Alkylphenol xenoestrogens with varying carbon chain lengths differentially and potently activate signaling and functional responses in GH3/B6/F10 somatomammotropes

Abstract

Background: Alkylphenols varying in their side-chain lengths [ethyl-, propyl-, octyl-, and nonylphenol (EP, PP, OP, and NP, respectively)] and bisphenol A (BPA) represent a large group of structurally related xenoestrogens that have endocrine-disruptive effects. Their rapid nongenomic effects that depend on structure for cell signaling and resulting functions are unknown.

Objectives: We compared nongenomic estrogenic activities of alkylphenols with BPA and 17beta-estradiol (E(2)) in membrane estrogen receptor-alpha-enriched GH3/B6/F10 pituitary tumor cells. These actions included calcium (Ca) signaling, prolactin (PRL) release, extracellular-regulated kinase (ERK) phosphorylation, and cell proliferation.

Methods: We imaged Ca using fura-2, measured PRL release via radioimmunoassay, detected ERK phosphorylation by fixed cell immunoassay, and estimated cell number using the crystal violet assay.

Results: All compounds caused increases in Ca oscillation frequency and intracellular Ca volume at 100 fM to 1 nM concentrations, although long-chain alkylphenols were most effective. All estrogens caused rapid PRL release at concentrations as low as 1 fM to 10 pM; the potency of EP, PP, and NP exceeded that of E(2). All compounds at 1 nM produced similar increases in ERK phosphorylation, causing rapid peaks at 2.5-5 min, followed by inactivation and additional 60-min peaks (except for BPA). Dose-response patterns of ERK activation at 5 min were similar for E2, BPA, and PP, whereas EP caused larger effects. Only E2 and NP increased cell number. Some rapid estrogenic responses showed correlations with the hydrophobicity of estrogenic molecules; the more hydrophobic OP and NP were superior at Ca and cell proliferation responses, whereas the less hydrophobic EP and PP were better at ERK activations.

Conclusions: Alkylphenols are potent estrogens in evoking these nongenomic responses contributing to complex functions; their hydrophobicity can largely predict these behaviors.

Keywords: ERK activation; bisphenol A; calcium oscillation; estradiol; hydrophobicity; non-genomic response; prolactin release.

Figure 1

Structural similarities among E₂, BPA, and para-substituted phenolic compounds [(CH₂)_n; n = 2, 3, 8, and 9 for the alkyl-phenols in our studies] with possible estrogenic activities. The phenol backbone of BPA and alkylphenols show structural similarity with the hydroxylated phenolic A-ring of E₂, which appears to be essential for ER binding (Tabira et al. 1999).

Figure 2

Rapid effect of E₂, BPA, and alkylphenols on cytosolic Ca oscillations. Time-dependent Ca levels (340:380 ratio) were recorded from a single representative responding cell for each tested compound. We added each estrogen at the 10-min time point at a concentration of 1 nM, as indicated by the arrows.

Figure 3

Quantitation of different aspects of estrogen-induced Ca responses. (A) The number of responding cells calculated as a percentage of total number of cells assessed in a given dish, recorded simultaneously (3–5 different dishes, 54–228 cells for each data point). (B) Changes in Ca oscillation frequency (average number of Ca peaks per minute, measured over 10 min) compared with the average baseline frequency before any estrogen treatment (Con). n = 11–33 individual cells for each point. (C) The integral of the Ca response to estrogens or vehicle-treated controls (Con), measured as the total area under the Ca peaks, recorded over 10 min. We subtracted the baseline pretreatment Ca levels. *p < 0.05 compared with vehicle control. #p < 0.05 compared with E₂

Figure 4

Concentration dependence of estrogen-induced rapid changes in PRL secretion. We measured PRL released into the medium by RIA after 1 min of treatment with E₂, BPA, EP, PP, OP, or NP at different concentrations (n = 12–24 for each data point). *p < 0.05 versus vehicle control (Con; n = 32).

Figure 5

Changes in ERK phosphorylation in response to E₂, BPA, EP, PP, OP, and NP. Values are the amount of pNP generated for each well, normalized to the CV value for cell number, presented as a percentage of vehicle-treated controls (Con), which we set to 100. (A) Time-dependent changes in the phosphorylation status of ERKs due to 1 nM estrogen treatments (n = 24 samples over three experiments). (B) Concentration-dependent changes in the phosphorylation status of ERKs after 5-min estrogen treatments (n = 32 samples over four experiments). *p < 0.05 versus control.

Figure 6

Proliferative response to E₂, BPA, and alkylphenols measured by the CV assay. We grew cells in the presence of different concentrations of E₂ or xenoestrogens for 3 days. Proliferation at day 3 is cell number as a percentage of vehicle control (Con). n = 24 samples over three experiments. *p < 0.05 versus control.

Figure 7

Rapid signaling responses elicited by various alkylphenols, BPA, and E₂, by increasing hydrophobicity. Responses are ordered by hydrophobicity indices; the numbers next to each symbol are carbon atoms in each compound’s side chain (where applicable). (A) Ca oscillation frequency (averaged for 1 fM to 1 nM treatments). *p < 0.05 compared with E₂, EP, and PP. (B) Integral (volume) of the Ca response (averaged for 1 fM to 1 nM treatments). *p < 0.05 compared with E₂ and PP. (C) PRL secretion activated by 100 pM treatment. *p < 0.05 compared with PP and NP. (D and E) ERK activation response to 1 nM treatments measured at 5 and 60 min, respectively. *p < 0.05 compared with OP; ^#p < 0.05 compared with NP. (F) Proliferative response (averaged for 10 pM to 10 nM treatments). *p < 0.05 compared with all other tested compounds; ^#p < 0.05 compared with EP and PP for OP, compared with PP for BPA.