Real-Time Growth Kinetics Measuring Hormone ... - ACS Publications

Article pubs.acs.org/crt

Real-Time Growth Kinetics Measuring Hormone Mimicry for ToxCast Chemicals in T‑47D Human Ductal Carcinoma Cells Daniel M. Rotroff,†,‡ David J. Dix,‡ Keith A. Houck,‡ Robert J. Kavlock,‡ Thomas B. Knudsen,‡ Matthew T. Martin,‡ David M. Reif,‡ Ann M. Richard,‡ Nisha S. Sipes,‡ Yama A. Abassi,§ Can Jin,§ Melinda Stampfl,§ and Richard S. Judson*,‡ †

Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, North Carolina 27514, United States ‡ Office of Research and Development, National Center for Computational Toxicology, United States Environmental Protection Agency, Research Triangle Park, North Carolina 27711, United States § ACEA Biosciences, Inc., 6779, Mesa Ridge Road, San Diego, California 92121, United States S Supporting Information *

ABSTRACT: High-throughput screening (HTS) assays capable of profiling thousands of environmentally relevant chemicals for in vitro biological activity provide useful information on the potential for disrupting endocrine pathways. Disruption of the estrogen signaling pathway has been implicated in a variety of adverse health effects including impaired development, reproduction, and carcinogenesis. The estrogen-responsive human mammary ductal carcinoma cell line T-47D was exposed to 1815 ToxCast chemicals comprising pesticides, industrial chemicals, pharmaceuticals, personal care products, cosmetics, food ingredients, and other chemicals with known or suspected human exposure potential. Cell growth kinetics were evaluated using real-time cell electronic sensing. T-47D cells were exposed to eight concentrations (0.006−100 μM), and measurements of cellular impedance were repeatedly recorded for 105 h. Chemical effects were evaluated based on potency (concentration at which response occurs) and efficacy (extent of response). A linear growth response was observed in response to prototypical estrogen receptor agonists (17β-estradiol, genistein, bisphenol A, nonylphenol, and 4-tert-octylphenol). Several compounds, including bisphenol A and genistein, induced cell growth comparable in efficacy to that of 17β-estradiol, but with decreased potency. Progestins, androgens, and corticosteroids invoked a biphasic growth response indicative of changes in cell number or cell morphology. Results from this cell growth assay were compared with results from additional estrogen receptor (ER) binding and transactivation assays. Chemicals detected as active in both the cell growth and ER receptor binding assays demonstrated potencies highly correlated with two ER transactivation assays (r = 0.72; r = 0.70). While ER binding assays detected chemicals that were highly potent or efficacious in the T-47D cell growth and transactivation assays, the binding assays lacked sensitivity in detecting weakly active compounds. In conclusion, this cell-based assay rapidly detects chemical effects on T-47D growth and shows potential, in combination with other HTS assays, to detect environmentally relevant chemicals with potential estrogenic activity.

■

INTRODUCTION Many xenobiotic chemicals are able to interact with molecular targets in the endocrine system, making screening for potential endocrine disrupting compounds (EDCs) an important issue.1,2 One possible effect of endocrine disrupting chemicals is perturbation of cell growth through pathways linked to cell cycle regulation.3,4 Activation of the estrogen receptor (ER) signaling pathway, for example, is one possible mechanism that underlies cell proliferation in hormonally sensitive tissues such as mammary and endometrial tissue.4,5 Furthermore, the role of steroid hormones in the regulation of some mammary tumors has been well established and has motivated the development of estrogen pathway-based chemotherapeutics.6−9 © XXXX American Chemical Society

The United States Environmental Protection Agency (U.S. EPA) formed the Endocrine Disruptor Screening Program (EDSP) following the Food Quality Protection Act (FQPA) amendment of the Federal Food, Drug, and Cosmetic Act (FFDCA) in order to determine whether certain substances may have an effect in humans that is similar to an effect produced by a naturally occurring estrogen or other endocrine effects (www.epa.gov/endo/). EDSP is a tiered program requiring chemical manufacturers to submit or generate data for regulatory decision-making. The Tier 1 screening battery (T1S) consists of both in vitro and in vivo assays to test chemicals for their potential Received: March 19, 2013

A

dx.doi.org/10.1021/tx400117y | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Figure 1. RTCA cell growth assay. 96-well E-plates have electrodes at the bottom of each well. Cells are seeded, and the impedance signal is altered by the confluence or morphological changes of cells in each well. Increases in cell density correspond to increased impedance (Z).

suggests that this ToxCast assay could be useful in identifying potentially estrogenic compounds. The present study analyzes a ToxCast assay that monitors cell growth kinetics in the estrogen responsive, human mammary ductal carcinoma cell line T-47D. The T-47D cell line expresses ERα and low levels of ERβ, it also expresses androgen, glucocorticoid, and progesterone receptors. 14,15 Proliferative responses of this cell line have been shown to be highly correlated to estrogenic effects observed in uterotrophic assays.14 An inventory of 1815 unique environmentally relevant chemicals along with positive and negative reference chemicals were tested. Cell growth kinetics were assessed using real-time cell analysis (RTCA) on the xCELLigence system, which offers a more comprehensive evaluation of cellular growth kinetics compared to traditional in vitro viability and proliferation assays. The assay uses electronic microsensors located at the bottom of the cell culture well to detect changes in cell number, morphology, and adhesion through electrical impedance measurement at the electrode−solution interface (Figure 1). Although, mechanisms may impact cell growth in T-47D cells other than by ER activation, these cells readily respond to estrogens and may provide an important confirmation of ER activity from chemicals observed in other in vitro assays. Here, we compare the results of this cell-impedance assay with other ER related assays in order to better understand the role of xenobiotics in ER mediated cell-level activity and further characterize the ability of in vitro HTS assays to detect these interactions.

for estrogen, androgen, thyroid or steroidogenic modes of action. The Tier 2 assay battery consists of more complex in vivo tests. The initial list of 67 chemicals for EDSP testing was finalized in 2009, and T1S testing is currently nearing completion. There are thousands of chemicals subject to the T1S, and it is estimated that the T1S will cost ∼$1 million per chemical.10,11 Because of the amount of resources required to meet the current test guidelines, the U.S. EPA launched the EDSP21 project evaluating new computational or in silico models, and higher throughput in vitro assays to prioritize and eventually replace more resource-intensive T1S assays.10 Previous studies have demonstrated that in vitro assay data can be used to accurately classify compounds relative to in vivo tests for certain modes of action for endocrine disruption.12,13 Many of the assays under consideration for EDSP21 testing come from the U.S. U.S. EPA’s ToxCast, and the U.S. government interagency Tox21 projects. ToxCast and Tox21 using high-throughput in vitro and in silico technologies to screen and prioritize large numbers of chemicals for further hazard testing. In conjunction with prioritization, the goals of ToxCast are to guide targeted testing strategies and to build predictive systems models that can eventually replace costly, low-throughput testing in EDSP and other regulatory programs. These efforts will support regulatory decision-making and help to ensure that resources are being used to test the chemicals with the highest likelihood of causing adverse health outcomes. In order to develop an effective prioritization scheme, each input data source needs to be well characterized in terms of sensitivity, specificity, and other technical characteristics. The data analysis presented herein

■

EXPERIMENTAL PROCEDURES

Chemical Selection. This study was conducted using data from three ToxCast chemical libraries and reference compounds totaling 1,819 unique chemicals. This inventory includes chemical structures B



Article

from the phase Iv2 ToxCast chemical library, the majority of which are active ingredients of current or former food-use pesticides, with a smaller number of industrial chemicals of environmental relevance included.16 Additional chemicals were added from the ToxCast phase II chemical library, spanning a much larger diversity of environmental chemicals and use-group categories including antimicrobials, fragrances, green chemistry alternatives, food-additives, toxicity reference compounds, and pharmaceuticals.17 Lastly, the remaining chemical structures were added from the ToxCast E1K chemical library; this chemical library was designated to be run on a subset of ToxCast endocrine related assays and included a relatively large subset of endocrine reference compounds. Individual chemical library information can be found at http:// www.epa.gov/ncct/toxcast/chemicals.html, with a downloadable structure file for the entire inventory available from the U.S. EPA DSSTox Web site (http://www.epa.gov/ncct/dsstox/sdf_toxcst.html). The majority of chemicals were shipped as 20 mM stock solutions in DMSO, whereas relatively few compounds that were insoluble at 20 mM were provided at lower concentrations, 2−15 mM. During screening, 2 μM MG132 (Tocris) was used as a positive control for cytotoxicity, and 200 and 500 pM 17β-estradiol (Tocris) (E2) was used as a positive control for estrogen response. To characterize the specificity of the assay, progesterone (Sigma), dexamethasone (Sigma), aldosterone (Sigma), and EGF (Life Technologies) were used as reference compounds for response to the progesterone receptor, glucocorticoid receptor, mineralcorticoid receptor, and EGF receptor pathways, respectively. Experimental Procedures. The xCELLigence system Multi-EPlate stations were used to measure the time-dependent response to chemicals described above by RTCA. Each compound was tested in an eight-point, 1:4 serial dilution series starting at a maximum final concentration of 100 μM. A maximum starting concentration of 0.5% DMSO was present in the 100 μM chemical samples and was diluted along with the test article dilution series. The screen was performed in biological duplicate using two separate, 96-well, E-Plates 96 for each dilution series (n = 2). Positive controls (MG132 and E2) and a negative control (assay media) were tested in quadruplicate on each testing plate. Then, 0.5% and 0.125% DMSO were tested in duplicates in each plate to serve as solvent controls for the 2 highest concentrations of testing compounds: 100 μM and 25 μM. Reference compounds were tested with 8 concentrations with 1:5 serial dilutions. All screening was carried out by ACEA Biosciences, Inc. (San Diego, CA). T-47D cells purchased from ATCC were maintained in RPMI1640 media supplemented with 10% characterized fetal bovine serum (FBS). Before screening, T-47D cells were preconditioned in assay medium: phenol red-free RPMI1640 supplemented with 10% charcoal-stripped FBS. Cells were then detached and seeded in E-Plates 96 in assay medium. After overnight monitoring of growth once every hour, compounds were added to T-47D cells and remained in the medium until the end of the experiment. Cellular responses were then recorded once every 5 min for the first 5 h, and once every hour for an additional 100 h. Data Processing. Raw impedance data from the xCELLigence system were collected for each time point before and after compound addition and referred to as the cell index (CI). Data were collated into .txt files and converted to normalized cell index (NCI) according to the following equation:

NCI(Ti ) =

(SMA) was used to smooth the data in order to combine replicates for time course analysis:

SMA t =

NCI t − 6 + NCI t − 5... + NCI t + 6 13

where, SMAt is the simple moving average at time t, and NCIt is the NCI at time t. The SMA was implemented because although each chemical was tested in duplicate on separate plates over a comparable time frame, the exact time points between plates were not identical. For this reason, a SMA was chosen to combine replicates since pairing of time points was not feasible. E2 was run in quadruplicate on each plate, and the maximum average E2 response on each plate was used as a positive control for all the test chemicals on that plate. Vehicle control (0.5% DMSO) wells were run in duplicate on each plate, and the maximum average NCI was used as a negative control for all the test chemicals on that plate. If a chemical sample was run on two different plates, then the maximum NCI values for the positive and negative controls were averaged. All smoothed NCI values for the treatment chemicals were then converted to a percentage of the positive control value. For cell loss, the NCI value at the time of compound administration was considered to represent complete (100%) viability. MG132 (2 μM), a proteasome inhibitor and known cytotoxic agent, was used as the positive control for cell loss and was tested in quadruplicate on each plate. The minimum average response on each plate was used as a positive control for cell loss for all the test chemicals on the corresponding plate. If a chemical sample was run on two different plates, then the minimum NCI values for MG132 were averaged. If an NCI value for MG132 fell below zero, the response was considered to be below the limit of detection and was replaced with the minimum value greater than zero across all plates. All smoothened NCI values were then converted to a percentage of positive control, which was considered to represent no (0%) viability. Concentration−Response Curves. All concentration−response curves were fit using nonlinear least-squares regression in the opensource statistical software R and the sfsmisc package.18,19 For the cell growth analysis, cytotoxicity was ascribed to a particular chemicalconcentration level when the response fell below 15% of the next lower test concentration. This concentration and all higher concentrations were flagged and excluded from the concentration−response data prior to curve fitting. If a minimum of four responses remained after cytotoxicity filtering, then concentration−response curves were fit for each time-point using the Hill-equation:

Y=T−

(T − B) 1+

W

X ( AC50 )

where, Y is the response variable (% E2), T is the upper asymptote, B is the lower asymptote, X is the concentration, AC50 is equal to the concentration at which 50% of activity (response) occurs, and W is the Hill-slope coefficient. The Hill-slope coefficient was constrained between 1 and 8. The minimum value of Y produced was recorded as the minimum efficacy (EMIN), and the maximum value of Y produced was recorded as the maximum efficacy (EMAX). If the EMAX reached 25% of the window between the maximum growth attained by the negative control (0.5% DMSO) and the maximum growth of the positive control (200 pM E2 for phase II chemicals, and 500 pM E2 for PhaseIv2 and E1K chemicals), then the chemical was assessed as active for T-47D cell growth. To ensure the quality of the controls, the ratio of active chemicals to the dynamic range of the negative and positive controls was calculated for each plate. Plates that were 3 standard deviations from the mean were substituted with the average negative and positive control values across all plates. This helped prevent false positives due to controls with a low dynamic range. After substituting with the average controls, the concentration−responses for these curves were refit. If the EMIN for a chemical was greater than the threshold for activity (25%), then the chemical’s AC50 was set to the minimum concentration. A final manual curation was performed where active and inactive calls could be adjusted if the systematic curve fitting did not accurately reflect the data.

CI(Ti ) , {i = 1, 2, 3, ...N } CI(Tk)

where CI(Tk) is the cell index at Tk, the last time point before compound addition, CI(Ti) is the cell index at Ti, the i-th measured time point, and N is the number of total measured time points. Wells flagged for technical issues such as contamination or electronic failure were removed from analysis and substituted with the average of the duplicate chemical’s time points before and after the missing value. Additional details regarding these wells can be found in Supporting Information. Data were then grouped by chemical, and a simple moving average C



Article

Figure 2. Example of growth curve translation to the concentration−response curve. In this example, the effect of treatment with genistein on cell proliferation was plotted over time and concentration. As the cell number increased, the impedance increased. The treatment had an apparent effect on cell growth or viability, which can be measured through the relative decrease in electrical impedance. The circles on the growth curve correspond to the time of the concentration−response curve plotted on the bottom. assay in order to have a comparable metric to the other in vitro ER assays, which were measured at a single time point. For the transactivation assay, all chemicals were tested in 8-point concentration−response format using 3-fold serial dilutions from a top concentration of 100 μM (phase I) or 200 μM (phase II) in Attagene Inc.’s cellular biosensor system (Factorial). This system combines libraries of cis- and transregulated transcription factor reporter constructs with a highly homogeneous method of detection, enabling simultaneous evaluation of multiplexed transcription factor activities. The ERα assay end point from the TRANS system (ATG_ERa_TRANS) and the corresponding estrogen receptor response element from the CIS system (ATG_ERE_ CIS) comprised the two ER transactivation assays used in all subsequent assay comparisons.20,21 The human ER binding assay (hER) (Caliper, a PerkinElmer company, Hanover, MD) assessed whether the test compound could potentially interrupt estradiol binding with the estrogen receptor.22,23 The hER biochemical assay was performed with human estrogen receptor proteins, tritium-labeled estradiol, and the test compound up to 50 μM. Changes in radioactivity between no test chemical and test chemical determined whether the chemical interfered with estradiol−estrogen receptor binding. For additional information regarding these assays, see Supporting Information. Chemicals with no AC50 value were assigned an arbitrary value of 1M. All AC50 values were transformed according to the following formula:

This step was performed for all 80 h concentration−response curves. Both systematic and manual calls for cell growth and cell loss are annotated in Supporting Information, Files 1 and File 2,respectively. An example of how the growth profiles translate into concentration− response curves can be seen in Figure 2. A separate but similar approach was taken to assess cell loss. Concentration−response curves were fit for each time-point using the same Hill equation as the growth curves but replaced with a negative Hill-slope coefficient. B was constrained to 0 (100% MG132), and the minimum value of Y produced was recorded as the EMIN. If the EMIN value dropped an arbitrary 25% below the starting value of 100%, then the chemical was considered active by cell loss. Replicate Concordance. All active and inactive chemicals at 80 h after compound administration that had at least two blind duplicates across all three chemical libraries were used to assess replicate concordance. If a chemical produced either all active or inactive responses across available replicates, then that chemical was given a concordance value of 1. If a chemical produced 50% active responses, then that chemical was given a concordance value of 0. For all other combinations, the concordance value was based on the number of active responses divided by the total number of replicates. This analysis was performed separately for cell growth and cell loss. ER Assay Comparison. Available data on ToxCast phase I and phase II, totaling 957 unique chemicals, from ER transactivation assays and ER binding were compared to the results from the cell growth analysis at 80 h. A single time point (80 h) was chosen for the cell growth

AC50transformed = − log10(AC50original) + 6 D



Article

Table 1. Chemical Library Results Summary and Replicate Concordance chemical no. chemical no. unique no. active chemicals - percentage active - replicate concordance no. active percentage active - replicate concordance library samples chemicals cell growth cell growth (%) -cell growth chemicals - cell loss cell loss (%) cell loss Phase Iv2 Phase IIab E1K total a

311 700 1000 2015a

293 675 880 1819a,b

16 81 107 204

5.5 12.0 12.2 11.2

0.96 0.78 0.88 0.84

74 130 156 360

25.3 19.2 17.7 19.8

0.96 0.82 0.92 0.89

Four additional reference chemicals were included. bSome compounds were run in more than one chemical library.

Figure 3. Growth curve profiles. Representative growth curves of various test compounds measuring changes in normalized cell index (NCI) over time (h). Biphasic growth curves can be seen for some endogenous and synthetic steroid hormones including corticosteroids (B,F), progestins (A,C), and some androgens (D,E), whereas linear growth curves are observed for compounds known to bind to the estrogen receptor (ER) (G−I). This transformed AC50 increases with potency and sets inactive compounds to 0. Compounds that were positive in the hER binding assay were tabulated as “Binding,” and all others were tabulated as “Non-Binding.” Pearson correlation coefficients (r) were calculated for the transformed

AC50 values between the growth assay and transcriptional activation assays for all, binding and nonbinding, categories independently. Conditional Probabilities. Conditional probabilities were calculated to determine whether or not chemicals classified as positive for ER E



Article

Table 2. Twenty-Five Most Potent and Efficacious Compounds in the Cell Growth Assay at 80 h CASRN 77-40-7 50-28-2 57-91-0 446-72-0 27193-86-8 80-05-7 84-16-2 57-63-6 72-33-3 84852-15-3 17924-92-4 105624-86-0 53-16-7 474-86-2 2971-36-0 521-18-6 50-27-1 56-53-1 486-66-8 68-22-4 1478-61-1 10161-33-8 140-66-9 57-83-0 104-43-8

chemical name bisphenol bb 17beta-estradiolb

classification

bisphenol endogenous steroid hormone 17alpha-estradiol endogenous steroid hormone genisteinb bisphenol dodecylphenol alkylphenol bisphenol ab bisphenol meso-hexestrol bisphenol steroid 17alpha-ethinylestradiolb mestranol steroid alkylphenol 4-nonylphenol, branchedb zearalenone phenol 5hpp-33 phenol estrone endogenous steroid hormone equilin endogenous steroid hormone hpte (2,2-bis(4-hydroxyphenyl)- bisphenol 1,1,1-trichloroethane) 5alpha-dihydrotestosterone endogenous steroid hormone estriol endogenous steroid hormone diethylstilbestrolb bisphenol daidzein bisphenol norethindrone steroid bisphenol af bisphenol 17beta-trenbolone steroid 4-(1,1,3,3-tetramethylbutyl) alkylphenol phenol progesterone endogenous steroid hormone 4-dodecylphenol alkylphenol

time since time since treatment treatment (h) when first active (h)

AC50 (μM)a

EMAX (% E2)

EMIN response cutoff for activity (% E2) (% growth above starting value)

80.0 79.8

46.2 44.9

0.283 ≤0.006

128.93 128.72

25.28 84.56

47.45 59.66

80.0

44.0

≤0.006

127.35

92.59

49.76

79.7 80.1 79.5 79.7 79.7 80.1 80.0 80.0 79.8 79.5

48.9 44.1 47.1 46.7 13.7 41.3 46.4 52.1 49.9 52.4

0.082 0.229 0.387 ≤0.006 ≤0.006 ≤0.006 0.304 0.022 0.386 ≤0.006

123.09 117.50 116.70 115.46 114.87 114.18 110.94 109.84 108.81 105.17

34.56 41.08 32.73 93.70 104.75 89.55 26.33 64.00 20.47 85.66

46.99 57.00 51.63 56.14 47.14 54.32 48.34 64.34 48.04 51.97

79.7

50.0

≤0.006

103.63

79.63

53.64

79.7

50.7

0.102

101.65

35.61

57.65

79.6

43.1

0.181

100.35

45.01

55.11

80.1

56.4

≤0.006

97.87

78.38

49.36

79.0 79.5 79.9 79.7 79.7 79.7

48.0 56.1 40.9 52.1 40.7 52.0

≤0.006 0.400 ≤0.006 0.118 ≤0.006 0.230

95.87 90.25 89.26 88.41 88.21 82.83

87.12 30.34 53.55 30.88 59.43 26.68

46.46 54.21 51.56 55.79 55.57 49.09

79.8

43.8

0.079

77.86

48.35

49.06

80.0

65.0

0.244

75.21

20.90

60.22

Compounds that showed activity at the lowest concentration tested are denoted with “≤”. bCompounds that were present in multiple chemical libraries, the one with the highest EMAX was included in this table. a

binding had increased probability of testing positive in the ATG_ERa_ TRANS, ATG_ERE_CIS, or the T-47D cell growth assay compared to the chemicals classified as ER nonbinding. Chemicals were split into two categories depending on whether they were active in the hER binding assay (Binding) or inactive in the hER binding assays (Non-Binding). For each of the categories (Binding or Non-Binding), the total number active in 0, 1, 2, or all 3 ER assays (T-47D cell growth, ATG_ERa_ TRANS, or ATG_ERE_CIS) was divided by the total number of active chemicals in that category to calculate conditional probabilities. Logistic Regression Model. In order to determine if increased potency and efficacy in the T-47D cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS increased the likelihood of being positive in the hER binding assay, logistic regression was performed using a generalized linear model implemented using the R statistical software. The AC50transformed and efficacy normalized to E2 from the cell growth assay at 80 h and both transactivation assays were used to model the binary dependent variable for ER receptor binding (Binding = 1; Non-Binding = 0). Odds ratios were then calculated by exponentiating the coefficient for each input variable.

multiple ToxCast libraries. The overall concordance of these replicates was 0.84 and 0.89 for cell growth and cell loss at 80 h, respectively (Table 1). A total of 204 of 1819 chemicals caused increased cell growth at 80 h after treatment. The efficacy values ranged from 28% to 156% of the E2 positive control, and the potency values spanned the full extent of concentrations tested here. With regards to inhibition of cell growth, a total of 360 chemicals were flagged for >25% drop in impedance at 80 h after treatment. Maximum inhibition ranged from 26% of the negative control down to 100%. Some chemicals caused both an increase in cell proliferation at lower concentrations with higher concentrations causing inhibition of cell growth. Growth curves were created for all compounds and can be found in Supporting Information, File 3. Figure 3 illustrates representative profiles seen throughout the chemical set. Glucocorticoids, mineralocortocoids, and progestins caused a distinctive time-dependent biphasic response profile with an initial increased impedance lasting approximately 24 h followed by a return to near baseline growth rates (Figure 3A−F), compared to the more linear growth curve seen by some androgens and estrogens (Figure 3G−I). It should be noted, however, that some estrogens at high concentrations also had tendencies to show time-dependent, biphasic growth characteristics (Supporting Information, File 3). A few chemicals, such as 5αdihydrotestosterone, produced both the initial increase followed

■

RESULTS A total of 1998 chemical samples including reference chemicals were screened for effects on cell growth. This chemical library included 1815 unique chemicals from the various ToxCast libraries and 4 embedded reference compounds, totaling 1819. A total of 264 replicated samples consisting of 69 chemicals were added as duplicates or triplicates or else were contained in F



Article

Table 3. Paraben Compounds: T-47D Cell Growth Potency and Efficacy at 80 h

a

CASRN

chemical name

time since treatment (h)

AC50 (μM)

EMAX (% E2)

response cutoff for activity (% growth above starting value)

5153-25-3 1219-38-1 17696-62-7 1085-12-7 94-26-8 94-18-8 94-26-8 94-13-3 120-47-8 120-47-8 99-76-3

2-ethylhexylparaben octylparaben phenylparaben heptylparaben butylparabena benzylparaben butylparabena propylparaben ethylparabena ethylparabena methylparaben

80.1 80.1 80.1 79.9 80.0 80.1 79.7 80.0 80.1 80.1 80.0

0.607 1.143 1.332 1.492 1.776 1.871 3.426 4.329 7.617 7.860 NA

82.55 90.91 65.78 82.94 99.23 69.70 116.36 88.88 63.35 65.69 40.98

54.08 54.32 54.08 50.21 51.39 61.94 53.61 49.14 52.77 52.78 49.14

Chemical was included in two different ToxCast chemical inventories.

(r = 0.98, p < 0.001), compared to chemicals that were positive in both ER transactivation assays but tested negative for ER binding (r = 0.59, p < 0.001) (Figure 4C). This suggests that chemicals active for both transactivation assays and ER receptor binding have an increased likelihood of activating the ER signaling pathway through similar mechanisms, i.e., through ER binding. A complementary analysis was performed between the T-47D cell growth assay and each transactivation assay. The cell growth assay results demonstrated a weak correlation coefficient of 0.42 (p < 0.001) and 0.35 (p < 0.001) across all chemicals for ATG_ERE_CIS and ATG_ERa_TRANS assays, respectively (Figure 4A,B). However, when nonbinding chemicals were excluded, the correlation coefficient increased to 0.70 (p < 0.001) and 0.72 (p < 0.001) for ATG_ERE_CIS and ATG_ERa_ TRANS assays, respectively (Figure 4A,B). Because these chemicals were detected as positive for ER binding and were active for both ER transactivation and T-47D cell growth at similar potencies, these chemicals would be expected to have a higher likelihood of activating the ER signaling pathway through a common mechanism. There were a total of 51 (5%) chemicals that were active in the T-47D cell growth assay and both transactivation ER assays out of the available 957 unique chemicals from ToxCast phase I and phase II chemical libraries. There were a total of 20 (2%) unique chemicals that were considered positive for ER binding, 15 (2%) of which were positive for all three cell growth and transactivation assays. Out of the 20 unique chemicals that were positive for binding in the hER binding assay, 3 were considered negative in all three cell growth and transactivation assays. Overall, this indicates that the hER binding assay is less sensitive, but highly specific. (Supporting Information, File 4). Table 4 shows the conditional probabilities of being active in the cell growth or transactivation assays if also positive in the hER binding assay. Compounds that were negative for ER binding have a decreased probability of being active in either the cell growth or transactivation assays (Table 4). Likewise, compounds active in the hER binding assay have a conditional probability of 0.75 for being active in all three cell growth and transactivation assays (Table 4). A logistic regression model revealed that the demonstrated potency (AC50) and efficacy (EMAX) for a test compound were highly associated with the likelihood of the compound being active in the ER binding assays (p = 3.04 × 10−7). Odds ratios were calculated using the logistic regression coefficients for both the AC50 and EMAX values to determine the contribution of each variable for predicting ER binding. Using the AC50 yielded an odds ratio (OR) of 14.57 (95% CI 3.50−109.04), and using the

by a significant linear growth increase at later time points. The time-dependent biphasic response observed for certain endogenous steroid hormones, such as aldosterone, progesterone, and corticosterone, can also be observed for some synthetic steroids. Treatment with triamcinolone, a synthetic corticosteroid, and norgesterel, a synthetic progestin, also caused time-dependent biphasic growth responses (Figure 3C,F). Although some estrogens caused biphasic growth curves, such as the synthetic estrogen, 17α-ethinylestradiol (Supporting Information, File 3), most of these responses had a more linear growth profile (Figures 2 and Figure 3G−I). The 25 most efficacious compounds can be seen in Table 2. Several phenolic compounds including bisphenol B, bisphenol A, 4,4-sulfonyldiphenol, dodecylphenol, 4-(1,1,3,3-tetramethylbutyl)phenol, and 4-nonylphenol (branched) were among the most efficacious. Though only butylparaben, with a four carbon side chain, was among the 25 most efficacious compounds, the 9 unique paraben chemicals all tested positive except for methylparaben. Results from paraben compounds can be seen in Table 3. Parabens with longer side chains appeared to be more potent inducers of proliferation than parabens with shorter side chains. Concordance among chemical responses across the subset of ToxCast ER assays was determined to test whether chemicals active in one ER assay produce similar responses in other ER assays. Pearson’s correlation was chosen as a metric for assessing the agreement among assays because in order to obtain a high correlation, both assays not only are required to have concordant activity (positive/negative responses) but also have to produce similar potencies. To assess concordance, the potency values (AC50), between the T-47D cell growth assay and two ToxCast transcriptional activation assays (ATG_ERa_TRANS, ATG_ERE_CIS), were compared (Figure 4C). The ATG_ERa_ TRANS and ATG_ERE_CIS assays displayed the highest correlation across all values with a correlation coefficient of 0.64, indicating that compounds active in one assay were likely to be active in the other and also have similar potency (AC50) values between the two assays. This is expected because both are measuring similar, but not identical, ER-mediated transcriptional activation. Although, both assays are measuring ER transcriptional activity, the ATG_ERE_CIS assay uses a full-length reporter construct, whereas the ATG_ERa_TRANS contains only the ligand binding domain. An additional comparison was made to determine if there was a difference in correlation between these two transcriptional assays and chemicals that were also positive in the ER binding assay. AC50 values for chemicals that were positive for ER binding were highly correlated G



Article

Table 4. Estrogen Receptor Binding Conditional Probabilities no. of assays detected as activea

no. chemicals detected as ER bindersb

conditional probability for binders

no. chemicals detected as ER non-bindersb

conditional probability for non-binders

0 1 2 3

3 1 1 15

0.15 0.05 0.05 0.75

664 152 93 36

0.70 0.16 0.10 0.04

a

Assays include T-47D cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS. bBased on the hER binding assay.

stronger indicator of ER receptor binding than efficacy for this set of assays (Table 5). Table 5. Logistic Regression Model Performance no. of coefficients input variable chemicalsa (β)

std. error

intercept

51

−16.90

4.84

mean efficacy (EMAX) mean potency (AC50) overall model

51

0.02

0.03

51

2.68

0.86

51

odds ratio (95% CI)

P-value

4.16 × 10−8 0.0004 (5.23 × 10−13 − 8.72 × 10−5) 1.02 0.53 (0.96 − 1.07) 14.57 0.0012 (3.50 − 109.04) 3.04 × 10−7

a

Only chemicals active in all three assays were used (cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS).

■

DISCUSSION The results from the present study add an important functional component to measuring ER activation, namely, T-47D cell growth kinetics and expand the chemical space to many compounds for which no estrogenic information has been thus far captured. Thirty-two (3%) unique compounds among 957 test compounds tested negative in all other ER assays and tested positive in the T-47D cell growth assay. This subset of chemicals may have induced T-47D cell growth through nonclassical ER signaling pathways.24,25 The other 172 active chemicals would be expected to have a higher likelihood of perturbing the estrogenic pathway through a more canonical, ligand-activated signaling network. One advantage to the real-time capabilities of the xCELLigence platform is the capacity to differentiate between chemical growth profiles that may be indistinguishable from testing strategies that use traditional proliferation assays with fewer time points. Time-dependent biphasic growth curves were observed in some cases, whereby increased impedance occurred, followed by decreased impedance around 30 h after compound administration, with a subsequent increase in impedance sustained throughout the remaining sampling time. Impedance can be impacted by alterations in cell morphology due to cytoskeletal changes and alterations in cellular adhesion machinery, in addition to cell proliferation and cell loss.26 The time-dependent biphasic growth curves appeared predominantly after exposures to progestins, mineralcorticoids, glucocorticoids, and some androgens. The MCF-7 cell line is also commonly used to investigate ER induced cell growth and is the cell line used in the E-SCREEN. Wang et al. demonstrated that these cell lines perform similarly.14 Both MCF-7 cells and T-47D cells are sensitive to androgens, but these activities may occur via two different mechanisms. Unlike T-47D cells, MCF-7 cells express aromatase, which converts testosterone to estrogens; however, T-47D cells express the androgen receptor.14

Figure 4. Comparison of cell growth results with additional er assays. Chemical AC50 values were log transformed so that the axis increases with potency. The potency values were compared with two transactivation assays (ATG_ERa_Trans and ATG_ERE_CIS). Chemicals positive for one of three ER binding assays were colored red, and chemicals testing negative in all three binding assays were colored red. Linear regression demonstrated a higher linear correlation between compounds that were also detected to bind to ER.

EMAX resulted in a weaker and nonstatistically significant OR of 1.02 (95% CI 0.96−1.07), signifying that chemical potency is a H



Article

tested positive in T-47D cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS assays. The paraben chemical class is composed of commonly used preservatives found in a wide range of household items, some of which include cosmetics, shampoos, pharmaceuticals, and food additives.40 Parabens have demonstrated activity in a variety of ER assays, and an increased paraben side chain length positively correlates with ER activation potency.35,41 This was highly consistent with the T-47D cell growth assay described herein, in which parabens with longer side chains achieved greater potency than parabens with shorter side chains (Table 3). Relative efficacies for most of the parabens were between 60% and 90% of estradiol, but no structurally related trend for efficacy was apparent. Parabens all demonstrated linear growth profiles, suggesting the activation of different pathways compared to the compounds associated with time-dependent, biphasic growth profiles. Studies have shown that parabens are metabolized by esterases in vivo and that varying amounts of intact parabens are absorbed upon dermal application of paraben-containing products.42 The current assay (similar to many in vitro ER assays) shows ER activity for these compounds, indicating a limited metabolic capacity. Within ToxCast, this assay is part of a larger battery of ER assays, including some with more significant metabolic capacity to more fully characterize the ER activity of these compounds. The ratio of ERα to ERβ is an important factor in determining the phenotypic response after estrogenic exposure.43−45 Genistein, a naturally occurring phytoestrogen, resulted in a growth response similar to that of 17β-estradiol, albeit less potent; however, genistein has been associated with decreased cell proliferation in certain cell lines, e.g., Caco2-BBe, MCF-7, and PC-3.46,47 Although, genistein may bind to ERα, it preferentially binds to ERβ.43 ERβ has been shown to counter the proliferative response typically associated with ERα activation.43,44 T-47D cells have a much higher basal ERα/ ERβ ratio, resulting in increased proliferation after genistein exposure.14,48,49 Sotoca et al.49 used a T-47D-ERβ recombinant cell line with enforced expression of ERβ and managed to attribute the observed antiproliferative effect to ERβ activation. This suggests that the high ratio of ERα/ERβ in T-47D cells is the likely cause for the observed proliferative response by genistein in this study. In addition to the impact of ERα to ERβ ratios, cofactor availability may be tissue- and cell type-dependent and could alter the observed response.50,51 None of the evaluated assays offer a single, complete solution to screening chemicals for ER activity. Every assay has strengths and limitations (e.g., issues with sensitivity toward weak compounds, different types of assay interference callusing false positive and false negative calls, and metabolic capacity) For reasons not entirely understood, the ATG_ERa_TRANS and ATG_ERE_CIS assays are prone to false positives under conditions of oxidative stress.21 As previously described, the T-47D cell growth assay is induced by chemicals other than estrogens. Lastly, the high-throughput hER binding assay is less sensitive compared to the T-47D cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS assays (Figure 4). The hER binding assays were exposed to chemical concentrations of up to 50 μM, and the T-47D cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS assays were tested with chemical concentrations up to 100 μM. This provides a partial explanation for why the hER binding assays did not detect the weakly active compounds, while detecting the most potent chemicals; however, many of the chemicals that were positive in the T-47D cell growth, ATG_ERa_TRANS, and ATG_ERE_CIS assays produced AC50 values < 50 μM and were not detected in the hER binding assay. The goal of the cross-assay

Two isoforms of the progesterone receptor (PR) have been identified, PRA and PRB, and both are expressed in T-47D cells with an approximate ratio of 0.5−0.9 PRA/PRB.27−29 Activation of PRA is associated with morphological changes, such as cell rounding and loss of cell adhesion, and is thought to contribute to increased malignancy in tumors expressing high ratios of PRA/PRB; whereas PRB is associated with a proliferative response.24,28 Biphasic proliferative responses have been observed in T-47D cells expressing only the PRB isoform, with increased numbers of cells retained in G2/M+S until approximately 15−48 h, upon which a late arrest in G1 occurred, roughly corresponding to the biphasic growth curves observed in the present study (Figure.3).24,27 Furthermore, PRB has been shown to initially upregulate cell cycle promoters, cyclin D1 and cyclin dependent kinase (CDK2), and subsequently upregulate cell cycle inhibitors (e.g., p21, p27klp1), resulting in a biphasic response corresponding with late G1 arrest following progesterone exposure.24,27 PR, AR, and the glucocorticoid receptor (GR) have been identified as inducers of serum and glucocorticoid regulated kinase-1 (SGK-1), which regulates cell proliferation and cell cycle progression and may explain the similar growth profile observed by androgens, glucocorticoids, and progestins.24,30,31 Although it is not clear what caused the timedependent biphasic growth profile, T-47D cells do express several steroid hormone receptors, and hormones are known to impact cell cycle mediators via numerous mechanisms.24,27,30−34 . In addition, electron transport chain inhibitors, pyridaben and rotenone, also potently induced cell growth in this assay. An ancillary experiment was conducted that demonstrated that, unlike estradiol, the growth induced by these compounds was not attenuated by the ER antagonist, ICI 182,780. The inability for ICI 182,780 to block the proliferative response indicates that these compounds are not inducing cell growth through an ER mechanism (see Supporting Information, Figure S1). Mechanisms regulating cell growth under hypoxic conditions in cancer cells have been previously described and may be responsible for the non-ER related cell growth observed in response to some chemical exposures, such as the electron transport chain inhibitors, pyridaben and rotenone.35−37 Many of the most efficacious compounds in the T-47D cell growth assay included steroids and phenol compounds (Table 2). The ability for phenolic compounds to bind and activate ER has been demonstrated in a number of studies.35,36 Out of the 25 most potent chemicals, 11 are steroids (either naturally occurring steroid hormones or pharmaceutical agents), 8 bisphenols (compounds with 2 phenol rings connected with a simple or complex linker), 4 alkylphenols, and 2 more complex compounds containing phenol groups. Therefore, the major classes of growth enhancers in this assay are steroids and phenols. One of these compounds (HPTE) is a major metabolite of the insecticide methoxychlor. Some phenolic compounds, such as bisphenol A and 4-nonylphenol are high-production volume (HPV) chemicals used in a wide variety of applications, including food and drink packaging, flame retardants, polycarbonate plastics, detergents, and polystyrene tubes.37−39 Potential for human exposure to these HPV chemicals is relatively high due to their use in a wide range of manufacturing applications. Bisphenol-A, for example, has been detected in blood and urine samples ubiquitously in the population.37 Although the estrogenic potential of compounds like bisphenol A and 4-nonylphenol has been previously studied, limited data exists on the potential estrogenicity for many chemicals tested in this chemical library, such as butam. Butam is an herbicide that I



Article

Notes

comparison is not to determine an optimal ER assay but to demonstrate that most chemicals detected as binding to ER and causing transcriptional activation of ER also have a high likelihood of causing increased cell growth in this ER responsive cell line. The cross-assay comparisons provide a better understanding of each assay’s strengths and weaknesses in order to better characterize an individual chemical’s estrogenic potential. When considered in the context of other ER in vitro assays, the T-47D cell growth assay provides an important functional measurement of estrogenic activity. This assay may also be an important indicator for nonclassical mechanisms of ER activation; the importance of these mechanisms in regulating proliferative responses are becoming increasingly apparent.52−54 Epidermal growth factor, which resulted in increased cell growth in the T-47D cell growth assay, has been shown to activate ER by the MAPK pathway and phosphorylation of ER without binding to the ligand binding domain of ER.55 It is possible that some of the discrepancies between active and inactive responses between the cell growth assay and the two transcriptional activation assays may be due to an estrogenic response not mediated by gene expression through the estrogen response element. Overall, the T-47D cell growth assay can provide important information regarding a chemical’s potential to interact with the ER signaling pathway. Our approach is to combine multiple ER assays (including the current one), plus physicochemical and chemical structure information to help classify chemicals as to their estrogenic potential. These data will then be combined with exposure information to develop a prioritization scheme for EDSP testing. Further work is needed to include a more comprehensive panel of assays, covering other key molecular initiating events in the ER pathway. Currently, efforts within the ToxCast program are underway to include assays that measure ER signaling before and after treatment of the test chemical with a liver S9 fraction to determine effects of xenobiotic metabolism and to include assays that measure receptor dimerization and nuclear translocation for both ERα and ERβ. The present analysis of T-47D cell growth provides significant information relating duration of exposure, potency, and efficacy to effects on cell growth or cell loss. In addition, the data-driven characterization of ER in vitro assays provides an important step toward developing a more comprehensive and integrated approach for testing chemicals for their potential modulation of ER signaling.

■

The views expressed in this article are those of the authors and do not necessarily reflect the views of policies of the U.S. Environmental Protection Agency. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. The authors declare the following competing financial interest(s): EPA authors declare they have no actual or potential competing financial interests. Y.A., C.J., and M.S. are employees of ACEA Biosciences, Inc. (San Diego, CA) conducting work under EPA contract EP-D-13-009.

■

ABBREVIATIONS AC50, activity concentration at 50% efficacy; CDK, cylclindependent kinase; CI, cell index; E2, 17β-estradiol; EDC, endocrine disrupting chemical; EDSP, endocrine disruptor screening program; EMAX, maximum efficacy; EMIN, minimum efficacy; ER, estrogen receptor; FFDCA, Federal Food, Drug, and Cosmetic Act; FQPA, Food Quality Protection Act; GR, glucocortocoid receptor; hER, human estrogen receptor; HPV, high production volume; HTS, high-throughput screening; NCI, normalized cell index; OR, odds ratio; PR, progesterone receptor; PRA, progesterone receptor subtype-A; PRB, progesterone receptor subtype-B; RTCA, real time cell analysis; SGK-1, serum and glucocorticoid regulated kinase-1; SMA, simple moving average; T1S, tier 1 screening; U.S. EPA, United States Environmental Protection Agency

■

(1) Tilghman, S. L., Nierth-Simpson, E. N., Wallace, R., Burow, M. E., and McLachlan, J. A. (2010) Environmental hormones: Multiple pathways for response may lead to multiple disease outcomes. Steroids 75, 520−523. (2) Soto, A. M., and Sonnenschein, C. (2010) Environmental causes of cancer: endocrine disruptors as carcinogens. Nat. Rev. Endocrinol. 6, 363−370. (3) Davis, D. L., Telang, N. T., Osborne, M. P., and Bradlow, H. L. (1997) Medical hypothesis: bifunctional genetic-hormonal pathways to breast cancer. Environ. Health Perspect. 105, 571. (4) Doisneau-Sixou, S. F., et al. (2003) Estrogen and antiestrogen regulation of cell cycle progression in breast cancer cells. Endocr. Relat. Cancer 10, 179−186. (5) Shiozawa, T., et al. (2004) Estrogen-induced proliferation of normal endometrial glandular cells is initiated by transcriptional activation of cyclin D1 via binding of c-Jun to an AP-1 sequence. Oncogene 23, 8603−8610. (6) Crowder, R. J., et al. (2009) PIK3CA and PIK3CB inhibition produce synthetic lethality when combined with estrogen deprivation in estrogen receptor−positive breast cancer. Cancer Res. 69, 3955−3962. (7) Jordan, V. C., and Morrow, M. (1999) Tamoxifen, raloxifene, and the prevention of breast cancer. Endocr. Rev. 20, 253−278. (8) Oh, D. S., et al. (2006) Estrogen-regulated genes predict survival in hormone receptor−positive breast cancers. J. Clin. Oncol. 24, 1656− 1664. (9) Vogel, V. G., et al. (2006) Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other disease outcomes: the NSABP study of Tamoxifen and Raloxifene (STAR) P-2 trial. J. Am. Med. Assoc. 295, 2727−2741. (10) U.S. EPA (2011) Endocrine Disruptor Screening Program for the 21st Century: EDSP21 Work Plan. http://www.epa.gov/endo/pubs/ edsp21_work_plan_summary%20_overview_final.pdf [accessed Jan 12, 2012], http://www.epa.gov/endo/pubs/edsp21_work_plan_ summary%20_overview_final.pdf. (11) U.S. EPA (2011) Response to Corrective Action Plan for OIG Report No. 11-p-0215, EPA’s Endocrine Disruptor Screening Program

ASSOCIATED CONTENT

* Supporting Information S

Concentration−response curve parameters and activity calls for cell growth (positive direction); concentration−response curve parameters and activity calls for cell loss (negative direction); growth curves; AC50 and EMAX values for cell growth, ER binding, ER transactivation assays. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*National Center for Computational Toxicology, Office of Research and Development, U.S. Environmental Protection Agency, 109 T.W. Alexander Drive (B205-01), Research Triangle Park, NC 27711. Phone: 919-541-3085. Fax: 919-5413513. E-mail: [email protected]. Funding

All funding was provided by the U.S. Environmental Protection Agency. J



Article

Should Establish Management Controls to Ensure More Timely Results. (12) Rotroff, D. M., et al. (2013) Using in vitro high throughput screening assays to identify potential endocrine-disrupting chemicals. Environ. Health Perspect. 121, 7. (13) Schenk, B., et al. (2010) The ReProTect Feasibility Study, a novel comprehensive in vitro approach to detect reproductive toxicants. Reprod. Toxicol. 30, 200−218. (14) Wang, S., et al. (2012) Proliferation assays for estrogenicity testing with high predictive value for the in vivo uterotrophic effect. J. Steroid Biochem. Mol. Biol. 128, 98−106. (15) Keydar, I., et al. (1979) Establishment and characterization of a cell line of human breast carcinoma origin. Eur. J. Cancer 15, 659−670. (16) Judson, R., et al. (2009) The toxicity data landscape for environmental chemicals. Environ. Health Perspect. 117, 685−695. (17) Kavlock, R., et al. (2012) Update on EPA’s ToxCast Program: providing high throughput decision support tools for chemical risk management. Chem. Res. Toxicol. 25, 1287−1302. (18) R Development Core Team (2011) R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria. (19) MaechlerM. (2011) sfsmisc: Utilities from Seminar fuer Statistik ETH Zurich, R Package, version 1.0-15, ETH, Zurich, Switzerland. (20) Romanov, S., et al. (2008) Homogeneous reporter system enables quantitative functional assessment of multiple transcription factors. Nat. Methods 5, 253−260. (21) Martin, M. T., et al. (2010) Impact of environmental chemicals on key transcription regulators and correlation to toxicity end points within EPA’s ToxCast program. Chem. Res. Toxicol. 23, 578−590. (22) Knudsen, T. B., et al. (2011) Activity profiles of 309 ToxCastTM chemicals evaluated across 292 biochemical targets. Toxicology 282, 1− 15. (23) Sipes, N. S., Martin, M. T., Kothiya, P., Reif, D. M., Judson, R. S., Richard, A. M., Houck, K. A., Dix, D. J., Kavlock, R. J., and Knudsen, T. B. (2013) Profiling 976 ToxCast Chemicals across 331 Enzymatic and Receptor Signaling Assays. Chem. Res. Toxicol. DOI: 10.1021/tx400021f. (24) Boonyaratanakornkit, V., et al. (2007) The role of extranuclear signaling actions of progesterone receptor in mediating progesterone regulation of gene expression and the cell cycle. Mol. Endocrinol. 21, 359−375. (25) Matés, J. M., Segura, J. A., Alonso, F. J., and Márquez, J. (2008) Intracellular redox status and oxidative stress: implications for cell proliferation, apoptosis, and carcinogenesis. Arch. Toxicol. 82, 273−299. (26) Atienza, J. M., Zhu, J., Wang, X., Xu, X., and Abassi, Y. (2005) Dynamic monitoring of cell adhesion and spreading on microelectronic sensor arrays. J. Biomol. Screen. 10, 795−805. (27) Groshong, S. D., et al. (1997) Biphasic regulation of breast cancer cell growth by progesterone: role of the cyclin-dependent kinase inhibitors, p21 and p27Kip1. Mol. Endocrinol. 11, 1593−1607. (28) McGowan, E. M., and Clarke, C. L. (1999) Effect of overexpression of progesterone receptor A on endogenous progestinsensitive endpoints in breast cancer cells. Mol. Endocrinol. 13, 1657− 1671. (29) Sartorius, C. A., et al. (1994) New T47D breast cancer cell lines for the independent study of progesterone B-and A-receptors: only antiprogestin-occupied B-receptors are switched to transcriptional agonists by cAMP. Cancer Res. 54, 3868−3877. (30) Amato, R., et al. (2009) Sgk1 activates MDM2-dependent p53 degradation and affects cell proliferation, survival, and differentiation. J. Mol. Med. 87, 1221−1239. (31) Shanmugam, I., et al. (2007) Serum/glucocorticoid-induced protein kinase-1 facilitates androgen receptor-dependent cell survival. Cell Death Differ. 14, 2085−2094. (32) Altucci, L., et al. (1996) 17beta-Estradiol induces cyclin D1 gene transcription, p36D1-p34cdk4 complex activation and p105Rb phosphorylation during mitogenic stimulation of G (1)-arrested human breast cancer cells. Oncogene 12, 2315.

(33) Neuman, E., et al. (1997) Cyclin D1 stimulation of estrogen receptor transcriptional activity independent of cdk4. Mol. Cell. Biol. 17, 5338−5347. (34) Roy, P. G., and Thompson, A. M. (2006) Cyclin D1 and breast cancer. Breast 15, 718−727. (35) Blair, R. M., et al. (2000) The estrogen receptor relative binding affinities of 188 natural and xenochemicals: structural diversity of ligands. Toxicol. Sci. 54, 138−153. (36) Gaido, K. W., et al. (1997) Evaluation of chemicals with endocrine modulating activity in a yeast-based steroid hormone receptor gene transcription assay. Toxicol. Appl. Pharmacol. 143, 205−212. (37) Calafat, A. M., et al. (2005) Urinary concentrations of bisphenol A and 4-nonylphenol in a human reference population. Environ. Health Perspect. 113, 391−395. (38) Shelby, M. D. (2008) NTP-CERHR monograph on the potential human reproductive and developmental effects of bisphenol A. NTP CERHR MON. Sep;(22):v, vii-ix, 1-64 passim. (39) Sonnenschein, C., and Soto, A. M. (1998) An updated review of environmental estrogen and androgen mimics and antagonists. J. Steroid Biochem. Mol. Biol. 65, 143−150. (40) Smith, K. W. (2012) Predictors and variability of urinary paraben concentrations in men and women, including before and during pregnancy. Environ. Health Perspect. 120, 1538−1543. (41) Okubo, T., Yokoyama, Y., Kano, K., and Kano, I. (2001) ERdependent estrogenic activity of parabens assessed by proliferation of human breast cancer MCF-7 cells and expression of ERalpha and PR. Food Chem. Toxicol. 39, 1225. (42) Darbre, P. D., and Harvey, P. W. (2008) Paraben esters: review of recent studies of endocrine toxicity, absorption, esterase and human exposure, and discussion of potential human health risks. J. Appl. Toxicol. 28, 561−578. (43) Kuiper, G. G. J. M., et al. (1997) Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 138, 863−870. (44) Gustafsson, J. A. (1999) Estrogen receptor beta–a new dimension in estrogen mechanism of action. J. Endocrinol. 163, 379−383. (45) Pravettoni, A., et al. (2007) Estrogen receptor beta (ERbeta) and inhibition of prostate cancer cell proliferation: studies on the possible mechanism of action in DU145 cells. Mol. Cell. Endocrinol. 263, 46−54. (46) Chen, A. C., and Donovan, S. M. (2004) Genistein at a concentration present in soy infant formula inhibits Caco-2BBe cell proliferation by causing G2/M cell cycle arrest. J. Nutr. 134, 1303−1308. (47) Peterson, G., and Barnes, S. (1996) Genistein inhibits both estrogen and growth factor-stimulated proliferation of human breast cancer cells. Cell Growth Differ. 7, 1345. (48) Strom, A., et al. (2006) Estrogen receptor beta inhibits 17betaestradiol-stimulated proliferation of the breast cancer cell line T47D. Proc. Natl. Acad. Sci. U.S.A. 103, 8298. (49) Sotoca, A. M., et al. (2008) Phytoestrogen-mediated inhibition of proliferation of the human T47D breast cancer cells depends on the ERalpha/ERbeta ratio. J. Steroid Biochem. Mol. Biol. 112, 171. (50) Shang, Y., Hu, X., DiRenzo, J., Lazar, M. A., and Brown, M. (2000) Cofactor dynamics and sufficiency in estrogen receptor−regulated transcription. Cell 103, 843−852. (51) Thenot, S., Charpin, M., Bonnet, S., and Cavailles, V. (1999) Estrogen receptor cofactors expression in breast and endometrial human cancer cells. Mol. Cell. Endocrinol. 156, 85−93. (52) Falkenstein, E., Tillmann, H. C., Christ, M., Feuring, M., and Wehling, M. (2000) Multiple actions of steroid hormonesa focus on rapid, nongenomic effects. Pharmacol. Rev. 52, 513−556. (53) Björnström, L., and Sjöberg, M. (2005) Mechanisms of estrogen receptor signaling: convergence of genomic and nongenomic actions on target genes. Mol. Endocrinol. 19, 833−842. (54) Watson, C. S., Jeng, Y. J., and Guptarak, J. (2011) Endocrine disruption via estrogen receptors that participate in nongenomic signaling pathways. J. Steroid Biochem. Mol. Biol. 127, 44−50. (55) Bunone, G., Briand, P. A., Miksicek, R. J., and Picard, D. (1996) Activation of the unliganded estrogen receptor by EGF involves the MAP kinase pathway and direct phosphorylation. EMBO J. 15, 2174. K


Real-Time Growth Kinetics Measuring Hormone ... - ACS Publications

Recommend Documents