ARTICLE IN PRESS

Int. J. Hyg. Environ. Health 212 (2009) 228–232

1438-4639/$ - se

doi:10.1016/j.ijh

�CorrespondE-mail addr

www.elsevier.de/ijheh

SHORT COMMUNICATION

Screening of estrogen-like activity of mineral water stored in PET bottles

Barbara Pinto�, Daniela Reali

Department of Experimental Pathology, Medical Biotechnology, Infectivology and Epidemiology, University of Pisa,

37 Via San Zeno, 56127 Pisa, Italy

Received 21 February 2008; received in revised form 24 June 2008; accepted 24 June 2008

Abstract

Bottled mineral water consumption is steadily rising in the World. Italy is the largest natural mineral waterconsumer in Western Europe, about 200L per capita per annum. Recently, research has concentrated upon emergingtoxicological problems such as the presence in drinking water of substances which interfere with the function of theendocrine system; defined as endocrine disruptors (EDs). The aim of this study was to assess the risk of exposure of thepopulation to chemicals with estrogen-like activity through mineral water consumption by monitoring the presence ofestrogenic compounds in mineral water bottled in polyethylene terephthalate (PET). A solid phase extraction (SPE)with C18 cartridges was carried out. The estrogenic activity of the extracts was assayed using a yeast assay expressingthe human estrogen receptor a (hERa). This preliminary study shows that more than 90% of the water samples did notexhibit any appreciable estrogenic activity. The highest estrogenic activity detected in water extracts was equivalent tothe activity induced by 23.1 ng/L of the natural hormone 17b-estradiol. Some mineral water samples showed toxicityon yeast cells.r 2008 Elsevier GmbH. All rights reserved.

Keywords: Natural mineral water; Estrogenic activity; Yeast assay

Introduction

Recently a significant increase in mineral waterconsumption has been recorded. We believe this is forreasons of safety and health (it is widely believed thatmineral water is more hygienic and better from anutritional point of view than tap water), and becausepeople prefer its taste. Indeed, disinfecting techniquesand the release of materials from network pipelines mayintroduce an unpleasant taste. Italy ranks first inmineral water consumption (about 200L per capitaper annum) not only in Europe, but also in other partsof the world, and mass consumption is by people who

e front matter r 2008 Elsevier GmbH. All rights reserved.

eh.2008.06.004

ing author. Tel.: +39 50 2213586; fax: +39 50 2213588.

ess: b.pinto@med.unipi.it (B. Pinto).

can be more vulnerable to damage (new mothers, babies,teenagers).

Currently, 80% of mineral water is sold in plasticcontainers, normally polyethylene terephthalate (PET).Glass bottles (24%) are primarily destined for restau-rants or biological feeding, while the cardboard packa-ging (Tetra Paks) seems destined to come off themarket or to be confined to commercial (1.5%) niches.

Studies carried out to date, in order to assess theadverse effects in humans for exposure to low doses ofchemicals through water consumption, have mainlyconsidered carcinogenic and mutagenic risks, but recentresearch has discovered an emerging aspect in environ-mental toxicology and in food safety, related to thepresence in drinking water of substances that possessbiological activity on the reproductive system (Wenzelet al., 2003).

ARTICLE IN PRESSB. Pinto, D. Reali / Int. J. Hyg. Environ. Health 212 (2009) 228–232 229

It is widely believed that the decline of the reproduc-tive functions in males, the increasing incidence of breastcancer in young women, and alterations of the thyroidfunction and neurobehavioural diseases observed in thepopulation of different countries may be, partly,attributable to exposure to several hormone-like envir-onmental chemicals, particularly during the intrauterinedevelopment or in critical periods in development.

Recent studies have focused attention on the migra-tion of chemicals from plastic containers to the water.Among these are plasticizers – additives used to impartflexibility and handling properties to several kinds ofplastics. Di(2-ethylexyl)phthalate (DEHP) is a widelyused plasticizer in the world for polyvinyl chlorideproducts but it is also present in PET (Balafas et al.,1999; Bosnir et al., 2007; Harris et al., 1997). DespitePET being a material characterized by elevated chemicalinactivity, various studies point out that differentstorage conditions (such as exposure to sunlight andhigh temperatures and the duration of the same) cancontribute to migration of chemicals from the bottles tothe water (Dabrowska et al., 2003; Evandri et al., 2000;Nawrocki et al., 2002; Westerhoff et al., 2008).

The release of substances with hormone-like activityfrom packaging material to water may represent a riskfactor for humans to develop chronic pathologies follow-ing long-term exposure to low doses of chemicals.Recently, it has been shown that the DEHP alters thegenic expression in rats and that, at appropriate concen-trations, it alters the development of the central nervoussystem in the fetus (Latini et al., 2006; Liu et al., 2005).

Moreover, it is important to consider that the actionof chemical contaminants on the endocrine systemgenerally takes place at lower doses than those necessaryto induce acute toxic effects. Some authors have shownthat arsenic, an ED that can be found in water forhuman consumption, both in tap water and mineralwater, acts on altering the hormonal function of theglucocorticoid system at a concentration notably lowerthan normally necessary to cause toxic effects (Kaltreideret al., 2001).

In this work the overall estrogenicity of organicextracts of nine commercial brands (brand ID #) ofItalian bottled natural mineral water was evaluated. Arecombinant yeast-based in vitro assay (YES) was usedto test for the ability of the extracts to bind the humanestrogen receptor a.

Materials and methods

A total of 30 samples of different commercial brands(n ¼ 9) of Italian mineral water packaged in plasticcontainers (PET) purchased from local markets wereanalysed for estrogenic activity using the Yeast Estrogen

Screen (YES). Ninety-three percent of the analysedsamples had total dissolved solids (TDS) values between39 and 276mg/L and a pH range of 7.21–8.2. Accordingto TDS values at 180 1C, the Italian legislation classifiesmineral natural water as low mineral content water(TDSo500mg/L), as medium mineral content water(500oTDSo1500mg/L), as high mineral content water(41500mg/L). More than 50% of mineral watercommercially sold in Italy is of low mineral content,about 25% is of medium mineral content.

Water samples were extracted by solid-phase extrac-tion (SPE) on C18 cartridges (Supelco, Bellefonte, USA)previously conditioned with 15mL methanol (CarloErba, Milan, Italy) followed by 15mL n-hexane (CarloErba, Milan, Italy), at a flow rate of 15mL/min. Thevolume was chosen assuming an average daily con-sumption of 1.5 L per capita. Depending on the climateand on physical activity the daily water requirement ofan individual can vary, but the daily assumption of1.5–2L for an adult and 1L for children is recom-mended by WHO (2000).

The cartridges were then dried under a gentle flow ofnitrogen for 1 h and eluted sequentially with 5mL ofmethanol and 5mL of n-hexane. Recovery efficiency(average of three determinations7SD) of estrogeniccompounds has been assessed with standard diethyl-stilbestrol (DES) (Sigma, Milan, Italy). Adsorbates weredissolved in dimethyl sulfoxide (DMSO) for the in vitroassay.

The yeast strain used was S. cerevisiae RMY326 (His3Leu2-3, 112 trp1-1 ura3-52/hER-TRP1-2m[pG/ER(G)],ERE-CYC-LacZ-URA3-2m[pUCDSS-ERE], HIS-3CEN/ARS[pRS423]). This strain contains the human estrogenreceptor a (hERa) and an estrogen-responsive element(ERE) bound to the reporter gene lacZ encoding for theenzyme b-galactosidase (Liu and Picard, 1998).

The activation of the receptor due to the formation ofa complex receptor–ligand causes expression of thereporter gene lacZ. The production of the enzymeb-galactosidase is measured with a spectrophotometer(OD420 nm) (Pinto et al., 2004; Garritano et al., 2006),normalised to the number of cells assayed and expressedas Miller units using the following formula (Miller,1972):

b� gal units ðM:U:Þ ¼ ð1000�OD420Þ=ðt� V �OD600Þ

t is the length of incubation (min); V is the volume ofculture used in the assay (mL). Water samples wereanalysed at final concentrations 100X and 200X.

Ultra-pure ‘‘endotoxin-free’’ water (Sigma-Aldrich,Milan, Italy) was used as a control of critical points inextraction procedure (blank). Vehicle was used as anegative control. For comparison, tap water samplesfrom local distribution networks (treated surface water,groundwater and spring water) were analysed.

ARTICLE IN PRESS

-11 -10 -9 -8 -70

200

400

600

800

1000

1200

Concentration log (M)

Mill

er u

nits

Fig. 1. Standard dose–response curve for 17b-estradiol (mean

of 14 standard curves obtained in separate experiments7SE).

B. Pinto, D. Reali / Int. J. Hyg. Environ. Health 212 (2009) 228–232230

Estrogenic activity was expressed as relative inductiveefficiency (RIE) calculated as the ratio between themaximal b-galactosidase activity obtained with eachwater extract and that elicited by 10 nM 17b-estradiol(E2)� 100, and as estradiol equivalents (ng/L EEQs).Estradiol equivalents (EEQs) of water samples weredetermined by interpolation from the estradiol standardcurve (mean of 14 standard curves obtained in separateexperiments). The standard curve was fitted usingGraphpad Prism (version 4.02) (Fig. 1).

Results and discussion

Ninety percent of samples elicited an estrogenicactivity lower than 10% of the activity induced by10 nM E2. Table 1 summarizes the highest values ofestrogenic activity measured in mineral water and tapwater extracts expressed as RIE and EEQs.

The highest estrogenic activity measured was 11.32%of E2, corresponding to 23.1 ng/L estradiol equivalents(EEQs) (range, 0.027–23.1 ng/L; mode, 9.5 ng/L). Or-ganic extracts from one brand represented 85.7% ofsamples with EEQs higher than 9.5 ng/L. Extracts of tapwater from treated river water, groundwater, and springwater showed estrogenic activity between 0.46 and17.2 ng/L EEQs.

All mineral water extracts showed a different degreeof inhibition of yeast cell growth at a final 200X

concentration, and 20% of them were toxic on the yeastcells (Table 2). The Sigma water used as a control for theextraction procedure did not show, in repeated experi-ments, cell toxicity.

Our results show a low estrogenic activity of Italianmineral natural water stored in PET bottles. Studies inprogress of other authors (Wagner et al., 2007) seem tounderline the presence in PET bottled water of activexeno-estrogens on both in vitro and in vivo tests. These

authors suggest that compounds may originate fromplastic packaging materials. Indeed, not all PETmaterials are of the same chemical quality. Qualitymay vary depending on the raw material as well as thetechnology used in bottle production that could dragtechnological contaminants. PET bottles are producedfrom different polymers to which correspond a numberof commercial trademarks. Italian law allows the use ofdifferent polymers at the same bottled water plant.

In previous studies carried out in our laboratory, therelease under dynamic flow conditions (15mL/min) ofestrogenic compounds from plastic tubes (silicon,polyvinyl chloride, Tygons, Teflons, polyethylene) towater was assessed in the YES assay. Several migratingcompounds were identified in all plastic materials usingchromatography–mass spectroscopy (GC/MS) and solidphase microextractions (SPME) techniques. A strongcorrelation between the chemical release and theestrogenic activity measured in vitro with the yeast testwas observed. Migrating compounds from plastic tubesincluded dibutyl phthalate (DBP), diisooctyl phthalate(DOP), ethyl hexanol, hexamethylcyclotrisiloxane,2,4-dichlorobenzoic acid, isobenzofuranone, methyl cyclo-pentanone, dimethylbenzene methanol, benzoisothiazolin,methylundecene (unpublished data).

The presence of xenobiotics in mineral water repre-sents a complex problem, because it can be contami-nated in different phases of the production process,from the supplying of the raw materials to handling,storing, and distribution (Biscardi et al., 2003).

The increasing diffusion of plastic containers forwater and soft drink packaging is due both to thegreatest practicality for the consumer and to cost cuttingfor the companies, but it may involve a potentialexposure of individuals consuming large amounts ofbottled mineral water to low doses of chemicalsmigrating from packaging materials or be a consequenceof storage conditions. Daily consumption and, conse-quently, the prolonged exposure to potentially dereg-ulating compounds of the endocrine system, is a factorthat must not be underestimated.

This study points out some interesting aspects thatneed further investigation, such as the inhibition of theyeast growth induced by some extracts, not attributableto the analytical method used for extraction. Celltoxicity was observed for water samples of the samelot of three different brands purchased from the sameretailer. Since toxicity might be attributable to thestorage conditions of the product, we think that thisaspect should be investigated further.

The national and international scientific communityhas long accepted the reports on the presence of EDs intap-drinking water from researchers from many coun-tries. However, currently there is no European Com-munity legislation specific to EDs in drinking water. Inorder to ensure food safety in a global trade it is

ARTICLE IN PRESS

Table 1. Estrogenic activity (highest value detected and range) measured in mineral water samples, tap water from surface water

(TW-SuW), tap water from groundwater (TW-GrW), and tap water from spring water (TW-SpW) expressed as RIE (% of 10 nM

E2) and EEQs (ng/L)

aBrand ID number TSD (mg/L) pH RIE (%) EEQs (ng/L)

1 39.2 7.4 7.22 (0.039–7.22) 17.6

2 75.5 7.8 5.69 (0.13–5.69) 13.9

3 137 8.2 3.82 (0.68–3.82) 8.41

4 179 7.8 2.51 (0.007–2.51) 4.1

5 152 7.9 2.34 (0.63–2.34) 3.5

6 250 7.68 11.32 (0.034–11.32) 23.1

7 276 7.21 1.58 (0.25–1.58) 0.9

8 899 6.26 2.72 (0.61–2.72) 4.8

9 1279 6.1 4.83 (0.49–4.83) 11.5

TW-SuW 340 7.5 7.04 (0.73–7.04) 17.2

TW-GrW o500 7.48 6.16 (1.76–6.16) 15.1

TW-SpW o500 7.53 0.57 (0.09–0.57) 0.46

Blank 0 7.0 0 0

aMineral water samples are ordered according to increasing total dissolved solids (TDS) value (mg/L) at 180 1C.

Table 2. Yeast cell growth inhibition induced by samples

(n ¼ 6) showing toxic effect

Water

sample

Yeast growth (%)

concentration 200X

Yeast growth (%)

concentration 100X

W1 14.370 22.972.67

W2 15.873.08 70.1711.14

W3 4.170.41 39.9711.10

W4 7.370.58 75.177.23

W5 16.0713.18 73.5710.68

W6 23.174.59 72.778.36

Blank 10071.84 10070.60

Yeast growth was measured as optical density at 600 nm (OD600). Cell

growth is expressed as a percentage of the OD600 measured for cells

treated with vehicle (DMSO) alone. Values are the means7SD.

B. Pinto, D. Reali / Int. J. Hyg. Environ. Health 212 (2009) 228–232 231

therefore opportune that the Public Health Servicespromote research and surveillance in this field.

Aknowledgments

The authors thank Dr. G. Valentini, ARCHALaboratories S.r.l., Pisa, for the analysis of migrants inplastic material (ARCHA laboratories S.r.l. is a certifiedresearch laboratory, UNI EN ISO 9001–2000, CERTI-QUALITY), and Dr. Geoffrey Phillips, Interdepart-mental Language Centre, University of Pisa, for Englishrevision.

References

Balafas, D., Shaw, K.J., Whitfield, F.B., 1999. Phthalate and

adipate esters in Australian packaging materials. Food

Chem. 65, 279–287.

Biscardi, D., Monarca, S., De Fusco, R., Senatore, F., Poli, P.,

Buschini, A., Rossi, C., Zani, C., 2003. Evaluation of the

migration of mutagens/carcinogens from PET bottles into

mineral water by Tradescantia/micronuclei test, comet

assay on leukocytes and GC/MS. Sci. Total Environ. 302,

101–108.

Bosnir, J., Puntaric, D., Galic, A., Skes, I., Dijanic, T., Klaric,

M., Grgic, M., Curkovic, M., Smit, Z., 2007. Migration of

phthalates from plastic containers into soft drinks and

mineral water. Food Technol. Biotechnol. 45, 91–95.

Dabrowska, A., Borcz, A., Nawrocki, J., 2003. Aldehyde

contamination of mineral water stored in PET bottles.

Food Addit. Contam. 20, 1170–1177.

Evandri, M.G., Tucci, P., Bolle, P., 2000. Toxicological

evaluation of commercial mineral water bottled in poly-

ethylene terephthalate: a cytogenetic approach with Allium

cepa. Food Addit. Contam. 17, 1037–1045.

Garritano, S., Pinto, B., Calderisi, M., Cirillo, T., Amodio-

Cocchieri, R., Reali, D., 2006. Estrogen-like activity of

seafood related to environmental contaminants. Environ.

Health 5 (9), (10 pp.) (Epub ahead of print).

Harris, C.A., Henttu, P., Parker, M.G., Sumpter, J.P., 1997.

The estrogenic activity of phthalate esters in vitro. Environ.

Health Perspect. 105 (8), 802–811.

Kaltreider, R.C., Davis, A.M., Lariviere, J.P., Hamilton, J.W.,

2001. Arsenic alters the function of the glucocorticoid

receptor as a transcription factor. Environ. Health Per-

spect. 109, 245–251.

Latini, G., Del Vecchio, A., Massaro, M., Verrotti, A., DE

Felice, C., 2006. In utero exposure to phthalates and fetal

development. Curr. Med. Chem. 13, 2527–2534.

Liu, J.W., Picard, D., 1998. Bioactive steroids as contaminants

of the common carbon source galactose. FEMS Microbiol.

Lett. 159, 167–171.

Liu, K., Lehmann, K.P., Sar, M., Young, S.S., Gaido, K.W.,

2005. Gene expression profiling following in utero exposure

to phthalate esters reveals new gene targets in the etiology

of testicular dysgenesis. Biol. Reprod. 73, 180–192 Epub.

ARTICLE IN PRESSB. Pinto, D. Reali / Int. J. Hyg. Environ. Health 212 (2009) 228–232232

Miller, J.H., 1972. Assay for beta-galactosidase. In: Miller,

J.H. (Ed.), Experiments in Molecular Genetics. Laboratory

Press, Cold Spring Harbour, New York, pp. 352–355.

Nawrocki, J., Dabrowska, A., Borcz, A., 2002. Investigation

of carbonyl compounds in bottled water from Poland.

Water Res. 36, 4893–4901.

Pinto, B., Picard, D., Reali, D., 2004. A recombinant yeast

strain as a short term bioassay to assess estrogen-like

activity of xenobiotics. Ann. Ig. 16, 579–585.

Wagner, M., Luther, E., Oehlmann, J., 2007. Xenoestrogens in

mineral water – first in vitro and in vivo evidence. In:

Proceedings ‘‘Second Workshop on Low Dose Effects of

Chemicals with special consideration to phthalates,’’ April

26–28, 2007, Berlin, Germany.

Wenzel, A., M +uller, J., Ternes, T., 2003. Study on endocrine

disruptors in drinking water. Final report. Available at

/ec.europa.eu/research/endocrine/pdf/drinking_water_en.

pdfS.

Westerhoff, P., Prapaipong, P., Shock, E., Hillahireau, A.,

2008. Antimony leaching from polyethylene terephthalate

(PET) plastic used for bottled drinking water. Water Res.

42, 552–556.

WHO, 2000. Bottled drinking water. Fact sheet no. 256, October

2000. Available at /www.who.int/inf-fs/fact256.htmlS.