Because women are exposed to environmental estrogens during life phases when different physiologic estrogens are prevalent, we studied the changes in pERK when each physiologic and environmental estrogen was present simultaneously. We previously detailed ERK (and other) responses of these cells to the physiologic estrogens E1
, and E3
(Watson et al. 2008
); those data can be directly compared with results of the present study. First, we examined the time-dependent changes in ERKs (–), showing that estrogenic stimulation caused a characteristic oscillating pattern with immediate (5 min), intermediate (10–30 min), and long-term (after 30 min) rises in ERK activation, similar to estrogen-induced fluctuations we reported previously (Bulayeva et al. 2004
; Bulayeva and Watson 2004
; Jeng et al. 2009
; Jeng and Watson 2009
; Kochukov et al. 2009
; Zivadinovic and Watson 2005
). We observed these oscillating patterns for all estrogens, although some were “trends” with peaks that did not achieve significance. The physiologic estrogens E2
, as well as BPA, tended to cause three oscillations during this 60-min time frame, whereas alkylphenols and E3
caused only two (missing the intermediate peak).
Figure 1 Time-dependent changes in pERK elicited by combinations of short-chain alkylphenols with E2, E1, or E3. GH3/B6/F10 cells were cotreated with 1 nM EP (A–C) or PP (D–F) and 1 nM E2 (A,D), E1 (B,E), or E3 (C,F). The pERK levels were measured (more ...)
Figure 3 Time-dependent changes in pERK elicited by combinations of BPA at two different environmentally relevant concentrations. GH3/B6/F10 cells were cotreated with 10−14 M BPA (A–C) or 10−9 M BPA (D–F) plus 1 nM E2 (A,D), E1 (more ...)
When pituitary cells were cotreated with XEs plus each of the physiologic estrogens (E2, E1, or E3) in combination, the usual 2.5- to 5-min pERK peak caused by individual estrogens was often abolished or blunted. Instead, the combination of compounds usually caused a pronounced early dephosphorylation and then created a new time-delayed, augmented phosphorylation peak, just as the actions of individual estrogens waned. These new large activations (although a bit smaller in the PP, OP, and NP combinations with E3) peaked instead in the intermediate 10- to 30-min time frame, demonstrating how combinations of estrogens with XEs that individually do not elicit a significant response at a given time point can cause a synergistic response. Then, as the ERK activation induced by individual estrogens again rose at 60 min, the level due to the combined estrogens usually declined, often far below the response to the individual estrogens. (The exceptions to this observation were combinations of PP plus E2 and 10−9 M BPA plus E1 or E2, when the intermediate pERK peak was sustained.) Although the later (30–60 min) response period was usually inhibited, it depended somewhat on the efficacy of the XE at that time point. That is, stronger individual XE responses (resembling physiologic estrogen responses) tended to predict the ability of the XE to inhibit the actions of physiologic estrogens when in combination. Thus, a three-peaked oscillation caused by physiologic estrogens was transformed to a single major intermediate peak of activation in most cases.
We used two doses of BPA to determine the effects on ERK activation caused by physiologic estrogens () in our temporal phase studies because BPA is a compound of high interest to the endocrine toxicology community due to its ubiquity and the controversy about its potential effects, and because both of the chosen concentrations are in the range of typical and prevalent exposures. More important, many estrogens, especially BPA, cause nonmonotonic nongenomic dose–response patterns, typically having multiple dose optima. Interestingly, the very low concentration of BPA (10−14 M) caused a two-phased oscillating response rather like the alkylphenol XEs ( and ). However, 1 nM BPA elicited a response more typical of the physiologic estrogen pattern (with three phases of activation).
Figure 2 Time-dependent changes in pERK elicited by combinations of long-chain alkylphenols with E2, E1, or E3. GH3/B6/F10 cells were cotreated with 1 nM OP (A–C) or NP (D–F) and 1 nM E2 (A,D), E1 (B,E), or E3 (C,F). The pERK levels were measured (more ...)
Although we could choose a single health-relevant level of physiologic estrogens based on their normal concentration ranges in animals and humans (Greenspan and Gardner 2004
), we felt it necessary to test many concentrations of the XEs to adequately describe the effects they might have at different common contamination levels, and to account for their nonmonotonic behavior. We chose the 5-min time point for this study because of its prominent and consistent appearance in all physiologic estrogen and XE treatments (–) and because its rapid time frame ensured that it would result in a distinctly nongenomic response.
For most alkylphenols acting alone (EP, PP, and NP; and ), significant estrogenic effects occurred at the higher doses (usually picomolar or higher concentrations), but OP showed more activity at the lower doses. The short-chain EP caused greater stimulations than did all physiologic estrogens. PP caused marked nonmonotonic dose responses, with effective doses peaking in both approximately picomolar and nanomolar ranges. Alkylphenols generally enhanced physiologic estrogen responses at lower doses but severely disrupted them at higher doses to far below vehicle control levels. EP was estrogenic over most of the tested concentration range and inhibited its paired physiologic estrogens throughout. PP was significantly estrogenic at two concentration ranges and lowered the estrogenicity of E2 and E1 beginning at the concentration at which it became estrogenic itself (approximately picomolar). However, PP enhanced the estrogenicity of E3 at all but the lowest tested concentrations (). This last result was the most significant departure from the general pattern and illustrates why each compound must be tested across its entire range to reveal such exceptions.
Figure 4 Concentration-dependent changes in pERK caused by short-chain alkylphenols. Cells were treated for 5 min with a combination of 1 nM E2 (A,D), E1 (B,E), or E3 (C,F) plus different concentrations of EP (A–C) or PP (D–F), and pERK was assayed. (more ...)
Figure 5 Concentration-dependent changes in pERK caused by the long-chain alkylphenols. Cells were treated for 5 min with a combination of 1 nM E2 (A,D), E1 (B,E), or E3 (C,F) plus different concentrations of OP (A–D) or NP (D–F), and pERK was (more ...)
BPA showed the most striking nonmonotonicity in its dose–response pattern (), as we have seen previously in this and other cell systems (Alyea and Watson 2009a
; Watson et al. 1999
). The lowest BPA doses were very estrogenic, followed by a (generally) 2 log dose range of ineffective concentrations (10–100 pM), followed finally by another dose range of estrogenic activity (at nanomolar or higher concentrations). Remarkably, at whatever doses BPA was most estrogenic, it had the most marked inhibitory effect on the estrogenicity of the paired physiologic estrogen. At doses where BPA was ineffective, the estrogenicity of the physiologic estrogen was spared (close to normal or, for E2
, at least clearly above vehicle control values) or even enhanced (for E1
Figure 6 Concentration-dependent alteration of physiologic estrogen-induced pERK by BPA. Cells were treated for 5 min with different concentrations of BPA with or without 1 nM E2 (A), E1 (B), or E3 (C). The blue horizontal bar indicates the pERK level and error (more ...)
To further characterize this response caused by all estrogens, we examined in more detail which of several ER proteins were responsible (). We again examined ERK activation after 5 min, as a distinct nongenomic response indicator. The selective ERα agonist PPT increased pERK levels over a wide range of concentrations (as low as 10−14
M), including those highly selective for ERα. The ERβ agonist DPN had no effect at any selective concentration; it is known to be rather nonselective at higher levels (Meyers et al. 2001
). The selective GPER agonist G1 increased pERK levels at relatively low concentrations (approximately picomolar) but decreased pERK at higher than nanomolar concentrations.
Figure 7 Concentration dependence of pERK (A–D) and proliferation (E–H) on selective ER agonists. pERK was measured in cells after 5 min of treatment with different concentrations of E2 (A), the ERα agonist PPT (B), the ERβ agonist (more ...)
Because ERK activation is often associated with the proliferative response, we also used these same receptor-selective compounds to evaluate their ability to evoke cell proliferation after 3 days of treatment (). At selective concentrations, both PPT and DPN caused cell proliferation, whereas G1 caused a decrease in cell numbers at all concentrations ≥ 100 fM. In our side-by-side comparison, ERK activation correlated with the ability to affect proliferation reasonably well for PPT and negatively for G1 but did not correlate well with cell proliferation caused by DPN (which also did not selectively activate ERK). Because E2 activates all forms of ERs, it has a composite (positive) profile for inducing cell proliferation.