Fate of natural estrogen conjugates in municipal sewage transport and treatment facilities

https://doi.org/10.1016/S0048-9697(02)00342-XGet rights and content

Abstract

The aim of this study was to investigate the fate of the conjugated forms of the three most common natural estrogens in the municipal aqueous environment. Levels of conjugated and free estrogens in (1) female urine; (2) a septic tank collecting domestic wastewater; (3) influents and effluents of six activated sludge sewage treatment plants (STPs) were measured. The analytical method was based on solid-phase extraction by using a Carbograph 4 cartridge and Liquid Chromatography-tandem Mass Spectrometry. On average, a group of 73 women selected to represent a typical cross section of the female inhabitants of a Roman condominium, excreted 106, 14 and 32 μg/day of conjugated estriol (E3), estradiol (E2) and estrone (E1), respectively. Apart from some E3 in pregnancy urine, free estrogens were never detected in urine samples. Estrogen sulfates represented 21% of the total conjugated estrogens. This situation changed markedly in the condominium collecting tank. Here, significant amounts of free estrogens were observed and the estrogen sulfate to estrogen glucuronated ratio rose to 55/45. A laboratory biodegradation test confirmed that glucuronated estrogens are readily deconjugated in unmodified domestic wastewater, presumably due to the large amounts of the β-glucuronidase enzyme produced by fecal bacteria (Escherichia coli). Deconjugation continued in sewer transit. At the STP entrance, free estrogens and sulfated estrogens were the dominant species. The sewage treatment completely removed residues of estrogen glucuronates and with good efficiency (84–97%) the other analytes, but not E1 (61%) and estrone-3-sulfate (E1-3S) (64%). Considering that (1) E1 has half the estrogenic potency of E2, (2) the amount of the former species discharged from STPs into the receiving water was more than ten times larger than the latter one and (3) a certain fraction of E1-3S could be converted to E1 in the aquatic environment, E1 appears to be the most important natural endocrine disrupter.

Introduction

Effluents from sewage treatments plants (STPs) can discharge into rivers estrogenic contaminants at levels sufficient to induce vitellogenin biosynthesis in male fish (Jobling et al., 1998). Several studies have shown that also birds, reptiles and mammals in polluted areas undergo alterations of the endocrine–reproductive system (Preziosi, 1998). At present, a multitude of chemicals have shown to be endocrine disrupters (Jobling et al., 1998). Among these, natural and synthetic estrogens are already effective at the lower ng/l level (Purdom et al., 1994, Routledge et al., 1998), while most other chemicals having an estrogenic effect are biologically active at the μg/l level. A toxicity identification and evaluation procedure using an ‘in vitro’ estrogen yeast assay identified the steroid estrogen component of the STP effluents as contributing the greatest proportion of the overall estrogenic activity (Desbrow et al., 1998). An input/output balance of estrogenic active compounds in a German STP (Körner et al., 2000) confirmed that xenobiotic estrogens in STP effluents amounted to only 0.7–4% of the overall estrogenic potency in ‘in vitro’ tests. The large majority of estrogenic material excreted from humans, and therefore released into sewers, is in biologically less active conjugated forms (glucuronides and sulfates). However, the occurrence of ‘free’ estrogens in STP effluents and rivers (Shore et al., 1993, Desbrow et al., 1998, Lee and Peart, 1998, Ternes et al., 1999a, Belfroid et al., 1999, Johnson et al., 2000, Baronti et al., 2000) indicates that estrogen metabolites are converted back into active forms somewhere between houses and STP outlets. A number of searchers suggested that de-conjugation occurs during the STP process (Desbrow et al., 1998, Tyler and Routledge, 1998, Allen et al., 1999, Ternes et al., 1999b). Escherichia coli, which is eliminated in large quantities in the feces and is able to synthesize large amounts of the β-glucuronidase enzyme (Dray et al., 1972), has been suggested to be responsible for the above transformation.

In Western countries, a large fraction of domestic wastewater is treated by activated sludge STPs. Thus, it is important to evaluate whether these act as chemical reactors producing free estrogens, or if they act in a positive role in removing free estrogens eventually formed upstream of STPs.

Detection and measurement of levels of free and conjugated estrogens in wastewater is difficult to perform, owing to the complexity of this matrix. By using a radioimmunoassay technique, Shore et al. (1993) measured free E2 concentration levels of up to 141 ng/l in raw sewage. This unexpectedly large amount raises the doubt that immunoassay techniques might overestimate E2 by cross-reactions. Other groups monitored free estrogens in aqueous environmental samples by more sensitive and selective techniques, such as gas chromatography (GC)-mass spectrometry (MS) or GC-tandem MS. However, analytical methodologies based on the GC technique for analyzing free estrogens are time-consuming and labor-intensive, as they require preparation of suitable estrogen derivatives. Moreover, for determining levels of conjugated estrogens by GC-MS, a further step of enzymatic hydrolysis is needed.

In two previous works (Johnson et al., 2000, Baronti et al., 2000), some of the present authors elaborated an analytical method based on solid phase extraction (SPE) with a graphitized carbon black (Carbograph 4) cartridge and Liquid Chromatography (LC)-tandem MS with an Electrospray (ES) ion source for estimating concentrations of the three natural estrogens, i.e. estriol (E3), estradiol (E2) and estrone (E1) and a synthetic estrogen, i.e. ethinyl estradiol (EE2), in both influents and effluents of the six major activated sludge STPs of the area of Rome. A simple prediction method (Johnson et al., 2000) suggested that the amounts of free estrogens entering the STPs were close to those that would be excreted by an ‘ideal’ population following complete deconjugation. However, data obtained by Nasu et al. (2001) showed the E2 concentration was still increasing from raw sewage to primary effluent within a Japanese STP. This suggested that estrogen deconjugation might also occur to a greater or lesser extent during the sewage treatment process. Baronti et al. (2000) found that the activated sludge treatment efficiently removed E3 (95%), E2 (87%), EE2 (85%), but not E1 (61%). In four events out of thirty, E1 outlet levels were even larger than inlet levels. Andreolini et al. (1987) observed that E1 is excreted in late-pregnancy urine preferentially as estrone-3-sulfate (E1-3S). On this basis, Johnson and Sumpter (2001) hypothesized that the anomalous behavior of free E1 was the result of the deconjugation of E1-3S surviving the sewer system during the activated sludge STP treatment. In addition, some E1 at the STP outlet could be originated from oxidation of E2 during sewage treatment. Johnson and Sumpter (2001) supposed also that, if the activated sludge does not contain sufficient amounts of the arylsulfatase enzyme, a certain amount of E1-3S could pass through the treatment system and enters the receiving water ecosystem.

The objective of this work was to enlighten the fate of estrogen conjugates in the municipal aqueous environment. For this purpose, we measured concentration levels of free and conjugated natural estrogens in female urine, in a domestic tank collecting wastewater from a block with approximately 250 inhabitants as well as in STP influents and effluents.

Section snippets

Reagents and chemicals

The following estrogen standards were purchased from Sigma–Aldrich (Oakville, ON, USA): E3, E2, 17α-E2, E1, E3-3-glucuronide (E3-3G), E2-3-glucuronide (E2-3G), E1-3-glucuronide (E1-3G), E3-16-glucuronide (E3-16G), E2-17-glucuronide (E2-17G), E3-3-sulfate (E3-3S), E2-3-sulfate (E2-3S), E1-3-sulfate (E1-3S), 4-octylbenzensulfonate (4-C8-LAS). 17α-E2 and C8-LAS were used as internal standards for quantifying free and conjugated estrogens, respectively. Stock solutions and working standard

Accuracy, precision and limits of quantification (LOQs)

For each type of aqueous matrix considered, analyte recoveries were determined by adding known and appropriate volumes of the working standard solutions to previously analyzed aqueous samples. Analyte addition was made with the criterion of at least quadruplicating the original concentrations. For each type of aqueous matrix, 5 recovery experiments were performed and results are shown in Table 1. Analyte recoveries ranged between 76 and 98% and were not matrix-dependent, provided that the

Acknowledgements

The authors are indebted to Andrew Johnson for helpful discussion during preparation of this manuscript.

References (32)

  • W. Körner et al.

    Input/output balance of estrogenic active compounds in a major municipal sewage plant in Germany

    Chemosphere

    (2000)
  • E.J. Routledge et al.

    Structural features of alkylphenolic chemicals associated with estrogenic activity

    J Biol Chem

    (1997)
  • C.H.L. Shackleton

    Profiling steroid hormones and urinary steroids

    J Chromatogr

    (1986)
  • T.A. Ternes et al.

    Behavior and occurrence of estrogens in municipal sewage treatment plants-I. Investigations in Germany, Canada and Brazil

    Sci Total Environ

    (1999)
  • T.A. Ternes et al.

    Behavior and occurrence off estrogens in municipal sewage treatment plants-II. Aerobic batch experiments with activated sludge

    Sci Total Environ

    (1999)
  • Y. Allen et al.

    The extent of estrogenic contamination in the UK estuarine and marine environments—further survey of flounder

    Sci Total Environ

    (1999)
  • Cited by (509)

    View all citing articles on Scopus
    View full text