.2 nm) 450 ( Absorbance

0.5 0.1

0 0 357911 pH Figure 3.27 Effects of the pH on the EE2 ELISA standard curve. The square indicates absorbance and the circle indicates IC50 values against pH.

3.3.3.3.2 Effect of water type and ionic strength

It is important to know whether the calibration curve constructed in standard buffer solutions can be used to quantify real samples, as the assay performance can be influenced by such aspH, inorganic and heavy metal ions in real water matrices. Hence, different types of water samples, in terms of source and chemical properties, were evaluated for any effects on the immunoassay. For this purpose, the water samples were collected from the lagoon at Tahbilk winery wetlands (pH 6.4), McWilliam’s winery reservoir (pH 6.7) and sea water from Maroubra Beach (pH 8.2).

In terms of the antigen-antibody interaction, real water matrices had no observable effect on the assay sensitivity, as illustrated in Figure 3.29. Nevertheless, real water matrices slightly interfered with enzyme activity, contributing to a decline in absorbance (Figure 3.28) for samples sourced from the lagoon at Tahbilk winery wetlands and McWilliam’s winery reservoir when compared to purified water as a control. This is probably due to the pH of these two water samples being below 7. These results indicated that neat real water samples without the sample preparation step did not interfere with the EE2 ELISA if pH was adjusted.

Table 3.11 Effect of ionic strength on EE2 ELISAs

Ab-EE2 ACT Ab-EE2 BUT

Compounds Colour IC50 Colour IC50 development (Abs) (μg L-1) development (Abs) (μg L-1) Control (10% EtOH) 0.503 0.3 0.433 0.7

1 M (NH4)2SO4 0.284 0.3 0.314 1.4 1M NaCl 0.420 0.3 0.393 0.7

1M MgSO4 0.196 0.2 0.359 2.1

1M CaCl2 0.146 0.2 0.178 >100

1M MnSO4 0.078 0.3 0.113 1.7 1M KCl 0.504 0.5 0.071 2.54

0.05M AlCl3 0.029 0.5 0.085 >100

0.1M CuSO4 0.030 >100 0.129 >100

0.01M CuSO4 0.371 0.03 0.136 15.51

0.001M CuSO4 0.446 0.3 0.147 1.26

0.1 M Fe2(SO4)3 0.048 >100 0.045 >100

0.01 M Fe2(SO4)3 0.337 0.1 0.295 5.01

0.001 M Fe2(SO4)3 0.383 0.2 0.231 1.06

0.8

McWilliam’s winery reservoir

0.6 lagoon at Tahbilk Winery wetlands

purified water

sea water from Maroubra Beach

0.4 Color Development (Abs)

0.2

0.0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.28 Matrix effects on colour development with different types of water (purified water, McWilliam’s winery reservoir, lagoon at Tahbilk winery wetlands and sea water from Maroubra Beach).

100

80 lagoon at Tahbilk Winery 70 wetlands

60 McWilliam’s winery reservoir 50 % Inhibition % purified water 40

30 sea water from Maroubra Beach 20

0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.29 Standard curve of EE2 concentration in different types of water (purified water, McWilliam’s winery reservoir, lagoon at Tahbilk Winery wetlands, and sea water from Maroubra Beach).

The chemical composition for treated sewage effluent Woolgoolga, NSW, indicated NH4 at -4 -1 -5 -1 -5 -1 2.72 × 10 mol L , NO3 at 1.13 × 10 mol L , PO4 at 6.63 × 10 mol L , SO4 at 8.23 × 10 -5 mol L-1, Cl- at 1.46 × 10 -3 mol L-1, Ca2+ at 2.72 × 10 -4 mol L-1, Mg2+ at 1.89 × 10 -4 mol L- 1, K+ at 3.63 × 10 -4mol L-1, Na+ 2.52 × 10 -3 mol L-1, Cu2+ at 1.73 × 10 -7 mol L-1, Fe 3+at 2.67 × 10 -6 mol L-1, Pb 2+at 8.68 × 10 -9mol L-1, Mn2+ at 1.27 × 10 -6 mol L-1, Ni 2+ at 5.96 × 10 -7 mol L-1, Zn2+ at 2.75 × 10 - mol L-1, Ar 2+at 6.00 × 10 -8mol L-1, Cd2+ at below 1.78 × 10 -10mol L-1, Cr3+ at below 1.92 × 10 -7 mol L-1with pH 8.1 (Johns and Mcconchie, 1994).

Studies relating to the influence of ionic strengths of dissolved salts in the water samples on the performance of the ELISAs were carried out. Common ions that are found in aquatic environments were tested at various concentrations as presented in Table 3.11. These include ionic salts such as (NH4)2SO4, NaCl, MgSO4, CaCl2, MnSO4, KCl, NH4Cl, CuSO4 and

Fe2(SO4)3.

For Ab-EE2-ACT, NaCl up to 1M was the only inorganic substance that showed little effect on the assay, while all the other inorganic substances either reduced the colour development or assay sensitivity. Furthermore, 1M MgSO4, 1M CaCl2, 1M MnSO4, 1M KCl and 0.05M

AlCl3, as interfering agents, diminished colour development, and subsequently the loss of sensitivity for Ab-EE2-ACT. The lowest concentration of CuSO4 tested (0.001M) had no effect on Ab-EE2-ACT, while the concentrations of CuSO4 from 0.01 to 0.1M reduced colour development, and increased the IC50 values. (NH4)2SO4, NaCl and MgSO4 at 1M did not affect the colour development of EE2-BUT ELISA, but only NaCl did not alter the assay sensitivity. The presence of 0.05M AlCl3 and 0.1 M Fe2(SO4)3 resulted in very low colour development for both antibodies. Additonally, Fe3+ gradually decreased colour development at higher concentrations, and loss of assay sensitivity for both antibodies (Table 3.11).

3.3.5.3 Effect of humic acid

Humic acids (HA) are a heterogenous mixture of small size molecules present in natural environments, especially in soils and water sediments. They come from plant and animal decay under moist conditions. HA are absorbents, ion exchangers, biochemical regulators and reservoirs (Schnitzer, 1977, Senesi and Miano, 1994, Boenigk et al., 2005).

In Austria, the concentration of organic material in lakes or rivers ranges from 10 μg L-1 to several hundred mg L-1 and even to several hundred g L-1(Boenigk et al., 2005). Humic acid 85 in water resources may act as a chelating agent for the estrogens in water samples, which could affect the analysis (Pfortner et al., 1998, Lai et al., 2000). Consequently, if naturally occurring chelators such as humic acid interfere in the determination of trace analytes, the extent of the interference could vary greatly through the year. Hence, serial dilutions of commercial HA were tested, with concentrations ranging from 0.01 to 10 mg L-1, to study the effect of HA.

A loss of sensitivity was observed with increasing concentration of humic acid, as shown in -1 -1 -1 -1 Figure 3.31. The humic acid at 0.01 mg L , 0.1 mg L , 1 mg L and 10 mg L showed IC50 values of 0.2, 0.3, 0.4 and 1.3 μg L-1 and maximum absorbances at 450 nm of 0.753, 0.895, 0.763 and 0.289, respectively (Figure 3.30). The calibration curve for purified water (control) -1 showed an IC50 value of 0.2 μg L (Figure 3.31) with an absorbance of 0.737. Hence humic acid at 0.01 mg L-1 was the only concentration which did not alter the sensitivity and colour development of the assay. Interference by humic acid presenting above 0.01 mg L-1 can cause an overestimation. This is due to EE2 forming a complex with humic acid, making EE2 unavailable for antibody binding. Such results were also observed in the previous studies by Deng et al. (2003) and Schneider et al. (2004, 2005).

1.0

Purified Water 0.8 Humic acid 10 mg L-1 Humic acid 1 mg L-1 Humic acid 0.1 mg L-1 0.6 Humic acid 0.01 mg L-1

0.4 Absorbance (540 nm) (540 Absorbance 0.2

0.0 0.01 0.1 1 10 -1 17α-ethynylestradiol (μg L )

Figure 3.30 Effects of humic acid on colour development of ELISA.

100

Purified Water

% Inhibition 40 Humic acid 10 mg L-1 30 Humic acid 1 mg L-1 Humic acid 0.1 mg L-1 20 Humic acid 0.01 mg L-1

0 0.01 0.1 1 10 17α-ethynylestradiol (μg L-1)

Figure 3.31 Effects of humic acid (HA) on the calibration curve of EE2.

3.3.6 Recoveries of EE2 from spiked purified and field water

Purified water and three water samples from, McWilliam’s winery reservoir, lagoon at Tahbilk Winery wetlands, and sea water from Maroubra Beach, were spiked with EE2 at 0.1, 0.3, 0.5, 1, 3, 10 and 15 μg L-1. The spiked samples were analyzed by the ELISA using Ab- EE2 ACT against E2-ACT 12 – HRP (Figure 3.32).

77 81 sea water from Maroubra Beach 15 87 75 lagoon at Tahbilk Winery wetlands 107 80 10 83 McWilliam’s winery reservoir 84 purified water 96 3 102 107 g L-1)

μ 117 105 1 93 106

96 103 0.5 93 105 Spiking Level ( 88 87 0.3 108 73

129 82 0.1 113 91

0 30 60 90 120 150 % Recovery

Figure 3.32 Average values (μg L-1) of spiking and recovery (%) from three water sources.

As illustrated in Figure 3.32, EE2 recoveries ranged between 73–107% for purified water, 83– 113% for samples from McWilliam’s winery reservoir, and 80–105% for samples from the lagoon at Tahbilk winery wetlands. A linear regression was applied to the data, as illustrated in Figure 3.33. The purified water (as a control) was also systematically included in the analysis and values that are lower and slightly above the assay detection limit was found in all cases, thus no false positives were detected. The regression coefficient (R2) of purified water and, samples from McWilliam’s winery reservoir, samples from a lagoon at Tahbilk winery wetlands and Maroubra Beach were 0.990, 0.997, 0.998 and 0.955 respectively (Figure 3.33). From the results, satisfactory recoveries were obtained and the assays have little matrix influence.

16 − − − − y = 1.133x R² = 0.955 (sea water from Maroubra Beach)

- - - - - y = 1.304x - 0.339, R² = 0.990 12 (Purified Water ) ) -1 gL

μ 8

------y = 1.172x - 0.118, R² = 0.997

SPIKE ( SPIKE (McWilliam's winery Reservoir)

...... y = 1.247x - 0.182 , R² = 0.998 (lagoon at Tahbilk winery wetland)

0 0481216 EE2 ELISA (μgL-1)

Figure 3.33 Correlation of ELISA and EE2 spikedinfour water sources: purified water, Mcwilliam’s winery reservoir, Tahbilk winery wetlands and sea water from Maroubra Beach.

3.3.7 Validation of ELISA with GC/MS

3.3.3.7 Determination of 17α-ethynylestradiol in water samples by the developed EE2 ELISA

The accuracy of the ELISA results was investigated by comparing with GC-MS results using real river water and real sewage water samples collected and extracted by Dr. Chunhua Li at the University of Sydney. The ELISA correlated relatively well with GC-MS with a correlation coefficient (R2) of 0.934 for the river samples spiked with EE2. There was a greater tendency for the ELISA to slightly overestimate than GC-MS, as shown in Figure 3.34. The under estimation by GC-MS is possible because GC-MS requires a clean-up/or derivatisation step for sample preparation, causing some loss of target compounds as reported in previous studies with estrone (Li et al., 2004). This suggests that the developed ELISA method is accurate.

y = 1.015x - 0.797 40 R² = 0.934 ) -1 30

20 EE2 GC-MS EE2 GC-MS (ngL

0 0 1020304050 EE2 ELISA (ng L-1)

Figure 3.34 Comparison of analytical results between the EE2 ELISA and GC-MS instrumental analysis for determination of EE2

3.4 Conclusion

The synthesis of hapten molecules with acetate and butyrate linkers, attached at the C3- position of 17α-ethynylestradiol, and used as a tracer and an immunogen for the development of a cdELISA method for the quantification of EE2 in water samples is presented in this chapter. The specific polyclonal antibodies were produced against conjugates of the haptens, 17α-ethynylestradiol-acetate or 17α-ethynylestradiol-butyrate, to a KLH carrier protein. The purified antibodies were then titrated to obtain the optimum working conditions for the

ELISA. Ab-EE2 ACT displayed a strong response with E2-ACT 12 – HRP with an IC50 value -1 of 0.2 μg L . Ab-EE2-BUT performed well with EE2-ACT 20 – HRP with an IC50 value of 0.5 μg L-1.

As expected, the heterologous system showed higher sensitivity compared to the homologous system. Additionally, a heterologous hapten and a heterologous linkage pair exhibited in a good dose response curve with high−sensitivity and good colour development, while the epitope ratio of hapten-enzyme conjugate had less significant impact on the the assay sensitivity.

For assay optimisation, the solvent for preparing standard solutions and samples can also affect colour development and assay sensitivity. Up to 10% methanol and ethanol did not alter the sensitivity of the assay, while concentrations above this shifted the curve to the right, reducing the assay sensitivity. Ethanol at 10% was subsequently chosen for routine analysis. Additionally, inclusion of Tween 20 in 1% BSA/PBS for enzyme diluents resulted in considerable inhibition of colour and reduced the IC50 of the assay, thus 1% BSA/PBS was chosen as the enzyme-conjugate diluent for routine analysis.

The assay developed from the two EE2-acetate and EE2-butyrate antibodies proved highly specific for EE2 and mestranol, exhibiting only negligible reactivities for endogenous compounds (CR < 5%). Substantial cross−reactivities existed only for synthetic ethynylestradiol conjugated at ring position 3 – ethynylestradiol 3-methyl-ether exhibiting cross reactivities of 200% for the developed ACT assay, and 700% for the developed EE2 BUT assay, respectively. Although, the cross−reactivity of ethynylestradiol 3 methyl ether with either Ab-EE2-ACT or Ab-EE2-BUT were considerably high, these substances are not EE2 metabolites and contamination levels lower than 0.01 μg L-1 in the environment have been reported in the literature. Using the two ELISAs simultaneously, both EE2 and mestranoo residues in environmental samples could be easily quantified.

The matrix interferences of the developed EE2 ELISA were studied by running calibration curves for three water samples, namely from a lagoon at Tahbilk winery wetlands, McWilliam’s winery reservoir and sea water from Maroubra Beach. The values obtained from these curves were similar to those found for the purified water calibration curve. Therefore, it can be concluded that EE2 ELISAs have negligible matrix interference. Humic acid, presence of inorganic salt, ionic strengths and pH were used as interfering agents for the determination of matrix effects. The assay exhibited a stable behaviour in the presence of humic acid and salt ions at natural concentrations. The optimum pH was 7 to 8.

Furthermore, the values obtained with the EE2 ELISA were well correlated with those derived from the GC-MS method with the correlation coefficient of 0.934. A slight overestimation by the ELISA method, however, was observed compared with GC-MS. However, given the complexity of both techniques and quantification at low parts per trillion levels, the data suggests that the developed EE2 ELISA as presented in this chapter is acceptably accurate and would be a valuable tool for monitoring EE2 and metranol residues in aquatic environment. 91

CHAPTER 4 DEVELOPMENT OF THE 17β-ESTRADIOL SPECIFIC ELISA (E2 ELISA)

4.1 Introduction

A sustainable and safe supply of drinking water from quality sources is an ongoing challenge for the water industry. The principle of water sustainability is to reuse wastewater for drinking water. It is an important concept because of the reducing water quality and the increasing pressure on potable freshwater reserves on a global scale. Wastewater treatment plants treat domestic and industrial wastewater and return the effluent to the environment via rivers, lakes streams and oceans. The inefficient Sewage Treatment Plants (STPs), (Gomes et al., 2003), may release endocrine disrupting compounds (EDCs) via the effluent to the aquatic ecosystem (Fox, 2001). EDCs encompass synthetic chemicals and natural chemicals. Pharmaceutical, pesticide, industrial chemicals and industrial by-products are grouped as synthetic chemicals, and natural chemicals include steroids, and plant constituents or phytoestrogens (Tylor et al., 1998) .

Among the large number of EDCs, steroidal estrogens are the most likely candidates for the investigation of efficiency of treatment facilities. One of the most potent estrogenic hormones is the natural steroid, 17β-estradiol (E2). The low molecular weight and the hydrophobic nature of E2 allow it to pass through biological membranes and cause a broad−range of effects during fetal and postnatal development in human and wildlife by alterations in endocrine function (Ying et al., 2002). Generally, mammals excrete relatively large quantities of estrogenic hormones in the form of sulfate or glucuronide conjugates, which can be converted back into its original form by glucuronidase or sulfatase enzymes (Orme et al., 1983). Furthermore, humans excrete as much as 0.002 to 0.1 mg of estrogenic hormone each day, while pregnant women produce and excrete approximately 30 mg per day of E2. In addition, oral contraceptive pills and hormone replacement therapy contains either natural or synthetic hormones in large quantities (Goodman and Gilman, 1996, Arcand-Hoy et al., 1998).

The concentrations of these estrogenic hormones in water sources are low, but their potency has considerable impact on aquatic organisms (Ying et al., 2002). E2 can bioaccumate in fish,

92 mainly in the bile, but also in both ovaries and testes of fish exposed to contaminated water (Gibson et al., 2005). E2 also has been observed to induce feminization of rainbow trout at 1 - 10 ngL-1 (Routledge et al., 1998a). Recent environmental concentrations of E2 measured in STP effluents cover a concentration range of between 0-15 ngL-1 (Desbrow et al., 1998, Belfroid et al., 1999, Snyder et al., 1999, Ternes et al., 1999a, Huang and Sedlak, 2001). The very low concentration of E2 found in the complexity of a waste water matrix is a challenge for quantification. Reliable instrumental methods are used widely for quantification of hormone-disrupting chemicals. Gas chromatography-mass spectrometry (GC-MS) (Heisterkamp et al., 2004, Yang et al., 2006), gas-chromatography tandem mass spectrometry (GC-MS/MS) (Belfroid et al., 1999, Ternes et al., 1999a) and liquid chromatography-mass spectrometry/ mass spectrometry (LC-MS/MS) (Heisterkamp et al., 2004) are generally employed, even though these instrumental methods have several potential drawbacks. These include the requirement of large sample sizes, expensive instrumentation, and the technical expertise needed for operation (Ying et al., 2002). More typically, these instrumental methodologies require incorporation of a sample enrichment and clean-up step which might impact the recovery. Both steps can lead to decreased precision. Therefore, there is a strong need for quantitative analysis of steroid hormones that is rapid, simple, and cost effective.

Alternatively, immunoassays can be used for the quantification of E2 in the low nanogram per litre range. For example, fluorescence immunoassays using DNA/dye conjugates as antibody multiple labels achieved a limit of detection (LOD) for E2 of 1.9 pg mL-1, which met the demand for high−sensitivity (Zhu et al., 2008). In addition, radioimmunoassay (RIA) exhibited a detection limit of 107 pg mL-1 for the determination of E2 in wastewater effluents (Snyder et al., 2003a). Consequently, Enzyme-Linked Immunosorbent Assay (ELISA) has become one of the most powerful techniques for routine analysis because of its reliability and efficiency. ELISA is also used for environmental monitoring. These assays have significant advantages, including their simplicity, rapidity, high−sensitivity, reproducibility, high selectivity, variable to quantitative and qualitative, specific for screening large number of samples, and lower demands on labour costs. However, the difficulties in terms of obtaining highly sensitive and specific antibodies, low matrix interference, and low cross-reactivity for structurally similar compounds are the shortcomings of ELISAs (Huang and Sedlak, 2001, Lee and Kennedy, 2007). Marcus and Dunford (1988) developed a simple enzyme-linked immunosorbent assay to identify E2 in serum extracts by using horseradish peroxidase- labelled estradiol-6-(0)-carboxymethyloxime as a tracer, with 2,2’-azino-his-(3- ethylbenzthiazoline sulfonic acid) diammonium salt (ABTS) as the chromogenic substrate. 93

The assay characteristics rival those of radio- or chemiluminescence immunoassays for estradiol. Furthermore, a polyclonal ELISA kit was used to determine E2 for water samples derived from the Llobregat River at Barcelona City with a limit of detection of 2.5 ng L-1 (Farré et al., 2007). A similar approach, with slightly lower sensitivity, was achieved with a monoclonal ELISA. This technique achieved a LOD of 0.05 - 0.1 μgL-1 for the determination of E2 from samples collected from STP (Japan) (Goda et al., 2005a, Hirobe et al., 2006). In another report, a detection limit was established at 7 -25 pgmL-1 for the determination of E2 by an ELISA from flushed dairy manure wastewater in Florida (Hanselman et al., 2004), as well as at 0.05 ngL-1 for E2 in eight surface water samples from the river Rhine in Germany (Hintemann et al., 2006).

In the past, several studies have been published using an ELISA technique for measuring E2 in different matrices (Sadeh et al., 1979, Zhu et al., 2008). Applicability to environmental samples were not considered and the majority of test kits for E2 were not assessed for ecological samples because the immunoassay test kits were optimized for clinical diagnostics (Marcus and Durnford, 1988c, Hanselman et al., 2004, Hintemann et al., 2006a, Zhu et al., 2008). This chapter describes the facile design and synthesis of the 17β-estradiol hapten with both acetate and butyrate linkers through its carbon number 3, as well as the raising of specific antibodies, formatting and characterizing the sensitivity of the competitive ELISA are reported. In addition, validation of its performance as a fast and effective water monitoring tool is also included.

4.2 Materials and Methods

4.2.1 Material and Instrument

4.2.1.1 Materials 17β-estradiiol was purchased from Sigma Aldrich (St. Louis, MO). Other materials were described in section 3.2.1.1 of Chapter 3.

4.2.1.2 Instruments As described in section 3.2.1.2 of Chapter 3.

4.2.2 Hapten Synthesis Two hapten conjugates for the 17β-estradiol were synthesized, by Ms Christine Tan (UNSW) at the Department of Pharmacology, the University of Sydney under the supervision of 94

A/Prof. Robin Allan and Dr. Alice Lee. 17β-Estradiol-acetate (denoted as E2-ACT) and 17β- estradiol-butyrate (denoted as E2-BUT) haptens were synthesized as follows.

4.2.2.1 Attachment of acetate linker onto C 3 of 17β-estradiol The general synthetic approach for the acetate linker is as follows: 17β-estradiol was first alkylated to attach a linker onto a phenolic hydroxyl group at C3 to give the ester compound 1 which was further hydrolyzed to give the acid 2. Compound 2 was purified via crystallization before being converted into the active ester 3.

Procedure for Synthesis of Ester compound (1), Scheme 1

Potassium carbonate (305.2 mg, 2.21 mmol) and ethylbromoacetate (246 μL, 2.21 mmol) were added to a solution of 17E-estradiol (500mg, 1.84 mmol) in acetone (10 mL). The reaction mixture was stirred under reflux for 24 h at 65 ºC with a calcium chloride drying tube. The reaction mixture was evaporated to dryness under vacuum, redissolved in ethylacetate (3 x 20 mL) and washed with citric acid (2 x 30 mL, 2 g/100 mL) and saturated brine (30 mL). The organic layer was dried over sodium sulfate and evaporated to dryness under vacuum, giving an oil for crude 1 with a yield of 0.64 g, Rf = 0.4 1 (dichloromethane/ethyl acetate, 12:1). H NMR (300 MHz, acetone-d6): δ 0.78 (s, 1H), 1.30 (t, J = 7.2 Hz, 3H), 3.73(t, J=8.1Hz, 1H), 4.26 (q, J = 7.2 Hz, 2H), 4.86 (s, 2H), 6.64 (d, J = 3.0 Hz, 1H), 6.70 (dd, J = 3, 8.4 Hz, 1H), 7.21 (d, J = 8.7 Hz, 1H).

Procedure for the Synthesis of Acid compound (2), Scheme 2

Compound 1 (320 mg, 0.894 mmol) was dissolved in THF/MeOH (3:1, 30 mL) and lithium hydroxide (8.94 mL, 1.0 M) was added. The reaction mixture was stirred at room temperature for 24 h. It was then acidified with citric acid (50 mL, 2 g/100 mL) and extracted with ethyl acetate (2 x 100 mL). The organic layer was washed with saturated brine (1 x 60 mL), dried over sodium sulfate and evaporated to dryness under vacuum, giving 236 mg (91% yield) of a white solid of crude 2, Rf = 0.06 (ethyl acetate). The crude 2 was used in the subsequent step 1 without further purification. H NMR (300 MHz, acetone-d6): δ 0.77 (s, 3H), 2.98 (s, 1H), 3.67 (t, J=8.4Hz, 1H), 4.64 (s, 2H), 6.63 (d, J = 2.7 Hz, 1H), 6.70 (dd, J = 2.7, 11.5 Hz, 1H), 7.19 (d, J = 8.7 Hz, 1H).

Procedure for the Synthesis of Active Ester (3), Scheme 3

Compound 2 (190 mg, 0.576 mmol) and N-hydroxysuccinimide (99.44 mg, 0.864 mmol) were dissolved in dry THF (20 mL) under nitrogen with constant agitation. N,N’- dicyclohexylcarbodiimide (178.27 mg, 0.864 mmol) was dissolved in dry THF (10 mL) and added to the above mixture. A white precipitate formed during the initial stage of the reaction and the mixture was left to stir overnight at room temperature. The reaction mixture was filtered through a sintered glass filter under vacuum to remove the urea by-product, then evaporated to dryness over vacuum and redissolved in ethyl acetate (5 mL). The mixture was allowed to sit in an ice bath for 1 h before filtering through a sintered glass filter under vacuum to remove as much of the urea by-product as possible. The mixture was then evaporated to dryness over vacuum, yielding 294 mg of a white fluffy solid. Rf = 0.64 (ethyl acetate). The crude 3 was dissolved in 1 mL ethylacetate and cyclohexane (3.5 mL) was added dropwise to initiate the crystallization process. The mixture was left overnight at 4 qC before the crystals were filtered under vacuum using a sintered glass filter and the crystals were rinsed first with the remaining solution mixture followed by ethylacetate/cyclohexane (2:3). The crystals were left to dry in vacuum for 1 h, yielding 135 mg (71% yield) a white 1 solid. H NMR (300 MHz, acetone-d6): δ 0.77 (s, 3H), 2.90 (s, 4H), 3.67 (t, J=8.4Hz, 1H), 5.12 (s, 2H), 6.72 (d, J = 2.7 Hz, 1H), 6.78 (dd, J = 2.7, 8,7 Hz, 1H), 7.22 (d, J = 8.4 Hz, 1H).

Scheme 1. Synthesis of compound 1

Scheme 2. Synthesis of compound 2

Scheme 3. Synthesis of compound 3

4.2.2.2 Attachment of butyrate linker onto C 3 of 17β-estradiol The general synthetic approach for the hapten with a butyrate linker is as follows: 17β- estradiol was first alkylated to attach a linker arm onto the phenolic hydroxyl group at position 3 of the steroid to give the ester compound 4 which was further hydrolyzed to give the acid compound 5. Compound 5 was purified via crystallization before being converted into an active ester 6.

Procedure for Synthesis of Ester compound (4), Scheme 4

To a solution of 17β-estradiol (500 mg, 1.84 mmol) in acetonitrile (20 mL) was added potassium carbonate (490.1 mg, 3.55 mmol) and ethyl-4-bromobuytrate (382 μL, 2.67 mmol). The reaction mixture was stirred under reflux for 48 h at 95 ºC with a calcium chloride drying tube. The reaction mixture was evaporated to dryness under vacuum, redissolved in dichloromethane (20 mL) and washed with citric acid (2 x 15 mL, 2 g/100 mL) and saturated brine (30 mL). The organic layer was dried over sodium sulfate and evaporated to dryness under vacuum, giving a quantitative yield of 741 mg of crude 4, including impurities and 1 residual solvent, Rf = 0.33 (dichloromethane/ethyl acetate, 12:1). H NMR (300 MHz, acetone-d6): δ 0.77 (s, 3H), 1.21 (t, J = 7.1 Hz, 3H), 2.02 (m, 2H), 2.47 (t, J = 7.3 Hz, 2H), 3.54 (t, J = 6.6 Hz, 2H), 4.08 (q, J = 7.1, 2H), 6.60 (d, J=2.7 Hz, 1H), 6.67 (dd, J=2.8, 8.7Hz, 1H), 7.17 (d, J=8.4Hz, 1H).

Procedure for the Synthesis of Acid compound (5), Scheme 5

Compound 4 (560 mg, 1.45 mmol) was dissolved in THF/MeOH (16 mL, 3:1) and lithium hydroxide (14.5 mL, 1.0 M) was added. The reaction mixture was stirred at room temperature for 2.5 days, and washed with ethyl acetate (15 mL) to remove unreacted organic material, and then washed with citric acid (50 mL, 10 g/100 mL) followed by saturated brine (1 x 30 97 mL) before being extracted with ethyl acetate (2 x 25 mL). The organic layer was dried over sodium sulfate and evaporated to dryness under vacuum, giving 1.02 g of a white solid of crude 5, Rf = 0.51 (ethyl acetate). The compound 5 (1.02 mg) was dissolved in ethyl acetate (15 mL) with heating and allowed to cool slowly to initiate the crystallization process. The mixture was left overnight at room temperature before the crystals were filtered under vacuum using a sintered glass filter. The crystals were left to dry in vacuum for 1 h, yielding 400 mg 1 (77% yields) of white solid. H NMR (300 MHz, acetone-d6): δ 0.77 (s, 3H), 2.05 (m, 2H), 2.49 (t, J = 7.5 Hz, 2H), 3.99 (t, J = 6.3 Hz, 2H), 6.62 (d, J = 2.7, 1H), 6.70 (dd, J = 2.7, 8.4 Hz, 1H), 7.16 (d, J = 8.9, 1H).

Procedure for the Synthesis of Active Ester (6), Scheme 6

Compound 5 (150 mg, 0.419 mmol) and N-hydroxysuccinimide (57.9 mg, 0.503 mmol) were dissolved in dry THF (6 mL) under nitrogen with stirring. N,N’-dicyclohexylcarbodiimide (32.39 mg, 0.16 mmol) was dissolved in dry THF (1 mL) and added to the above mixture. The mixture was left to stir for 6 days at room temperature while monitoring for active ester formation. The reaction mixture was filtered through a sintered glass filter under vacuum to remove the urea derivative, prior to being evaporated to dryness over vacuum and subsequently redissolved in ethyl acetate (5 mL). The mixture was allowed to sit in an ice bath for 1 h before filtering through sintered glass under vacuum to remove as much of the urea derivative as possible. The mixture was then evaporated to dryness over vacuum, yielding 203 mg of a white fluffy solid of crude 6 (showing 5-10% of acid). Rf = 0.49 (ethyl 1 acetate/hexane, 1:1). H NMR (300 MHz, acetone-d6): δ 0.77 (s, 3H), 2.09 (m, 2H), 2.85 (t, J = 7.4 Hz, 2H), 2.88 (s, 4H), 3.67 (t, J = 8.4 Hz, H), 4.07 (t, J=6.2Hz, 2H), 6.66 (d, J = 2.7, 1H), 6.73 (dd, J = 2.7, 8.7 Hz, 1H), 7.17 (d, J = 8.1, 1H).

Scheme 4. Synthesis of compound 4

Scheme 5. Synthesis of compound 5

Scheme 6. Synthesis of compound 6

4.2.2.3 Preparation of conjugates of hapten and carrier proteins or enzyme As described in section 3.2.2.3 of Chapter 3.

4.2.3 Antibody Production and Characterization

4.2.3.1 Immunization and antibody production The main principles of immunization and antibody production for 17β-estradiol-acetate (E2- ACT) and 17β-estradiol-butyrate (E2-BUT) were essentially the same as those of 17α- ethinylestradiol-acetate (EE2-ACT) and 17α-ethynylestradiol-butyrate (EE2-BUT). These were described in 3.2.3.1.

4.2.3.2 Purification of Rabbit IgG As described in section 3.2.3.2 of Chapter 3.

4.2.3.3 Determining antibody concentration As described in section 3.2.3.3 of Chapter 3.

4.2.3.4 Determining optimum working concentration by checkerboard titration As described in section 3.2.3.4 of Chapter 3.

4.2.3.5 Determining Sensitivity

4.2.3.5.1 Preparation of Standard Solution Stock solutions of 17β-estradiol were prepared in anhydrous ethanol at approximately 100, and 10 mg L-1. To construct a calibration curve, a stock E2 standard solution was freshly prepared in glass tubes by diluting 1/100 from 10 mg L-1 of stock solution, giving a 100 μg L- 1 standard solution (in 10% EtOH). Then the 100 μg L-1 standard solution was serially diluted 1 in 3 with 10% EtOH to obtain 33.3, 11.1, 3.3, 1.1, 0.37, 0.12, 0.041, and 0.014 μg L-1 standards.

4.2.3.5.2 Preparation of enzyme conjugate The working solution were prepared by diluting in 1% BSA/PBS as described of the enzyme conjugates for an ELISA in section 3.2.3.5.2 of Chapter 3

4.2.3.5.3 Direct Competitive ELISA protocol The direct competitive ELISA protocol for 17β-estradiol ELISA was the same as described in section 3.2.3.5.3 of Chapter 3.

4.2.3.5.4 Determination of standard curve parameter As described in section 3.2.3.5.4 of Chapter 3.

4.2.3.6 Optimization of EE2 ELISA condition

4.2.3.6.1 The effect of enzyme diluents on assay performance As described in section 3.2.3.6.1 of Chapter 3.

4.2.3.6.2 The effect of organic solvents on the ELISA As described in section 3.2.3.6.2 of Chapter 3

4.2.3.7 Determining specificity As described in section 3.2.3.7 of Chapter 3.

4.2.3.8 Matrix interferences study As described in section 3.2.3.8 of Chapter 3.

4.2.4 Spike and recovery study

The spike and recovery study was conducted to assess the matrix interference of the real water samples for the developed E2 ELISA as well as to assess its performance as an analytical

100 method. The spiked samples were assayed in the E2 ELISA to quantify the 17β-estradiol, which was calculated by intrapolating the E2 concentration from the percent inhibition of a calibration curve (Table 4.1).

Table 4.1 The concentrations of 17α-ethynylestradiol (EE2), 17β-estradiol (E2), and estrone (E1) in the spiked samples. Spike 17α-Ethynylestradiol 17β-Estradiol Estrone Total No. (μg L-1 ) (μg L-1 ) (μg L-1 ) (μg L-1 ) 1 0.5 - - 0.5 2 - - 0.5 0.5 3 - 0.5 - 0.5 4 0.5 0.5 0.5 1.5 5 - 1.0 - 1.0

4.2.5 Validation of the ELISA with GC/MS

Validation of the ELISA with GC/MS for the real-world water samples was conducted in collaboration with Dr Chun Hua Li (a PhD candidate) of the University of Sydney, as described in section 3.2.5 of Chapter 3

4.3 Results and Discussion

4.3.1 Hapten design and synthesis

Based on principles of hapten design for the ELISA (Szurdoki et al., 1995, Lee and Kennedy, 2007), an appropriate hapten should preserve as closely as possible of the target structure in size, shape (3D) and electronic properties for antibody production. A linker or spacer arm of the hapten, which is used to join the hapten to a carrier protein, should be of an appropriate length and should not elicit antibody recognition. According to Lee and Kennedy (2007), the optimal linking group or the preferable length of spacer arm is an alkyl chain of about four to six atoms such as propionic, succinic and caproic acids, because a spacer arm longer than eight carbon may lead to hapten folding, resulting in the production of low affinity antibodies for the target analytes.

101

For E2, the previous studies have focused on designing and synthesising a hapten with the linker attached to the 6-position of E2 (Kuss and Goebel, 1972, Garza and Rao, 1983, Tiefenauer and Andres, 1990). These synthesized derivatives have been used for developing immunoassays to improve the specificity of the antibodies by designing special site of attachment of carrier molecules on the hapten molecule (Exley, 1972b, Exley and Choo, 1974b, Exley and Abuknesha, 1978., Marcus and Durnford, 1988a). Li and co-workers (2004) showed that the attachment of a linker arm to the –OH at the C3-position of the steroid molecules resulted in the production of high−affinity antibodies, providing good sensitivity and specificity. In this experiment, the acetylene substituent group on carbon number 17 of estrogenic hormones was chosen as an epitope for antibody recognition, because all estrogenic hormones share the same backbone structure of the 4 rings, differing only in their substituent groups on carbon 17 and 18. Therefore, the polyclonal antibodies produced against such approach are able to differentiate 17β-estradiol from other steroids such as estrone and estriol by using an acetylene group as substituent on carbon number 17, and can significantly reduced the cross-reactivity with these structurally similar steroid molecules. Ms Christine Tan (UNSW) had successfully synthesized two hapten molecules, one of which has an acetate linker, and the other with a butyrate linker with short chain lengths of 2-4 carbons consisting of saturated carbon bonds as this will ensure sufficient flexibility of the hapten molecule after it is conjugated to the carrier molecule, but not long enough to result in folding into the carrier protein. The saturated carbon chain also minimizes immune recognition of the linker. .

4.3.1.1 Optimal Concentration of Enzyme Conjugates

The enzyme (peroxidase) conjugates used in E2-ELISA were prepared and used for screening CBT as describe in section 3.3.1.1 and Figure 3.4 in chapter 3.

102

3.5

3.0 EE2-ACT12-HRP EE2-ACT20-HRP 2.5 EE2-BUT12-HRP EE2-BUT20-HRP E1-HS-HRP 2.0

1.5

1.0

Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000 Dilution of hapten-enzyme conjugates

Figure 4.1 Titration curves of E2-ACT antibody against five hapten-enzyme conjugates (EE2- ACT12-HRP, EE2-ACT20-HRP, EE2-BUT12-HRP, EE2-BUT2HRP and E1-HS-HRP).

3.5 E2-ACT12-HRP E2-ACT20-HRP 3.0 E2-BUT12-HRP E2-BUT20-HRP

2.5 E2-OX-HRP E2-HS-HRP

2.0

1.5

1.0

Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000 Dilution of hapten-enzyme conjugates

Figure 4.2 Titration curves of E2-ACT antibody against six hapten-enzyme conjugates (E2- ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT2—HRP, E2-HS-HRP and E2-OX- HRP).

103

3.0

EE2-ACT12-HRP EE2-ACT20-HRP 2.5 EE2-BUT12-HRP EE2-BUT20-HRP E1-HS-HRP 2.0

1.5

1.0 Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000 Dilution of hapten-enzyme conjugates

Figure 4.3 Titration curves of E2-BUT antibody against five hapten-enzyme conjugates (EE2- ACT12-HRP, EE2-ACT20-HRP, EE2-BUT12-HRP, EE2-BUT20-HRP and E1-HS-HRP).

3.5 E2-ACT12-HRP E2-ACT20-HRP 3.0 E2-BUT12-HRP E2-BUT20-HRP 2.5 E2-OX-HRP E2-HS-HRP

2.0

1.5

1.0 Absorbance (450 nm) (450 Absorbance 0.5

0.0 100 1000 10000 100000 1000000 10000000 10000000 Dilution of hapten-enzyme conjugates

Figure 4.4 Titration curves of E2-BUT antibody against six hapten-enzyme conjugates (E2- ACT12-HRP, E2-ACT20-HRP, E2-BUT12-HRP, E2-BUT20-HRP, E2-HS-HRP and E2-OX- HRP),

104

According to past research, a good titre is related to low spacer arm recognition, resulting in high affinity for the desired epitope (Li et al., 2004, Schneider et al., 2004). In addition, the heterologous systems, consisting of hapten heterology, bridge heterology and linker attachment heterology, were employed for developing the most sensitive and competitive ELISA.

The E2-ACT antibody at 10 μg mL-1 was used in CBT to study the optimal enzyme-conjugate dilutions. As discussed in Chapter 3, the optimum dilutions were selected at a wavelength of 450 nm where absorbance was around1.0-1.5. The optimum dilution for EE2-ACT 12 –HRP, EE2- ACT 20 – HRP, EE2- BUT 12 – HRP, EE2- BUT 20-HRP, E2-ACT 12 –HRP, E2- ACT 20 – HRP, E2- BUT 12 – HRP, E2- BUT 20-HRP, E2-OX-HRP, E2-HS-HRP and E1- HS-HRP were 1/81000, 1/81000, 1/9000, 1/9000, 1/ 6561000, 1/6561000, 1/9000, 1/9000, 1/3000, 1/9000, 1/ 81000, respectively, as shown in Figures 4.1 and 4.2.

The E2-BUT antibody at 10 μg mL-1 was used in CBT using the same conditions for CBT as for the E2-ACT antibody. Figures 4.3 and 4.4 showed that two enzyme conjugates, E2-OX- HRP and E2-HS-HRP, exhibited a low absorbance, while other enzyme conjugates displayed good response with the E2-BUT antibody. The optimum dilutions for EE2-ACT 12 –HRP, EE2- ACT 20 – HRP, EE2- BUT 12 – HRP, EE2- BUT 20-HRP, E2-ACT 12 –HRP, E2- ACT 20 – HRP, E2- BUT 12 – HRP, E2- BUT 20-HRP, E2-OX-HRP, E2-HS-HRP and E1- HS-HRP were 1/3000, 1/9000, 1/3000, 1/3000, 1/6561000, 1/6561000, 1/9000, 1/3000, 1/1000, 1/1000 and 1/9000, respectively. The Ab-E2-BUT displayed the strongest response when reacting against heterologous linkage pair such as E2-ACT12-HRP and E2-ACT20- HRP.

Ab-E2-ACT generated high titre in the homologous systems with E2-ACT12-HRP and E2- ACT20-HRP.The optimized concentrations of the enzyme conjugate in the homologous systems was higher than those in the heterologous systems by 100-fold. On the other hand, the use of bridge heterology approach with the hapten with a butyrate spacer arm (Ab-E2- BUT) improved the colour development by 1000-fold. The hapten-enzyme conjugates, which produced the high titre (1/81000 to 1/ 6561000), were preceded to the assay sensitivity study.

105

4.3.2 Assay Sensitivity

The optimum concentrations of enzyme conjugates was established against purified antibodies, using competitive direct ELISAs, for assay sensitivity. The affinity of the antibodies was studied by running standard curves against 17β-estradiol. That is, 4 antibodies (two antibodies against each of E2-ACT and E2-BUT haptens) x 11 enzyme conjugate combinations were optimized and evaluate for their sensitivity. The sensitivity of different bleeds of each antibody was also investigated. These antibodies demonstrated improved sensitivity with booster injections, showing affinity maturation of the antibodies specific to the analyte. Polyclonal antibody production in vivo provides low level stimulation of the immune system initially, and generates a stronger response after subsequent booster immunization, resulting in high-affinity antibody. (Siklodi et al., 1995, Janeway et al., 2004).

As shown in Table 4.2, the absolute homologous systems evaluated were Ab-E2 ACT x E2- ACT 12- HRP or E2-ACT 20-HRP, and Ab-E2 BUT xE2-BUT 12 HRP or E2-BUT 20- HRP. -1 These homologous assays exhibited good IC50 values at 0.6 - 1.7 μgL for Ab-E2 ACT, while Ab-E2-BUT demonstrated less sensitivity for the absolute homologous system (i.e., both hapten and linkage are the same). The homologous hapten with heterologous linker conbination showed slightly lower sensitivity than their respective absolute homologous system. For example, the IC50 values of Ab-E2 ACT x E2-BUT 12 HRP or EE2-BUT 20- HRP was 2.1 – 3.7 μgL-1. While Ab-E2 BUT x E2-ACT 12- HRP or E2-ACT 20-HRP -1 exhibited approx. 3-fold better IC50 values ( 0.6 – 1.4 μgL ).

As expected, the heterologous assays significantly improved the assay sensitivity. The heterologous hapten systems such as Ab-E2 ACT or Ab-E2 BUT used with EE2-ACT 12 – HRP, EE2- ACT 20 – HRP, EE2- BUT 12 – HRP, and EE2- BUT 20-HRP, demonstrated -1 -1 increased sensitivity with IC50 at 0.5 - 1.3 μgL and 0.5 - 15.6 μgL for Ab-E2 ACT or Ab- E2 BUT, respectively. The absolute heterologous assays (i.e., both hapten and linkage are different) such as Ab-E2 ACT x EE2-BUT 12-HRP or EE2-BUT 20-HRP pairs, as well as Ab-E2 BUT x EE2-ACT 12-HRP or EE2-ACT 20-HRP pairs demonstrated good sensitivity.

Ab-E2-BUT x EE2-ACT 12-HRP or EE2-ACT 20-HRP pair exhibited an IC50 value of 0.5 μg L-1. The Ab-E2-ACT presented the best sensitivity when pairing with EE2-ACT 12-HRP or EE2-ACT 20-HRP (heterologous hapten with a homologous linkage system).

106

Very low concentration (dilution of 1/1000) of the heterogolous linkage systems was used to prepare standard curves for both antibodies pairing with E2-HS-HRP, E2-OX-HRP and E1- HS-HRP. The poor colour development was observed and no further investigation for sensitivity of these assays was performed.

For the two epitope densities of EE2-ACT 12 –HRP, EE2- ACT 20 – HRP, EE2- BUT 12 – HRP, EE2- BUT 20-HRP, E2-ACT 12 –HRP, E2- ACT 20 – HRP, E2- BUT 12 – HRP, E2- BUT 20-HRP, the ratio of a steroid to enzyme did not affect the assay sensitivity. It was obvious in this study that heterology was the key factor that influences the assay sensitivity.

In summary, for Ab-E2-ACT, the assay using the heterologous hapten with homologous -1 linkage system demonstrated the greatest sensitivity exhibiting an IC50 value of 0.5 μg L (Ab-E2-ACT × EE2-ACT 12-HRP). The heterologous hapten with heterologous linkage -1 system showed the greatest sensitivity with an IC50 value at 0.5 μg L for Ab-EE2-BUT (Ab- E2-BUT × EE2-ACT 20-HRP).

-1 Table 4.2 The IC50 (μg L ) for eleven different enzyme conjugates with the E2 antibodies.

-1 IC 50 of E2 Antibodies ( μgL ) Enzyme Ab-E2 BUT Conjugates Ab-E2 ACT 1st 2nd 3rd 1st 2nd 3rd bleeding bleeding bleeding bleeding bleeding bleeding EE2-ACT 12-HRP 1.5 0.6 0.5 13.7 0.5 0.5 EE2-ACT 20-HRP 0.5 0.5 0.5 10.0 0.9 0.5 EE2-BUT 12-HRP 1.3 0.6 0.6 15.6 3.2 0.8 EE2-BUT 20-HRP 1.1 0.8 1.1 6.7 3.2 0.8 E2-ACT 12-HRP 0.7 1.7 0.6 1.0 0.6 0.8 E2-ACT 20-HRP 0.9 0.8 0.6 0.7 1.4 0.9 E2-BUT 12-HRP 2.4 2.1 2.4 1.7 1.4 6.3 E2-BUT 20-HRP 3.7 2.9 2.9 NA* NA* NA* E2-OX-HRP NA* NA* 7.7 NA* 7.6 NA* E2-HS- HRP NA* NA* 1.9 NA* NA* 1.7 E1-HS-HRP NA* NA* 2.9 NA* 1.9 1.9 * The assay exhibited very low colour development, and did not show the correct standard curve.

107

4.3.2.1 Standard curve parameters and precision (IC80, IC50, IC 20, and maximum absorbance)

Nine point calibration curves for Ab-E2-ACT and Ab-E2-BUT are shown in Figures 4.5 and 4.8. The results were averaged over 21 analyses with Ab-E2-ACT and 11 analyses with Ab- E2-BUT that were conducted on different days. The %CV of %inhibition decreased as the E2 concentration increased for both antibodies. The precision, in term of assay sensitivity was relatively moderate with % CV ranging from 1.6-81% and 2.2-81% for Ab-E2-ACT and Ab-

E2-BUT, respectively (Figure 4.6 and 4.9). With Ab-E2-ACT and Ab-E2-BUT, the IC50 values were 0.38 + 0.07 μgL-1 (% CV = 20%), and 0.55 + 0.22 μgL-1 (% CV = 40%) respectively. The limit of detection (LOD), which is defined as the lowest concentration of an analyte that produces a signal which can be distinguished from zero for a given sample with a stated degree of confidence (Henniona and Barcelo, 1998), were also determined by using the concentration of standard solution causing 20% inhibition of colour development (IC20) and -1 by Ab-E2-ACT and Ab-E2-BUT, the IC20 values were 0.04 + 0.01 μgL (%CV = 23%), and 0.05 + 0.03 μgL-1 (%CV = 52%), respectively. The amount of E2 that could inhibit 80% of -1 the antibody binding (IC80) was 5.8 + 2.0 μgL (%CV = 34%) for Ab-E2-ACT, and 4.8 + 2.4 μg L-1 (%CV = 49%) for Ab-E2-BUT (Figure 4.7 and 4.10).

The average maximum absorbance was higher for the Ab-E2-ACT (1.015+ 0.509 with % CV= 50%) than the Ab-E2-BUT (0.684 + 0.149 with % CV= 22%). The % CV for the maximum absorbance ranged between 33 to 68% for Ab-E2-ACT, and 26 to 74% for AB-E2- BUT (Table 4.3).

Table 4.3 Standard curve parameters and precision of the Ab-E2-ACT Parameters Concentration (μgL-1) S.D. a %CV b -1 IC50 0.38 μgL 0.07 20 -1 IC20 0.04 μgL 0.01 23 -1 IC80 5.80 μgL 1.97 34 Maximun absorbance 1.015 0.509 50 a standard deviation b percent coefficient variation

108

% Inhibition

100 Colour development 1.4

90 1.2 80

70 1

60 0.8 50 0.6

% Inhibition % 40

30 0.4 nm) 450 ( Absorbance

20 0.2 10

0 0 0.01 0.1 1 10 E2 t ti ( L 1 )

Figure 4.5 The relationship between % inhibition and colour development of a nine point calibration curve assayed by the E2-ACT antibody. Each data point is an average of 21 analyses. The square indicates absorbance against the E2 concentration and the circle indicates % inhibition. 90

%CV 40

0 0.01 0.1 1 10 100 E2 concentration (μgL-1)

Figure 4.6 The % CV of the absorbance ( ) and % inhibition ( ). Each point is an average of 21 analyses by the E2-ACT antibody.

109

10 ) -1

g L 1 μ

0.1 E2 concentration (

0.01 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Number of assay

Figure 4.7 A Plot of IC80 ( ), IC50 ( ) and IC20 ( ) values of 21 analyses for the E2-ACT antibody. The solid lines indicate average values. The dotted lines indicate the upper and lower limits showingan average E2 concentration ± standard deviation.

Table 4.4 Standard curve parameters and precision for the E2-BUT antibody Parameters Concentration (μgL-1) S.D. a %CV b

IC50 0.6 0.2 40

IC20 0.05 0.03 52

IC80 4.8 2.7 49 Maximun absorbance 0.684 0.149 22 a standard deviation b % coefficient variation

110

Colour Development 100 0.7 % Inhibition 90 0.6 80

70 0.5

60 0.4 50 0.3

% Inhibition % 40

30 0.2 nm) 450 ( Absorbance

20 0.1 10

0 0 0.01 0.1 1 10 E2 concentration ( μg L-1 ) Figure 4.8 The relationship between % inhibition and colour development of a nine point calibration curve assayed by the E2-BUT antibody. Ecach data point is an average of 11 analyses. The square indicates absorbance and the circle indicates % inhibition.

100

50 % CV 40

0 0.01 0.1 1 10 E2 concentration ( μg L-1 )

Figure 4.9 The % CV of the absorbance ( ) and % inhibition ( ) of 11 analysis by the EE2- BUT antibody

111

10 ) -1

g L 1 μ

0.1 E2 concentration (

0.01 0 1 2 3 4 5 6 7 8 9 10 11 Number of assay

Figure 4.10 A Plot of IC80 (), IC50 ( ) and IC20 values of 11 analyses by the E2-BUT antibody. The solid lines indicate the average values. The dotted lines show the upper and lower limits (i.e., average E2 concentration + standard deviation)

4.3.3 Characteristic of EE2 ELISA

4.3.3.1 Assay Specificity

To investigate the specificity of these ELISAs, structurally similar natural and synthetic estrogens that may occur in STP effluents or in river water were selected. They were 17α- Ethynylestradiol, estriol, estrone, estradiol dipropionate, estradiol dipropionate, progesterone, 17α-estradiol, medroxyprogesterone, ethynylestradiol-3-methyl ether, and 17α- ethynylestradiol-3-cyclopentyl ether.

The results of the specificity or cross reactivity study are presented in Table 4.5.

112

-1 Table 4.5 The IC50 (μgL ) and cross-reactivity (CR %) for selected estrogens with two E2 antibodies.

Compounds Structure Ab-E2-ACT Ab-E2- BUT

IC50 CR IC50 % CR IC50 CR IC50 % CR (μgL-1) (%) (mol L-1) (μgL-1) (%) (mol L-1)

-9 -9 17β- estradiol OH 0.4 100 1.5 × 10 100 0.5 100 1.8×10 100 CH 3

-7 -7 17α-ethynylestradiol OH >100 < 0.4 3.4×10 0.4 >100 <0.5 3.4×10 0.5 CH 3 CH

-8 -8 OH Estriol CH 16.8 2.4 5.8×10 2.5 75.7 0.7 2.6×10 0.7 3 OH

113

O -7 -7 Estrone CH3 >100 < 0.4 3.7×10 0.4 >100 <0.5 3.7×10 0.5

Estradiol dipropionate O >100 < 0.4 2.6×10-7 0.6 36.5 1.4 9.5×10-8 1.9 O CH 3 CH 3

H C 3 O

-7 -7 Progesterone O CH >100 < 0.4 3.2×10 0.5 >100 <0.5 3.2×10 0.6 3 CH 3

CH 3

-7 -7 17 α- estradiol OH >100 < 0.4 3.7×10 0.4 >100 <0.5 3.7×10 0.5 CH 3

114

-7 -7 Medroxyprogesterone O >100 < 0.4 2.9×10 0.5 >100 <0.5 2.9×10 0.6

CH CH 3 3

CH 3

OH CH -7 -7 3 Ethynylestradiol-3- methyl CH >100 < 0.4 3.2×10 0.5 >100 <0.5 3.2×10 0.6 ether (Mestranol)

H C 3 O

OH -7 -7 CH 17α -ethynylestradiol 3- 3 >100 < 0.4 2.7×10 0.5 >100 <0.5 2.7×10 0.7 CH cyclopentyl ether (Quinestrol) O

115

These two antibodies, Ab-E2-ACT and Ab-E2-BUT, proved to be very specific for E2, and only negligible cross-reactivity of <2.5% were observed for the nine other structurally-related compounds. The IC50 for each of these compounds, 17α-ethynylestradiol, estradiol dipropionate, estrone, progesterone, 17α- estradiol, medroxyprogesterone, ethynylestradiol-3- methyl ether, and 17α−ethynylestradiol−3−cyclopentyl ether, was above 100 μg L-1, except -1 for estriol with the E2-ACT antibody for which the IC50 was 16.8 μg L . Hence, substantial cross−reactivity existed only for estriol exhibiting the cross−reactivity of 2.4%, but is low enough to consider the assay being very specific to E2.

The E2-BUT antibody exhibited a similar cross-reactivity pattern as the E2-ACT antibody. The cross−reactivity of the E2-BUT antibody for 17α-ethynylestradiol estrone, progesterone, 17α-estradiol, medroxyprogesterone, ethynylestradiol-3-methyl ether, and 17α− ethynylestradiol−3−cyclopentyl ether was negligible (<0.5%). Only estriol and estradiol dipropionate presented notable cross−reactivity of 0.7 and 1.4%, respectively.

In summary, both E2 antibodies exhibited excellent specificity. The cross−reactivity of the developed E2 ELISA for EE2 and typical metabolites such as estrone was sgnificantly lower than the previously reported studies that presented cross−reactivity of 1 -15% (Goda et al., 2000, Hintemann et al., 2006). In Hintemann and co-workers’ study, the cross-reactivity of the E2 ELISAs for estriol was below 3%. Interestingly, our E2 ELISAs exhibits a very similar patterns to the this assay (Hintemann et al., 2006b).

116

4.3.3.2 Assay Optimization

4.3.3.2.1 Effects of Solvents

1.4

5% EtOH 1.2 10% EtOH 20% EtOH 1.0 MilliQ water

0.8

0.6

0.4 Absorbance (450 nm) (450 Absorbance

0.2

0.0 0.01 0.1 1 10 17β-estradiol (μgL-1)

Figure 4.11 Effect of solvent on absorbance for Ab-E2-ACT (MilliQ Water, 5% EtOH, 10% EtOH, and 20% EtOH ).

100

50 5% EtOH

% Inhibition % 40 10% EtOH

30 20% EtOH MilliQ water 20

0 0.01 0.1 1 10 17β-estradiol (μgL-1)

Figure 4.12 Standard curves of E2 in different solvents (MilliQ Water, 5%, EtOH, 10% EtOH, and 20% EtOH) using Ab-E2-ACT.

117

1.0

5% MeOH

0.8 10% EtOH

20% MeOH

0.6

0.4 Absorbance (450 nm) (450 Absorbance

0.2

0.0 0.01 0.1 1 10 17β-estradiol (μgL-1)

Figure 4.13 Effects of solvents (5%, MeOH, 10% MeOH, and 20% MeOH) on absorbance using Ab-E2-ACT.

100

50 5% MeOH 40 % Inhibition % 10% MeOH 30 20% MeOH 20

0 0.01 0.1 1 10 -1 17β-estradiol (μgL )

Figure 4.14 Standard curves of E2 in different solvents (5%, MeOH, 10% MeOH, and 20% MeOH) using Ab-E2-ACT.

118

0.8

10% acetone

20% acetone

0.6 10% acetonitrile

0.4

0.2 Absorbance (450 nm) (450 Absorbance

0 0.01 0.1 1 10 -1 17β-estradiol (μg L ) Figure 4.15 Effect of solvents (10% acetone, 20% acetone, 10% acetonitrile, and 20% acetonitrile) on absorbance using Ab-E2-ACT.

100

50 10% acetone

% Inhibition % 40 20% acetone

10% acetonitrile 30 20% acetonitrile 20

0 0.01 0.1 1 10 -1 17β-estradiol (μg L ) Figure 4.16 Standard curves of E2 in different solvents (10% acetone, 20% acetone, 10% acetonitrile, and 20% acetonitrile) using Ab-E2-ACT

119

1.4

5% EtOH 1.2 10% EtOH 20% EtOH 1.0 Milli Q water

0.8

0.6

0.4 Absorbance (450 nm) (450 Absorbance

0.2

0.0 0.01 0.1 1 10 17β-estradiol (μgL-1)

Figure 4.17 Effect of solvents (MilliQ Water, 5% EtOH, 10% EtOH, and 20% EtOH) on absorbance using Ab-E2-BUT.

100

50 5% EtOH % Inhibition % 40 10% EtOH

30 20% EtOH Milli Q water 20

0 0.01 0.1 1 10 17β-estradiol (μg L-1)

Figure 4.18 Standard curves of E2 in different solvents (MilliQ Water, 5%, EtOH, 10% EtOH, and 20% EtOH) using Ab-E2-BUT.

120

0.8

5% MeOH

0.6 10% MeOH 20% MeOH

0.4

0.2 Absorbance (450 nm) (450 Absorbance

0.0 0.01 0.1 1 10 -1 17β-estradiol (μgL ) Figure 4.19 Effect of solvents (5%, MeOH, 10% MeOH, and 20% MeOH) on absorbance using Ab-E2-BUT.

100

50 5% MeOH

% Inhibition % 40 10% MeOH 30 20% MeOH 20

0 0.01 0.1 1 10 -1 17β-estradiol (μgL ) Figure 4.20 Standard curves of E2 in different solvents (5%, MeOH, 10% MeOH, and 20% MeOH) using Ab-E2-BUT.

121

0.7

0.6 10% acetone

20% acetone 0.5 10% acetonitrile

20% acetonitrile 0.4

0.3

0.2 Absorbance (450 nm) (450 Absorbance

0.1

0.0 0.01 0.1 1 10 -1 17β-estradiol (μgL ) Figure 4.21 Effect of solvents (10% acetone, 20% acetone, 10% acetonitrile, and 20% acetonitrile) on absorbance using Ab-E2-BUT.

100

50 10% acetone

% Inhibition % 40 20% acetone 10% acetonitrile 30 20% acetonitrile

0 0.01 0.1 1 10 17β-estradiol (μg L-1)

Figure 4.22 Standard curves of E2 in different solvents (10% acetone, 20% acetone, 10% acetonitrile, and 20% acetonitrile) using Ab-E2-BUT.

122

1.0 MilliQ Water

5% of the solvents

10% of gthe solvents 0.8 20% of the solvent ) -1 g L

μ 0.6 ( 50 IC

0.4

0.2

0.0 Ethanol Methanol Acetone Acetonitrile

Organic Solvent

Figure 4.23 Effect of organic solvents on the calibration curve for Ab-E2-ACT.

4.0

MilliQ Water

5% of the solvents

10% of the solvents 3.0

) 20% of the solvents -1 g L μ ( 50

IC 2.0

1.0

0.0 Ethanol Methanol Acetone Acetonitrile Organic Solvent Figure 4.24 Effect of organic solvents on the calibration curve for Ab-E2-BUT 123

As described in section 3.3.3.2.1, the solvents used for preparing standard solutions and samples also can affect the ELISA performance, and this experiment studied the effects of types of organic solvents and their concentratrion in aqueous solution on ELISA performance.

As indicated by Figures 4.23 and 4.24, use of methanol and ethanol up to 10% for preparing standard solutions and samples did not affect the assay sensitivity. For the assay using Ab-E2- -1 ACT, 5 and 10% ethanol presented the IC50 values at 0.33 and 0.39 μg L (Figure 4.12), -1 respectively and 5 and 10% of methanol showed the IC50 valuesat 0.32 and 0.44 μgL (Figure

4.14) respectively. For the assay using Ab-E2-BUT, the IC50 of 5 and 10% ethanol were 0.33 -1 -1 and 0.47 μg L , while the IC50 of 5 and 10% methanol were 0.37 and 0.38 μgL , respectively (Figure 4.18 and 4.20). Figures 4.11, 4.13, 4.17 and 4.19 demonstrated that ethanol and methanol at 20% in water retained good colour development, but reduced the assay sensitivity for both Ab-E2-ACT and Ab-E2-BUT. Poor colour development and assay sensitivity were observed when using acetone or acetonitrile at 10 and 20% for both Ab-E2-ACT and Ab-E2- BUT, (Figures 4.15 and 4.21). Acetone and acetonitrile at 10% and 20% displayed both low -1 colour development (0.566 - 0.797Abs), and reduced sensitivity (IC50 at 0.62 -0.92 μgL ) for Ab-E2-ACT (Figure 4.15 and 4.16). In the same way, Ab-E2-BUT observed low colour -1 development (0.261 - 0.603Abs) and poor sensitivity (IC50 at 1.13 - 3.92 μgL ) (Figure 4.21 and 4.22) when using acetone or acetonitrile at 10 or 20%, respectively. It was therefore concluded that no significant effects on the dose-response of 17β-estradiol were observed when 5 to 10% of methanol or ethanol were used as the solvents. Since absolute solubility of E2 in aqueous environment of an ELISA is important for quantification, 10% ethanol was chosen for further studies.

4.3.3.2.2 Effect of enzyme conjugate diluents

As discussed in section 3.3.3.2.2, enzyme-conjugate diluents, such as PBS, can reduce non−specific interactions, and also Tween 20 is generally used as a non−specific binding preventer for immunoassay (Wedge and Svenneby, 1986, Mohammed and Esen, 1989). Therefore, in order to determine their influence on the E2 ELISA, four enzyme-conjugate diluents were tested as described in section 3.3.3.2.2

124

Table 4.6 Effect of enzyme conjugate diluents on E2 ELISAs.

Ab-E2-ACT Ab-E2-BUT Assay diluents Maximum IC50 Maximum IC50 Absorbance (μg L-1) Absorbance (μg L-1) 1% BSA/PBS 1.942 0.4 0.514 0.5 1% BSA/PBS+0.05% Tween 20 1.479 0.5 0.395 0.6 1% BSA/PBS+0.1% Tween 20 0.441 1.4 0.371 0.6 1% BSA/PBS+0.25% Tween 20 0.465 1.0 0.325 1.1

As demonstrated in Table 4.6, increasing Tween 20 from 0.05 to 0.25% reduced both the colour development and assay sensitivity of both antibodies. The results for both E2 antibodies are in good agreement with the previous studies in section 3.3.3.2.2. When Tween

20 was added to the enzyme diluents, the IC50 value for both assays rose by approximately two-fold when an enzyme diluent of 1% BSA/PBS with 0.25% Tween 20 was used. As Shan and co-worker reported (1999), Tween 20 has considerable effects on assay performance, especially on the colour development because of non−specific hydrophobic interactions between Tween 20 and non−polar estradiol in an aqueous environment interfering with specific analytes and antibody interactions. Since, Tween 20 did not improve the sensitivity and reduce colour development; it was not used in further studies.

3.3.3.3 Matrix Interferences

3.3.3.3.1 Effect of pH

Bergmeyer (1974) stated that a change in pH can affect the rate of enzyme reaction, as well as affecting the structure and activity of proteins which are susceptible to pH. Hence, the determination of pH effect on assay performance was previously described in the section -1 3.3.3.3.1. The control (pH 7.2) showed an IC50 value of 0.53 μgL with colour development of 0.787 Abs, while pH 3, 5 and 11 gave low colour development (<0.35 Abs). A noticeable gradual declining of the colour development and lose of assay sensitivity was observed when the pH was raised above 7.2. The pH 3, 8, 9 and 11 reduced the assay sensitivity, while pH 5 stayed at a value similar maintain the sensitivity as oin the control (pH 7.2), as illustrated in Figure 4.25. The optimum pH is determined to be between 7 and 8, and this suggests that the sample must be pH adjusted prior to assay.

125

3.5 IC50 (μg L-1) 0.9 Color development 0.8 3

0.7 2.5 0.6 ) -1 2 0.5 g L μ

1.5 0.4 IC 50 ( 0.3 1 nm) 450 ( Absorbance 0.2

0.5 0.1

0 0 357911 pH

Figure 4.25 Effect of pH on the E2 ELISA standard curve. The square indicates absorbance against the pH and the circle indicates IC50 value against pH.

4.3.3.3.2 Effect of water type and ionic strength

As described in section 3.3.3.3.2, water samples were collected from the lagoon at Tahbilk Winery wetlands (pH 6.4), McWilliam’s Winery Reservoir (pH 6.7), and sea water from Maroubra Beach (pH 8.2), and were screened for the effect of water matrices on the E2 ELISAs.

In terms of the antigen-antibody interaction, water matrices had no observable effects on assay sensitivity, as illustrated in Figure 3.29. Nevertheless, real water matrices slightly interfered with enzyme activity, contributing to a decline in absorbance for samples sourced from the lagoon at Tahbilk Winery wetlands and at McWilliam’s Winery Reservoir when compared to purified water as a control (Figure 4.26). Results from these studies are in accordance with the effect of water matrices for the EE2 ELISA in section 3.3.3.3.2, which

126 indicated that raw water samples without the sample preparation step did not interfere with the E2 ELISA (Figure 4.27).

1.0

purified water

0.8 McWilliam’s winery reservoir

lagoon at Tahbilk Winery wetlands sea water from Maroubra Beach 0.6

Absorbance (450 nm) (450 Absorbance 0.4

0.2

0.0 0.01 0.1 1 10 17β-estradiol (μg L-1)

Figure 4.26 Matrix effects on absorbance for the E2 ELISA with different types of water samples (purified water, McWilliam’s winery reservoir, lagoon at Tahbilk Winery wetlands, and sea water from Maroubra Beach).

100

50 purified water

% Inhibition % McWilliam’s winery reservoir 30 lagoon at Tahbilk Winery 20 wetlands sea water from Maroubra 10 Beach

0.01 0.1 1 10 -1 17β-estradiol (μgL )

Figure 4.27 Standard curves of E2 in different types of water (purified water, McWilliam’s winery reservoir, lagoon at Tahbilk Winery wetlands, and sea water from Maroubra beach.

127

The influence of the ionic strength of dissolved salts in water samples on the performance of E2 ELISAs was investigated as described previously in section 3.3.3.3.2. As demonstrated in

Table 4.7, 1 M (NH4)2SO4 and 1M NaCl did not alter the colour development, while other compounds considerably reduced colour development for both Ab-E2-ACT and Ab-E2-BUT.

For Ab-E2-ACT, 17β-estradiol in solutions of 1 M (NH4)2SO4 and 1 M NaCl as interfering -1 agents performed acceptable IC50 values at 0.2 μgL . In addition, colour development of E2- ACT ELISA was diminished, subsequently leading to the loss of sensitivity when 17β- estradiol was serially diluted in solvent containing, 1 M MgSO4, 1 M CaCl2, 1 M MnSO4, 1 M

KCl, 0.05 M AlCl3, 0.001 to 0.1 M CuSO4, and 0.001 to 0.1 M Fe2(SO4)3. Furthermore, all the tested interfering agents, which were diluted in 17β-estradiol solutions, affected antibody activities resulting in some degree of loss of assay sensitivity for the E2-BUT ELISA.

Table 4.7 Effect of ionic strength on the E2 ELISAs.

Ab-E2-ACT Ab-E2-BUT

Compounds Colour IC50 Colour IC50 development (Abs) (μgL-1) development (Abs) (μgL-1) Control (10% EtOH) 0.949 0.3 0.829 0.6 1 M (NH4)2SO4 0.832 0.2 0.792 3.8 1 M NaCl 0.856 0.2 0.768 1.5 1 M MgSO4 0.116 0.4 0.172 2.3 1 M CaCl2 0.180 0.4 0.281 1.2 1 M MnSO4 0.098 0.6 0.171 0.2 1 M KCl 0.308 0.3 0.151 0.45 0.05 M AlCl3 0.054 0.2 0.079 >100 0.1 M CuSO4 0.201 >100 0.177 92.04 0.01 M CuSO4 0.464 >100 0.272 5.43 0.001 M CuSO4 0.435 0.45 0.275 1.36 0.1 M Fe2(SO4)3 0.066 >100 0.063 >100 0.01 M Fe2(SO4)3 0.358 0.39 0.515 >100 0.001 M Fe2(SO4)3 0.348 0.22 0.468 14.89

4.3.5.3 Effect of humic acid

As previous described in section 3.3.5.3, humic acid (HA) was used as an interfering agent in order to study the effect of organic materials on the performance of the E2 ELISA. A commercial HA was serially diluted 0.01 to 10 mg L-1 for this study.

128

A loss of sensitivity was observed with concentration of humic acid higher than 1 mg L-1 -1 -1 -1 -1 (Figures 4.29). Humic acid at 0.01 mg L , 0.1 mg L , 1 mgL and 10 mg L resulted in IC50 values of 0.3, 0.3, 0.8 and 1.0 μgL-1 and maximum absorbance at 0.868, 0.858, 0.775 and 0.457 Abs, respectively (Figure 4.28). The calibration curve of the control (purified water -1 with no interfering agent) gave an IC50 value of 0.3 μg L with colour development at 0.878 Abs. It can be seen that the humic acid at 0.01 and 0.1 mg L-1 did not affect the sensitivity and colour development of the assay, compared to the control. The acceptable level of humic acid in environmental samples is between 0.01 to 0.1 mg L-1 for the E2 ELISA. Concentrations of humic acid in natural water and in waste water rarely exceed 0.1mgL-1 generally. Thus, effects of humic acid in water samples on the ELISA performance would be minimum.

The results of the matrix effects on the E2 ELISA are in good agreement with the previous study of the EE2 ELISA in section 3.3.5.3. This suggests that other matrix components in surface and wastewater samples could potentially interfere with the antibody-antigen interaction, and validation of each environmental matrix to be analysed is important (Deng et al., 2003, Schneider et al., 2004, Schneider et al., 2005).

129

Purified water 0.8 Humic acid 0.1 mg L-1

Humic acid 10 mg L-1

0.6 Humic acid 1 mg L-1

Humic acid 0.01 mg L-1

0.4 Absorbance (540 nm) (540 Absorbance 0.2

0 0.01 0.1 1 10

17β-estradiol (μgL-1) Figure 4.28 Effects of humic acid on the absorbance of the E2 ELISA.

100

50 Purified water 40 Humic acid 10 mg L-1 % Inhibition % Humic acid 1 mg L-1 30 Humic acid 0.1 mg L-1 Humic acid 0.01 mg L-1 20

0 0.01 0.1 1 10

-1 17β-estradiol (μg L )

Figure 4.29 Effects of humic acid (HA) on the calibration curve.

130

4.3.6 Recoveries of EE2 from spiked purified and field water

Purified water and three water sources, Mcwilliam Winery Reservoir, Lagoon of Tahbilk Winery Wetland, and Maroubra Beach sea water were spiked with E2 at 0, 0.5 and 1 μgL-1 concentrations. The spiked samples were analyzed by the E2 ELISAs using Ab-E2 ACT against EE2-ACT 12-HRP.

As illustrated in Figure 4.30, E2 recovery ranged between 84 – 113% for purified water, 68- 100% for McWilliam Winery Reservoir, 62 – 100% for Lagoon of Tahbilk Winery Wetland and 93 – 110% for Maroubra Beach sea water, and linear regression was applied to the spike and recovery graph as illustrated in Figure 4.31. This study used purified water (control sample) as a standard curve. It can be seen that the recoveries of sample from lagoon at Tahbilk Winery wetland and McWilliam Winery Reservoir were underestimated, due to the matrix interference from sources. The regression coefficient (R2) of purified water, McWilliam’s Winery Reservoir, Lagoon at Tahbilk Winery wetlands and sea water from Maroubra Beach was 0.991, 0.984, 0.976 and 0.986, respectively. However, as can be seen from the results, satisfactory recoveries were obtained and the assays had negligible matrix dependency.

131

93 1 78 84 84

110 0.5 90 68 105 g L-1) μ 110 0.5 62 72 113

100 101 sea water from Spiking Level ( 0 101 Maroubra Beach 100 lagoon at Tahbilk Winery wetlands 100 McWilliam’s winery 0 100 reservoir 101 purified water 100

0 20 40 60 80 100 120 % Recovery

Figure 4.30 Average values (μgL-1) of spiking and recovery (%) from three water sources.

0.8 ) -1 0.6 g L

μ ...... y = 1.377x R² = 0.991 (purified water)

- - - y = 1.278x R² = 0.976 SPIKE ( 0.4 (lagoon at Tahbilk Winery wetlands)

____ y = 1.251x R² = 0.984 (McWilliam's Winery Reservoir) 0.2 _ .. _ .. y = 1.006x R² = 0.986 (sea water from Maroubra Beach)

0 0 0.2 0.4 0.6 0.8 1 E2 ELISA (μgL-1)

Figure 4.31 Average values (μgL-1) of spiking and spiking level (μgL-1) from three water sources: purified water, Mcwilliam’s winery Reservoir, lagoon at Tahbilk Winery wetlands, and sea water from Maroubra Beach by the E2 ELISA.

132

4.3.7 Validation of ELISA with GC/MS

4.3.3.7 Determination of 17β-estradiol in water samples by the developed E2 ELISA

Results obtained from the developed E2 ELISA were used to determine the accuracy by comparing with the reference measurements using GC-MS. River water and effluent samples from the sewage treatment plants were collected and extracted by Dr Chun Hua Li of the University of Sydney, as described previously in section 3.3.3.7. All samples were spiked with E2 at 0, 20 and 50 ng L-1. The values obtained with the E2 ELISA were well correlated to those of GC-MS with a correlation coefficient of 0.936, although there was a slight tendency for the ELISA values to be slightly lower than GC-MS, as shown in Figure 4.32 the data suggested that the developed E2 ELISAs were accurate.

50 y = 1.159x - 1.118 R² = 0.936

40 ) -1 30

20 E2 GCMS (ngL E2 GCMS

0 01020304050 E2 ELISA (ngL-1)

Figure 4.32 Comparison of analytical results between the E2 ELISA and GC-MS for determination of E2 spiked in river water.

133

4.4 Conclusion

A competitive direct enzyme-linked immunosorbent assay (cd ELISA) specific to E2 was developed with hapten molecules synthesised with the acetate and butyrate linkers attached at the 3-position of 17β-estradiol. The specific polyclonal antibodies were raised against a conjugate of 17β-estradiol-acetate hapten to keyhole limpet hemocyanin (KLH), and 17β- estradiol-butyrate hapten and KLH.

The purified antibodies were then titrated to obtain the optimum working conditions for the ELISA. Of all the antibody/enzyme conjugate combinations tested, Ab-E2 ACT displayed a -1 strong response with EE2-ACT 12-HRP with an IC50 value of 0.5 μg L , and Ab-E2-BUT -1 well performed with EE2-ACT 20 HRP with an IC50 value of 0.5 μg L . As expected, the heterologous assays exhibited higher sensitivity compared to the homologous assays. Additionally, hapten heterologous system and linkage heterologous system presented high−sensitivity and good colour development, while the epitope ratio had little influence on the sensitivity of E2 ELISA.

The assays developed from these two antibodies proved to be highly specific for E2, showing only negligible cross-reactivity for the endogenous compounds. Of the water miscible solvents tested, methanol and ethanol up to 10% can be used without affecting the assay sensitivity. For the purpose of routine analysis, 10% ethanol was chosen. Addition of Tween 20 in assay diluents resulted in considerable colour inhibition and reduced sensitivity, thus it was excluded from the reagent.

The matrix interferences of the developed E2 ELISA were investigated using three water samples collected from the lagoon at Tahbilk Winery wetlands, McWillaim’s Winery Reservoir, and sea water from Maroubra Beach. No significant effects were observed, suggesting the ELISA is fairly tolerant of sample matrices.

Effects of humic acid, inorganic salt, ionic strengths and pHs on the assay performance were also investigated. The assay displayed acceptable robust behaviour in the presence of humic acids up to 0.1 mg L-1. Salt ions have major effects on assay performance, but only at high concentrations that are not naturally present in water. The optimum pH was determined to be approximately 7.

134

Furthermore, the analyses of spiked samples by E2 ELISA were well correlated with those derived from the GC-MS. Although the ELISA method provided slightly underestimated values compared to the GC-MS method, probably due to a loss during several transfer between three different laboratories. However, given the complexity of both techniques and analysis of ultra low concentrations, the data suggest that the developed E2 ELISA as presented in this thesis is acceptably accurate.

135

CHAPTER 5 DEVELOPEMTNOF TESTOSTERONE ELISA

(T ELISA)

5.1 Introduction

The androgenic hormones, such as testosterone and progesterone, are the most potent hormones occurring naturally in the human body. These natural hormones are released into the environment mainly through sewage treatment plants (STPs). These endocrine-disrupting compounds have received increasing attention in recent years (Chang et al., 2008) – for example the masculinisation of fish due to exposure at concentrations as low as 1 ngL-1 (Jenkins et al., 2001, Orlando et al., 2004). Clearly, these substances possess activities that are able to interfere with normal endocrine functions by inducing decreased fertility and hermaphroditism in aquatic organisms (Fan et al., 2007, Lafont and Mathieu, 2007).

Natural and synthetic androgens have been used as growth promoters in human and veterinary therapy, which can subsequently be discharged into the environment via STPs (Orlando et al., 1999, Orlando et al., 2004). The development of sensitive and reliable method to monitor these compounds at ultra low concentrations in complex water matrices for environmental risk assessment remains a challenge. Many studies generally use instrumental methods such as high−performance liquid chromatography (HPLC), liquid chromatography-mass spectrometry (LC-MS), liquid chromatography-tandem mass spectrometry (LC-MS/MS), gas chromatography-mass spectrometry (GC-MS) and gas chromatography-tandem mass spectrometry (GC-MS/MS) to quantify natural hormones in water samples because of modern-day instrument reliability (Snyder et al., 1999, Kelly, 2000, Ingrand et al., 2003, Vulliet et al., 2007). However instrumental methods have several shortcomings; these include extensive cleanup (even with the mass spectrometric analysis, large sample volume requirement (in some cases), capital costs of expensive instrumental hardware acquisition, technical expertise in operation and lengthy analytical times (Goda et al., 2000; Huang and Sedlak, 2001). For high−throughput screening, Enzyme-Linked Immunosorbent Assay (ELISA) is a good alternative which promises speed, simplicity, high-throughput and cost- effectiveness for quantitative analysis. It is one of the most sensitive methods available because of antibodies posses high−affinity for their antigens, and an ELISA fulfills most analytical criteria. The limitations of ELISA are the challenges of generating highly−specific

136 and sensitive antibodies or antibodies of desirable characteristics, with low matrix interference and cross−reactivity (Henniona and Barcelo, 1998, Lee and Kennedy, 2000).

Quantification of testosterone by ELISA has been conductedd particularly for clinical diagnostic evaluation of plasma, serum and in humans urine analysis (Yap et al., 1996, Al- Dujaili, 2005, Lu et al., 2006), Using such ELISAs, but without adequate validation for environmental matrices, few environmental studies have been reported for detection of androgenic hormones in environmental samples (Chang et al., 2008, Streck, 2009). This study aims to develop a highly sensitive and specific direct ELISA and validation of its performance for monitoring testosterone as a major androgen in aquatic environment.

5.2 Materials and Methods

5.2.1 Materials and Instrumentation

5.2.1.1 Materials Testosterone was purchased from Sigma Aldrich (St. Louis, MO, USA). Other materials were described in section 3.2.1.1 of Chapter 3.

5.2.1.2 Instrumentation As described in section 3.2.1.2 of Chapter 3.

5.2.2 Preparation of conjugates of haptens and carrier proteins or enzymes As described in section 3.2.2.3 of Chapter 3, conjugation of haptenic analogues of testosterone−3-O-carboxymethyl-oxime (T-CMO) and progesterone−3- O-carboxymethyl- oxime (P-CMO) to the carrier proteins (KLH) and horseradish peroxidase were performed by using a N-hydroxysuccinimide (NHS) and N,N′-dicyclohexylcarbodiimide (DCC) mediated reaction. The structures of these haptens are shown in Figure 5.1.

137

CH O CH CH 3 3 3 CH 3

CH 3 CH 3

OH OH O O N O N O Testosterone−3-O-carboxymethyl-oxime Progesterone−3- O-carboxymethyl-oxime (T-CMO) (P-CMO)

Figure 5.1 Structures of androgenic haptens: Testosterone−3-O-carboxymethyl-oxime (T-CMO) and Progesterone−3-O-carboxymethyl-oxime (P-CMO).

5.2.3 Antibody Production and Characterisation

5.2.3.1 Immunisation and antibody production

The immunisation regime and antibody production for testosterone-3-O-carboxymethyl- oxime (T-CMO) are essentially the same as those of 17α-ethynylestradiol-acetate (EE2-ACT) and 17α-ethynylestradiol-butyrate (EE2-BUT) described in section 3.2.3.1.

5.2.3.2 Purification of rabbit IgG As described in section 3.2.3.2 of Chapter 3.

5.2.3.3 Determining antibody concentration

As described in section 3.2.3.3 of Chapter 3. Two polyclonal antibodies (T-CMO 1 and T- CMO 2) were purified by protein A-affinity chromatography and the resultant antibody (IgG) was tested for efficacy.

5.2.3.4 Determining optimum working concentration (Checkerboard Titration) As described in section 3.2.3.4 of Chapter 3.

5.2.3.5 Sensitivity Determination of T-ELISA

5.2.3.5.1 Preparation of standard solutions The standard solution of testosterone was prepared as described in section 3.2.3.5.1 of Chapter 3.

138

5.2.3.5.2 Preparation of enzyme conjugate The working strengths of the enzyme conjugates for the T-ELISA were prepared by diluting enzyme conjugates in 1% BSA/PBS as described in section 3.2.3.5.2 of Chapter 3.

5.2.3.5.3 Competitive Direct ELISA protocol The competitive direct ELISA protocol for T-ELISA was the same as described in section 3.2.3.5.3 of Chapter 3.

5.2.3.5.4 Determination of the standard curve parameters As described in section 3.2.3.5.4 of Chapter 3.

5.2.3.6 Determining specificity As described in section 3.2.3.7 of Chapter 3.

5.2.3.7 Matrix interferences study As described in section 3.2.3.8 of Chapter 3.

5.2.4 Spike and recovery study

The spike and recovery study was conducted to assess the matrix interference of real water samples for the testosterone-ELISA. The spiked samples were assayed in the T-ELISA, and the results were calculated by intrapolating the testosterone concentration from the percent inhibition calibration curve.

Table 5.1 The concentrations of EE2, E2 and Testosterone (T) used to prepare the spiked samples.

Spike No. 17α-Ethynylestradiol 17β-Estradiol Testosterone (μgL-1 ) (μgL-1 ) (μgL-1 ) 1 0.5 0.5 0.0 2 0.0 0.0 0.1 3 0.0 0.0 0.5 4 0.5 0.5 0.5 5 0.0 0.5 1.0

139

5.3 Results and Discussion

5.3.1 Hapten selection and conjugation to carrier proteins and an enzyme Testosterone−3-O-carboxymethyl-oxime (T-CMO) was chosen as the hapten to give rise to differentiation between testosterone from other androgenic and estrogenic steroids such as 17β-estradiol, estrone and progesterone. It can significantly reduce the cross−reactivity with these structurally similar steroid molecules.

5.3.1.1 Optimal concentration of enzyme conjugates

Purified antibody was titrated against two enzyme conjugates (T-CMO-HRP and P-CMO- HRP) using a direct immunoassay. Briefly, an optimum condition of each enzyme-conjugate was assessed by the checker board titration (CBT) against immobilised antibodies and selecting a concentration/dilution that yielded approximately 1–1.5 absorbance units.

As expectedly, Ab-T-CMO generated high−titres in the homologous system using T-CMO- HRP. The optimised dilution factor of homologous system was higher than the heterologous system by 80-fold (Figure 5.2). Good colour development does not always indicate high sensitivity, and assay sensitivity of each assay was evaluated using a competitive ELISA with an immobilized antibody format (Lee and Kennedy, 2007).

3.0

2.0

1.0 Absorbance (450nm) Absorbance

0.0 100 1000 10000 100000 1000000 10000000 10000000 Dilution of hapten-enzyme conjugates

Figure 5.2 Titration curves of Ab-T-CMO against two hapten-enzyme conjugates. The circles indicate T-CMO-HRP and the square indicate P-CMO-HRP.

140

5.3.2 Assay Sensitivity

Using optimised concentrations of enzyme conjugates, the sensitivity of each antibody /enzyme-conjugate combination was investigated using full standard curves of testosterone.

As summarised in Figure 5.3, the homologous hapten assays (i.e., Ab-T-CMO × T-CMO- HRP) exhibited ten−fold greater sensitivity than the heterologous hapten systems (i.e., Ab-T-

CMO × P-CMO-HRP). The IC50 values of the heterologous hapten assays, Ab-T-CMO × T- CMO-HRP and Ab-T-CMO × P-CMO-HRP, were 0.3 and 3.1 μgL-1 , respectively, although the LOD of both assays, which was determined by the concentration of testosterone causing -1 20% inhibition of colour development (IC20), was very similar (0.1 μgL ). The IC80 values, which determined the upper limit of the linear range, for Ab-T-CMO × T-CMO-HRP and Ab- T-CMO × P-CMO-HRP were 1.5 and >100 μgL-1, respectively. The homologous assay displayed a good slope or dynamic range of 45% over a ten−fold difference in concentration within the linear range. While the heterologous assay displayed a somewhat flatter slope of only 29%, and hence would lack the needed quantifiability (Lee and Kennedy, 2007). The results suggest that the homologous assay based on Ab-T-CMO × T-CMO-HRP exhibited sensitivity closer to the desirable ng L-1 level monitoring of androgenic testosterone in aquatic systems (Barel-Cohen et al., 2006, Kudak and Namienik, 2008), thus it was selected for further characterisation.

141

100

% Inhibition % 40

30 T-CMO-HRP 20 P-CMO-HRP 10

0 0.01 0.1 1 10

-1 Testosterone (μg L ) Figure 5.3 Testosterone standard curves against two enzyme conjugates, T-CMO-HRP and P-CMO-HRP.

5.3.2.1 Standard curve parameters and precision (IC80, IC50, IC20 and maximum absorbance)

Nine-point calibration curves for Ab-T-CMO averaged over five analyses conducted on different days are shown in Figure 5.4. The % CV of % inhibition decreased as testosterone concentration increased for the assay, as expected, while the % CV of absorbance remained steady. The precision as determined by % CV over the calibration range varied from 2.9 – -1 51.4%. The IC50 value was 0.46 + 0.15 μgL (% CV = 33%), and the IC20 value was 0.06 + -1 -1 0.01 μgL (% CV = 15%) (Table 5.2). The IC80 was 3.77 + 1.12 μgL (% CV = 30%) for Ab- T-CMO. The average maximum absorbance was on the low side by our standard (0.497 + 0.295 with % CV= 59%) (Figure 5.6), however, did not affect the assay performance as an analytical method. The % CV for the absorbance at nine testosterone concentrations were greater than acceptable, ranging between 44 to 75%, showing the assay was sensitive to environmental conditions such as temperature of reagents and laboratory(Figure 5.5).

142

Table 5.2 Standard curve parameter and precision for Ab-T-CMO.

Parameters Value S.D. a %CV b

IC 50 0.46 0.15 33

IC 20 0.06 0.01 15

IC 80 3.77 1.12 30 Maximun absorbance 0.497 0.295 59 a standard deviation b percent coefficient variation

100 % Innhibition 0.5 Absorbance

80 0.4

60 0.3

40 0.2 % Inhibition % Absorbance (450 nm) (450 Absorbance

20 0.1

0 0.0 0.01 0.1 1 10 Testosterone concentration (μg L-1)

Figure 5.4 The relationship of % inhibition and colour development of Ab-T-CMO (average of five analyses).

143

100

% Innhibition

80 Absorbance

60 % CV 40

0 0.01 0.1 1 10 Testosterone concentration (μg L-1)

Figure 5.5 The % CV for absorbance ( ) and % inhibition ( ) by Ab-T-CMO (average of five analyses).

10.00 ) -1 g L μ

1.00

0.10 Testosterone concentration (

0.01 0 1 2 3 4 5 Number of assay

Figure 5.6 Plot of IC80 ( ), IC50 ( ) and IC20 ( ) values for Ab-T-CMO (average of five analyses). The middle line indicates the average value. The dotted line shows the upper and lower limit (average T concentration ±standard deviation).

144

Table 5.3 The cross-reactivity of the testosterone ELISA.

Compounds Structure Ab-T-CMO 1

IC50 %CR IC50 %CR (μg L-1) (mol L-1) -9 CH OH Testosterone 3 0.4 100 1.4×10 100

CH 3

-7 O CH 3 >100 <0.01 3.2×10 < 0.4 Progesterone CH 3 CH 3

OH -7 CH 3 >100 <0.01 3.7×10 <0.4 Estradiol

O -7 Estrone CH3 >100 <0.01 3.7×10 <0.4

OH -7 CH >100 <0.01 3.5×10 <0.4 3 OH Estriol

OH -7 CH 3 >100 <0.01 3.4×10 <0.4 17α–Ethynylestradiol CH

-7 O >100 <0.01 2.9×10 <0.5 CH CH 3 3

CH Medroxyprogesterone 3

145

5.3.3 Assay Specificity

The specificity of the homologous ELISA using Ab-T-CMO was evaluated by conducting a crossreactivity study with structurally similar natural and synthetic compounds that may occur in STP effluent or in river water. Progesterone, estradiol, estrone, estriol, 17α-ethynylestradiol and medroxyprogesterone were used in this study. The results are shown in Table 5.3., The -1 IC50 value for each of the test compounds was above 100 μg L with % cross-reaction being <0.01%, suggesting this immunoassay exhibits excellent specificity for testosterone. The structural relationship between androgens (testosterone vs progesterone) and androgen vs estrogens do not affect the cross reactivity. Testosterone−3-O-carboxymethyl-oxime showed high specificity to testosterone, even though it uses the same backbone as progesterone. Similarly, testosterone−3-O-carboxymethyl-oxime exhibit excellent specificity to testosterone, which have nineteen carbons, and differ from estrogen, which have eighteen carbons.

5.3.4 Matrix Interferences

Matrix constituents can affect an immunoassay’s performance in three ways: 1) inhibiting the enzyme activity of an enzyme conjugate, 2) affecting antibody binding, and 3) both of the above (Lee & Kennedy, 2007). The matrix effects of real water samples on the testosterone ELISA were determined from both the colour development and assay sensitivity. This is because water matrices may affect the enzyme-conjugate leading to change in colour development. In terms of sensitivity, effect of matrix interference may affect antibody-antigen interactions, causing a change in sensitivity by shifting of standard curve to left or right.

Matrix effects were studied by using pH solutions ranging from pH 2 to 11, by inorganic and metal ions, as well as ionic strengths. Humic acid (HA), which may present in environmental samples, was also used as a potential interfering agent due to its capacity to act as an absorbent and ion exchange. HA at 0.01 mg L-1 to 10 mg L-1 were tested in this study.

146

5.3.4.1 Effect of pH

As previously discussed, pH can affect assay performance by inhibiting enzyme activity or affecting the structure and activity of the antibody. This study tested the pH values of 2, 3, 4, 5, 7, 9 and 11 because those pH values are obvious to affect either the enzyme or antibody- antigen binding.

-1 As illustrated in Figure 5.7, the control (pH 7.2) gave an IC50 value of 0.38 μg L with max absorbance at 0.676, while pH values of 2, 3, 4, 5 and 11 gave max absorbance below 0.480. This suggests that enzyme activity was indeed affected. The fluctuation of colour development (ranging from 0.096–0.676) and the IC50 value (ranging from 0.21–0.87) were noticeable when the pH was increased to greater than 7. At pH 9, colour development was unaffected, whereas the assay sensitivity was decreased by almost two−fold. This suggests that pH 9 affects the antibody-antigen interaction but not the enzyme activity (as enzyme conjugation contributes to colour development). The solvent at pH values of 3, 4, 5 and 9 reduced assay sensitivity, Contrarily, it was overestimated at pH 2 and 11 when compared to the sensitivity of the control (pH = 7.2). This suggests that this ELISA is particularly sensitive to pH and adjusting pH to 7–8 prior to analysis would be important for the validity of the results.

147

0.9 0.8

0.7

0.6

) 0.5 -1 g L μ 0.4

IC 50 ( 0.3 0.3 Absorbance ( 450 nm) 450 ( Absorbance 0.2 IC 50 value

Color Development 0.1

0 0 357911 pH Figure 5.7 Effect of pH on the T-ELISA standard curve. The square indicates absorbance against pH and the circle indicates IC50 value against pH.

5.3.4.2 Effect of water type and ionic strength

Real water matrices have great potential to influence assay sensitivity, and their effects were evaluated by comparing calibration curve constructed in real samples with that of purified water. Three water samples collected from the lagoon at Tahbilk winery wetlands (pH 6.4), McWilliam’s winery reservoir (pH 6.7) and sea water from Maroubra Beach (pH 8.2) were used. Since ion salts are one of the major components in the water matrix, T-ELISA was used to study the effect of ionic strength. The ion salts studied were (NH4)2SO4, NaCl, MgSO4,

CaCl2, MnSO4, KCl, NH4Cl, CuSO4 and Fe2(SO4)3 .

As illustrated in Figure 5.8, the enzyme activity was slightly influenced by real water matrices, contributing to a slight decrease in absorbance for the sample sourced from McWilliam’s winery reservoir when compared to purified water as a control. Additionally, real water matrices had an observable effect on the assay sensitivity for the sample sourced from the lagoon at Tahbilk winery wetlands. The IC50 values for McWilliam’s winery reservoir, the lagoon at Tahbilk winery wetlands, and Maroubra Beach sea water (Figure 5.9)

148 were 0.95, 1.73 and 1.17 μgL-1 , respectively, compared to the control (purified water) which -1 showed an IC50 value of 1.16 μgL . Interesting, the IC50 values of purified water and sea water were similar, indicating that sea water did not interfere with the ELISA, probably because little matrix intereference present in sea water. Only salt ion and ionic strength are the potential interference.

0.9

purified water

McWilliam’s winery reservoir

lagoon at Tahbilk winery wetlands 0.6 Maroubra Beach sea water

0.3 Absorbance ( 450 nm) 450 ( Absorbance

0.0 0.01 0.1 1 10 -1 Testosterone (μg L )

Figure 5.8 Matrix effects on absorbance for the T−ELISA with different types of water (purified water, McWilliam’s winery reservoir, lagoon at Tahbilk winery wetlands, and Maroubra Beach sea water).

149

100 purified water

90 McWilliam’s winery reservoir

80 lagoon at Tahbilk winery wetlands 70 Maroubra Beach sea water 60

% Inhibition % 40

0 0.01 0.1 1 10 Testosterone (μg L-1)

Figure 5.9 Standard curve of testosterone concentrations in different types of water (purified water, McWilliam’s winery reservoir, Lagoon of Tahbilk winery wetlands, and Maroubra Beach: sea water).

(NH4)2SO4 and NaCl at 1 M did not alter the assay sensitivity, while the other compounds led to a considerable underestimation of testosterone. Moreover, the colour development in T- ELISA was diminished by all the tested compounds, leading to the subsequent loss of assay sensitivity, (Table 5.4). Thus ionic strength of samples can affect the enzyme-conjugate activity. The ion salts and ionic strengths studied greatly exceed those naturally found in aquatic environment and such concentrations are rarely found in nature. However, this result supported the above matrix effect study using real water samples, and reinforces that sample preparation is important prior to assay.

150

Table 5.4 Effect of ionic strength on T-ELISA. Ab-T-CMO Compounds -1 Colour development (Abs) IC50 (μgL ) Control (10% EtOH) 0.557 0.28 1M NaCl 0.338 0.27

1M CaCl2 0.113 0.40 1M KCl 0.332 0.27

1M MgSO4- 0.148 0.32

1M (NH4)2SO4 0.252 0.26

0.001 CuSO4 0.213 1.35

0.001 Fe2(SO4)3 0.274 >100

0.01M Fe2(SO4)3 0.204 21.40

5.3.4.3 Effect of humic acid

Humic acid (HA) in the natural environment can act as a chelating agent by precipitating organic compounds in water (Boenigk et al., 2005). In Austria, the presence of humic acid in lakes and rivers ranged from 0.01 mg L-1 to several hundred g L-1 (Pfortner et al., 1998, Lai et al., 2000). Naturally occurring humic acid can potentially affect the determination of trace analytes if interacting with them. Hence, commercial humic acid was used to study its effects on the T-ELISA.

The IC50 values and colour development in the testosterone ELISA using humic acid as an interfering agent are shown in Figures 5.10 and 5.11. The higher concentration of humic acid shifted the curves to the right, reducing the assay sensitivity. Humic acid at 0.01 mg L-1, 0.1 -1 -1 -1 mg L and 1 mg L resulted in IC50 values of 0.48, 0.71 and 0.81 μg L and a maximum absorbance at 0.758, 0.552 and 0.587, respectively (Figure 5.10 and 5.11). Humic acid at 0.1 mg L-1 did not show effects on the ELISAs for ethynylestradiol and estradiol ELISAs; however, this is not the case for the T-ELISA. Humic acid even at 0.01 mg L-1 affected both assay sensitivity and colour development of T-ELISA.

These results are in good agreement with the above studies for matrix interference of various water types, and it is also apparent that the three different types of water samples contain a heterogenous mixture of organic materials at a level which affects either enzyme activity or

151 the activity of the antibody in the T-ELISA. Therefore, it is concluded that T-ELISA has appreciable matrix dependency, and sample preparation is desirable to remove any potential interference prior to assay.

0.8

Purified water Humic acid 1 mg L-1 0.6 Humic acid 0.1 mg L-1 Humic acid 0.01 mg L-1

0.4

Absorbance (540 nm) (540 Absorbance 0.2

0.0 0.01 0.1 1 10

Testosterone (μg L-1)

Figure 5.10 Effect of humic acid on absorbance of the T-ELISA.

152

100

% Inhibition 40 Purified water

30 Humic acid 0.1 mg L-1

20 Humic acid 1 mg L-1

10 Humic acid 0.01 mg L-1

0 0.01 0.1 1 10 Testosterone (μg L-1)

Figure 5.11 Effect of humic acid on the calibration curve of testosterone.

5.3.5 Recoveries of testosterone from spiked, purified and field water samples

The matrix effects of three water samples were investigated by the spike and recovery study. The water samples obtained from the lagoon at Tahbilk winery wetlands, McWilliam’s winery reservoir and Maroubra Beach were spiked with testosterone at 0.1, 0.5 and 1 μg L-1, in three replicates. The spiked samples were then analyzed by the T-ELISA.

As demonstrated in Figure 5.12, the recoveries of spiked purified water and the water samples from Tahbilk, McWilliams and Maroubra Beach were 89–100%, 61–100%, 52–100% and 92–130%, respectively. The recovery results showed underestimated testosterone concentrations for the environmental samples. The correction coefficient (R2) of purified water, and water samples from Tahbilk, McWilliam’s and Maroubra Beach were 0.999, 0.991, 0.999 and 0.988, respectively (Figure 5.13). It was shown that the matrix substances in raw water may interfered with the antibody-antigen binding, leading to lower sensitivity, and sample preparation for simultaneous cleanup and concentration prior to measurement is expected to improve assay performance.

153

93 sea water from Maroubra 61 Beach 1.0 65

) McWilliam’s winery reservoir

-1 90 g L

μ lagoon at Tahbilk winery 110 93 wetlands 0.5 61 purified water 93

130 52 0.1 78 98

100 Spiking testosterone level testosterone Spiking ( 100 0.0 100 100

0 20 40 60 80 100 120 140 160 % Recovery Figure 5.12 Average values (μg L-1) of spiking and recovery (%) from three water sources by the T-ELISA.

_____ y = 1.098x, R² = 0.991 1.0 (lagoon at Tahbilk winery wetlands)

- - - - y = 1.061x, R² = 0.999 0.8 (McWilliam’s winery reservoir ) ) -1 0.6 g L μ

0.4 SPIKE (

_ .. _ .. y = 1.103x , R² = 0.999 0.2 ( Purified water)

...... y = 1.030x. R² = 0.988 (sea water from Maroubra beach)

0.0 0.0 0.2 0.4 0.6 0.8 1.0 -1 Testosterone ELISA (μg L )

Figure 5.13 Comparison of the spiking of testosterone in four water sources.

154

5.4 Conclusion

Testosterone-ELISA was developed using polyclonal antibodies raised against a testosterone hapten linked to a carrier protein via a carboxymethyloxime linker. Unlike E2 and EE2 ELISAs, which the heterlogous system improved assay sensitivity, the homologous system using a testosterone-HRP conjugate showed better sensitivity than the heterologous system using a progesterone hapten–HRP conjugate. This assay demonstrated an IC50 value of 0.46 μg L-1 and a LOD of 0.06 μg L-1.

The assay developed from the testosterone antibody proved to be highly−specific for testosterone. The assay was susceptible to many components of sample matrices. The pH 7 was determined to be the optimum for the T-ELISA. T-ELISA was also significantly influenced by ionic strengths and matrix components such as inorganic and metal ions. The colour development was particularly affected by the presence of heterogeneous substances (e.g. organic compounds and salt ions) in the water samples, and generally leading to loss of sensitivity. Hence, enhanced sample preparation was deemed necessary to improve assay performance.

155

CHAPTER 6 VALIDATION OF YEAST SCREENING ASSAY FOR ENDOCRINE DISRUPTING HORMONES

6.1 Introduction

As discussed in the general introduction, both natural and synthetic hormones are suspected to induce intersexuality in fish. For example, a decrease of spermatozoa fertilizing ability in adult rainbow trout (Salmo gairdneri) (Billard et al., 1981), a concomitant inhibition of testicular growth of male rainbow trout (Oncorhynchus mykiss) (Jobling et al., 1996c) and the delay in progression of spermatogenesis of exposed male wild roach (Rutilus rutilus; a cyprinid fish) living in rivers in the United Kingdom (Jobling et al., 2002a). Brain and co- workers (2007) stated that fish populations, however, are concurrently exposed to several EDCs, which are expected to contribute to the total estrogenic potency and ecological health. Their research indicated that mixtures of estrogenic chemicals can pose a significant threat to egg production of fathead minnow (Pimephales Promelas), even when the components are present at low and individually ineffective concentrations. Effluent from wastewater treatment plants are common sources of estrogenic chemicals coming into water bodies (Sumpter and Purdom, 1994, Desbrow et al., 1998, Jobling et al., 2002a). Therefore, toxin identification and toxicity evaluation for the nature of the mixture of chemicals are necessary to identify cause-effect relationships.

The previous chapters described the development of main endocrine disrupting steroids: E2, EE2 and testosterone. Total estrogenic and androgenic activities should also be screened for proper risk assessment. Among the screening assays developed, such as human breast cancer cell line (MCF-7), the estrogen receptor (ER)-mediated chemical activated luciferase gene expression (ER-CALUX) assay, vitellogenin gene expression in hepatocyte cultures, the recombinant yeast screen is considered to be the simplest, in term of easy−of−handling, high− sensitivity and cost−effectiveness (Gaido et al., 1997a, Aneck-Hahn et al., 2005)

The recombinant yeast screen reported in Gaido and co-workers (1997) relies on the expression of the enzyme, β-galactosidase, upon a response to a hormone stimulus. The yeast, Saccharomyces cerevisae, was used as the recipient organism because of its rapid growth rate, sexual cycle and commercial availability of a range of plasmids and promoters. Furthermore, the yeast recombinants contained the human estrogen, androgen and progesterone used to

156 assess chemicals that can interact directly with steroid hormone receptors, and have the potential for detection at extremely low concentrations. The estrogen receptor assay can respond to a concentration range from 10-12 to 10-8 M of the steroids: 17β-estradiol, diethylstilbesterol, estriol and coumestrol. The androgen receptor assay is sensitive to a concentration range from 10-10 to 10-8 M of testosterone and dihydrotestosterone (Gaido et al., 1997a, Schultis et al., 2002, Urbatzka et al., 2007) .

In this study, the yeast screening assays for the determination of estrogen and androgen activities have been established and validated using water samples collected from creeks around NSW.

6.2 Materials and Methods

6.2.1 Material and Instrument

6.2.1.1 Materials The standards estrone (E1), 17β-estradiol (E2), and 17α-ethinylestradiol (EE2) (>98% purity) and analytical grade general chemicals purchased from Sigma-Aldrich (St Louis, MO) were: potassium phosphate monobasic anhydrous, ammonium sulphate, potassium hydroxide pellets, anhydrous magnesium sulfate, Iron (III) sulfate pentahydrate (Fe2(SO4)3), L-leucine, free base L-histidine, free base adenine, hydrochloride L-argenine, L-methionine, L-tyrosine free base, L-isoleucine monohydrochloride, L-lysine, L-phenylalanine, inositol, L-glutamic acid free acid, L-valine, L-serine hydrochloride, thiamine hydrochloride, pyridoxine, D- pantothenic acid hemicalcium salt, d-biotin, anhydrous D-(+)-glucose anhydrous; mixed anomer, anhydrous copper (II) sulfate (CuSO4), L-aspartic acid free acid, L-threonine, glycerol. Chlorophenolred-β-D galactopyranoside (CPRG) was supplied by Boehirnger Mannheim (Lewes, East Sussex, UK).

For the preparation of the medium component, 0.2 μm pore size filters were obtained from Whatman (Maidstone, Kent, UK)

For the assay procedure, 96-well flat bottom microplates were obtained from Nunc (Demark).

6.2.1.2 Instrument The Avanti JA 20 centrifuge (Beckman Coulter) was used for preparing the stock culture and AD 340 microplate reader (Beckman Coulter) was used for measurements of absorbance. 157

6.2.1.3 The recombinant yeast expressing the human estrogen receptor (hER)

The DNA sequence of the human estrogen receptor was integrated into the yeast (Saccharomyces cerevisiae) genome (Kevin et al., 1997b). This yeast also contained expression plasmids carrying estrogen-responsive elements regulating the expression of the reporter gene lacZ (encoding the enzyme, β-galactoctosidase). Thus, when an active ligand (in this case, estradiol or an estrogen-mimicking substance) was successfully bound to the receptor, β-galactosidase was synthesized and secreted into the medium, and as a result the chromogenic substrate chlorophenol red β-D-galactopyranoside (CPRG) changes colour from yellow to red (Figure 6.1). The product was measured at 570 nm, which changed from an absorbance unit of about one (yellow indicating no activity) to about two (red indicating maximum activity).

- SO3 chlorophenol red β-D-galactopyranoside (CPRG) Na+ (yellow colour) O

CH2OH

OH O O Cl

H H β-D-galactosidase

H OH

- SO3 Na+

O CH2OH

Galactose OH O OH Chlorophenol red (red colour) OH Cl

H H -O Cl H OH

Figure 6.1 Chemical reaction of CPRG

6.2.2 Preparation of the medium components

All components except for CPRG (Boehringer Mannheim, Lewes, East Sussex, UK) were research grade biochemical suitable for cell culture, purchased from Sigma-Aldrich (St. Louis, MO).

158

Minimal medium (pH 7.1) was prepared by adding 13.61 g KH2PO4, 1.98 g (NH4)2SO4, 4.2 g KOH, 0.2g MgSO4, 1 mL Fe2(SO4)3 solution (40 mg/ 50 mL H2O; purchased from Aldrich, 50 mg L-leucine, 50 mg L-histidine, 50 mg adenine, 20 mg L-argenine-HCl, 20 mg L- methionine, 30 mg L-tyrosine, 30 mg L-isoleucine, 30 mg Lysine-HCl, 25 mg L- phenylalanine, 100 mg L-glutamic acid, 150 mg L-valine, and 375 mg L-serine to 1 L of MilliQ purified water. The mixture was dissolved on a heated stirrer and 45 mL aliquots of the dissolved solution were dispensed into glass bottles. The minimal medium was sterilized at 121 oC for 10 min, and stored at room temperature.

Vitamin solution was prepared by adding 8 mg thiamine, 8 mg pyridoxine, 8 mg pantothenic acid, 40 mg inositol and 20 mL biotin solution (2 mg/ 100 mL H2O) to 180 mL double- distilled water. The solution was filtered using 0.2 μm pore size Whatman PURADISC filters, and 10 mL aliquots were stored in sterilized glass bottles at 4 oC. A 20% (w/v) solution of D- (+) glucose was sterilized in 20 mL aliquots at 121 oC for 10 min and stored at room temperature. Stock solutions of L-aspartic acid (4 mg mL-1) and L-threonine (24 mg mL-1) were sterilized in 20 and 5 mL aliquots, respectively, at 121 oC for 10 min. L-Aspartic acid stock solutions were stored at room temperature and L-threonine stock solutions were stored o at 4 C. A 20 mM CuSO4 solution was prepared and filtered through 0.2 μm pore size disposable filters. The solution was stored in 5 mL aliquots at room temperature in sterilized glass bottles. A 10 mg mL-1 stock solution of CPRG was prepared in sterile distilled water and stored at 4 oC in sterilized glass bottles.

Growth medium was prepared by adding 5 mL glucose solution, 1.25 mL L-aspartic acid solution, 0.5 mL vitamin solution, 0.4 mL L-threonine solution, and 125 μL CuSO4 solutions to 45 mL of single-strength minimal medium in a sterile conical flask. The growth medium was then inoculated with 125 μL 10x yeast stock and incubated at 28 oC for approximately 24 h on an orbital shaker (250 rpm with a 50 mm throw) until the absorbance at 640 nm of 1 was reached.

The assay medium was prepared by adding 0.5 mL of the CPRG solution to 50 mL fresh growth medium. The medium was seeded with 2 mL yeast from a 24 h yeast culture with an absorbance at 640 nm of 1.

159

6.2.3 Preparation of stock culture

The preparations of yeast stock were carried out in a type II laminar flow cabinet. The yeast inoculum in the growth medium was incubated at 28 oC for approximately 24 h on an orbital shaker (250 rpm with a 50 mm throw) until the absorbance at 640 nm of 1 was reached. The 24 h culture was transferred to sterile 50 mL centrifuge tubes and centrifuged at 4 oC for 10 minutes at 2,000g. The supernatant was decanted and the pellet was resuspended in 5 mL minimal medium containing 15% glycerol (i.e., 8 mL sterile glycerol added to 45 mL minimal medium). Aliquots of 10x stock culture (0.5 mL) was transferred to 1.2 mL sterile cryovials and stored at -20 oC for a maximum of 4 months.

6.2.4 Preparation of standard solutions

Working standard solutions for each studied compound (100 μg L-1) was prepared weekly for generating standard curves and for sample spiking. These standard solutions were prepared from each individual concentrated stock standard solution.

6.2.4.1 Stock standard solution

Stock standards of 1000 mg L-1 was prepared by dissolving10.0 mg (+ 0.0001 g) for each of the studied compounds, E2 and testosterone, in 10 mL (volumetric flask) absolute methanol.

6.2.4.2 Intermediat standard solution

The intermediate stock standard solutions of 100 mg L-1 were prepared by diluting 1.0 mL of 1000 mg L-1 solution to a total volume of 10.0 mL absolute methanol using a volumetric flask. The solution was transferred to brown bottles to minimize light degradation, and stored at 4 oC.

6.2.5 Assay Procedure

The yeast assays were carried out in a type II laminar flow cabinet to minimize aerosol formation. Each assay plate contained at least triplicate sample wells, at least one row of blanks (200 μL assay medium only), as well as a calibration curve of 17β-estradiol (0.02μgL-1 - 100 μgL-1). Stock solution of each test chemical was serially diluted in ethanol, in a 96-well optically flat-bottomed microtitre plate. Aliquots of 10 μL were then transferred to identical 160 sterile plates and the ethanol was allowed to evaporate in room temperature. Aliquots of 200 μL per well of the seeded assay medium containing CPRG and yeast were then added to testing wells using a multichannel pipette. Calibration curves of 17β-estradiol as a positive control was included every time the yeast screen assay was carried out. The plates were sealed with autoclave tape and shaken thoroughly for 2 min on a microtitre plate shaker prior to incubation at 32 oC in a naturally ventilated incubator. The plates were shaken vigorously on the plate shaker for 2 min after 24 and 48 h, to disperse the growing cells. On the fourth day, after incubating for 72 h (3 days), the plate were shaken for 2 min, left for 1 h to allow the yeast to settle and then read at 570 nm (for absorbance) and 620 nm (for turbidity).

6.2.6 Response of the yeast screen bioassay

The ability and the degree of the testing compound to potentially interact with an estrogen receptor in yeast estrogen screen assay (YES assay) and interact with an androgen receptor in yeast androgen screen assay (YAS assay) were assessed during the candidate’s major project in 2006 (as part of Masters by coursework program), by incubating the test compound in the growth medium at the determined optimal concentration. The ability of each chemical was initially assessed and the concentration range producing a full dose response curve was then generated to assess estrogenic activity equivalent to 17β-estardiol or androgenic activity equivalent to testosterone

6.2.7 Spike and recovery study

The spike and recovery experiment was conducted to assess the quantitative potential of the yeast screening assays. The spiked samples were assayed in the yeast screening bioassays to quantify the 17β-estradiol-equivalent activity (EEQs). The EEQs were calculated by extrapolating the activity from the absorbance of the dose-response curve. Testosterone- equivalent activity (TEQs) was calculated by extrapolating the activity from the absorbance of the dose-response curve to quantify the testosterone-equivalent activity (Table 6.1).

161

Table 6.1 The ratio of 17α-ethynylestradiol (EE2), 17β-estradiol (E2), estrone(E1), estriol (E3) and testosterone (T) used to make the spike samples. Spike no. EE2 spiked E2 spiked E1 spiked E3 spiked T spiked ( μgL1) (μgL-1) (μgL-1) (μgL-1) (μgL-1)

1 0.5 0.1 0.0 0.0 0.1 2 0.1 0.5 0.1 0.0 0.0 3 0.3 0.0 0.3 0.0 0.0 4 0.0 0.5 0.0 0.0 0.0 5 0.1 0.1 0.3 0.0 0.0 6 0.0 0.0 1.0 0.0 0.0 7 2.0 0.0 0.0 0.0 0.5 8 1.0 0.5 0.5 0.0 0.5 9 0.0 5.0 5.0 0.0 10.0 10 0.5 0.5 1.0 10.0 0.0 11 0.0 0.0 0.0 10.0 0.0 12 0.0 0.5 1.0 10.0 0.0

6.2.8 Water and water samples

The real-world water samples studied were conducted in collaboration with Ms Chun Hua Ly (a PhD candidate) of University of Sydney, NSW (Australia). All water samples were collected by this candidate.

162

Field water samples from New South Wales (NSW) and Victoria (VIC), Australia The location details for collecting field water samples are shown in Table 6.2

Table 6.2 Sampling locations in South Creek in Sydney

GPS position Description NS04 200m to the confluence point of South Creek with Hawkesbury- Nepean NS081 Eastern Creek after River Stone STP, before the confluence point with South Creek NS083 Eastern Creek before River stone STP and after Quakers Hill WTP NS014 South Creek at Richmond Road Bridge, before the confluent joint with Eastern Creek NS023 South Creek at Eighth Av Bridge after St Mary’s STP NS026 South Creek at Great Western HWY Bridge, before St Mary’s STP NS037 South Creek at Luddenham Bridge. EC01 Emigrant Creek, close to the dam EC03 Emigrant Creek, EC09 Emigrant Creek, EC11 Emigrant Creek, beginning of the creek MW02 Wodonga Creek, discharge point of the Wodonga WTP MW01 Wodonga Creek, 250m up the discharge point MW03 Wodonga Creek, 500m down the discharge point MW 04 Effluent from the Wodonga WTP MW05 Warragatta Creek, 800m up stream MW06 200m down stream of discharge point of the effluent from Wongaratta WTP WR01 Wilson River at Lismore Park, before the Lismore WTP WR02 Wilson River after the Lismore WTP

163

Figure 6.2 Map of sampling sites, South Creek, Sydney Basin, NSW (Li et al., 2007)

6.2.9 Analysis of estrogenic and androgenic hormones in water samples by the YES and YAS assays

As part of the collaborative monitoring study between the University of NSW and the University of Sydney, water samples were collected in three different regions of New South Wales, Australia (by Dr. Chun hua Li of the University of Sydney). The three sites of study were South Creek in Sydney Basin, Emigrant Creek at Ballina and Wilsons River at Lismore, the Northern River area, NSW, and in the Murray River at Albury-Wodonga at the boundary of NSW with Victoria (Figure 6.2). The Murray River was confluent by two creeks, the Wodonga Creek and Wangaratta Creek. The samples were concentrated and re-constituted in either dichloromethane or methanol for analysis by GC/MS (by C.H. Li of the USyd), LC/MS/MS (by C.H. Li of the USyd), ELISA (by the author, UNSW), CALUX (by Miss 164

Jenny Keep of the Department of Agriculture, Lismore) and yeast screening assays. Briefly, 8 L of water sample was extracted using SM2-Biobeads solid phase extraction method to a final volume of 10 mL (the sample preparation was carried out by C.H. Li of the USyd).

The YES and YAS bioassays were also conducted as described in section 2.2.5 with a 17β- estradiol or testosterone calibration curve. Each sample was assayed in at least four replicate wells.

6.3 Results and Discussion

6.3.1 The performance of Yeast Assays

The sensitivity of YES and YAS bioassays was compared by measuring the response to 17β- estradiol and testosterone, respectively. The test compounds interacting with a hormone receptor was assessed at a concentration that may produce a maximal response. The ability of each chemical was initially assessed and the concentration range producing a full dose response curve was then generated to assess estrogenic activity equivalent to17β-estradiol and androgenic activity equivalent to testosterone. The compounds were serially diluted in three- fold steps from 100 μg L-1. Figure 6.3 shows the standard curve of 17β-estradiol, which gives the limit of detection at 0.33 μg L-1, which is equivalent to 0.003 ng L-1 17-estradiol. The lowest amount of testosterone that responded in the YAS assay was 3 μg L-1, which is equivalent to 0.03 ng testosterone (Figure 6.4).

6.3.2 Assay Specificity

The specificity of the YES and YAS assays were periodically re-assessed by comparing the response of testing compounds with 17β-estardiol as a standard. Figure 6.3 demonstrates the specificity of the YES screen. Natural hormones, 17β-estradiol, estrone and estriol, and synthetic hormone, 17α-ethynylestradiol showed potent estrogenic activity on the human estrogen receptor (hER) assay. These results are in agreement with numerous reports which have shown that these hormones can compete with estradiol for binding to the estrogen receptor (Céspedes et al., 2004, Viganò et al., 2008). The highest estrogenic activity was observed with 17α-ethinylestradiol and the lowest was with estriol. 17α-Ethynylestradiol, 17β-estradiol, and estrone exhibited lowest activity at 0.33 μg L-1, 0.41 μg L-1 and 1.23 μg L-1

165

, respectively, and estriol at 11.1 μg L-1. Moreover, this assay did not respond to testosterone even at 100 μg L-1, which is structurally similar to estradiol. This result is not in agreement with some previous reports which showed the ability of androgen such as testosterone to bind to and activate the estrogen receptor at high−concentrations (Zava and Mcguire, 1978, Pelissero et al., 1993). It can be therefore concluded that 17β-estradiol, 17α-ethynylestradiol, estrone and estriol shows estrogenic activity while testosterone does not show estrogenic activity at 100 μg L-1.

Furthermore, the specificity of the human androgen receptor (hAR) assay was assessed by comparing the response of the testing compounds to bind to the androgenic receptor with that of testosterone. The test compounds interacting with an androgen receptor was assessed at a concentration that may produce a maximal response. Figure 6.4 illustrates the specificity of the YAS assay with the testing compounds. Testosterone showed the highest activity at concentration >3.7 μg L-1, followed by progesterone showing approximately 11.1 μg L-1. Estrone, estriol and 17β-estradiol also unexpectedly showed some minimal activity, but were orders of magnitude less than that of testosterone. As reported by Sohoni and Sumpter (1998), testosterone had the strongest response activity in the YAS assay, and both progesterone and estrone showed moderate activity. In addition, 17α-ethynylestradiol did not show any response even at 100 μg L-1 in the YAS screen.

1.6

EE2

1.2 E2

T 0.8

0.4 Absorbance ( 570 nm) 570 ( Absorbance

0 0.01 0.1 1 10 Concentration ( μg L-1) Figure 6.3 The specificity of hER yeast screen; Plot of ( ) E1, ( ) EE2, ( ) E2, ( -- --) Estriol and (--- --) Testosterone 166

1.6

E1 EE2 1.2 E2 E3 T Prog 0.8

Absorbance ( 570 nm) 570 ( Absorbance 0.4

0 0.01 0.1 1 10 Concentration ( μg L-1)

Figure 6.4 The specificity of hAR yeast screen;; Plot of ( ) Progesterone, (-- --) Testosterone, ( ) EE2, ( ) E2, ( -- --) E3 and (--- --) E1

6.3.3 EEQs and TEQs of spiked purified water

The recovery of this method was determined by spiking a mixed standard solution in ethanol (refer to Table 6.3). The spiked samples were assayed in the yeast screen bioassays to quantify the 17β-estradiol-equivalent activity (EEQs) and testosterone-equivalent activity (TEQs). Figures 6.5 and 6.7 demonstrate the equivalent trend. The following equations were applied to estimate EEQ values from spiking levels: x Estimated EEQ value = (concentration of E2 × Δ1) + (concentration of EE2 × Δ 2.6) + (concentration of E1 × Δ 0.26) + (concentration of E3 × Δ 0.026) + (concentration of T × Δ 0) x Estimated TEQ value = (concentration of E2 × Δ 0.15) + (concentration of EE2 × Δ 0) + (concentration of E1 × Δ 0.23) + (concentration of E3 × Δ 0.14) + (concentration of T × Δ 1)

The spiked sample containing 0.5 μg L-1 of each of 17α-ethynylestradiol and 17β-estradiol shows an estrogenic activity equivalent to > 5 μg L-1 17β-estradiol. A sample spiked with 17α-ethynylestradiol at 0.1 μg L-1 and estriol at 10 μgL-1 showed an estrogenic activity 167 equivalent to 0.4 μg L-1 17β-estradiol. The sample spiked EE2 at 0.5 μg L-1, E2 at 0.1 μg L-1 and T at 0.1 μg L-1 showed EEQ of 0.77 μg L-1 17β-estradiol and TEQ to 0.16 μg L-1. The water spike containing 0.5 μg L-1of each of E1 ,E2, and T, and 1 μgL-1 of EE2 also showed an estrogenic activity equivalent to 2.9 μg L-1 17β-estradiol and 0.85 μg L-1 testosterone. E1 spiked at 1 μg L-1 was equal to the 17β-estradiol activity at 0.85 μg L-1, as well as E3 spiked at 10 μg L-1 showed an estrogenic activity equivalent to 0.4 μg L-1 17β-estradiol. It can be noticed that EE2 was most estrogenic and responded to the highest activity in the YES bioassay. Any samples containing this compound greater than 0.1 μg L-1 will respond with high−estrogenic activity. Estriol showed the lowest estrogenic activity among the tested estrogens in the YES bioassay, showing a 38− fold lower activity compared to that of 17β- estradiol. Samples containing only 17β-estradiol and estrone can generally be more or less quantifiable.

The correlation between the estimated EEQs and TEQs of spiked samples and the measured EEQs/TEQs by yeast assays are shown in Figure 6.5 and 6.7, respectively. The estimated EEQ of the spiked samples with the measured EEQ using the hER assay showed R2 = 0.687, n = 12, while the predicted TEQ of the spiked samples with the measured TEQ using hAR assay showed better correlation with a R2 value of 0.931 (n = 12). Moreover, the correlation between the measured TEQs by the yeast assay and the testosterone spiked level was less than 1.25 μg L-1 with a R2 value of 0.838 (n = 11) (Figure 6.6). From these results, it is clear that these bioassays are only suitable for screening purposes and further developments are needed to improve the quantification capability.

168

Table 6.3 The EEQs and TEQ of the spike samples

Spike EE2 E2 E1 E3 T Total Total Estimate Equivalent Estimate Equivalent no. spiked spiked spiked spiked spiked estrogen Androgen Equivalent estradiol Equivalent testosterone ( μg L-1) ( μg L-1) ( μg L-1) ( μg L-1) ( μg L-1) ( μg L-1) ( μg L-1) estradiol activity Testosterone activity activity Activity ( μg L-1) ( μg L-1) ( μg L-1) ( μg L-1) 1 0.5 0.1 0.0 0.0 0.1 0.6 0.1 1.4 0.8 0.1 0.2 2 0.1 0.5 0.1 0.0 0.0 0.7 0.0 0.8 0.8 0.1 > 0.003 3 0.3 0.0 0.3 0.0 0.0 0.6 0.0 0.9 0.6 0.1 > 0.003 4 0.0 0.5 0.0 0.0 0.0 0.5 0.0 0.5 0.5 0.1 0.0 5 0.1 0.1 0.3 0.0 0.0 0.5 0.0 0.4 0.7 0.1 > 0.003 6 0.0 0.0 1.0 0.0 1.0 1.0 1.0 0.3 0.8 1.2 1.1 7 2.0 0.0 0.0 0.0 0.5 2.0 0.5 5.2 2.8 0.5 > 0.003 8 1.0 0.5 0.5 0.0 0.5 2.0 0.5 3.2 2.9 0.7 0.8 9 0.0 5.0 5.0 0.0 10.0 10.0 10.0 6.3 8.0 11.9 10.1 10 0.5 0.5 1.0 10.0 0.0 12.0 0.0 2.3 5.6 1.7 > 0.003 11 0.0 0.0 0.0 10.0 0.0 10.0 0.0 0.3 0.4 1.4 > 0.003 12 0.0 0.5 1.0 10.0 0.0 11.5 0.0 1.0 1.6 1.7 > 0.003

y = 0.982x + 0.279 R² = 0.687 6 ) -1 g L μ

2 EEQ by EEQ by hER assay (

0 02468 Estimated EEQ value ( μg L-1) Figure 6.5 YES assay validations of spiked samples

169

1.2

1 ) -1

g L 0.8 μ

0.6

y = 0.894x - 0.012 R² = 0.838 0.4

0.2 TEQ by TEQ by hAR assay (

0 0 0.25 0.5 0.75 1 1.25 Spiking levels ( μg L-1)

Figure 6.6 The correlation between YAS assay and spiking levels.

12 ) -1 9 y = 0.934x + 0.020 g L R² = 0.931 μ

3 TEQ by TEQ by hAR assay (

0 036912 Estimated TEQ ( μg L-1)

Figure 6.7 YAS assay validations of spiked samples.

6.3.4 Estrogenic and androgenic potency in Creek samples

6.3.4.1 Analysis of estrogenic hormones in water samples by YES assay

As previously described in section 3.2.5, 8 L water samples collected and extracted with Bio- Beads were assayed using the YES and YAS assays. The data were divided into 3 groups

170 according to the levels of EEQs, < 0.5-−5, 5.0 – 10.0 and > 10.0 ng L-1. The percentile of distribution for water samples with EEQ values in each group are illustrated in Figure 6.8. The measured EEQs with the YES assay were less than 0.5 ng L-1 – 5 ng L-1 in 52 samples, and 5 −10 ng L-1 in 5 samples, and only 3 samples were detected EEQs value higher than 10 ng L-1. It can be seen that most of the creek samples in the assay (95%) showed estrogenic activities of less than 10 ng L-1, which are in good agreement with previous national studies. For example, Mispagel and co-worker had, using the YES assay, determined estrogenic activity in WWTP effluents in South Australia showing 7.9 ng L-1 EEQ and in Victoria, Australia showing 8 − 16 ng L-1 EEQ (Mispagel et al., 2005, Mispagel et al., 2009), and STPs in south Queensland, Australia (<1−4.2 ng L-1 EEQ) were determined, using bioassay- derived estrogenic activity (Leusch et al., 2006). Meanwhile, the results were generally near the lower end of the range observed overseas, as YES assays were applied to determine estrogenicity of effluents in Japan (5 −15 ng L-1 EEQ), the USA (44−151 ng L-1EEQ), Sweden (< 0.1− 15 ng L-1 EEQ) and Switzerland (0.1−90 ng L-1 EEQ) (Matsui et al., 2000, Tilton et al., 2002, Svenson et al., 2003, Rutishauser et al., 2004).

> 10 5 -1 gL n 5.0 - 10 8

< 0.5 - 5.0 87

0 20 40 60 80 100 EEQ by EEQ by hER assay, The percentile of sample distribution

Figure 6.8 Levels and distribution of estrogenic activity (n = 60)

The total estrogenic potencies of the water samples collected from different locations are listed in Table 6.4. Water samples were collected from the Wilson River, Lismore City in September 2007. The EEQ at the discharge point of the effluent from Lismore Wastewater Treatment Plant was 0.42 ng L-1, while the EEQ of Wilson River at Lismore Park was 0.33 ng L-1. In Wangaratta Creek, the EEQ in the effluent stored in a sediment pool inside the Wangaratta STP was 1.09 ng L-1 and in the creek itself, it was 0.25 ng L-1. In addition, the total EEQ levels for the water samples in Wodonga Creek ranged between 0.20 to 6.28 ng L-1. The water quality of this creek is likely to be impacted both by runoff and by effluents from the nearby waste water treatment plant. The EEQ of the direct effluents from the Wodonga 171

STP (site MW04) was 6.28 ng L-1; at the discharge point into the creek (site MW01) the EEQ levels decreased to 4.56 ng L-1, and at a site 250 m downstream of the discharge point, the EEQ levels decreased further to 0.29 ng L-1. At 500 m upstream from the discharge point the EEQ was only 0.20 ng L-1. This can be explained by the dilution effect of effluents were influencing the levels of EEQs at each site.

Moreover, the total EEQ levels for the South Creek water samples ranged between 0.57 −1.84 ng L-1, indicating low estrogenic potency. The water in the upstream of South Creek mainly comes from the runoff into the catchment and the EEQ levels for these sites were 1.84 ng L- 1and 1.11 ng L-1 (site NS37 and NS26). At downstream discharge point of St. Mary STP (site NS23) and further down from STP (site NS14), the EEQ levels were measured at 1.28 and 0.68 ngL-1, respectively. The EEQ level of the immediate downstream of Riverstone STP (site NS081) was 0.58 ng L-1, whereas the EEQ level of site NS083, which is located upstream of Riverstone STP, but downstream of Quakers Hill STP, was 1.32 ng L-1. The EEQ measured in these four sites (NS23, NS14, NS081 and NS083) were also impacted by both effluents and runoff. Furthermore, the EEQ downstream in South Creek (site NS04) represented the total EDC levels brought into the Hawkesbury River. The EEQ level was measured at 0.57 ng L-1, also showing a low level of estrogenic potency.

6.3.4.2 Analysis of androgenic hormones in water samples by the YAS assay

The androgenic activity in 50 samples was investigated by the YAS assay. The results showed the percentile of distribution for androgen in levels of TEQs of 4 groups as demonstrated in Figure 6.9. The measured TEQ with the YAS assay were < 0.01-−0.5 ng L-1 in 43 samples, 0.5 – 5 ng L-1 in 6 samples, and no sample with detected TEQ in the 5 −10 ng L-1 range. Only 1 sample was detected with TEQ > 10 ng L-1. It can be seen that most of the South Creek samples in the assay (98%) showed androgenic activity of less than 0.5 ng L-1, which was the lower range observed by Leusch and co-worker (2006) (< 6.5 – 736 ng L-1 TEQ).

172

Table 6.4 Total estrogenic potencies (by YES assay) of water samples collected from creeks, expressed as the equivalent estradiol

Sampling date Location Site of sampling and EEQ levels (E2 ng L-1) 29/09/2007 Wilson River WR01 WR02 0.33 0.42 5/10/2006 Wangaratta Creek MW05 MW06 0.25 1.09 5/10/2006 Wodonga Creek MW01 MW02 MW03 MW04 4.56 0.29 0.20 6.28 21/12/2005 South Creek NS04 NS081 NS083 NS14 0.57 0.58 1.32 0.68 NS23 NS26 NS37 1.28 1.11 1.84

>10 2 -1 gL

n 5.0-10.0 0

0.5 - 5 12

< 0.01 - 0.5 86 EEQ by hAR EEQ by hAR assay, 020406080100

The percentile of sample distribution

Figure 6.9 Levels and distribution of androgenic activity (n = 50)

The total TEQ levels for the South Creek water samples ranged between 0.00 ngL-1 to 0.89 ng L-1 (Table 6.5), indicating a very low androgenic potency. In >85% of the samples, the TEQs were lower than 0.20 ng L-1, showing consistently low levels of androgenic potency. NS04 was the most downstream site in South Creek; represented only 0.05 ng L-1 of TEQ levels brought into the Hawkesbury River. Sampling sites NS37 and NS26 are on the upstream of South Creek; the waters mainly came from the runoff of catchment (Figure 6.2), presented TEQ levels at 0.15 and 0.89 ng L-1 respectively. The TEQ level of 0.06 ng L-1 was monitored at site NS23 (downstream of discharge point of St. Mary STP), and further down from STP 173

(site NS14), TEQ levels measured at 0.17 ng L-1. The TEQ level of the immediate downstream from Riverstone STP (site NS081) did not detect any androgenicity, whereas the TEQ level of site NS083, which is located upstream of Riverstone STP, but downstream of Quakers Hill STP, was 0.07 ng L-1. The TEQ values among the sites did not show an obvious relationship to the distance to STPs, although the TEQ level at site NS26 appeared slightly high.

Like South Creek, the water quality in Wodonga Creek and Wangaratta Creek is likely to be impacted by runoff and effluents from wastewater treatment plants. The TEQ level of 0.27 ng L-1 was monitored at Wangaratta STP, while in the creek it was not detected. The TEQ of the direct effluents from the Wodonga STP (site MW04) was 0.03 ng L-1; at the discharge point into the creek (site MW01) the TEQ levels decreased to 0.01 ng L-1, and at a site 250 m downstream of the discharge point, the TEQ levels remained at 0.01 ng L-1. At 800 m upstream from the discharge point the EEQ was 0.64 ng L-1.

Table 6.5 Total androgenic potencies (by YAS assay) of water samples collected from creeks, expressed as the testosterone (ng T L-1) equivalent.

Sampling date Location Site of sampling and TEQ levels (ng L-1) 29/09/2007 Wilson River WR01 WR02 0.03 0.02 5/10/2006 Wangaratta Creek MW05 MW06 0.27 0.00 5/10/2006 Wodonga Creek MW01 MW02 MW03 MW04 0.01 0.01 0.64 0.03 21/12/2005 South Creek NS04 NS081 NS083 NS14 0.05 0.00 0.07 0.17 NS23 NS26 NS37 0.06 0.89 0.15

6.3.5 Determiination of estrogenic and androgenic compounds

6.3.5.1 Estrogenic compounds determined by ELISA, CALUX and yeast assay

The ELISA measurements were validated using the yeast screen assay, and comparisons of the yeast screen assay and the chemically-activated luciferase gene expression (CALUX) 174 were investigated in order to confirm the accuracy of both assays. The significance of this study is to screen water levels in NSW and identify ‘hot spots’ for potential hormone contamination.

The estradiol equivalency values (EEQs) were predicted by summation of measured individual E1, E2 and EE2 levels determined with the developed ELISA, as well as the EEQ levels of environmental water samples measured by the YES assay. On the same samples, CALUX bioassay was also performed by Ms Jenny Keep from DPI in Wollongbar. Hence, it is interesting to compare the performance of these two bioassays with those of CALUX for the same samples.

The correlation between the predicted EEQs by the ELISA and the measured EEQs by the yeast assay are shown in Figure 6.10. The predicted EEQ from the ELISA with the measured EEQ using the YES assay showed relatively good agreement, with a R2 = 0.775 (n = 58). It can be seen that the hER assay presented higher EEQs than the predicted EEQs from the ELISA. This may be the result because the EEQ measured by the bioassays may be result from a combination of micropollutants, other than estrogens, present in the sample extracts that exhibit similar estrogenic activities. Some if these may include coumestrol, 4- nonylphenol, hydroxylated PCB, zearalenone, bisphenol-A, tributyltin. These compounds have been reported as estrogenic compounds, which have been regularly found in surface water (Tylor et al., 1998, Roepkea et al., 2005, Kudak and Namienik, 2008).

Moreover, the comparison of the yeast screen assay and CALUX presented a slightly low correlation of R2 = 0.749 (n = 58), as shown in Figure 6.11. The discrepancy between the results could be from the time delay in analysis by the two bioassays. The analyses by ELISA and hER assay of the same sample extracts were performed within 4 months, whereas those by CALUX assay were collected at least 12 months ahead. Additionally, comparison of both screening assays was limited to samples with EEQ values less than 2.5 ng L-1. The results were moderately scattered and result in a poor correlation with R2= 0.304 (n = 49) (Figure 6.12). Results from this study highlights the significant different between EEQ values of hER assay and CALUX assay of EEQ with values of less than 2.5 ng L-1. The asymmetric error such as the measurement errors, effect of humidity and temperature to solution or sample, was also presented in this experiment.

175

10 y = 0.660x + 0.391, R² = 0.775 ) -1

gL 8 n

EEQ by EEQ by YES assay ( 2

0 024681012 Estimate EEQ by ELISA (ng L-1) Figure 6.10 Correlation of EEQs, n = 58. The EEQ level predicted with the summation of detected individual estrogen levels was correlated with the measured EEQs in water samples by the yeast assay.

10 )

-1 8

y = 0.667x + 0.346 6 R² = 0.749

4 EEQ by CALUX by EEQ CALUX (ngL 2

0 0246810

EEQ by YES assay (ng L-1)

Figure 6.11 Correlation between EEQ by the yeast screen assay (YES) and EEQ by CALUX

176

2.0 )

-1 1.5 gL n

1.0

EEQ by CALUX EEQ by CALUX ( 0.5 y = 0.514x + 0.608, R² = 0.304

0.0 0.0 0.5 1.0 1.5 2.0

EEQ by YES assay (ng L-1)

Figure 6.12 Correlation between EEQ by YES assay and EEQ by CALUX. The dispersion at 0 to 2 ng L-1

6.3.5.2 Androgenic compounds determined by ELISA and yeast assay

The TEQ level of environmental water samples measured with the YAS assay were compared with the TEQ predicted by the determined T level using the ELISA. The correlation between the predicted TEQs by the testosterone ELISA and the measured TEQs by the yeast assay are shown in Figure 6.13. The predicted TEQ from the ELISA with the measured TEQ using the YAS assay showed relatively good agreement, with a R2 value of 0.931 (n = 49). When the comparison of both screening assays were restricted to samples with TEQ values of less than 1.5 ng L-1, resulting in low scattering, g a correlation of R2 = 0.702 (n= 32) was obtained (Figure 6.14). The result suggests that testosterone is probably the predominant micropollutant showing androgenic activity.

177

10 y = 0.934x + 0.020

) R² = 0.931 -1

gL 8 n

TEQ by YAS assay ( assay YAS by TEQ 2

0 0246810 -1 Estimated TEQ (ng L )

Figure 6.13 Correlation of TEQs, n = 49. The ELISA estimated TEQ level correlated with the YAS measured TEQs in water samples.

1.5 ) -1 y = 0.897x - 0.044 ngL R² = 0.702 1

0.5 TEQ by YAS assay ( assay YAS by TEQ

0 0 0.5 1 1.5 Estimated TEQ ( ng L-1)

Figure 6.14 Correlation between the TEQ (0−1.5 ng L-1 ) obtained by the YAS assay and estimated TEQ by ELISA.

178

6.4 Conclusion

The recombinant yeast screen assays, YES and YAS assays, have been considered to be sensitive and reliable methods for screening for estrogenic and androgenic potencies (Gaido et al., 1997a, Céspedes et al., 2004, Schultis and Metzger, 2004, Urbatzka et al., 2007). The estradiol and testosterone equivalence in river water were analyzed from the creek samples. These samples were equivalent to < 0.5 – 25 ng L-1 17β-estradiol and <0.01 – 12 ng L-1 testosterone. The estrogenic activities of the creek water in rural and urban areas of NSW were similar to those previously determined in a national study, and were lower than those reported by international studies. Of the two types of activity tested, the androgenic activity of the samples was lower than other studies.

Good agreement between the measured EEQs/TEQs and the predicted EEQs/TEQs (Figures 6.10 and 6.12) demonstrated the reliability of the ELISA assay in determining the estrogen and androgen levels. The discrepancy between the predicted and the measured EEQs was probably due to the overestimated E2 and EE2, which was most likely due to the presence of other compounds that have an estrogenic effect. Moreover, the total EEQ levels measured by YES assay in South Creek ranged between 0.57 − 1.84 ng L-1, showing the consistently low and stable level of estrogenic potencies since these differed little upstream compared to downstream. The consistency and low contribution of effluents to the estrogenic activities was also observed in the monitoring of Wilson River, Lismore City, Wangaratta Creek and Wodonga Creek. Although the low and rather consistent EEQ levels in creek samples were shown, they were probably contributed by both agriculture and human activities. Occasional high EEQs observed in these waters were most likely from agricultural practices.

Finally, the continuing increase in population in Sydney and increasing agricultural practices closer to urban areas to reduce the cost of transportation, and to maintain the freshness of produce, results in more usage of domestic and industrial water. Given the serious health problems associated with these EDCs, the potential impact of estrogens from the effluents should be monitored regularly, and agricultural practices including live stock activity should be controlled as a measure to reduce estrogenic potency in recycled drinking water.

179

CHAPTER 7 CONCLUSION

For monitoring trace levels of estrogens, sensitive and reliable analytical methods are needed to detect and quantify very low parts per trillion levels which might interfere with the endocrine function of aquatic and terrestrial organisms (Agasan et al., 1994, Fox, 2001, Coille, 2002, Farré et al., 2007). Instrumental methods such as GC-MS and LC-MS/MS have been widely used for the detection of trace levels of estrogens because of their high− precision and accuracy (Ternes et al., 1999a, Céspedes et al., 2004, Heisterkamp et al., 2004, Yang et al., 2006, Zhang et al., 2006). However, these methods suffer from drawbacks such as expensive instrumentation (high initial capital costs), high running costs as well as the requirement for highly trained staff to operate them properly. In addition, the required extensive sample preparation and derivatisation is time consuming, thus substantially decreasing the throughput of instrumental analysis (Huang and Sedlak, 2001). An immunochemical assay is an alternative method which can be used as a water quality monitoring tool because of their high−throughput capacity at a lower cost, and the greatly reduced reliance for highly trained staff. Moreover, it’s convenient and rapid performance is ideal for screening purposes.

In the present work, the facile synthesis of hapten molecules for four potent estrogens and androgens that are potential endocrine disruptors and the development of a series of ELISA tests for detecting parts per trillion concentrations with reduced matrix interference and low cross−reactivity are described. The performance and efficacy of the ELISA tests developed in this study were validated by using them to monitor target EDCs in water samples from selected urban and rural areas of New South Wales, Australia. Concluding remarks for this thesis are highlighted in the following:

1. Synthetic estrogen haptens were designed and synthesised based on 17α-Ethynylestradiol haptens with the attachment of linkers with varying lengths at the C3 position. The specific polyclonal antibodies were produced against conjugates of a haptenic molecule (17α- ethynylestradiol-acetate and 17α-ethynylestradiol-butyrate) and a carrier protein, KLH. The antibodies showed specific recognition of 17α-ethynylestradiol and mestranol without significant cross−reactivity to six other major estrogenic compounds (estriol, estrone, estradiol dipropionate, progesterone, 17α-estradiol and medroxyprogesterone) that are commonly found in treated and untreated water. The ELISAs had a limit of detection (LOD) 180

-1 -1 of 0.07 μg L for 17α-ethynylestradiol and an IC50 value of 0.20 μg L . The LOD of the EE2 ELISA covering the preconcentration step was 0.13 ng L-1in water samples. These ELISAs showed good sensitivity compared to previously developed EE2 ELISAs, which showed an -1 IC50 of 0.14 ng mL (Schneider et al., 2004). The developed 17α-ethynylestradiol ELISA (EE2 ELISA) was used to monitor of rural and urban water resources and nearby water treatment plants in NSW, Australia. The ELISA results correlated well with those obtained by an independent analyses using GC-MS (R2 = 0.934) for spiked samples (Li et al., 2007).

Similarly, specific polyclonal antibodies were raised against KLH conjugates of 17β- estradiol-acetate hapten (E2-ACT-KLH) and 17β-estradiol-butyrate hapten (E2-BUT-KLH). The developed E2 ELISA was highly specific to 17β-estradiol with a LOD of 0.05 μg L-1 in water (without preconcentration). The LOD of the E2 ELISA covering the preconcentration step was 0.03 ng L-1in water samples. These E2 ELISAs exhibited excellent specificity, showing very little cross−reactivity against structurally similar steroids, while the previously developed E2 ELISA presented cross−reactivity of 25% to other cross reactants such as estrone (Hintemann et al., 2006a, Farr´E et al., 2007, Caron et al., 2010). Furthermore, the E2 ELISA exhibited high matrix tolerance. E2 values measured in spiked samples were in good agreement with the results obtained independently by GC-MS (R2 = 0.936).

Additionally, in a study involving the production of a specific polyclonal antibody against testosterone, a hapten was raised against KLH conjugates of testosterone-3-O-carboxymethyl- oxime (T-CMO). This ELISA was developed for environmental analysis with respect to the -1 desired sensitivity and specificity. The LOD and IC50 values for this assay were 0.07 μg L and 0.4 ng mL-1, respectively. A cross reactivity study revealed an excellent specificity. Unlike the other assays developed in this project, this assay was more matrixes sensitive. Increasing concentrations of organic substances caused a decrease in assay sensitivity, and the presence of humic substances at 0.01 mg L-1 led to an overestimation of testosterone. It was determined that a sample preparation step prior to assay would be crucial for the success of this system.

2. The validation of ELISAs suggested an acceptable correlation with the data derived from the GC-MS method and a reasonable correlation with the biological assay (i.e., YES assay), though a slightly higher estimation of E2 and EE2 by the biological assays was observed, most likely a result of other contaminating estrogenic substances. It can be concluded that the estrogenic activities predicted by ELISA for the selected estrogens cannot be used for the 181 estimation of the total estrogenic potencies in the surface water matrix. This suggests that other compounds possessing estrogenic potency are likely to be present in water, which should be address in future research. However, satisfactory agreement in results for testosterone was obtained from a comparison of T ELISA and YAS assays. Good agreement was obtained between the predicted androgenic activities and the measured estrogenic activities using the YAS assay, giving a correlation coefficient (R2) of 0.931. Testosterone is probably the predominant micropollutant responsible for the observed androgenic activity in the rural and urban water bodies in NSW. Contrary to the conventional method (GC-MS), the developed ELISAs for specific quantification of steroidal hormones, such as E2, EE2 and testosterone, may have the drawback of slight overestimating levels because of cross−reactivity. ELISA, however, offers considerable advantages over the conventional analytical procedure in terms of low detection limits, simple technique, rapid measurement and high sample turnover due to less sample volume, and acceptable operating costs. As a screening tool for water quality management, these attractive features would allow ELISAs to be used for routine monitoring of endocrine disrupting compounds in water.

As shown in this thesis, the main factors effecting the development of an optimized ELISA test are the hapten design and antibody affinity and specificity. The reduction of experimental error and elimination of potential sources of imprecision are relevant to the optimization process. The present work has presented the facile synthesis of specific haptens and developed an optimized ELISA for the selected endocrine disrupting steroids, namely E2, EE2, mestranol and testosterone. As proved in this thesis, the study was able to establish a good detection limit in this manually performed ELISA. Further advances are expected from implementing these assays on automated systems such as immunosensors for environmental samples. Furthermore, sensitivity of the current system can easily be enhanced by using other detection systems such as fluorescent or chemiluminescent probes – which calls for further assay development using the existing antibodies.

3. These developed ELISAs were used to access the quality of influents and effluents of various Sewage Treatment Plants (STPs) and nearby water resources in New South Wales. E2 and EE2 were detected in the South Creek water samples, and very low testosterone levels were detected. The consistency of estrogen/androgen levels upstream compared to those downstream from the discharge point indicated that the discharge from Wastewater Treatment Plants (WTPs) did not appreciably increase the estrogenic potencies. Moreover, this consistent occurrence of estrogen/androgen in South Creek was similar to the results obtained 182 in rural area of northern NSW (Emigrant Creek), suggesting that even though the current wastewater treatment by these WTPs is reasonably efficient, it does not necessarily completely remove EDCs.

Inevitably, regular monitoring of EDCs is highly recommended to ensure the levels and the contribution of estrogenic potency is maintained below the ecotoxicologically significant level of 10 ng L-1 (Fenske et al., 2005, Huschek and Hansen, 2005), and to continually maintain a regime of water quality improvement and to help limit peak values of these endocrine disrupting steroids. In addition, agricultural practices, including livestock husbandry, should be controlled so that these activities occur a certain distance away from water sources as a precautionary measure to reduce estrogenic potency in our water resources.

183

REFERENCES

Abe, M., Hoshi, T. & Tajima, A. 1987. Characteristics of transmural potential changes associated with the proton-peptide co-transport in toad small intestine J. Physiol, 394, 481−499. Adams, N. R. 1998. Clover phyto-oestrogen in sheep in Western Australia. Pure Appl. Chem., 70, 1855-1862. Agasan, A. L., Stewart, B. J. & Watson, T. G. 1994. Development of a radioimmunoassay method for ethynylestradiol in plasma using a monoclonal antibody. J. of Immuno. Met., 177, 251-260. Al-Dujaili, E. A. S. 2005. Deveopment and validation of a simple and direct ELISA method for the determination of conjugate (glucoronide) and non-conjuagted testosterone excretion in urine. Clinica Chimica Acta, 364, 172-179. Allinson, M., Shiraishi, F., Salzman, S. A. & Allinson, G. 2010. In Vitro and Immunological Assessment of the Estrogenic Activity and Concentration of 17b-Estradiol, Estrone, and Ethynylestradiol in Treated Effluent from 45 Wastewater Treatment Plants in Victoria, Australia. Arch Environ Contam Toxicol, 58, 576-586. Andrew, M. N., Dunsta, R. H., O’connor, W. A., Zwieten, L. V., Nixon, B. & Macfarlane, G. R. 2008. Effects of 4-nonylphenol and 17α-ethynylestradiol exposure in the Sydney rock oyster, Saccostrea glomerata: Vitellogenin induction and gonadal development. Aquat. Toxicol. , 88, 39–47. Aneck-Hahn, N. H., de Jager, C., Bornman, M. S. & du Toit, X. 2005. Oestrogenic acticity using a recombinant yeats screen assay (RCBA) in South African laboratory water sources. Water SA, 31, 253-256. Arcand-Hoy, L. D., Nimrod, A. C. & Benson, W. H. 1998. Endocrine modulating substances in the environment: Estrogenic effects of pharmaceutical products. Int J Toxicol 17, 139–158. Barel-Cohen, K., Shore, L. S., Shemesh, M., Wenzel, A., Mueller, J. & Kronfeld-Schor, N. 2006. Monitoring of natural and synthetic hormones in a polluted river. J Environ Management, 78, 16–23. Baronti, C., Curini, R., D'Ascenzo, G., Di corcia, A., Gentili, A. & Samperi, R. 2000. Monitoring Natural and Synthetic Estrogens at Activated Sludge Sewage Treatment Plants and in a Receiving River Water. Env. Sci. Tech. , 34, 5059-5066. Barrett, J. 1996. Phytoestrogen: friends or foes? . Environ Health Perspect, 104, 478-482. Barton, H. A. & Andersen, M. E. 1998. Endocrine Active Compounds: From Biology to Dose Response Assessment. Cri. Rev. Toxicol., 28, 363-423. Belfroid, A. C., Van Der Horst, A., Vethaak, A. D., Schafer, A. J., Rijs, G. B. J., Wegener, J. & Cofino, W. P. 1999. Analysis and Occurance of estrogenic hormones and their glucuronides in surface water and wate water in the Netherlands. The Sci. Total Env., 225, 101-108. Bergman, A. & Olsson, M. 1985. Pathology of Baltic gray seal and ringed seal females with special reference to adrenocortical hyperplasia: Is environmental pollution the cause of a widely distributed disease syndrome? Finn. Game Res, 44, 47–62. Bergmeyer, H. U. 1974. Methods of Enzymatic Analysis, New York, Academic Press. Berson, S. A. & Yalow, R. S. 1996. Quantitative aspects of the reaction between insulin and insulin-binding antibody. J. Clin. Invest, 38. Bignert, A., Litzen, K., Odsjo, T., Olsson, M., Persson, W. & Reutergardh, L. 1995. Time- related factor influence the concentration of DDT, PCBs, and shell parameter in eggs of Bactic guillemot (Uria aalge). Environ Poll, 89, 27-36.

184

Billard, R., Breton, B. & Richard, M. 1981. On the inhibitory effect of some steroids on spermatogenesis in adult rainbow trout (Salmo gairdneri). Can. J. Zoo, 59, 1479– 1487. Boenigk, J., Wiedlroither, A. & Pfandl, K. 2005. Heavy metal toxicity and bioavailability of dissolved nutrients to bacterivorous flagellate are linked to suspended particles physical properties. Aquat. Toxicol, 71, 249-259 Brady, J. F., Lemasters, G. S., Williams, R. K., Pittman, J. H., Daubert, J. P., Cheung, M. W., Skinner, D. H., Turner, J., Rowland, M. A., Lange, J. & Sobek, S. M. 1995. Immunoassay Analysis and Gas Chromatography Confirmation of Atrazine Residues in Water Samples from a Field Study Conducted in the State of Wisconsin. J. Agric. Food Chem, 43, 268-274. Brain, J. V., Harris, C. A., Scholze, M., Kortenkamp, A., Lamoree, M., Pojana, G., Jonkers, N., Marcomeni, A. & Sumpter, J. P. 2007. Evidence of estrogenic mixture effects on the reproductive performance of fish Environ Sci Technol, 41, 337- 344. Brostoff, J. & Gamlin, L. 1989. The Complete Guide to Food Allergy and Intolerance, Bloomsbury, London Brotons, J. A., Olea-Serrano, M. F., Villalobos, M., Pedraza, V. & Olea, N. 1995. Xenoestrogens released from lacquer coatings in food cans. Environ Health Perspect, 103, 608-612. Brouwer, A., Reijinder, P. J. H. & Koeman, J. H. 1989. Polychlorinated biphenyl (PCB)- contaminated fish induces vitamin A and thyroid hormone deficiency in the common seal (Phoca vitulina). Aquat Toxicol, 15, 99-105. Cargouet, M., Perdiz, D., Mouatassim-Souali, A., Tamisier-Karolak, S. L. & Levi, Y. 2004. Assessment of river contamination by estrogenic compounds in Paris area (France). Sci Total Environ, 324, 55-66. Caron, E., Sheedy, C. & Farenhorst, A. 2010. Development of competitive ELISAs for 17β- estradiol and 17β-estradiol +estrone+estriol using rabbit polyclonal antibodies. J. Env. Sci. Heal. B 45, 145–151. Céspedes, R., Petrovic, M., Raldúa, D., Saura, Ú., Piña, B., Lacorte, S., Viana, P. & Barceló, D. 2004. Integrated procedure for determination of endocrine-disrupting activity in surface waters and sediments by use of the biological technique recombinant yeast assay and chemical analysis by LC–ESI-MS. Anal. Bioanal. Chem., 378, 697-–708. Chaloupka, K., Krishnan, V. & Safe, S. 1992. Polynuclear aromatic hydrocarbon carcinogens as antiestrogens in MCF-7 human breast cancer cells: role of the Ah receptor. Carcinogenesis, 13, 2233-2239. Chang, H., Wu, S., Hu, J., Asami, M. & Kunikane, S. 2008. Trace analysis of androgens and progestogens in environmental waters by ultra-performance liquid chromatography– electrospray tandem mass spectrometry. J. Chromatogr. A, 1195 44–51. Chang, S. & Jang, N. 2005. Fate and transport of EDCs ( estrone and 17β-estradiol) membrane bioreactor used for water reuse. Further Urban Wastewater sys.- decentralisation and reuse, 461− 468. Chimchirian, R. F., Suri, R. P. S. & Fu, H. 2007. Free synthetic and natural estrogen hormones in influent and effluent of three municipal wastewater treatment plants. Water Env. Res., 79, 969 − 974. Coille, I., Reder, S., Bucher, S. & Gauglitz, G. 2002. Comparison of two fluorescence immunoassay methods for the detection of endocrine disrupting chemicals in water. Biomol Engineer, 18, 273–280. Coille, K. 2002. Comparison of two immunoassay methods for the detection of endocrine disrupting chemicals in water. Biomolecular Engineering 18, 273-280. Colborn, T. & Clement, C. (eds.) 1992. An " in culture" bioassay to assess the estrogenicity of xenobiotics (e-screen). Chemically-induced alterations in sexual and fuctional developement: The wildlife/human connection, Princeton, N.J: Princeton Scientific. 185

Colborn, T., vom Saal, F. S. & Soto, A. M. 1993. Developmental effects of endocrine- disrupting chemicals in wildlife and humans. Environ. Health Perspect. , 101, 378 − 384. Crisp, T. M., Clegg, E. D., Cooper, R. L., Wood, W. P., Anderson, D. G., Baetcke, K. P., Hoffmann, J. L., Morrow, M. S., Rodier, D. J., Schaeffer, J. E., Touart, L. W., Zeeman, M. G. & Patel, Y. M. 1998. Environmental endocrine disruption: An effects assessment and analysis. Env. Health Perspec., 106, 11-56. Crowther, J. R. 2000. The ELISA Guidebook, New Jersey, Humana press. Damstra, T., Barlow, S., Bergman, A., Kavlock, R. & Van Der Kraak, G. 2002. Global Assessment of the State-of-the-Science of Endocrine Disruptors. Available: http:// www.who.int/pcs [Accessed 2009, January 19]. Danzo, B. J. 1997. Environmental xenobiotics may disrupt normal endocrine function by interfering with the binding of physiological ligands to steroid receptors and binding protein. Env. Health Perspec., 105, 372-377. Dasmahapatra, A. K. & Medda, A. K. 1982. Effect of Estradiol Dipropionate and Testosterone Propionate on the Glycogen, Lipid, and Water Contents of Liver, Muscle, and Gonad of Male and Female (Vitellogenic and Nonvitellogenic) Singi Fish (Heteropneustes fossilis Bloch). General Comp. Endocri., 48, 476-484. Deng, A., Himmelsbach, M., Zhu, Q. Z., Frey, S., Sengl, M. & Buchberger, W. 2003. Residue analysis of the pharmaceuticals diclofenac in different water types using ELISA and GC-MS. Env. Sci. Tech., 37, 3422-3429. Desbrow, C., Routledge, E., Brighty, G. C., Sumpter, J. P., Waldock, M. & 1998. Identification of Estrogenic Chemicals in STW Effluent. 1. Chemical Fractionation and in Vitro Biological Screening. Environ. Sci. Technol., 32, 1549 −1558. Dorabawila, N. & Gupta, G. 2005. Endocrine disrupter-estradiol in Chesapeake Bay tributaries. J. Har. Material, 120, 67-71. Doyle, C. J. & Lim, R. P. 2002. The effect of 17β-oestradiol on the gonopodial development and sexual activity of Gambusia holbrooki. EnvironToxicol Chem, 21, 2719-2724. Exley, D. 1972. Specificities of Antibodies to Oestrogens. J. Steroid Biochem. Mol. Biol., 3, 497−501. Exley, D. & Abuknesha, R. 1977. The preparation and purification of a beta-D-galactosidase- oestradiol-17beta conjugate for enzyme immunoassay. FEBS letters, 79, 301−304. Exley, D. & Abuknesha, R. 1978. A highly sensitive and specific enzyme-immunoassay method for oestradiol-17β. FEBS letters, 9, 162−165. Exley, D. & Choo, Q. L. 1974. Specificity of antisera to oestrone determined by different radioimmunoassay methods. J. Steroid Biochem, 5, 497−500. Fan, Z., Casey, F. X. M., Hakk, H. & Larseng.L 2007. Persistence and fate of 17β-estradiol and testosterone in agricultural soils. Chemos, 67, 886-895. FarrÉ, M., Kuster, M., Brix, R., Rubio, F., LÓpez De Alda, M. J. & BarcelÓ, D. 2007. Comparative study of an estradiol enzyme-linked immunosorbent assay kit, liquid chromatography–tandem mass spectrometry, and ultra performance liquid chromatography–quadrupole time of flight mass spectrometry for part-per-trillion analysis of estrogens in water samples. J. Chromatogr. A, 1160, 166-175. Fawell, J. K., Sheahan, D., James, H. A., Hurst, M. & Scott, S. 2001. Oestrogen and oestrogenic activity in raw and treated water in seven trent water. Wat. Res., 35, 1240- 1244. Fenske, M., Maack, G., Schafers, C. & Segner, H. 2005. An Environmentally Relevant Concentration of Estrogen Induces Arret of Male Gonad Development in Zebrafish, Danio Rerio. Env. Toxicol. Chem., 24, 1088–1098. Feyk, L. A. & Giesy, J. P. 1998. Xenobiotic modulation of endocrine function in birds’, In Priciples and Processes Evaluating Endocrine Disruption in Wildlife, Pensacola, Florida, USA, SETAC Press. 186

Fox, G. A. 2001. Effect of endocrine disrupting chemicals on wildlife in Canada: past, present and future. Wat. Quality Res. J. Canada, 36, 233-251. Fraser, I. S. (ed.) 1998. Estrogens and Progestrogens in Clinical Practice, London: Churchill Livingstone. Furuichi, T., Kannan, K., Zuzuki, K., Tanaka, S., Giesy, J. P. & Masunaga, S. 2006. Occurence of estrogenic compounds in and removal by a swine farm waste treatment plant. Environ Sci Tech, 40, 7896-7902. Gabet, V., Miège, C., Bados, P. & Coquery, M. 2007. Analysis of estrogens in environmental matrices. Trends Analyt. Chem., 26, 1113-1131. Gadd, J., Stewart, C. & Sikes, E. 2005. Estrogenic activity and known environmental estrogen in sewage effluent, Hamilton, New Zealand. Aus J Ecototoxicol, 11, 149-154. Gaido, K. W., Leonard, L. S., Lovell, S., Gould, J. C., Babai, D., Potier, C. J. & Mcdonnell, D. P. 1997. Evaluation of Chemicals with Endocrine Modulating activity in a Yeast- Based Steroid Hormone Receptor Gene Transcription Assay. Toxicol. App. Pharma., 143, 205-212. Garza, G. A. & Rao, P. N. 1983. Chromic Anhydride-3,5-Dimethylpyrazole Complex: An efficient reagent for oxidation of steriodal estrogens to 6-Oxo-Derivatives. Steroids, 42, 469−474. Gibbs, P. E. & Bryan, G. W. (eds.) 1994. Biomonitoring of Coastal Waters and Estuaries, Boca Raton, FL: CRC Press. Gibson, R., Smth, M. D., Spary, C. J., Tyler, G. R. & Hill, E. M. 2005. Mixtures of estrogenic contaminants in bile of fish exposed to wastewater treatment works effluents. Env. Sci. Tech., 39, 2461-2471.. Goda, Y., Hirobe, M., Kobayashi, A., Fujimoto, S., Lke, M. & Fujita, M. 2005. Production of monoclonal antibody and development of enzyme-linked immunosorbent assay for alkyl ethoxylates. Analytica Chemica Acta., 528, 47-54. Goda, Y., Kobayashi, A., Fukuda, K., Fujimoto, S., Ike, M. & Fujita, M. 2000. Development of the ELISAs for detection of hormone-disrupting chemicals. Water Sci. Tech., 42, 81-88. Gomes, R. L., Scrimshaw, M. D. & Lester, J. N. 2003. Determination of endocrine disrupters in sewage treatment and receiving waters. Trends Analyt. Chem., 22, 697-707. Gooding, M. P., Wilson, V. S., Folmar, L. C., Marcovich, D. T. & Lebalnc, G. C. 2003. The biocide tribytyltin reduces the accumulation of testosterone as fatty acid esters in the Mud snail (Ilyanassa obsoleta). Env. Health Perspec., 111, 426-430. Goodman, L. S. & Gilman, A. (eds.) 1996. The Pharmacological Bases of Therapeutic, USA: Macmillan. Goodrow, M. H. & Hammock, B. D. 1998. Hapten design for compound-selective antibodies: ELISAS for environmentally deleterious small molecules. Analyti. Chem. Act., 376. Gower, D. B. (ed.) 1975. Biochemistry of steroids hormones., UK: Blackwell, Oxford. Gray, M. A. & Metcalfe, C. D. 1997. Induction of testis-ova in Japanese medaka (Oryzias latipes) exposed to p-nonylphenol. Environ Toxicol Chem, 16, 1082-1086. Greene, R. R. 1941. A comparison of the clinical effectiveness of estradiol diporpionate and estradiol benzoate. The J.Clini.Endocri., 1, 559-561. Gronen, S., Denslow, N., Manning, S., Barnes, S., Barnes, D. & Brouwer, M. 1999. Serum vitellogenin levels and reproductive impairment of male Japanese medaka (Oryzias latipes) exposed to 4-tert-octylphenol. Environ Health Perspect, 107, 385-390. Guillette, L. J. 1995. Endocrine disrupting environmental contaminants and developmental abnormalities in embryos. Human Eco. Risk Ass., 1, 25-36. Gutendorf, B. & Westendorf, J. 2001. Comparison of an array of in vitro assays for the assessment of the estrogenic potential of natural and synthetic estrogens. Toxico., 166, 79–89.

187

Hanselman, T. A., Graetz, D. A. & Wilkie, A. C. 2004. Comparison of Three Enzyme Immunoassay for Measuring 17β-Estradiol in Flushed Dairy Manure Wastewater. J. Environ. Qual., 33, 1919−1923. Hardell, L., Van Bavel, B., Lindstrom, G., Carlberg, M., Dreifaldt, A. C., Wijkstrom, H., Starkhammer, H., Eriksson, M., Hallquist, A. & Kolmert, T. 2003. Increased concentrations of polychlorinated biphenyls, hexachlorobenzene and chlordenes in mothers to men with testicular cancer. Env. Health Perspec., 111, 930-934. Hardly, M. E. 1996. Endocrinology., Upper Saddle River (NJ):Prentice-Hall. Hayes, T. B., Haston, K., Tsui, M., Hoang, A., Haeffele, C. & Vonk, A. 2002. Feminization of male frogs in the wild. Nature, 419, 895-896. Hegari, A. H. & Andersson, J. T. 2007. Limitation to GC-MS Determination of Sulfur- COntaining Polycyclic Aromatic Compounds in Geochemical, Petroleum, and Environmental Invetigations. E. Fuels, 21, 3375-3384. Heisterkamp, I., Gandrass, J. & Ruck, W. 2004. Bioassay-directed chemical analysis utilizing LC–MS: a tool for identifying estrogenic compounds in water samples? Anal. Bioanal. Chem., 378, 709--715. Hemmer, M. J., Hemmer, B. L., Bowman, C. J., Kroll, K. J., Folmar, L. C., Marcovich, D., Hoglund, M. & Denslow, N. D. 2001. Effects of p-nonylohenol, methoxychlor and endosulfan on vitellogenin induction and expression in sheephead minnow (Cyprinodon variegates). Environ Toxicol Chem, 20, 336-343. Henniona, M. C. & Barcelo, D. 1998. Strengths and limitations of immunoassays for effective and efficient use for pesticide analysis in water samples: A review. Analyti. Chem. Act., 362, 3-34. Hintemann, T., Schneider, C., Schöler, H. F. & Schneider, R. J. 2006. Field study using two immunoassays for the determination of estradiol and ethinylestradiol in the aquatic environment. Wat Res 40, 2287-2294. Hirobe, M., Goda, Y., Okayasu, Y., Tomita, J., Takigami, H., Ike, M. & Tanaka, H. 2006. The use of enzyme-linked immunosorbent assays (ELISA) for the determination of pollutants in environmental and industrial wastes. Water Sci. Tech., 54, 1-9. Hood, E. 2005. Are EDCs Blurring Issues of Gender? Environ Health Perspect., 113, 670- 677. Huang, C. H. & Sedlak, D. L. 2001. Analysis of Estrogenice Hormoness in Municipal Wastewater Effluent and Surface Water Using Enzyme-Linked Immunosorbent Assay and Gas Chromatography/Tandam Mass Spectrometry. Env. Toxicol. Chem., 20, 133- 139. Huschek, G. & Hansen, P. D. 2005. Ecotoxicological Classification of The Berlin River System Using Bioassays in Respect to The European Water Framework Directive. Env. Moni. Ass., 121, 15–31. Ingerslev, F. & Sørensen, B. H. 2003. Evaluation of Analytical Chemical Methods for Detection of Estrogen in the Environment. Danish Environmental Protective Agency. Ingrand, V., Herry, G., Beausse, J. & Renée De Roubin, M. 2003. Analysis of steroid hormones in effluents of wastewater treatment plants by liquid chromatoraphy–tandem mass spectrometry. J. Chromatogr. A, 1020, 99–104. Itoh, Y., Taknashi, K., Itoh, S. & Yishizawa, I. 2001. High-performance liquid chromatographic separation of potential hydroxylated metabolites of estradiol 17- sulfate by female rat liver microsomes. Ana. Sci., 17, 659-661. Jacobson, J. L., Jacobson, S. W. & Humphrey, H. E. B. 1990. Effects of exposure to PCBs and related compounds on growth and activity in children. Neurotox. Terato., 12, 319- 326. James, W. H. 1997. Reproductive effects of male dioxin exposure. Environ Health Perspect, 105, 162-163.

188

Janeway, C., Travers, P., Walport, M. & Ahlomchikvv, M. 2004. Immunobiology: The Immune System in Heatlh and Diesease Garland Science Textbooks. Jenkins, R. J., Angus, R. A., Mcnatt, H., Howell, W. M., Kemppainen, J. A., Kirk, M. & Wilson, E. M. 2001. Identification of androstenedione in a river containing paper mill effluents Environ. Toxicol. Chem, 20, 1325-1331. Jobling, S., Beresford, N., Nolan, M., Rodgers, G. T., Brighty, G. C., Sumpter, J. P. & Tyler, C. R. 2002. Altered sexual maturation and garmete production in wild roach (Rutilis rutilis) living in rivers that receive treated sewage effluents. Biol. Reproduction, 66, 272-281 Jobling, S., Casey, D., Rodgers, G. T., Oehlman, J., Schulte, O. U., Pawlowski, S., Baunbeck, T., Turner, A. P. & Tyler, C. R. 2003. Comparative response of mollusks and fish to environmental oestrogen and an oestrogenic effluents. Aquat. Toxicol, 65, 205-220. Jobling, S., Nolan, M., Tyler, C. R., Brighty, G. & Sumpter, J. P. 1996a. Widespread sexual disruption in wild fish. Env. Sci. Tech., 32, 2498-2506 Jobling, S., Sheahan, D., Osborne, J. A., Matthiessen, P. & Sumpter, J. P. 1996b. Inhibition of Testicular Growth in Rainbow Trout (Oncorhynchus Mykiss) Exposed to Estrogenic Alkylphenolic Chemicals. Env. toxicol. Chem., 15, 194-202. Johns, G. G. & Mcconchie, D. M. 1994. Irrigation of Bananas with Secondary Treated Sewage Effluent. I. Field Evaluation of Effect on Plant Nutrients and Additional Elements in Leaf, Pulp and Soil. Aust. J. Agric. Res, 45, 1601-1617. Johnson , K. L., Cumming, A. M. & Birnbaum, L. S. 1997. Promotion of Endometriosis by polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls. Environ Health Perspect, 105, 750-755. Kavlock, J. R., Daston, G. P., Derosa, C., Fenner-Crisp, P., Kaattai, S., Lucier, G., Luster, M., Mac, M. J., Miller, R., Moore, J., Scott, G., Sheehan, M., Sink, T. & H.A., T. 1996. Reseach need for the risk assessment of health and environmental effect of endocrine disruptor: a report of the U.S. sponsored workshop. Env. Health Perspec., 104, 715- 740. Keam, S. J. & Wagstaff, A. J. 2003. Ethinylestradiol/drospirenone: a review of its use as an oral contraceptive, Treatment in Endocrinol, 2, 49-70. Kelce, W. R., Stone, C. R., Laws, S. C., Gray, L. E., Kemppainen, J. A. & Wilson, E. M. 1995. Persistant DDT metabolite p,p’-DDE is a potent androgen receptor anatagonist. Nature, 375, 581-585. Kelly, K. 2000. Analysis of Steroids in environmental water samples using solids-phase extraction and ion-trap gas chromatography-mass spectrometry and gas chromatography-tandem mass spectrometry. J. Chromatogr. A, 872, 309-314. Khan, S. J., Wintgens, T., Sherman, P., Zaricky, J. & Schafer, A. I. 2004. Removal of hormones and pharmaceuticals in the advanced water recycling demonstartion plant in Queensland, Australia. Water Sci Technol, 50, 15-22. Kile, D. E. & Chiou, C. T. 1989. Water solubility enchancement of DDT and trichlorobenzene by some surfactants below and above the critical micelle concentration Env. Sci. Tech., 23, 832-838. Kim, S. D., Cho, J., Kim, I. S., Venderford, B. J. & Snyder, S. A. 2007. Occurance and removal of pharmaceuticals and endocrine disruptors in South Korean surface, drinking, and waste waters. Wat Res, 41, 1013-1023. Kjaer, J., Olsen, P., Bach, K., Barlebo, H. C., Ingerslev, F., Hansen, M. & Sorensen, B. H. 2007. Leaching of of estrogenic hormones from manure-treated structured soils. Environ. Sci. Technol., 41, 3911-3917. Krishnan, A. V., Stathis, P., Permuth, S. F., Tokes, L. & Feldman, D. 1993. Bisphenol-A: An estrogenic substance related from polycarbonate flasks during autoclaving. Endocrinol, 132, 2279-2286.

189

Kuch, H. M. & Ballschmiter, K. 2001. Determination of Endocrine-Disrupting phenolic compounds and estrogens in surface and drinking water by HRGC-(NCI)-MS in the Picogram per Liter Range. Env. Sci. Tech., 35, 3201-3206. Kudak, B. & Namienik, J. 2008. Environmental Fate of Endocrine Disrupting Compounds— Analytical Problems and Challenges. Critic Revi Ana Chem 38, 242− 258. Kundu, N., Keenan, S. & Slaunwhite, W. R. J. 1977. Production of Antisera Against Contraceptive Steroids. Steroids, 30, 85−98. Kuss, E. & Goebel, R. 1972. Determination of estrogens by radioimmunoassay with antibodies to estrogen-C6-conjugates: I. Synthesis of estrone-, estradiol-17β-, and estriol-6-albumin conjugates. Steroids, 19, 509−518. Lafont, R. & Mathieu, M. 2007. Steroids in aquatic invertebrates. Ecotoxicology 16, 109-130. Lagana`, A., Bacaloni, A., De Leva, I. , Faberi, A., Fago, G. & Marino, A. 2004. Analytical methodologies for determining the occurrence of endocrine disrupting chemicals in sewage treatment plants and natural waters. Ana Chem Acta., 501, 79−88 Lai, K. M., Johnson, K. L., Scrimshaw, M. D. & Lester, J. N. 2000. Binding of waterbourne steroid estrogens to solid phases in river and estuarine system. Environ. Sci. Technol., 34, 3890-3894. Lee, N. A. & Kennedy, I. R. 2007. Immunoassay. In: PICO, Y. (ed.) Food Toxicants Analysis: Techniques, Strategies and Developments. Valencia, Spain: Elsevier. Leusch, F. D. L., Chapman, H. F., Van De Heuvel, M. R., Tan, B. L. L., Gooneratne, S. R. & Tremblay, L. A. 2006. Bioassay-derived androgenic and estrogenic activity in municipal sewage in Australia and New Zealand. Ecotox. Env. Safety, 65. Li, C., Tyler, T., Zwieten, L. V., Lee, N. A. & Kennedy, R. I. 2007. Optimisation of analytical method for estrogen in surface water and primary risk assessment in South Creek. Int. J. Water, 3. Li, Z., Wang, S., Lee, N. A., Allan, R. D. & Kennedy, I. R. 2004. Development of a solid- phase extraction—enzyme-linked immunosorbent assay method for the determination of estrone in water. Analyti. Chem. Act., 503, 171-177. Lindner, H. R., Perel, E., Friedlander, A. & Zeitlin, A. 1972. Specificity of antibodies to ovarian hormones in relation to the site of attachment of the steroid hapten to the peptide carrier. Steroids, 19, 357--375. López De Alda, M. J. & Barceló, D. 2001. Review of analytical methods for the determination of estrogens and progestogens in waste waters. J. Analyt. Chem., 371, 437-447. Lu, H., Conneely, G., Crowe, M. A., Aherne, M., Pravda, M. & Guilbault, G. G. 2006. Screening for testosterone, methyltestosterone, 19-northestosterone residues and their metabolites in bovine urine with enzyme-linked immunosorbant assay (ELISA). Analyti. Chem. Act., 570, 116-123. Marcus, G. J. & Durnford, R. 1988. Estradiol assay by microtiter plate enzyme immunoassay. J. Steroid Biochem. Mol. Biol., 29, 207-212. Matějíček, D. & Kubáň, V. 2007. High performance liquid chromatography/ion-trap mass spectrometry for separation and simultaneous determination of ethynylestradiol, gestodene, levonorgestrel, cyproterone acetate and desogestrel. Analyti. Chem. Act., 588, 304–315. Matani, K., Fujioka, M. & Kataoka, H. 2005. Fuly automated analysis of estrogens in environmental waters by in-tube solid-phase microextraction coupled with liquid chromatography-tandam mass spectrometry. J Chromatogr A, 22, 218-224. Matsui, S., Takigami, H. & Matsuda, T. 2000. Estrogen and estrogen mimics contamination in water and the role of sewage treatment. Water Sci Technol 42, 173–179. Matthiessen, P., Arnold, D., Johnson, A. C., Pepper, T. J., Pottinger, T. G. & Pulman, K. T. 2006. Contamination of headwater streams in the United Kingdom by oestrogenic hormones from livestock farms. Sci Total Environ, 367, 616–630. 190

Matthiessen, P. & Gibbs, P. E. 1998. Critical appraisal of the evidence for tributyltin- mediated endocrine disruption in Mollusks. Env. Toxico. Chem., 17, 37-43. Mclachlan, J. A. & Arnold, S. F. 1996. Environmental estrogens. Amer. Sci., 84, 452-461. Melamed, M., Castano, E., Notides, A. C. & Sasson, S. 1997. Molecular and kinetic basis for the mixed agonist/antagonist activity of estriol. Mol. Endocrinol., 11, 1868-1878. Metzger, D., Losson, R., Bornert, J. M., Lemoine, Y. & Chambon, P. 1992. Promoter specificity of the two transcriptional activation functions of the human oestrogen receptor in yeast. Nucleic Acids Res, 20, 2813-2817. Meulenberg, E. P., Mulder, W. H. & Stoks, P. G. 1995. Immunoassay for pesticides, Advance ACS Abstracts. Mian, I., Bradwell, A. & Olson, A. 1991. Structure, function and properties of antibody binding sites. J. Mol. Biol, 217, 133-151. Mispagel, C., Allinson, G., Allinson, M., Shiraishi, F., Nishikama, M. & Moore, M. R. 2009. Obeservation on the estrogenic activity and concentration of 17b-estradiol in the discharge of 12 wastewater treatment plants in south australia. Arch Environ Contam Toxicol, 56, 631-637. Mispagel, C., Shiraishi, F., Allinson, G. & Allinson, M. 2005. Estrogenic Activity of Treated Municipal Effluent from Seven Sewage Treatment Plants in Victoria, Australia. Mohammed, K. & Esen, A. 1989. A blocking agent and a blocking step are not needed in ELISA, immunostaining dot-blot and Western blots. J. Immunol. Methods, 117, 141- 145. Mooradian, A. D., Morley, J. E. & Korenman, S. G. 1987. Biological actions of androgens. Endocr. Rev., 8. Morteani, G., Moller, P., Fuganti, A. & Paces, T. 2006. Input and fate of anthropogenic estrogens and gadolinium in surface water and sewage treatment plants in the hydrological basin of Prague (Czech Republic). Environ Geochem Health, 28, 257- 264. Mortensen, A. S. & Arukwe, A. 2007. Effects of 17b-ethynylestradiol on hormonal responses and xenobiotic biotransformation system of Atlantic salmon (Salmo salar). Aqaut. Toxicol., 85, 113–123. Muller, M., Rabenoelina, F., Balaguer, P., Patureau, D., Lemenach, K., Budzinski, H., Barceló, D., López De Alda, M., Kuster, M., Delgenes, J. P. & Hernandez-Raquet, G. 2008. Chemical and biological analysis of endrocrine-disrupting hormones and estrogenic activity in an advance sewage treatment plant. Env. toxicol. Chem., 27, 1649-1658. Nagler, J. J. & Cry, D. G. 1997. Exposure of male American plaice (Hippoglossoides platessoides) to contaminated marine sediments decrease the hatching success of their progeny. Env. Toxico. Chem., 16, 1733-1738. Nahm, M. H. & Hoffmann, J. W. 1990. Heteroantibody: phantom of immunoassay. Clin. Chim., 36, 829. Nakamura, S., Sian, T. H. & Daishima, S. 2001. Determination of estrogen in river water by gas-chromatography-negative-ion chemical-ionization mass spectrometry. J. Chromatogr. A, 919, 275 - 282. Newbold, R. R., Banks, E. P., Bullock, B. & Jefferson, W. N. 2001. Uterine adenocarcinoma in mice treated neonatally with genistein. Cancer Res, 61, 4325-4328. Nikolayenko, I. V., Galkin, O. Y., Grabchenko, N. I. & Spivak, M. Y. 2005. Preparation of highly purified human IgG,IgM, and IgA for immunization and immunoanalysis. Ukra. Bio. Acta 2, 3-11. Oberdorster, E. & Cheek, A. O. 2001. Gender benders at the beach: Endocrine disruption in marine and estuarine organisms. Environ. Toxicol. Chem, 20, 23–36.

191

Oehlmann, J., Schulte-Oehlmann, U., Tillmann, M. & Markert, B. 2000. Effect of endocrine disruptors on prosobranch snails (Mollusca: Gastropoda) in laboratoty. Part I: bisphenol A and Octylphenol as xeno-estrogens. Ecotoxicol., 9, 383-397. Ohtake, F., Takeyama, K. I., Matsumoto, T., Kitagawa, H., Yamamoto, Y., Nohara, K., Tohyama, C., Krust, A., Mimura, J., Chambon, P., Yanagisawa, J., Fujii-Kuriyama, Y. & Kato, S. 2003. Modulation of oestrogen receptor signaling by association with the activated dioxin receptor. Nature, 425, 545-550. Orlando, E. F., Denslow, N., Folmar, L. C. & Guillette, L. J. J. 1999. Comparison of the reproductive physiology of largemount bass, Micropterus salmoides, collected from the Escambia and Blackwater River in Florida Environ. Health Perspect, 107, 199- 204. Orlando, E. F., Kolok, A. S., Binzcik, G. A., Gates, J. L., Horton, M. K., Lambright, C. S., .Gray, L. E., Soto, A. M. & Guillette, L. J. 2004. Endocrine-Disrupting Effect of Cattle Feedlot Eefluent on An Aquatic Sentinel Species, the Fathead Minnow. Environ. Health Perspect, 112, 353-358. Orme, M. L., Back, D. J. & Breckenridge, A. M. 1983. Clinical pharmacokinetics of oral contraceptive steroids. . Clin Pharmacokinet, 8, 95–136. Oubiña, A., Ballesteros, B., Bou Carrasco, P., Galve, R., Gascón, J., Iglesias, F., Sanvicens, N. & Marco, M. P. 2000. Chapter 7 Immunoassays for environmental analysis. Techn. Instru. Analyti.Chem., 21, 287-339. Pacáková, V., Loukotková, L., Bosáková, Z. & Štulik, K. 2009. Analysis for estrogen as environmental pollutants- A review. J. Sep. Sci, 32, 867-882. Palmer, J. R., Hatch, E. E., Rosenberg, C. L., Hartge, P., Kaufman, R. H., Titus-Ernstoff, L., Noller, K. L., Herbst, A. L., Rao, R. S., Troisi, R., Colton, T. & Hoover, R. N. 2002. Risk the breast cancer in women exposure to diethylstilbestrol in utero: preliminary results (United States). Cancer Causes Contr., 13, 753-758 Pelissero, C., Flouriot, G., Floucher, J. L., Bennetau, B., Donogues, J., Legac, F. & Sumpter, J. P. 1993. Vitellogenin synthesis in cultured hepatocutes; An in vitro test for the oestrogenic potency of chemicals. J. Steroid Biochem. Mol. Biol., 44, 263-272. Pfortner, P., Weller, M. G. & Niessner, R. 1998. Immunologocal method for the detection of nitroaromatic residue covalently bound to humic acids. Fresenius J Anal Chem, 360, 192-198. Phillips, B. & Harrison, P. 1999. Overview of the endocrine disrupters issue. Env. Sci. Tech., 12, 1-26. Porter, W. P., Green, S. M., Debbink, N. L. & Carlson, I. 1993. Groundwater pesticides: interactive effects of low concentrations of carbamates aldicarb and methomyl and the triazine metribuzin on thyroxine and somatotropin levels in white rats. J. Toxico. Env. Health, 40, 15-34. Propper, C. R. 2006. The study of endocrine-disrupting compounds: Past approaches and new directions. Intergr. Comp. Biol., 45, 194-200. Reddy, S., Rudel, R. A., Swartz, C. & Brownawell, B. J. 2004. Transport and fate of steriods estrogens in sunsurface environments. ACS National Meeting. Philadelphia, USA. Rehmann, K., Schramm, K. W. & Ketteup, A. 1999. Applicability of a yeast oestrogen screen for the detection of oestrogen-like activities in environmental samples. Chemos., 38, 3303-3312. Reijnders, P. J. H. 1986. Reproductive failure in common seals feeding on fish from polluted coastal waters. Nature, 324, 456-457. Rhind, S. M. 2002. Endocrine disrupting compounds and farm animal: their properties, action and route of exposure. Domestic Animal Endocrinol., 23, 179-187. Richardson, M. L. & Bowron, J. M. 1985. The fate of pharmaceutical chemicals in the aquatic environment. J Pharm Pharmacol, 37, 1-12.

192

Robinson, C. D., Brown, E., Craft, J. A., Davies, I. M., Moffat, C. F., Piries, D., Robertson, F., Stagg, R. M. & Struthers, S. 2003. Effects of sewage effluent and ethynylestradiol upon molecular markers of estrogenic exposure, maturation and reproductive success in the sand goby (Pomatoschistus minutus, Pallas). Aquat Toxicol, 24, 119-134. Roepke, T. A., Snyder, M. J. & Cherr, G. N. 2005. Estradiol and endocrine disrupting compounds adversely affect development of sea urchin embryos at environmentally relevant concentrations. Aquat. Toxicol.y, 71, 155−173 Routledge, E. J., Shbahan, D., Desbrow, C., Brighty, G. C., Waldock, M. & Sumpter, J. P. 1998. Identification of oestrogenic chemicals in STW effluent.2.In vivo response in Trout and Roach. Env. Sci. Tech., 32, 1559-1565 Routledge, E. J. & Sumpter, J. P. 1997. Strucrual features of alkylphenolic chemicals associated with estrogenic activity. J. bio. chem., 272, 3280-3288. Roux, K. 1999. Immunoglobulin structure and function as revealed by electron microscopy. Int Arch Allergy Immunol, 120, 85-99. Rutishauser, B. V., Pesonen, M. & Escher, B. I. 2004. Comparative analysis of estrogenic activity in sewage treatment plant effluents involving in vitro assays and chemical analysis of steroids. Environ Toxicol Chem, 23, 857–864. Sadeh, D., Sela, E. & Hexter, C. S. 1979. Novel Enzyme Immunoassay for 17b-estradiol. J. Immunol. Methods, 28, 125−131. Safe, S. 1995. Environmental and dietary estrogens and human health: Is there a problem? Env. Health Perspec., 103, 346−351. Schneider, C., Schöler, H. F. & Schneider, R. J. 2005. Direct sub-ppt detection of the endocrine disruptor ethinylestradiol in water with a chemiluminescence enzyme- linked immunosorbent assay. Analyti. Chem. Act., 551, 92−97. Schneider, C., Schöler, H. F. & Schneider, R. J. 2004. A novel enzyme-linked immunosorbent assay for ethynylestradiol using a long-chain biotinylated EE2 derivative. Steriods, 69, 245-253. Schultis, T. & Metzger, J. W. 2004. Detremination of oestrogenic activity by LYES-assay (yeast oestrogen screen-assay assisted by enzymatic digestion with lyticase. Chemos., 75, 1649-1655. Schultis, T., Spengler, P., Konig, A. & Metzger, J. W. 2002. Determination of oestrogenic activity of environmental samples by means of binding to oestrogen receptor ER-α and ER-β and fluorescenc polarization. Vom Wasswe, 98, 1-12. Schultz, I. R., Skillman, A., Nicolas, J. M., Cyr, D. G. & Nagler, J. J. 2003. Short-term exposure to 17α-ethinylestradiol decreases the fertility of sexually maturing male rainbow trout (Oncorhynchus mykiss). Env. Toxico. Chem., 22, 1272-1280. Schuurs, A. H. W. M. & Van Weemen, B. K. 1977. Enzyme-immunoassay. Clin. Chem. Acta, 81. Shan, G., Stoutamire, D. W., Wengatz, I., Gee, S. J. & Hammock, B. D. 1999. Deveopment of immunoassay for the pyrethoid insecticide esfenvalerate. J. Agric. Food Chem, 47, 2145-2155. Sharpe, M. 1993. Declining sperm counts in men-is there an endocrine cause? J. Endocrinol., 136, 357-360. Siklodi, B., Barna-Vetro, I. & Solti, L. 1995. Latent immunisation to produce high affinity monoclonal antibodies to porgesterone. Hybridoma, 14, 79-84. Singer, H., Muller, S., Tixier, C. & Pillonel, L. 2002. Triclosan: occurrence and fate of a widely used biocide in the aquatic environment: field measurements in wastewater treatment plants, surface waters, and lake sediments. Environ Sci Tech, 36, 4998-5004. Slikker, R. W., Lipe, G. W. & Newport, G. D. 1981. High-performance liquid chromatographic analysis of oestradiol-17β and metabolites in biological media. J. Chromatogr. A, 224, 205-219.

193

Smith, S. 2003. Population Growth: Implications for Australia and Sydney. NSW Parliamentary Lib. Res. Service Snyder, S., Vanderford, B., Pearson, R., Quinñones, O. & Yoon, Y. 2003. Analytical Methods Used to Measure Endocrine Disrupting Compounds in Water. Practice Periodical of Hazardous. Toxic. Radi. Waste Manag., 7, 224-234. Snyder, S. A., Keith, T. L., Verbrugge, D. A., Snyder, E. M., Gross, T. S., Kannan, K. & Giesy, J. P. 1999. Analytical method for detection of selected estrogenic compounds in aqueous mixtures Env. Sci. Tech., 33, 2814-2820. Sohoni, P. & Sumpter, J. P. 1998. Several environmental oestrogens are also anti-androgens. J. Endocrinol., 158, 327-339. Soini, E. & Hemila, I. 1979. FIuoroimmunoassay- Present status and key problems. Soto, A. M., Chung, K. L. & Sonnenschein, C. 1994. The pesticide endosulfan, toxaphene, and dieldrin have oestrogenic effect on human oestrogen-sensitive cells. Environ Health Perspect., 92, 167-173 Streck, G. 2009. Chemical and biological analysis of estrogenic, progestagenic and androgenic steroids in the environment. Trends Analyt. Chem., 28, 635 −652. Sumpter, J. P. & Purdom, C. E. 1994. Estrogenic effects of effluents from sewage treatment work. Chem Ecol, 8, 275-285. Svenson, A., Allard, A. S. & Ek, M. 2003. Removal of estrogenicity in Swedish municipal sewage treatment plants. Wat. Res., 37, 4433–4443. Swerdloff, R. S., Wang, C. & Bhasin, S. 1992. Developments in the control of testicular function. Baillieres Clin. Endocrinol. Metab., 6, 451-483. Szurdoki, F., Bekheit, H. K. M., Marco, M. P., Goodrow, M. H. & Hammock, B. D. 1995. New Frontiers in Agrochemical Immunoassay. In: STANKER, L., SKERRITT, J. H. & KURTZ, D. A. (eds.). Arlington, VA: AOAC International. Tang, T., Li, P., Luo, L., Shi, D., Li, J. & Cao, Y. 2009. Development and validation of a HPLC method for determination of levonorgestrel and quinestrol in rat plasma. Biomed Chromatogr. Ternes, T. A., Anderson, H., D., G. & Bonerz, M. 2002. Determination of estrogens in sludge and sediments by liquid extraction and GC/MS/MS. Ana. Chem., 74, 3498-3504 Ternes, T. A., Kreckel, P. & Mueller, J. 1999a. Behaviour and occurrence of estrogens in municipal sewage treatment plants-II. Aerobic batch experiments with activated sludge. The Sci. the Total Env., 225, 91-99. Ternes, T. A., Stumpg, M. M. J., Haberer, K., Wilken, R. D. & Servos, M. 1999b. Behavior and occurance of oestrogen in municipal sewage treatment plant-I. Investigations in Germany, Canada and Brazil. The Sci. Total Env., 225, 81-90. Tiefenauer, L. X. & Andres, R. Y. 1990. Biotinyl-Estradiol Derivatives in Enzyme Immunoassays: Structural Requirements for Optimal Antibody Binding. J. Steroid Biochem., 35, 633−639. Tilton, F., Benson, W. H. & Schlenk, D. 2002. Evaluation of estrogenic activity from a municipal wastewater treatment plant with predominantly domestic input. Aquat Toxicol, 61, 211–224. Tylor, C. R., Jobling, S. & Sumpter, J. P. 1998. Endocrine disruption in wildlife: A critical review of the evidence. Critic. Rev. Toxicol., 28, 319-361. Urbatzka, R., Van Cauwenberge, A., Maggioni, S., Viganò, L., Mandich, A., Benfenati, E., Lutz, I. & Kloas, W. 2007. Androgenic and antiandrogenic activities in water and sediment samples from the river Lambro, Italy, detected by yeast androgen screen and chemical analyses. Chemos., 67, 1080-1087. Van den Berg, M., Birnbaum, L., Bosveld, A. T.C, Brunström, B., Cook, P., Feeley, M., Giesy, J. P., Hanberg, A., Hasegawa, R., Kennedy, S. W., Kubiak, T., Larsen, J.C., Rolaf van Leeuwan, F.X., Dijien Liem, A.K., Nolt, C., Peterson, R.E., Poellinger, L., Safe, S., Schrenk, D., Tillitt, D., Tysklind, M., Younes, M., Wærn, F. & Zacharewski, 194

T. 1998. Toxic equivalency factors ( TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife. Env. Health Perspect., 106, 775-792. Van Wyk, J. H., Pool, E. J. & Leslie, A. J. 2003. The effect of anti-androgenic and estrogenic disrupting contaminants on breeding gland (nuptial pad) morphology, plasma testosterone levels and plasma vitellogenin level in male Xenopus laevis (African clawed frog). Arch Environ Contam Toxicol, 44, 247-256. Viganò, L., Benfenati, E., Cauwenberge, A. V., J.K., E., Erratico, C., Goksoyr, A., Kloas, W., Maggioni, S., Mandich, A. & R., U. 2008. Estrogenicity profile and estrogenic compounds determined in river sediments by chemical analysis, ELISA and yeast assays. Chemos, 73, 1078-1089. Vreudgenhil, H. J. I., Slijper, F. M. E., Mulder, P. G. H. & Weisglas-Kuperus, N. 2002. Effects of perinatal exposure to PCBs and dioxins on play behavior in Dutch children at school age. Environ Health Perspect, 110, A593-A598. Vulliet, E., Baugros, J. B., Magdeleine, M., Waton, F. & Loustalot, M. F. 2007. Analytical methods for the determination of selected steroid sex hormones and corticosteriods in wastewater. Anal Bioanal Chem 387, 2143-2151. Vulliet, E., Wiest, L., Baudot, R. & Grenier-Loustalot, M. F. 2008. Multiresidue analysis of steroids at sub-ng L−1 levels in surface and ground-waters using liquid chromatography coupled to tandem mass spectrometry. J Chromatogr A 1210, 84-91. Walker, C. H., Hopkin, S. P., Sibly, R. M. & Peakall, D. B. 1996. Principles of Ecotoxicology, London, Taylor & Francis. Warhurst., A. M. 1995. An Environmental Assessment of Alkylphenol Ethoxylates and Alkylphenols. Available: http://website.lineone.net/~mwarhurst/aperpt.pdf. [Accessed 14 May 2008]. Warner, K. E. & Jenkins, J. J. 2007. Effects of 17a-ethynylestradiol and Biphenol A on Vertebral Development in The Fathead Minnow (Pimephales Promelas). Env. toxicol. Chem., 26, 732-737. Weber, G., Schaumann, J., Carl, C. & Chwarz, S. 1989. Steriods. 24. A Novel Efficient Approach to the Synthesis of 6-Oxo Ethinylestradiol and its 3-Isopropanesulfonate. Steroids, J. fur Praktische Chemie (Leipzig), 331, 223−230. Wedge, E. & Svenneby, G. 1986. Effect of the blocking agent bovine serum albumin and tween 20 in different buffers on immunoblotting of brain proteins and marker proteins. J. Immunol. Methods, 88, 233-237. Wester, P. W. 1991. Histopathological effects of environmental pollutants β-HCH and methylmercury on reproductive organs in fresh-water fish. Comp. Biochem. Physiol., 100C, 237-239. Wild, D. 2000. Immunoassay Handbook, London, London: Nature Publishing Group. Wild, S. R. & Jones, K. C. 1992. Organic chemical entering agricultural soils in sewage sludge: screening for their potential to transfer to crop plant and livestock. Sci Total Environ, 119 , 85-119 Woods, M. 2007. Assessment of xenoestrogens in the Australian freshwater environment: use, development and validation of in-vitro modes., Adelaide, Australia, Sanson Institute, Division of health Science. University of South Australia. Wright, A. P. H., Carlstedt-Duke, J. & Gustafsson, J. A. 1990. Ligand specific transactivation of gene expression by a derivative of the human glucocorticiod receptor expressed in yeast. J. Biol. Chem, 265, 14763-14769. Xie, S., Xiang, B., Zhang, M. & Deng, H. 2010. Determination of medroxyprogesterone in water samples using dispersive liquid-liquid microextraction with low solvent consumption. Micro Acta, 168, 253–258. Xu, H., Yang, J., Wang, Y., Jiang, Q., Chena, H. & Song, H. 2008. Exposure to 17b- ethynylestradiol impairs reproductive functions of both male and female zebrafish (Danio rerio). Aqaut. Toxicol., 88, 1-8. 195

Yang, L., Luan, T. & Lan, C. 2006. Solid-phase microextraction with on-fiber silylation for simultaneous determinations of endocrine disrupting chemicals and steroid hormones by gas chromatography–mass spectrometry. J. Chromatogr. A, 1104, 23−32. Yap, B. K., Kazlauskas, R. & Elghazi, K. 1996. Profiling of urinary testosterone and luteinizing hormone in exercise-stressed male athletes, using gas chromatography- mass spectrometry and enzyme immunoassay techniques. J. Chromatogr. B. Biolmed. Appl, 687, 117-125. Yellayi, S., Naaz, A., Szewczykowski, M. A., Sato, T., Woods, J. A., Chang, J., Segre, M., Allred, C. D., Helferich, W. G. & Cooke, P. S. 2002. The phytoestrogen genistein induces thymic and immune changes: A human health concern? Pro the National Aca Sci 99, 7616-7621. Ying, G. G., Kookana, R. & Waite, T. D. 2004. Australian Water Conservation and Reuse Research Program: Endocrine Disrupting Chemicals (EDCs) and Pharmaceuticals and Personal Care Products (PPCPs) in Reclaimed Water in Australia. Ying, G. G., Kookana, R. S. & Kumar, A. 2008. Fate of estrogens and xenoestrogens in four sewage treatment plants with different technologies. Environ Toxicol Chem, 27, 87-94. Ying, G. G., Kookana, R. S. & Ru, Y. J. 2002. Occurrence and fate of hormone steroids in the environment. Env. Inter., 28, 545-551. Yolanda, P. (ed.) 2007. Food Toxicants Analysis: Techniques, Strategies and Development: Elsevier B.V. 91-133 Young, W. F., Whitehouse, P., Johnson, I. & Sorokin, N. 2002. Proposed Predicted-no-effect- concentrations (PNECs) for Natural and Synthetic Steroid Oestrogens in Surface Waters. Briston, UK: RanD Technical Zava, D. T., Blen, M. & Duwe, G. 1997. Estrogenic activity of natural and synthetic estrogens in human breast cancer cells in culture. Environ Health Perspect, 105(Suppl 3), 637- 645. Zava, D. T. & Mcguire, W. L. 1978. Androgen action through oestrogen receptor in human breast cancer cell line. Endocrinol, 103, 624-631. Zhang, S. M., Mada, S. R., Sharma, S., Torch, M., Mattison, D., Caritis, S. & Venkataramanan, R. 2008. Simultaneous quantitation of 17α-hydroxyprogesterone caproate, 17α-hydroxyprogesterone and progesterone in human plasma using high- performance liquid chromatography-mass spectrometry (HPLC–MS/MS). J Pharm Biomed Anal, 48, 1174. Zhang, Z. L., Bhibberd, A. & Zhou, J. L. 2006. Optimisation of derivatisation for the analysis of estrogenic compounds in water by solid-phase extraction gas chromatography-mass spectrometry. Anal. Chi Acta, 577, 52-61 Zhu, S., Zhang, Q. & Guo, L. H. 2008. Part-per-trillion level detection of estradiol by competitive fluorescence immunoassay using DNA/dye conjugate as antibody multiple labels. Analyt. Chem. Acta, 624, 141–146. Zuehlke, S., Duennbier, U. & Heberer, T. 2005. Determination of estrogenic steroids in surface and wastewater applying liquid chromatography-electrospray tandem mass spectrometry. J. Sep. Sci., 28, 52-58.

196