In Vitro Transformation of Chlorinated Parabens by the Liver S9 Fraction: Kinetics, Metabolite Identification, and Aryl Hydrocarbon Receptor Agonist Activity

Abstract

We investigated the kinetics of in vitro transformation of a dichlorinated propyl paraben (2-propyl 3,5-dichloro-4-hydroxybenzoate; Cl₂PP) by the rat liver S9 fraction and assessed the aryl hydrocarbon receptor (AhR) agonist activity of the metabolite products identified in HPLC and GC/MS analysis and by metabolite syntheses. The results indicated that the chlorination of Cl₂PP reduced its degradation rate by approximately 40-fold. Two hydroxylated metabolite products showed AhR agonist activity of up to 39% of that of the parent Cl₂PP when assessed in a yeast (YCM3) reporter gene assay. The determination of the metabolic properties of paraben bioaccumulation presented here provides further information on the value of risk assessments of chlorinated parabens as a means to ensure human health and environmental safety.

Alkyl esters of 4-hydroxybenzoic acid (parabens) are widely used as preservatives in pharmaceuticals and personal care products, because they are stable over a wide pH range, are sufficiently soluble in water, and also exhibits a broad spectrum of antimicrobial activity.¹⁾ In Japan, parabens are permitted for use in all cosmetic products at concentrations up to 1% (w/w).

Compounds, including parabens, that contain phenolic hydroxyl groups exhibit favorable chlorination kinetics; thus, parabens are easily chlorinated by chlorine in tap water and by sodium hypochlorite, a commonly used bleaching agent.²⁾ In recent years, the increasing use of parabens has led to concerns related to the potential for adverse effects associated with their widespread occurrence in the environment. Several monitoring studies have now detected chlorinated parabens in effluents, pool water, and even river water.^3,4)

Organochlorine compounds such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and polychlorinated biphenyls (PCBs) are hydrophobic and persist in the environment for significant periods of time. This exposure and persistence may adversely affect humans and other organisms. The toxicity of TCDD and PCBs is associated with their interaction with the aryl hydrocarbon receptor (AhR), which is a ligand-dependent transcription factor.⁵⁾ Activation of AhR by TCDD and dioxin-like compounds can induce various biological responses, including activation of the CYPs system (particularly CYP1A1) and disruption of normal hormone signaling pathways, and can lead to developmental and reproductive toxicities, immunotoxicity, and mutagenicity.⁶⁾ Recently, chlorinated parabens were suggested to effect toxicity through interaction with AhR.⁷⁾ Parabens are hydrolyzed mainly to 4-hydroxybenzoic acid in mammals; this metabolite can thereafter be conjugated with sulfate, glucuronide, or glycine, prior to excretion in urine.⁸⁾ However, not all metabolite products and metabolic reactions that can arise from paraben exposure in mammals have been chemically determined. Moreover, the biological activity especially the endocrine disrupting properties, of these metabolites, are most often not known. The main reason for this is the lack of chemical standards to support identification of all metabolite products arising from bio-transformations of parabens.

The aim of the present study was to carry out an assessment of the metabolic kinetics of in vitro transformation of chlorinated paraben by a rat liver S9 fraction and to identify associated metabolite products and their toxicities. AhR agonists only need to meet minimal requirements in hydrophobicity, size, and planar shape to fit into the receptor-binding pocket to induced endocrine disrupting properties.^9,10) Here, we focused particularly on paraben metabolites containing a phenolic hydroxyl group. AhR agonist activity was measured using a yeast reporter gene assay. For our model compound, we selected the dichlorinated derivative (Cl₂PP) of propyl paraben, the most frequently used paraben in pharmaceuticals and personal care products (Fig. 1). Cl₂PP has been detected at levels up to 0.18 nM in pool and river water in Japan.^3,7) Cl₂PP shows AhR agonist activity (EC₂₅=1300 nM) and more acute toxicity toward invertebrates (EC₅₀=33 µM) than that by the corresponding parent paraben.^7,11) Cl₂PP is a common and toxic paraben in the water environment and it is therefore important to evaluate its metabolic fate.

Fig. 1. Chemical Structure for Dichlorinated Propyl Paraben (Cl₂PP) and Its Transformed Products (1)–(6) by a Rat Liver S9 Fraction

Results and Discussion

Kinetics

To investigate the degradation rate of propyl paraben and its chlorinated derivative ln(C/C₀) values were plotted against reaction time, as shown in Fig. 2. Correlation coefficients (r²) were 0.864 and 0.915, respectively. From these values, it was established that the metabolic reaction for the paraben and its chlorinated derivative is first order. The rate constant k and the half-life t_1/2 for the paraben were 0.800 per min and 0.87 min, respectively; for the chlorinated derivative, the corresponding values were 0.0195 per min and 36 min, respectively. The degradation rate of chlorinated paraben was 1/40-fold that of the paraben in the rat liver S9 fraction treatment; the results suggested that chlorination of the paraben may enhance accumulation of the paraben in the biota.

Fig. 2. Time–Course Plots of Propyl Paraben (PP) and Cl₂PP in the Presence of the Liver S9

Syntheses and Identification of Metabolites

The reaction products of Cl₂PP after exposure to in vitro transformation by a rat liver S9 fraction were assessed by GC/MS. The total ion chromatogram (TIC) and the MS data for trimethylsilyl (TMS) derivatives of the reaction products, and the associated synthetic compounds (which fulfill structural criteria for endocrine disrupting properties) are shown in Figs. 3 and 4. In the MS spectrum of peak 2, the molecular ion peaks (M⁺) at m/z 374 and m/z 376 presented the typical isotopic pattern of a mono-chlorinated compound and indicated the substitution of an –OH group with a Cl atom in the TMS derivative of Cl₂PP. We therefore identified peak 2 as propyl 3-chloro-4,5-dihydroxybenzoate (2) and completed its synthesis (see Experimental).

Fig. 3. GC/MS Total Ion Chromatograms (TIC) of the Extractives of Metabolites of 60 min Metabolic Reaction

Fig. 4. Mass Spectra and Retention Time (R.T.) of Metabolites 2–6

Numbers refer to the peak numbers of Fig. 3.

The molecular ion peaks (M⁺) at m/z 408, observed in the spectra of peaks, 3, 5, and 6 indicated the substitution of an –OH group with a H atom in the TMS derivative of Cl₂PP. The MS spectrum of peak 4 was dominated by the mass fragment ions at m/z [M⁺−15], resulting from the loss of a methyl moiety in the derivatives rather than by the molecular ion peaks (M⁺). Each of the ion peaks at [M⁺−15] and (M⁺) for all products also presented the typical isotopic pattern corresponding to dichlorinated compounds. We therefore completed the synthesis of four compounds that fit these analytical data. These were isopropyl 3,5-dichloro-2,4-dihydroxybenzoate (3), 2-hydroxypropyl 3,5-dichloro-4-hydroxybenzoate (4), 2-hydroxy-1-methylethyl 3,5-dichloro-4-hydroxybenzoate (5) and 3-hydroxypropyl 3,5-dichloro-4-hydroxybenzoate (6).

The identification and quantification of peaks on the TIC was confirmed by comparing their gas chromatography (GC) retention times, mass spectra, and the peak area to those of commercial (1) or synthetic standards (2)–(6). The relative abundances of metabolites are 48% for (1), 32% for (5), 8.7% for (2), 5.7% for (6), 3.3% for (4), and 2.0% for (3). Maximum abundances of individual metabolite were observed after 120 min reaction with the rat liver S9 fraction.

AhR Agonist Activity of Metabolites

Table 1 shows the AhR agonist activities of the metabolites as measured using the yeast reporter gene assay. Two hydroxylated metabolite products (4) and (6) of the dichlorinated propyl paraben showed AhR agonist activity, when assessed in the yeast (YCM3) reporter gene assay, of up to 18 to 39% of that of Cl₂PP.

Table 1. AhR Activities of Cl₂PP and Its Metabolites (1)–(6) in the Yeast Reporter Gene Assay

Cmpd	AhR assay
	EC25 (nM)	RAb)
1	Not active
2	Not active
3	Not active
4	3400±1400	0.39
5	Not active
6	7800±300	0.18
Cl2PP	1900±230c)	1.0
BNFa)	6.84±1.2c)	520

a) Positive control. b) Relative activity for Cl₂PP. c) Data reported in ref. 7.

The S9 rat liver preparation contains a range of enzyme activities, which may eliminate AhR agonist activity through, for example, conjugating phenolic hydroxyl groups (e.g., by glucuronidation) and/or hydroxylating aromatic rings or side-chains. Conjugation of the phenolic hydroxyl group and hydroxylation of the aromatic ring eliminates AhR agonist activity but the hydroxylation of the side chain induces agonist activity. Our results indicated that the AhR agonist activity of Cl₂PP remained after exposure to the rat liver S9 fraction.

The AhR agonist activity of halogenated aromatics is structure-dependent. The presence of a halogen on the aromatic ring, for example, is required for receptor activation, and factors that influence the planarity of the molecule can determine affinity for receptor binding.⁵⁾ In the metabolic reaction conducted in this, compounds (4) and (6) retained the 2,3-dichloro-4-hydroxy benzoic unit and the linear propyl side chain present in Cl₂PP; as such, they retained AhR agonist activity induced by Cl₂PP. On the other hand, hydroxylation at the aromatic ring as observed in compounds (2) and (3), as well transformation from a linear to branched-chain are believed to minimize planarity and thus reduce binding to AhR. Molecular docking of TCDD congeners to a model of the human AhR-ligand binding domain has suggested that hydrophobic contacts would strengthen the H-bond network, which in turn would energetically favor ligand binding.¹²⁾ Carboxylic groups also enhance polarizability and can favor interaction between the AhR and ligands as in the case of compound (1).

AhR is widely expressed in mammalian tissues and it affects a number of fundamental cellular processes. Modulation of AhR function by xenobiotics, even at low concentrations, may ultimately result in a variety of adverse effects such as hepatotoxicity, immunotoxicity, neurotoxicity, dermal toxicity, teratogenicity, and carcinogenicity in many different species.¹³⁾ Moreover, activation of AhR by xenobiotics modulates a number of endocrine systems, most notably by interfering with estrogen signaling. Consequently, there has been an extensive analysis of the role of AhR in estrogen-dependent breast cancer cells. Ohtake et al. investigated AhR-estrogen receptor (ER)α crosstalk in vitro and in vivo using 3-methylcholanthrene (3MC) as an AhR ligand.¹⁴⁾ Inhibitory effects were observed for some responses when 17β-estradiol (E2) and 3MC were co-administered; it was concluded that 3MC inhibited E2-induced estrogenic responses. Another xenobiotic, the polyaromatic hydrocarbon 6-methyl-1,3,8-trichlorodibenzofuran (MCDF), which is related to the industrial by-product dioxin, has been shown to be a weak agonist of AhR when tested in a murine reporter cell line (0.1% of the activity of β-naphthoflavone (BNF)).¹⁵⁾ Studies in vitro have demonstrated that MCDF has a proteasome-dependent degradative effect on ER protein levels.^16,17) Therefore, our current data warrant further studies to investigate AhR-ER crosstalk in the presence of Cl₂PP. Some chlorinated parabens occur widely in aquatic environments and, as such, organisms may be continuously exposed. The determination of the metabolic properties of chlorinated parabens can provide information on their potential for bioaccumulation and whether a xenobiotic is likely to be converted to a more potent or less potent toxin. It is probable that risk assessments of chlorinated parabens to ensure human health and environmental safety will require analysis of all their metabolite products in the manner exemplified here.

Experimental

Materials

Propyl paraben was purchased from Tokyo Chemical Industry, Tokyo, Japan. Cl₂PP was synthesized using previously reported procedures (purity (GC)>99%).⁴⁾ The male Sprague-Dawley rat liver S9 mix induced with phenobarbital and 5,6-benzoflavone was purchased from Kikkoman Corporation, Noda, Japan. D-Glucose 6-phosphate disodium salt (G-6-P-Na₂) and β-nicotinamide-adenine dinucleotide phosphate, oxidized form, monosodium salt (β-NADP⁺-Na) were purchased from Oriental Yeast Company Limited, Osaka, Japan. Dimethyl sulfoxide (DMSO) was purchased from Dojindo Laboratories, Kumamoto, Japan and was of fluorometric grade. N,O-Bis(trimethylsilyl)-trifluoroacetamide (BSTFA) was purchased from Wako Pure Chemical Industries, Ltd., Osaka, Japan. All chemicals used were of analytical grade or HPLC grade.

Propyl 3-chloro-4,5-dihydroxybenzoate (2) was synthesized via 3 steps from chlorinated hydroxybenzoate (a–c in Chart 1). Isopropyl 3,5-dichloro-2,4-dihydroxybenzoate (3) was synthesized via the 2 steps from commercially available hydroxybenzoic acid (d–e in Chart 1). 2-Hydroxypropyl 3,5-dichloro-4-hydroxybenzoate (4) was synthesized by esterification of 3,5-dichloro-4-hydroxybenzoic acid (1) suspended in benzene with 1-bromo-2-propanol in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) as the catalyst under an atmosphere of nitrogen (f in Chart 1). 2-Hydroxy-1-methylethyl 3,5-dichloro-4-hydroxy-benzoate (5) was synthesized by chemoselective benzoylation of 1,2-propanediol using 1.5 M equivalent of triphenylphosphine (Ph₃P) and diethyl azodicarboxylate (DEAD)¹⁸⁾ (g in Chart 1). 3-Hydroxypropyl 3,5-dichloro-4-hydroxy-benzoate (6) was synthesized by a transesterification of chlorinated 4-hydroxybenzoate¹⁹⁾ (h in Chart 1). Analytical data for synthetic metabolites and their intermediate substances are as follows:

Chart 1

(a) HNO₃/AcOH, r.t., 30 min; (b) Fe/NH₄Cl, EtOH, reflux, 6 h; (c) HCl, NaNO₂, H₂O, 0°C/heat; (d) SO₂Cl₂/AcOH, r.t., 1 h; (e) 2-PrOH/HCl, reflux, 6 d; (f) DBU, benzene, reflux, 24 h; (g) Ph₃P, DEAD, THF, r.t., 14 h; (h) Zn(CH₃COO)₂, 90°C, 2 h.

Compound (2b)

Yield 89%. ¹H-NMR (CDCl₃) δ: 1.04 (3H, t, J=7.5 Hz), 1.77–1.84 (2H, m), 4.30 (2H, t, J=6.8 Hz), 8.28 (1H, d, J=2.5 Hz), 8.60 (1H, d, J=2.0 Hz). MS m/z: 261 (M⁺+2), 259 (M⁺).

Compound (2c)

Yield 96%. ¹H-NMR (500 MHz, CDCl₃) δ: 1.02 (3H, t, J=7.5 Hz), 1.72–1.79 (2H, m), 4.20 (2H, t, J=10 Hz), 6.72 (1H, d, J=8.0 Hz), 7.40 (1H, d, J=2.0 Hz). MS m/z: 231 (M⁺+2), 229 (M⁺).

Compound (2)

Yield 57%. Purity (GC)>99% ¹H-NMR (CDCl₃) δ: 0.93 (3H, t, J=7.3 Hz), 1.63–1.70 (2H, m), 4.11 (2H, t, J=6.8 Hz), 7.26 (1H, d, J=1.5 Hz), 7.37 (1H, d, J=1.5 Hz). High resolution-time-of-flight (HR-TOF)-MS (M⁺−H): 229.0348 (Calcd for C₁₀H₁₀O₄³⁵Cl: 229.0268), 231.0313 (Calcd for C₁₀H₁₀O₄³⁷Cl: 231.0238).

Compound (3b)

Yield 63%. ¹H-NMR (CDCl₃) δ: 7.80 (1H, s).

Compound (3)

Yield 12%. Purity (GC)>99% ¹H-NMR (CDCl₃) δ: 1.39 (6H, d, J=6.3 Hz), 5.24–5.32 (1H, sept, J=1.3 Hz), 7.78 (1H, s). HR-TOF-MS (M⁺–H): 262.9872 (Calcd for C₁₀H₉O₄³⁵Cl₂: 262.9870), 264.9844 (Calcd for C₁₀H₉O₄³⁵Cl³⁷Cl: 264.9840), 266.9818 (Calcd for C₁₀H₉O₄³⁷Cl₂: 266.9820).

Compound (4)

Yield 67%. Purity (GC)>99% ¹H-NMR (CDCl₃) δ: 1.26 (3H, d, J=5.5 Hz), 4.12–4.30 (3H, m), 7.94 (2H, s). HR-TOF-MS (M⁺−H): 262.9842 (Calcd for C₁₀H₉O₄³⁵Cl₂: 262.9870), 264.9841 (Calcd for C₁₀H₉O₄³⁵Cl³⁷Cl: 264.9840), 266.9812 (Calcd for C₁₀H₉O₄³⁷Cl₂: 266.9820).

Compound (5)

Yield 51%. Purity (GC) 97% ¹H-NMR (CDCl₃) δ: 1.48 (3H, d, J=6.5 Hz), 3.77–3.85 (2H, m), 5.22–5.29 (1H, m, J=6.0 Hz), 8.00 (2H, s). HR-electron ionization (EI)-MS (M⁺): 263.9955 (Calcd for C₁₀H₁₀O₄³⁵Cl₂: 263.9956), 265.9928 (Calcd for C₁₀H₁₀O₄³⁵Cl³⁷Cl: 265.9927), 267.9904 (Calcd for C₁₀H₁₀O₄³⁷Cl₂: 267.9897).

Compound (6)

Yield 55%. Purity (GC)>99% ¹H-NMR (CDCl₃) δ: 2.00 (2H, t, J=3.3 Hz), 3.72–3.77 (2H, q, J=6.0 Hz), 4.45–4.48 (2H, t, J=6.0 Hz), 7.94 (2H, s). HR-TOF-MS (M⁺–H): 262.9879 (Calcd for C₁₀H₉O₄³⁵Cl₂: 262.9870), 264.9849 (Calcd for C₁₀H₉O₄³⁵Cl³⁷Cl: 264.9840), 266.9810 (Calcd for 266.9820: C₁₀H₉O₄³⁷Cl₂).

Yeast Reporter Gene Assay

AhR agonist activity of compounds was measured using a reporter assay based on the recombinant yeast YCM3 strain.²⁰⁾ After the yeast cells were pre-incubated at 30°C overnight in a modified synthetic dextrose medium lacking tryptophan and leucine, a solution (1.25 µL) of the test compound in DMSO was mixed with 125 µL of synthetic galactose medium (2% galactose, lacking tryptophan) and 1.25 µL of yeast cell suspension and then was incubated for 18 h at 30°C in a 96-well culture plate.⁷⁾ After incubation, cell density was determined by reading the absorbance at 600 nm (A₆₀₀) using a microplate reader. The cells (10 µL) were digested by zymolyase 20 T, and then chlorophenol red-β-galactopyranoside (CPRG) was added. The colorimetric reaction was stopped by addition of Na₂CO₃. β-Galactosidase activity was assessed by measuring the absorbance at 578 nm (A₅₇₈) and calculated using the equation: β-galactosidase activity={(1000×A₅₇₈)/[A₆₀₀×0.01 (cell volume, mL)×reaction time (min)]}.

BNF (β-naphthoflavone) was used as a positive control. AhR agonist activity was recorded as the 25% effective concentration of the maximal response (EC₂₅). The dose–response curve of each compound was derived using logistic regression analysis with the following equation: Y=Bottom+(Top−Bottom)/(1+Ks/X^slope), where X is the sample concentration (nM) or sample concentration ratio, Y is the response (from bottom to top, with a sigmoid shape), Ks is the substrate constant, and slope is the Hill coefficient. Ks and slope were determined by the Solver function of Microsoft Excel 2010.

Statistical Analysis

Data are expressed as the mean±standard deviation (S.D.) from a minimum of three assays performed in duplicate.

Metabolic Reaction and Extraction

DMSO solution (2 mL) of Cl₂PP was incubated with 48 mL of the S9 mix solution (2.5 mL of S9 mix, 35 mg of G-6-P-Na₂, 70 mg of β-NADP⁺-Na, 50 mL of Z buffer, and 200 µL of N-(2-hydroxyethyl)piperazine-N′-2-ethanesulfonic acid (HEPES) buffer, pH 7.2) at 37°C and 100 rpm. Its initial concentration (C₀) Cl₂PP was 100 µM. Reaction samples (100 µL) were filtered through a 0.45 µm membrane filter (Minisart RC, Sartorius Japan K.K., Tokyo) and were subjected to HPLC for Cl₂PP quantification. Reaction samples (2 mL) also were adjusted to pH 2 with 2 M hydrochloric acid and extracted in ethyl acetate (3×1.2 mL). The ethyl acetate extracts were combined and dried over anhydrous sodium sulfate. After removing the solvent under reduced pressure, 10 µL of BSTFA and 100 µL of ethyl acetate were added to the test tubes containing the extracts, and the solution was allowed to stand for 1 h at room temperature to complete trimethylsilylation. After the reaction was complete, samples were examined using GC/MS with a mass-selective detector (HP 5973N Series; Agilent Technologies, Palo Alto, CA, U.S.A.) on an HP-5 fused-silica capillary column (J&W Scientific, Folsom, CA, U.S.A.), according to our previously described method.⁴⁾

Acknowledgment

This work was supported by the Grant-in-Aid for Scientific Research (C) No. 26410195 from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.

Conflict of Interest

The authors declare no conflict of interest.

References

• 1) Soni M. G., Carabin I. G., Burdock G. A., Food Chem. Toxicol., 43, 985–1015 (2005).
• 2) Gallard H., von Gunten U., Environ. Sci. Technol., 36, 884–890 (2002).
• 3) Terasaki M., Takemura Y., Makino M., Environ. Chem. Lett., 10, 401–406 (2012).
• 4) Terasaki M., Makino M., J. Health Sci., 55, 631–635 (2009).
• 5) Nguyen L. P., Bradfield C. A., Chem. Res. Toxicol., 21, 102–116 (2008).
• 6) Giesy J. P., Kannan K., Crit. Rev. Toxicol., 28, 511–569 (1998).
• 7) Terasaki M., Yasuda M., Makino M., Shimoi K., Environ. Sci. Water Res. Technol., 1, 375–382 (2015).
• 8) Ye X., Bishop A. M., Reidy J. A., Needham L. L., Calafat A. M., Environ. Health Perspect., 114, 1843–1846 (2006).
• 9) Esser C., Rannug A., Stockinger B., Trends Immunol., 30, 447–454 (2009).
• 10) Kawanishi M., Kondo M., Shiizaki K., Chu W. L., Terasoma Y., Yagi T., Environ. Sci. Technol., 42, 6897–6902 (2008).
• 11) Terasaki M., Makino M., Tatarazako N., J. Appl. Toxicol., 29, 242–247 (2009).
• 12) Xing Y., Nukaya M., Satyshur K. A., Jiang L., Stanevich V., Korkmaz E. N., Burdette L., Kennedy G. D., Cui Q., Bradfield C. A., Toxicol. Sci., 129, 86–97 (2012).
• 13) Hankinson O., Annu. Rev. Pharmacol. Toxicol., 35, 307–340 (1995).
• 14) Ohtake F., Baba A., Takada I., Okada M., Iwasaki K., Miki H., Takahashi S., Kouzmenko A., Nohara K., Chiba T., Fujii-Kuriyama Y., Kato S., Nature (London), 446, 562–566 (2007).
• 15) Schulz V. J., Smit J. J., Huijgen V., Bol-Schoenmakers M., van Roest M., Kruijssen L. J., Fiechter D., Hassing I., Bleumink R., Safe S., van Duursen M. B., van den Berg M., Pieters R. H., Toxicol. Sci., 128, 92–102 (2012).
• 16) Pearce S. T., Liu H., Radhakrishnan I., Abdelrahim M., Safe S., Jordan V. C., Cancer Res., 64, 2889–2897 (2004).
• 17) Zhang S., Kim K., Jin U. H., Pfent C., Cao H., Amendt B., Liu X., Wilson-Robles H., Safe S., Mol. Cancer Ther., 11, 108–118 (2012).
• 18) Mitsunobu O., Kimura J., Iizumi K., Yanagida N., Bull. Chem. Soc. Jpn., 49, 510–513 (1976).
• 19) Hanayama N., Sakanashi K., Wakamatsu S., Patent WO 9218459 A1 (1992).
• 20) Miller C. A. 3rd, Toxicol. Appl. Pharmacol., 160, 297–303 (1999).