REPOSITORY > RESULTS

Abstract

The presence of steroid estrogens in the environment has been studied intensively for two decades because of their possible deleterious effects on living organisms at very low (ng/L) concentrations. To determine concentrations of steroid estrogens and the estrogenic potential of environmental samples, sophisticated and sensitive analytical techniques as well as biological methodologies are employed.
In the first part of this doctoral thesis, sample preparation and chemical analysis using gas chromatography with mass selective detection (GC-MSD) was optimised, validated and applied to environmental samples. Hospital effluent and connected wastewater treatment plant (WWTP) influent and effluent were sampled over six consecutive days to determine the levels and inter-day variations of three naturally occurring steroid estrogens: estrone (E1), 17β-estradiol (E2), estriol (E3), and synthetic 17α-ethinylestradiol (EE2). After solid phase extraction, interferences were removed with a silica gel clean-up step and the samples analyzed by GC-MSD. The determined inter-day concentrations in hospital effluent were between 8.6 and 31.3 ng/L for E1, < limit of detection (LOD) to 4.2 ng/L for E2 and 6.4 to 385.5 ng/L for E3. In the WWTP influent concentrations were 18.9 to 49.7 ng/L for E1, 2.4 to 12.7 ng/L, for E2 and <limit of quantification (LOQ) to 63.9 ng/L for E3. Reduced levels were found in the WWTP effluent: <LOD to 7.1 ng/L for E1, <LOQ for E2 and <LOD to 5.2 ng/L for E3. EE2 was detected in only one influent sample. Interestingly, the E1:E2:E3 ratio in the hospital effluent (1:0.1:9.4) was comparable to that present in the urine of pregnant women (1:0.3:20) indicating them as the most likely source of steroid estrogens. In WWTP influent the ratio was similar to that in the non-pregnant population. Calculated estradiol equivalents (EEQ) were 33.4, 22.4, 1.7 ng(EEQ)/L in the hospital effluent, WWTP influent and WWTP effluent, respectively. During wastewater treatment, 99 % of the steroid estrogens were removed from the water phase. However, there were some cases, where removal was <80 % and even <40 %. In addition, variability of removal rates (39 % up to 99 %) was observed. Results revealed E3 as being one of the most important steroid estrogens, accounting for up to 92 % of the total EEQ present in hospital samples and 37 % and 46 % in WWTP influent and effluent samples, respectively. In this study, hospitals were recognised as an important point source of steroid estrogens. Additionally, in hospital and WWTP samples, concentrations of steroid estrogens show daily variations, e.g. - from 6.4 to 385.5 ng/L for E3. Clearly, careful sampling strategies must be adopted to ensure proper risk assessment of the impact of steroid estrogens in the environment. In addition, the low potency steroid estrogens that contribute towards overall estrogenicity of the sample, e.g. E3, should be incorporated into environmental monitoring programs.
In the second part of this Thesis an in vitro estrogenicity ER-Calux® assay was modified and integrated with instrumental analysis (GC-MSD) to create a novel single protocol. This integrated protocol enables simultaneous determination of steroid estrogens concentrations and the total estrogenic potential of environmental samples. For integration purposes the commonly used dimethyl sulphoxide (DMSO) in the ER-Calux® assay was replaced by ethyl acetate, which is more compatible with gas chromatography and does not affect the performance of the ER-Calux® assay. The integrated protocol was initially tested using a standard mixture of authentic estrogens. The calculated estrogenicity values of pure standards based on GC-MSD and the ER-Calux® assay showed good correlation (r2=0.96; α=0.94). The result remained the same when spiked wastewater extracts were tested (r2=0.92, α=1.02). When applied to “real” wastewater influent and effluent samples the results proved (r2=0.93; α=0.99) the applicability of the method. The main advantages of this newly developed protocol are that the same sample can be analysed by both GC-MSD and ER-Calux® and method complexity together with material consumption and labour is significantly reduced. The methods can be applied as either a complete or sequential analysis where the ER-Calux® assay is used as a pre-screening method prior to chemical analysis.
However, for both, GC-MSD and ER-Calux®, time consuming sample preparation (sample extraction) is necessary. In the last part of the study the ER-Calux® assay was modified with the aim to simplify and shorten sample preparation. The ER-Calux® assay was modified in order to be able to test raw wastewater samples without the need to perform sample extraction. This method (NE-(ER-Calux®) assay) has been compared to the ER-Calux® assay and chemical analysis by GC-MSD, described above. In the case of spiked tap and wastewater samples there was no statistical difference between the results obtained by NE-(ER-Calux®) and the other two methods. The “real” influent and effluent samples from seven municipal WWTPs and receiving surface waters were tested. With some exceptions, higher estrogenic potential was determined by NE-(ER-Calux®) assay in comparison to ER-Calux® assay. The results of NE-(ER-Calux®) were 2.1-32.4 ng(EEQ)/L, 0.8-14.2 ng(EEQ)/L and 1.7-12.3 ng(EEQ)/L higher in influent, effluent and river downstream samples, respectively. The results of biological methods were in agreement with the results of chemical analysis. NE-(ER-Calux®) was also used to determine the intra-day dynamics of estrogenic potential in one WWTP. Influent and effluent samples were taken hourly over 12 h periods and analysed within 24 h. The results showed that the estrogenic potential increased in the influent from 12.9 ng(EEQ)/L reaching a peak at 40.0 ng(EEQ)/L at 16:00. The average hourly values (27.1 ng(EEQ)/L) were comparable to the estrogenic potential of the time proportional sample (25.2 ng(EEQ)/L), which was 62 % of the highest estrogenic potential measured. The reduction in the estrogenic potential was 92 % to 98 % according to hourly sampling and 96 % in the case of the time proportional sample. The results indicate the daily variability in the estrogenic potential, which should improve our understanding of the risk posed by environmental steroid estrogens in the future. Significant part in achieving these results is due to the ability of NE-(ER-Calux®) assay to test raw samples without extensive sample preparation. Since sample extraction was not necessary, sample preparation time was greatly reduced (95 %). This advantage is of particular importance in environmental studies involving large numbers of samples for screening and monitoring purposes.